USF Libraries
USF Digital Collections

Counting or discounting television information

MISSING IMAGE

Material Information

Title:
Counting or discounting television information an examination of viewer perceptions of old-age from a cognitive processing perspective of cultivation effects
Physical Description:
Book
Language:
English
Creator:
Pasteur, Lynda
Publisher:
University of South Florida
Place of Publication:
Tampa, Fla
Publication Date:

Subjects

Subjects / Keywords:
Media effects
Elderly
Aging
Accessibility model
Priming
Dissertations, Academic -- Mass Communications -- Masters -- USF   ( lcsh )
Genre:
non-fiction   ( marcgt )

Notes

Summary:
ABSTRACT: Cultivation theory (Gerbner & Gross, 1976) suggests that the cumulative effect of heavy exposure to television's underrepresentation and negative portrayal of people 65+ as sexless, insignificant, and comical can cause people to assume such television-like perceptions of the age group in the real world. This study's purpose was to explore not only television's cultivation effect on viewers' perceptions of the number and nature of people 65+, but also the psychological processes that are expected to mediate this effect. As an extension of Shrum, Wyer, and O'Guinn's (1998) study on the role of source discounting in cultivation, this quasi-experiment employed three experimental conditions-no-priming, source-priming, and relation-priming-to manipulate participants' awareness of television as the source of the information they retrieve to make judgments about people 65+.The experimental conditions were expected to moderate the cultivation effect like they did in Shrum, Wyer, and O'Guinn's (1998) study; in the no-priming condition, but not in the source-priming or relation-priming conditions, heavy viewers were to report more television-like perceptions of people 65+ than light viewers.The results of this study revealed six major findings: first, heavy television viewing does not cultivate viewers to underestimate the 65+ population in the U.S.; second, heavy television viewing cultivates viewers to perceive people 65+ as sexless (specifically, "not sexually attractive" and "not sexually passionate") and comical (specifically, "unintentionally funny"); third, priming is not necessary to induce source discounting of television information for judgment-making about the number and nature of the elderly in the real world; fourth, whether people are prompted to recognize television as an information source, they will discount television information when making judgments about the number of people 65+, and they will count television information when making judgments about the nature of people 65+; fifth, high perceived reality of television encourages heavy viewers to perceive people 65+ as "comical"; sixth, high direct experience with people 65+ discourages heavy viewersto perceive people 65+ as "insignificant in society."
Thesis:
Thesis (M.A.)--University of South Florida, 2008.
Bibliography:
Includes bibliographical references.
System Details:
Mode of access: World Wide Web.
System Details:
System requirements: World Wide Web browser and PDF reader.
Statement of Responsibility:
by Lynda Pasteur.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains 114 pages.

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001993782
oclc - 317153925
usfldc doi - E14-SFE0002412
usfldc handle - e14.2412
System ID:
SFS0026730:00001


This item is only available as the following downloads:


Full Text
xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam Ka
controlfield tag 001 001993782
003 fts
005 20090326160224.0
006 m||||e|||d||||||||
007 cr mnu|||uuuuu
008 090326s2008 flu s 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0002412
035
(OCoLC)317153925
040
FHM
c FHM
049
FHMM
090
P90 (Online)
1 100
Pasteur, Lynda.
0 245
Counting or discounting television information :
b an examination of viewer perceptions of old-age from a cognitive processing perspective of cultivation effects
h [electronic resource] /
by Lynda Pasteur.
260
[Tampa, Fla] :
University of South Florida,
2008.
500
Title from PDF of title page.
Document formatted into pages; contains 114 pages.
502
Thesis (M.A.)--University of South Florida, 2008.
504
Includes bibliographical references.
516
Text (Electronic thesis) in PDF format.
520
ABSTRACT: Cultivation theory (Gerbner & Gross, 1976) suggests that the cumulative effect of heavy exposure to television's underrepresentation and negative portrayal of people 65+ as sexless, insignificant, and comical can cause people to assume such television-like perceptions of the age group in the real world. This study's purpose was to explore not only television's cultivation effect on viewers' perceptions of the number and nature of people 65+, but also the psychological processes that are expected to mediate this effect. As an extension of Shrum, Wyer, and O'Guinn's (1998) study on the role of source discounting in cultivation, this quasi-experiment employed three experimental conditions-no-priming, source-priming, and relation-priming-to manipulate participants' awareness of television as the source of the information they retrieve to make judgments about people 65+.The experimental conditions were expected to moderate the cultivation effect like they did in Shrum, Wyer, and O'Guinn's (1998) study; in the no-priming condition, but not in the source-priming or relation-priming conditions, heavy viewers were to report more television-like perceptions of people 65+ than light viewers.The results of this study revealed six major findings: first, heavy television viewing does not cultivate viewers to underestimate the 65+ population in the U.S.; second, heavy television viewing cultivates viewers to perceive people 65+ as sexless (specifically, "not sexually attractive" and "not sexually passionate") and comical (specifically, "unintentionally funny"); third, priming is not necessary to induce source discounting of television information for judgment-making about the number and nature of the elderly in the real world; fourth, whether people are prompted to recognize television as an information source, they will discount television information when making judgments about the number of people 65+, and they will count television information when making judgments about the nature of people 65+; fifth, high perceived reality of television encourages heavy viewers to perceive people 65+ as "comical"; sixth, high direct experience with people 65+ discourages heavy viewersto perceive people 65+ as "insignificant in society."
538
Mode of access: World Wide Web.
System requirements: World Wide Web browser and PDF reader.
590
Advisor: Kenneth C. Killebrew, Ph.D.
653
Media effects
Elderly
Aging
Accessibility model
Priming
690
Dissertations, Academic
z USF
x Mass Communications
Masters.
773
t USF Electronic Theses and Dissertations.
4 856
u http://digital.lib.usf.edu/?e14.2412



PAGE 1

Counting or Discounting Te levision Information: An Examination of Viewer Perceptions A bout Old-Age From a Cognitive Processing Perspective of Cultivation Effects by Lynda Pasteur A thesis submitted in partial fulfillment of the requirements for the degree of Master of Arts School of Mass Communications College of Arts and Sciences University of South Florida Major Professor: Kenne th Killebrew, Ph.D. Kelly Werder, Ph.D. Kelli Burns, Ph.D. Date of Approval: April 3, 2008 Keywords: media effects, elderly, aging, accessibility model, priming Copyright 2008, Lynda Pasteur

PAGE 2

i TABLE OF CONTENTS LIST OF TABLES ............................................................................................................. iv ABSTRACT ...................................................................................................................... ...v INTRODUCTION ...............................................................................................................1 LITERATURE REVIEW ....................................................................................................6 Old-Age Stereotypes ................................................................................................6 Television’s Depiction of People 65+ .....................................................................8 People 65+ as Sexless ................................................................................14 People 65+ as Insignificant ........................................................................16 People 65+ as Comical ..............................................................................19 Television’s Demographic Repr esentation of People 65+.....................................21 Cultivation Theory .................................................................................................26 Criticisms of Cultivation Theory ...............................................................31 Cognitive Subprocesses in th e Cultivation Effect .................................................32 Perceived Reality of Television .................................................................33 Learning Theory of Cultivation .................................................................35 First-Order Cultivation Effects Models .................................................................37 Weighing and Balancing Model of Cultivation .........................................38 Accessibility Model of Cultivation ............................................................40

PAGE 3

ii Television Viewing Influe nces Accessibility ................................43 Accessibility Mediates Cultivation ................................................44 Motivation to Process Informa tion Moderates Cultivation ...........46 Ability to Process Information Moderates Cultivation ..................47 Television Exemplars are Not Discounted ....................................48 Hypotheses .............................................................................................................50 Cultivation Within Priming Conditions .....................................................50 Cultivation Across Pre-Existing Conditions .............................................54 METHODOLOGY ............................................................................................................58 Sample....................................................................................................................58 Procedure ...............................................................................................................58 Dependent Measures ..................................................................................60 Demographic Measures .............................................................................61 Television Measures ..................................................................................61 Direct Experience Measures ......................................................................63 Data Analysis .............................................................................................63 RESULTS ....................................................................................................................... ...65 Sample....................................................................................................................65 Cultivation Within Priming Conditions .................................................................67 Cultivation Across Other Pre-Existing Conditions ................................................73 Exploratory Analysis .............................................................................................80 DISCUSSION .................................................................................................................... 83

PAGE 4

iii Cultivation Effects .................................................................................................83 Priming Effects ......................................................................................................85 Effects of other PreExisting Conditions ...............................................................87 Limitations of the Study.........................................................................................89 CONCLUSION ..................................................................................................................93 Directions for Future Research ..............................................................................96 LIST OF REFERENCES ...................................................................................................98 BIBLIOGRAPHY ............................................................................................................105 APPENDICES .................................................................................................................107 Appendix A: No-Priming Questionnaire .............................................................108 Appendix B: Source-Priming Questionnaire .......................................................110 Appendix C: Relation-Priming Questionnaire .....................................................112

PAGE 5

iv LIST OF TABLES Table 1: Priming Groups ................................................................................................51 Table 2: Perceived Reality Groups ................................................................................54 Table 3: Direct Experience Groups ................................................................................55 Table 4: Television Viewing Level Statistics ................................................................66 Table 5: Frequency of Experimental Conditions ...........................................................66 Table 6: H1, H2, & H3 Statistics ...................................................................................67 Table 7: H4-P1 t-Test for No-Priming ...........................................................................70 Table 8: H6 t-Test for Source-Priming ..........................................................................72 Table 9: H6 t-Te st for Relation-Priming ........................................................................73 Table 10: Perceived Reality Level Statistics ...................................................................74 Table 11: H8-P3 t-Test.....................................................................................................76 Table 12: Direct Experi ence Level Statistics ...................................................................77 Table 13: H10-P2 t-Test ...................................................................................................79 Table 14: H1, H2, & H3Test of Between Subjects Effects .............................................81 Table 15: GPA and Perceived Reality t-Test ...................................................................82

PAGE 6

v COUNTING OR DISCOUNTING TELEVISION IN FORMATION: AN EXAMINATION OF VIEWER PERCE PTIONS OF OLD-AGE FROM A COGNITIVE PROCESSING PERSPECT IVE OF CULTIVATION EFFECTS Lynda Pasteur ABSTRACT Cultivation theory (Gerbner & Gross, 1976) suggests that the cumulative effect of heavy exposure to television’s underrepresentation and negati ve portrayal of people 65+ as sexless, insignificant, and comical can cau se people to assume such television-like perceptions of the age group in the real world. This study’s purpose was to explore not only television’s cultivation effect on viewers’ perceptions of the number and nature of people 65+, but also the psychologi cal processes that are expect ed to mediate this effect. As an extension of Shrum, Wyer, and O’ Guinn’s (1998) study on the role of source discounting in cultivation, this quasi-experiment employed three experimental conditions—no-priming, source-priming, and relation-priming—to manipulate participants’ awareness of television as the so urce of the information they retrieve to make judgments about people 65+. The experi mental conditions were expected to moderate the cultivation effect like they did in Shrum, Wyer, and O’Guinn’s (1998) study; in the no-priming condition, but not in the source-priming or relation-priming conditions, heavy viewers were to report more television-like percep tions of people 65+ than light viewers. The results of this st udy revealed six major findings: first, heavy

PAGE 7

vi television viewing does not cul tivate viewers to underestimat e the 65+ population in the U.S.; second, heavy television viewing cultivat es viewers to perceive people 65+ as sexless (specifically, “not se xually attractive” and “not sexua lly passionate”) and comical (specifically, “uninten tionally funny”); third, priming is not necessary to induce source discounting of television information for j udgment-making about the number and nature of the elderly in the real world; fourt h, whether people are prompted to recognize television as an information source, they will discount television information when making judgments about the number of people 65+, and they will count television information when making judgments about the nature of people 65+; fifth, high perceived reality of television encourag es heavy viewers to percei ve people 65+ as “comical”; sixth, high direct experience with people 65+ discourages heavy viewers to perceive people 65+ as “insigni ficant in society.”

PAGE 8

1 INTRODUCTION America’s baby boomers, those born between 1946 and 1965, are growing older, and the average age of a United States reside nt is aging with them. According to U.S. Census Bureau (2004) interim predictions, 86.7 million U.S. residents will be age 65 and older (65+) by the year 2050, accounting for 20.7% of the total population. In fact, the bureau estimates that between 2000 and 2050, th e number of U.S. residents age 65 to 84 will increase by 113% while the number of residents age 85 and older will rise by 388%. In light of this imminent demographi c shift in the U.S. population, scholarly research dedicated to the effects of the ma ss media’s portrayal of people 65 and older is both timely and essential. Research of old-ag e portrayals on televisi on, in particular, is warranted for three reasons. First, Ameri cans watch a lot of television. Despite the increasing popularity of streaming video th rough new media such as the Internet, cell phones, and MP3 players, average television vi ewing time continues to rise. According to Nielsen Media Research (2006), the 2005-2006 television year boasted a record-high average of eight hours and 14 minutes per day for total household television viewing; the average amount of television watched by an individual reached a record high of four hours and 35 minutes per day, which amounts to about 32 hours of television viewing per week. Nielsen Media Research (2007) also estimates that the total number of households in the United States owning at least one te levision set will grow from 111.4 million in 2008 to 163.7 million in 2050, amoun ting to a 47% increase.

PAGE 9

2 Second, television has a well-documented reputation for negatively stereotyping people 65 and older and underrepresenting th e age group (Davis & Kubey, 1982; Harris, 2004; Signorielli, 1985). Since the early 1970s researchers from various disciplines— sociology, gerontology, psyc hology, and mass communication—have used content analysis to identify common themes in tele vision’s depiction of people 65+ that cut across multiple television program genres as well as television commercials. The studies reveal that television portrays people age 65+ both negatively (Aronoff, 1974; Peterson & Ross, 1997) and positively (Miller, Leyell, & Mazachek, 2004; Roy & Harwood, 1997) in terms of their physical appearance, health, in telligence, and overall personality. However, television negatively portrays the age group in three consistent ways : people 65 and older are sexless, insignificant, and comical (Cassata & Irwin, 1997; Greenberg, Graef, Fernandez-Collado, Korzenny, & Atkin, 1980; Harwood & Giles, 1992). In terms of demographic representation, nearly all of the research conducted on the relative percentage of age 65+ television characters reveals that they are underrepresented in relation to their true percen tage in the total population (G erbner, Gross, Signorielli, & Morgan, 1980; Signorielli, 2004). The third rationale for studying old-age de pictions on televisi on relates to the medium’s potential power for socializati on. Gerbner and Gross ( 1976) posited that a process called cultivation molds television viewers’ understanding of the world. The cumulative effects of heavy viewing make them believe that the real world, often termed social reality is actually identical to the world th ey see on the television screen. Hence, heavy exposure to underrepresentations and ne gative portrayals of people 65 and older on television could affect peopl e’s judgments about the demographic set size and trait

PAGE 10

3 characteristics of the age group. In other wo rds, television could be molding viewer’s perceptions about the number a nd nature of people age 65+. However, cultivation theory has been heav ily criticized, and it remains a topic of debate in media effects circles to this day. One major criticism is that the theory fails to account for the mediating processes that ta ke place between television viewing and subsequent judgment-making about social reality. To counter this argument, cultivation process models have evolved. The earliest m odels proposed that cultivation occurs in four stages; learning from television yields construction of conceptions about social reality, fostering first-order beliefs that eventually lead to generalized second-order attitudes (Gerbner, Gross, Morgan, & Signorielli 1980; Hawkins, Pingree, & Adler, 1987). Although the linearity of this proces s-oriented model of cultivation was not supported by the research, results did demonstr ate that second-order attitudes do not stem from first-order beliefs, and the tw o should be tested independently. The accessibility model of first-order cultivation effect s (Shrum, 1999b; 2002) emerged from the learning theory of cultivat ion. Suggesting that pe ople typically rely on heuristic, or unlabored, processing strategies when making decisions about social reality, Shrum’s model is based theo retically on two highly supporte d principles of social cognition research, the availability heuristic (Tversky & Kahneman, 1973) and construct accessibility (Wyer & Srull, 1986). Tversky and Kahneman’s availability heuristic proposes that when individuals make judgments about the probability or the frequency of some object or occurrence in nature, they assign higher frequency estimates to those things that come to mind with the greatest ease. Construct accessibility conceptualizes an individual’s long-term memory as an asso rtment of mental storage bins; when an

PAGE 11

4 informed decision needs to be made, people se arch the appropriate bin by starting from the top and moving down so that the most recent information is what is called upon (Wyer & Srull, 1986). The availability heuristic and construct acce ssibility are the basis for the first two of Shrum’s five propositions in the accessibi lity model of cultiva tion, which include: 1) television viewing influences accessibility, 2) accessibility mediates the cultivation effect, 3) television exemplars are not disc ounted, 4) motivation to process information moderates the cultivation effect, and 5) ab ility to process information moderates the cultivation effect (Shrum, 2002, pp. 80-85). The model posits that if people are motivated and able to process a message, they will engage in systematic processing (central, highinvolvement) and discount the television sour ce such that no cultivation effect occurs. However, when either motivation or ability is absent, people will engage in heuristic processing (peripheral, low-i nvolvement) of a message, and the presence or absence of source-priming will determine whether the information source (i.e., television) is discounted. If, as cultivation theory suggests, portrayals of peopl e 65 and older on television can ultimately manipulate viewers’ beliefs a bout the age group, it wi ll be imperative to validate the psychological processe s that are expected to mediate this effect. The current study will explore television’s cultivation of perc eptions about people 65+ by testing the third cognitive subprocess proposed by Sh rum’s accessibility model of cultivation effects, source discounting (Shrum, 2002). As an extension of Shrum, Wyer, and O’Guinn’s (1998) study on the role of source discounting in cultiva tion, this study will employ an experiment involving three

PAGE 12

5 priming conditions—no-priming, source-priming, and relation-priming—to determine the effects of both television vi ewing and priming conditions on first-order social reality perceptions about the number and nature of people 65+. It is hypothesized that the findings of this study will be concurrent with those of Shrum et al. (1998). Priming conditions are expected to moderate the cult ivation effect such that heavy television viewers will report more television-like pe rceptions of people age 65+ than light television viewers under no-priming conditions but not under source-priming or relationpriming conditions. Since it is also important to recognize th at not all television viewers believe that the medium offers them an accurate depict ion of reality (Hawki ns, 1977; Potter, 1986), this study predicted that perc eived reality of television content would influence the cultivation effect. Real life direct experience with people ag e 65+ was also expected to influence participants’ perceptions of the number and nature of the age group.

PAGE 13

6 LITERATURE REVIEW Old-Age Stereotypes Seminal studies on stereotyping of particular population groups re veal that next to gender-based prejudice, ageism is one of th e most ubiquitous forms of discrimination in America. In fact, one important study by K ite, Deaux, and Miele (1991), which sought to understand the interrelationship of stereotypes attached to the two general categories of age and gender, claimed that age stereotyping is extensively widespread. Despite the fact that so much research has been conducted to recognize and interpret gender stereotypes, the results of Kite et al.’ s (1991) study suggest that age stereotypes are even more prominent in American culture. In reference to the 65+ p opulation, Kite Deaux, and Mi ele (1991) asserted that “there is general agreement that the st ereotype is multidimensional and includes characteristics such as ill, tired, grouchy, unlikely to participat e in activities, unhappy, undesirable for company, and physically unattra ctive” (p. 20). This notion was one of the many phenomena their study intended to test. The sample consisted of college students (n=98) and members of a senior commun ity (n=100), and the study examined both group’s beliefs about 35-year-old men and women as well as 65-year-old men and women. Results of the study suggested that when considering traits, role behaviors, and

PAGE 14

7 physical characteristics, the 65-year-old ta rgets of both genders were consistently perceived more negatively than the 35 -year-old male and female targets. Providing evidence for the existence of negative perceptions of old-age among America’s youth, the younger subjects had less favorable perceptions of the 65-year-old targets compared to observations of the older subjects; the 35-year-old targets were rated similarly by both the younger and older resear ch participants. Another notable finding was that although the college-ag ed subjects were prone to ne gative perceptions of the 65year-old targets, the subjects from the senior community “did not reverse this trend” by rating the 35-year-old targets less favorably than the 65-year-old targets, who more closely matched their age (Kite, Deaux, & Miele, 1991, p. 25). Another influential stud y on stereotyping of people 65 and older conducted by Hummert, Garstka, Shaner, and Strahm ( 1994) examined perceptions of the age group across young, middle-aged, and el derly adults in a two-part study. Part one sought to gather a more comprehensive trait list for older adults. Since th e extensive list of 99 positive and negative traits of old-age co mpiled by Schmidt and Boland (1986) was only representative of young adult perceptions, Hummert et al. (1994) looked to the middleaged and elderly participants in their study to identify additional traits of seniors. In part two, the young, middle-aged, and elderly partic ipants were asked to organize the people 65+ traits list that they had compiled in th e first part of the study, ranking the traits according to their assessments of how well each one described the age group. Results of the study supported Hummert, Garstka, Shaner, and Strahm’s (1994) hypothesis that complexity of old-age percep tions increases with age. The middle-aged group of participants could account for far fewe r older adult stereotypes than the elderly

PAGE 15

8 group. Similarly, the middle-aged group had a more complex view of older adult stereotypes than did the young participant gr oup, who came up with the least amount of older adult stereotypes. Signi ficant correlations among percei ved older adult stereotypes among the three age groups’ trait lists revealed seven stereotypes of people 65 and older: Perfect Grandparent, Golden Ager, John Wa yne Conservative, Severely Impaired, Shrew/Curmudgeon, Despondent, and Recluse. It is important to note that the st udies conducted by Kite, Deaux, and Miele (1991) and Hummert, Garstka, Shaner, a nd Strahm (1994) uncovered both positive and negative stereotypes of people 65 and older. “Although the elderly are thought to be grouchy, critical, miserly, and hard of h earing,… they are also viewed as likable, intelligent, and experienced” (Kite et al., 1991, p. 25). Therefore, it is hasty to assume that all stereotypes of people 65 and older ar e adverse; sometimes, albeit less often, older individuals are viewed in a favor able light (Kite et al., 1991). Television’s Depiction of People 65+ Over the past fifty years, nu merous research studies have used content analysis to determine whether television paints a favorab le or unfavorable picture of people 65 and older. Like the research on old-age ster eotyping, research conducted to determine television’s depiction of the age group ha s yielded mixed results ; television’s age 65+ characters are portrayed bot h positively and negatively in terms of their physical appearance, health, intelligen ce, and overall personality. Two of the first and most cited scholars in research on televi sion’s depiction of people 65 and older reported very different results. The first scholar, Petersen (1973),

PAGE 16

9 sampled 30 half-hour time slots from primetime commercial tele vision broadcast in 1972 on the three major networks. Image of pe ople age 65+ was operationalized as the sum of rating scores on a checklist of 21 pair s of attributes (e.g., “friendly/unfriendly”; “liked/disliked”; “strong/weak”), and the attr ibutes were rated by each investigator on a seven-point scale (p. 572). Th e middle position was considered “neutral,” and the three positions on each side were collapsed to be considered “favorable” and “unfavorable.” Although Petersen expected to find that old people had an unfavorable image on television, her results were qui te the opposite; only 18.2% of th e summed attribute ratings were classified as unfavorable A total of 58.7% were favorab le, and 23.1% were neutral. What’s more, 92.9% of the age 65+ characters analyzed were described as “active,” and 82.1% were in “good health” as well as “independent” (p. 573). The other research scholar, Aronoff (1974) studied network prime-time dramatic programming from 1969 to 1971. In stark cont rast to Petersen’s (1973) findings, his results showed that television portrayed the el derly in a mostly negative manner. In fact, the elderly were more likely than characters of any other social age category (childadolescent, young adult, settled adul t, and elderly) to be featured as one of the “bad guys” and to be portrayed as failures (p. 86). In a prominent study that followed the seminal work of Petersen (1973) and Aronoff (1974), Harris and Fei nberg (1977) analyzed sample s of television dramas, comedies, game shows, news programs, a nd commercials broadcast on the three major networks in 1976. The research team found th at on the television programs as well as commercials, health problems increased a nd physical activity d ecreased with age. Twenty-five percent of the age 60 to 70 ch aracters on the television programs were

PAGE 17

10 depicted as being in poor health. In the comm ercials, 35% of the characters in the 60 to 70 age group had health problems. The re search team concluded the following: “Although it is true that in real life the in cidence of health problems is higher among older age groups, it is nonethel ess noteworthy that televisi on has chosen to dramatize subject matter where old people fail (health) rather than subject matter where they are successful...” (p. 466). Sampling dramatic television programmi ng broadcast over an eight-year period, between 1969 and 1976, Signorielli and Ger bner (1978) analyzed prime-time and weekend daytime network television. The rese arch team posited that in terms of their attractiveness, rationality, efficiency, and happiness, the elderly were portrayed less positively than young and middle-aged adults. In 1980, Gerbner, Gross, Signorielli, and Morg an studied an ever larger sample of prime-time and weekend daytime television dramas. The ten-year sample was drawn from network programming broadcast fr om 1969 to 1978. Like Aronoff (1974), the researchers found that older characters were often depicted as “bad guys” and were relatively less successful when compared to their younger counter parts. Additionally, about two-thirds of the elderl y television characters in the sa mple were disrespected by others, and the age group was portrayed as eccentric and foolish. Gerbner et al. (1980) concluded, “In every case, heavier viewing ma kes a consistently negative contribution to the public’s image of the persona l characteristics of the elderly, and the quality of their lives. We did not find watching television to be associated with any positive images of older people” (p. 47).

PAGE 18

11 Nevertheless, Greenberg (1980) came up w ith contradictory results when studying television programming broadcast from 1975 to 1977. Unlike examinations of television programming broadcast during similar time periods (Gerbner, Gross, Signorielli, & Morgan, 1980; Harris & Feinberg, 1977; Signorielli & Gerbner, 1978), Greenberg’s content analysis posited that people 50 and older were depicted in a primarily positive fashion. Specifically, the research team studied the positive social behavior of altruism, or unselfishness, and found that people 50 and older both performed and received acts of humanity at comparable levels to the other age categories. Cassata, Anderson, and Skill (1983) also found mostly positive portrayals of age 55+ characters when they analyzed a sample of daytime network serial dramas broadcast in 1978. Their results suggested that out of th e older characters, age 55+, observed in the sample, 92.9% were portrayed as healthy and over 98% had a “pleasant” physical appearance and demeanor (p. 40). In a ddition, Cassata et al (1983) executed a personality assessment of the 58 characters age 55+ in much the same way as Petersen (1973). Fifteen attributes of opposing positive and negative polarity were presented on a five-point scale where the middle response was considered “neutral,” and the two positions on each side were collapsed to be considered “favorable” and “unfavorable.” While 18.2% of the summed attrib ute ratings in Petersen’s ( 1973) study were classified as unfavorable, even fewer, less than 9%, we re unfavorable in Ca ssata et al.’s (1983) study. Additionally, 51.7% of the older character s were placed in the wealthy or upper middle class socioeconomic categories, and 87% of them were judged to be “friendly” (p. 42).

PAGE 19

12 In an analysis of Saturday morning cartoons, Bishop and Krause (1984) sampled 24 hours of programming broadcast on netw ork television during 1981. When comparing young, adult, and old characters, the res earch team concluded that on cartoon programming, “old” characters were most likely to be portrayed as unhealthy, unattractive, and bad. The researchers noted that “it is striking that, when age was mentioned in these cartoons, the remarks were nearly always negative and easily classified into the stereotypes so frequently found in attitudinal research on aging” (p. 93). Over a decade later, a study conducted by Peterson and Ross (1997) also examined the level of favorability of televi sion characters in three age ranges, but the sample consisted of 1,437 television commercia ls broadcast on the three major networks, one local station, and five cable companies in 1991. The study analyzed not only age 65 and older characters, but also age 45 to 64 characters and characters age 44 and younger. The research pair discovered that the propor tion, 46%, of characte rs age 65+ who made undesirable appearances in television ads was a great deal higher than that of the age 45 to 64 category, 34%, and the age 44 and younger category, about 17%. The results of Peterson and Ross’ (1 997) study suggested that the oldest characters on television commercials ar e indeed the most negatively portrayed. Conversely, two studies published in the same year described the nature of old-age portrayals in television commer cials as relatively positive. A total of 778 commercials were sampled by Roy and Harwood (1997) from a 30-hour block of television programming broadcast on the three major networks in 1994. One hundred and forty-two of the commercials were discarded because they did not feature people, and the human

PAGE 20

13 characters in the remaining 636 commercial s were analyzed. Roy and Harwood’s results suggested that the commercials in their sample featured older adult characters as strong (96.9%), active (100%), happy (93.7%), and lucid (100%). The other 1997 study of television comme rcials, conducted by Hajjar, sampled just over 61 hours of daytime television br oadcast in 1995. Elderl y characters, those assumed to be over 60 years old, were cla ssified as being depi cted favorably or unfavorably according to the presence of one of more positive (“good,” “friendly,” “credible,” etc.) or negative (“inactive,” “una ttractive,” “dependent,” etc.) personality attributes (p. 238). If a character was not obvi ously portrayed as favor able or unfavorable, he or she was labeled neutra l. Hajjar’s results showed that the sampled commercials exhibited more favorable (48%) than unfa vorable (8%) depictions of old-age. A recent content analysis of 50 years of television commercials found the elderly to be generally portrayed in desirable ways Miller, Leyell, and Mazachek (2004) studied commercials from the 1950s to the 1990s to discover whether American advertisements negatively stereotype older persons, as ma ny scholars have argued. After obtaining a convenience sample of 1,662 commercials from multiple databases containing television advertisements from portions of time within th e five decades to be studied, the research team coded 69 commercials portraying a total of 101 elderly people. M iller et al. (2004) categorized each older individual by age-gr oup, “young-elderly” (age 60 to 74) or “oldelderly” (age 75+), and then identified se ven stereotype clusters reminiscent of the Hummert, Garstka, Shaner, and Strahm’s ( 1994) classifications, including: “Productive Golden Ager,” Adventurous Golden Ager ,” “Perfect Grandpa rent,” “John Wayne

PAGE 21

14 Conservative,” “Despondent,” “Shrew/Cur mudgeon,” and “Mildly Impaired” (pp. 325326). Of the aforementioned stereotypes, both positive and negati ve in nature, the favorable Adventurous Golden Ager was found to be the most common. In fact, Miller, Leyell, and Mazachek (2004) posited that 78.2% of the 101 elderly people in their sample of television commercials were portrayed as one of the positive stereotypes (Perfect Grandparent, John Wayne Conservative, A dventurous Golden Ager, or Productive Golden Ager) while only 11.9% were portray ed as one of the negative stereotypes (Despondent, Mildly Impaire d, or Shrew/Curmudgeon). Even more recently, Robinson and Anderson (2006) sampled 45 hours of animated children’s programming from five network and cable sta tions, including 121 different episodes of 41 different programs The study analyzed a total of 82 older characters, age 55+, and found that 59% of th em had positive persona lity traits and 41% had negative personality traits. However, the four most commonl y portrayed traits included both positive and negative attributes. At 37%, the most common trait was intelligent, followed by angry (28%), ha ppy (27%), and senile/crazy (22%). People 65+ as Sexless Clearly, both positive and negative depictions of age 65+ people’s physical appearance, health, intelligence, and overa ll personality are present on television. However, research studies have consisten tly found that the medium portrays the age group as essentially void of positive roman tic involvement and basically sexless.

PAGE 22

15 In Harris and Feinberg’s (1977) analysis of television programming, the research team placed all of the characters in the samp le into age groups and rated each character’s romantic involvement. They reported that th e age distribution of the positive romantic involvement observed in the study was very heavily skewed to wards the younger age groups; characters under age 30 accounted fo r 84% of all of the positive love relationships. On the other hand, character s over age 60 were not shown having any positive romantic involvement at all. Even wh en individuals in this age category were shown in the context of marriage, the relatio nship was never portray ed as sexually active and there were no elements of love to speak of. Signorielli and Gerbner (1978) as well as Signorielli (1983) reported results similar to Harris and Feinberg’s. In Signorie lli’s (1983) sample of prime-time dramatic programming broadcast betw een 1969 and 1981, older characters, especially women, were found to be significantly less likely to engage in romantic relationships. In fact, while 60% of young women and 50% of mi ddle-aged women were depicted as romantically involved, only 9% of older women were portrayed as such. Likewise, Signorielli and Gerbner (1978) reported that in their sample of dramatic television programming, there were almost no depictions of elderly characters being romantically involved. Greenberg, Graef, Fernandez-Collado, Ko rzenny, and Atkin (1980) sampled two weeks of prime-time programs broadcas t in 1977 and 1978 and analyzed television characters in terms of whether they were initiators of intimate sexual references, targets of intimate sexual references, both, or neither. A total of 156 initiators of intimate sexual references and 146 targets of those references were observed. Only 15% of the initiators

PAGE 23

16 and targets were characters ag e 50 to 64, and characters age 65 + were never the initiator or target of any inti mate sexual references. Research shows that not only positive ro mantic involvement and sex, but also displays of affection are lacking in televisi on’s characterizations of people 65 and older. Greenberg, Edison, Korzenny, Fernandez-Co llado, and Atkin (1980) sampled over 3,549 characters in prime time and Saturday morning programs between 1975 and 1977. The research team discovered that characters be tween age 20 and 34 were the most likely to show signs of affection, while increasing ag e equaled decreasing aff ectionate displays. People 65+ as Insignificant In much the same way that research studies have found se xless portrayals of people 65 and older on television, time and ag ain they have supported the notion that television depicts people 65+ as shallow characters who play trivial roles in social society. Northcott (1975) sampled 41 prime time dram atic television shows, 35 total hours of programming, broadcast on the three majo r networks in 1974 and examined agerelated differences between the “major” and “minor” characters. Characters playing major roles were central to the program’s plot and appeared frequently throughout the show. Minor characters were not essential to the plot, and they existed only to support the storyline being played out by the major charac ters. Only 28.6% of age 64+ characters in Northcott’s sample played major roles. What ’s more, the results suggested that when older people did appear in the programs, they were typically featured as dependents on younger people, who were portrayed as more capable and good-looking than their elders.

PAGE 24

17 Northcott concluded that television positioned members of the age 64+ population in stark contrast to competent adult males and attractive, youthful adult females. Harris and Feinberg’s (1977) analysis of television programming suggested that characters age 60+ on television seriously lack depth; they have a relatively shallow range of emotions compared to character s in younger age categories. Hence, the researchers posited that television tends to highlight old people’s surface-level flaws instead of presenting them as deeply emo tional or full of knowledge from experience. Harris and Feinberg concluded that televi sion programming presents a “remarkably onedimensional” portrait of the elderly, char acterizing them as useless and generally incompetent (p. 467). Hiemstra, Goodman, Middlemiss, Vosco, a nd Zeigler (1983) examined a total of 136 television commercials broadcast during we ekdays, weekends, and evenings in 1981 and studied how the older characters, age 50+, were portrayed in terms of family relationships. Despite the fact that most ol der people have children, grandchildren, and perhaps even great-grandchildren, and they are thus members of a family in some manner, the research team discovered that pe ople 50 and older were usually depicted as “nondescript adults with no obser vable family ties” (p. 117). Perhaps the reason that research shows television only shallowly depicts people 65 and older is that the medium almost alwa ys casts the age group in minor, not major, roles. In a study published in 1987, Swayne a nd Greco analyzed a total of 814 television commercials taken from 36 hours of network television programmi ng broadcast in 1985. Examining the level of involvement of ch aracters age 65+ in the commercials, the researchers classified each one as a “major “minor,” or “background” character. Major

PAGE 25

18 roles were assigned to age 65+ character s who spoke on-camera and were visible throughout the duration of the ad. Minor characters did not speak on-camera and were only visible during half of the ad while ba ckground characters did not speak and only made a brief appearance in the ad. The data revealed that only 31.6% of people 65 and older in the sampled commercials played majo r roles; well over half, 56%, played minor roles. Eight years later, Robinson and Skill ( 1995) studied television programming and found similar results in terms of the prominence of age 65+ role-portrayals. From 181.5 hours of network programming comprising 260 fi ctional series epis odes broadcast in 1990, the researchers randomly sampled 100 episodes comprising 67.5 hours of programming for analysis. Using the same rule s for classifying majo r and minor roles as Northcott (1975), Robinson and Skill perfor med their content analysis under the assumption that the proportion of age 65+ ch aracters cast in majo r roles in their 1990 sample would be greater than the proporti on found in Northcott’ s 1974 sample. However, the results revealed that the opposite was true; there was a dr astic decrease in the amount of television program charac ters age 65+ cast in major ro les from 28.6% in Northcott’s study to only 2.8% in Robinson and Skill’s study. Another research study on the role-p ortrayals of people 65 and older on television, conducted by Cassata and Irwin (1997), also sugges ted that the number of age 65+ characters cast in major roles on televi sion is seriously dwindling. In Cassata and Irwin’s 45-hour sample of network day time programming broadcast in 1994, only one percent of all the characters who played major roles were age 65 and older. The

PAGE 26

19 researchers concluded with the following stat ement about the substa nce and depth of oldage characterizations on daytime television: In conclusion, whereas the overall pr ofile of the older character on soap operas is positive, it is important to note that our study revealed that a larger proportion of older characte rs are minor characters, that is, characters whose absence would not be detrimental to the storylines. If one considers the role that these portrayals may play in cultivating viewers attitudes, then one wo uld expect that viewer s would have a positive impression of older people but, at th e same time, might question their significance in our lives (p. 229). People 65+ as Comical As well as being consistently portrayed as sexless and insignificant, television’s characters age 65+ are habitually featured as amusing characters. When television features characters 65 and older in any kind of real depth, it casts them in comical roles on both programming (Gerbner, Gross, Signor ielli, & Morgan, 1980; Signorielli, 1983; Signorielli & Gerbner, 1978) a nd commercials (Francher, 1 973). The older characters do not intend to be funny; rather, they are a subj ect of amusement because of their irrational and eccentric behavior. Bell (1992) examined five prime time netw ork television dramas that did feature elderly characters in major roles, including: Murder, She Wrote; Matlock; Jake and the Fat Man; In the Heat of the Night; and The Golden Girls Bell posited that these shows, as part of a “golden age” of television programming broadcast in the midto late-eighties and early nineties, portrayed el derly characters in a richer, more respected light (p. 306). But many scholars (Gerbner, 1997; Harw ood, 2000; Harwood & Giles, 1992; Kubey, 1980) argued that television shows like the The Golden Girls put elderly characters in the

PAGE 27

20 spotlight only to mock and deride them for engaging in behaviors considered to be uncharacteristic of the age group. Kubey (1980) was one of the first resear ch scholars to propose that by placing members of social groups, incl uding people 65 and older, in si tuations that are seemingly atypical, television shows could be using humor to propagate certain so cial stereotypes of those groups. In other words, Kubey posited th at comical depictions of social groups do not counter negative social labe ls, but they serve to make those labels even stronger. Hence, reverse stereotyping of people 65 and older on television (e.g., showing an elderly woman getting her legs waxed or an elderly couple skydiving ), which is intended to be funny, reinforces typecasting of the age group as being unconcerned about their physical appearance and unwilling to enga ge in wild, risky conduct. Following Kubey’s (1980) seminal work, Harwood and Giles (1992) examined humor in the television show The Golden Girls arguing that in terms of television’s role in bolstering social stereotype s, “the greatest threat to prosocial outcomes lies in the discounting of ‘counter-stereotypical’ message s that is facilitated by humor” (p. 415). The Golden Girls was a half-hour sitcom that feat ured four elderly women living in Florida: Dorothy, Rose, Blanche (all age 55-65), and Sophia, a woman in her mideighties (Harwood & Giles, p. 405). Harwood and Giles randomly sampled six episodes of The Golden Girls that aired during the 1990-1991 season of th e show and analyzed the ch aracter-based humor that each of the four elderly women brought to th e show. In their analysis, none of the characters were portrayed mo re counter-stereotypically th an Blanche, a man-hunter who appeared to be as sexually active in her later years as she was when she was a young

PAGE 28

21 adult, if not more so. Unabashed by erot ic innuendos and overt promiscuity, Blanche reveled in her sexuality. The research team suggested that a cursor y glance at Blache’s character might offer a refreshing depiction of people 65 and older. However, the comical nature of the show’s references to Blanche’ s sex life, which were explicitly intended to be amusing to both the other characters on the show and the audience, only accentuates the humor of “such an extreme portrayal of hypersexuality” in a character age 65+ (p. 423). Reverse stereotyping Blanche as a sex addi ct is supposed to be hilarious. Thus, this further perpetuates the first of the three negative ways that television consistently portrays people 65 and older, as sexless beings void of romantic involvement. Television’s Demographic Repr esentation of People 65+ Scholarly research provide s conflicting reports of how television depicts people 65 and older in terms of their physical a ppearance, health, intelligence, and overall personality; however, there is a consensus that television not only negatively portrays people 65+ as sexless, insignificant, and co mical, but also that the medium grossly underrepresents the age group in comparison wi th its stake in th e total population. Gerbner (1972) was one of the first to look at the demographic representation of old-age on television. He studied prime time and weekend daytime network programming broadcast in 1967-1969 and found that the t ypical television character was male, American, middle or upper class, unmarried, a nd in the prime of his life. Thus, Gerbner reported that there was an obvious absence of both the young and the old on television. Children, adolescents, and old people combined only made up about 10% of the total 762 major characters analyzed in his sample.

PAGE 29

22 Aronoff (1974), who analyzed 2,741 majo r characters of prime time network television drama broadcast between 1969 and 1971, found that less than five percent of the characters were elderly. Similarly, No rthcott’s (1975) sample of 41 prime time network dramatic television programming br oadcast in 1974 contained a total of 464 major and minor characters, but only seven were age 64+. This amounts to 1.5% of the total population of characters analyzed, despit e the fact that, according to Northcott, the U.S. Census Bureau’s estimate for this ag e group was 10% at the time of the study. From their sample of randomly selected drama, comedy, game shows, and news programs aired in 1976, Harris and Feinberg (1977) discovered th at out of the 312 speaking television characters observed in th e study, only 24 were members of the 60to 70-year-old age group, and only two were classi fied as 70+; nearly 8% of the characters were 60 to 70, and less than 1% were over 70. In the research team’s random sample of television commercials, an additional 198 char acters were evaluated; 10.6% was 60 and older. Signorielli and Gerbner’s (1978) sample of prime time and weekend daytime network television programming aired betw een 1969 and 1976 also showed a dramatic underrepresentation of older people. The resear chers analyzed a total of 9,131 characters, 1,898 major characters and 7,233 minor character s, and found that only four percent could be classified as elderly. Greenberg, Korzenny, and Atkin (1980) analyzed a sample of television programming from a three-year time pe riod, 1975 through 1977, and found that people 65 and older were sparsely represented on th e small screen. Only about 100 characters, out of 3,549 analyzed over th e entire three year study belonged to the over-65 age

PAGE 30

23 bracket, More specifically, characters over age 65 made up only 4% of the 1975-1976 sample, 3% of the 1976-1977 sample, and 2% of the 1977-1978 sample. Adults of retirement age were greatly underrepresente d, and there was an ove rrepresentation of people in their 20s, 30s, and 40s ; these characters accounted for two-thirds of the sample even though they only made up one-third of th e U.S. population at the time, according to Greenberg, Korzenny, and Atkin. Gerbner, Gross, Signorielli, and Mo rgan’s (1980) extensive study, which analyzed a total of 16,688 characters on 1,365 television programs from 1969 to 1978, also found seniors to be severely underrepr esented. In fact, both the younger and older age groups were underrepresented in the res earch team’s sample. Individuals under the age of 18 comprised only 8% of the sample’s characters, even though they made up about 30% of the U.S. population at that time; only 2.3% of the sample characters were over 65, despite the fact that they represented 11% of the population at that time, according to Gerbner, Gross, Signorielli, and Morgan. Gerbner and Signorielli (1982) sampled a total of 878 prime-time network programs broadcast from 1969 to 1981 and also found that the very young as well as the very old were underrepresented on tele vision. Characters over 65 made up only two percent of the 14,037 major and minor char acters analyzed, even though the age group made up 11% of the U.S. population at that time, according to Gerbner and Signorielli. In Hiemstra, Goodman, Middlemiss, Vosc o, and Zeigler’s (1983) sample of 136 commercials broadcast in 1981, only 3% of the characters were 60 and older.

PAGE 31

24 Likewise, in their analysis of Saturday morning cartoons broadcast on network television during 1981, Bishop and Krause (1984) studied a total of 378 characte rs of varying age and found that only 25 characters, or 7%, c ould be categorized as old people. Elliott (1984) examined 20 consecutive weeks of 13 daytime television serial dramas aired in 1979, monitoring the presen ce of 65+ characters on a total of 260 episodes. The total number of characters an alyzed in the study was 723, but only 46 of them were classified as members of the age 60 to 69 group, and 12 fell into the age 70+ category. Hence, Elliot’s study re vealed that 6.4% of the char acters in her sample were between 60 and 69 while 1.6% were 70+, equa ling a sum of 8% for characters 60+. Swayne and Greco’s (1987) wide-rangi ng content analysis of 814 commercials provided insight into the underrepresentati on of older individu als in television advertising. Since the exact ag e of any of the individuals in television advertisements could not be determined, the research team created a framework for defining persons age 65 and older, which many research scholars modeled thereafter. Certain age-specific criteria were used, including: appearance of retirement, extensive grey hair and wrinkles of the skin around the eyes and/or hands, us e of ambulatory aids such as canes or wheelchairs, parent of a son or daughter who was middle-aged or older, and evidence of grandchildren or great-grandch ildren. Using the presence of one or more of these as a means for considering a person to be elderly, Swayne and Greco’s results revealed that no more than 7.1% of ads on any of the thre e networks depicted elderly people, even though that age category represented 12% of the U.S. population at that time, according to the research team.

PAGE 32

25 Robinson and Skill (1995), who studied television programs aired in 1990, monitored the ages of 1,228 adult speaking char acters and found that people 65 and older made up only 2.8% of the sample. Perhaps more interesting is that Robinson and Skill noted a decrease in the percentage of ch aracters over 65 reported by Greenberg, Korzenny, and Atkin (1980), 4.5% in pr ogramming broadcast from 1975, and the matching percentage found in their study, onl y 2.8% in programming broadcast in 1990. The percentage of characters 50-64 also decreased from 19% in 1975 (Greenberg, Korzenny, & Atkin) to 16.3% in 1990 (Robinson and Skill). These finding are especially interesting in light of the f act that in the national demographic population, the percentage of people 50+ increased, not d ecreased, between 1975 and 1990. Two 1997 studies found that older adults ag e 60+ were not present in realistic numbers on both television commercials and pr ogramming in the mid-nineties. Sampling 778 television commercials aired in 1994, Roy and Harwood (1997) found that adults age 60 and older were not present in realistic num bers. According to th e research team, the U.S. Census Bureau estimate for the per centage of the population age 60+ in 1994 was 16.74%. However, adults 60 and older only made up 6.9% of the characters in Roy and Harwood’s data set; 246 out of the total 3,547 characte rs analyzed were 60+. Similarly, in Hajjar’s (1997) sample of television programming broadcast in 1995, only 355 out of the total 4,617 characters observe d, or nearly 8%, were over 60. Two other 1997 studies found similar results for the 65+ age category. Positing that underrepresentation of old-age in televisi on advertising truly does exist, Peterson and Ross (1997) discovered that indi viduals age 65 and older only made up a little over 8% of the population of the sample advertisements in their study. Cassata and Irwin’s (1997)

PAGE 33

26 examination of characters 65 and older on television programming also yielded comparable results; only 3% of the 328 total characters analyzed were age 65+. A recent longitudinal study conducted by Si gnorielli (2004) analyzed a sample of prime-time television drama programming from 1993 to 2002. Signorielli studied the major, leading, and supporting characters from each sample year and found that not only was the percentage of older pers ons very low, but also there was a consistent decrease in the number of characters age 65+ from year to year. In fact, age 65+ characters comprised only 1% of the characters anal yzed for the final sample year, 2002. In the body of literature published on th e demographic representation of the television’s characters 65 a nd older, two well cited studies stand in contrast to the revelation that the medium underrepresents the age group. Age 65+ characters in Petersen’s (1973) sample of television progr ams broadcast in 1972 accounted for 13% of the sample, overestimating the size of the age group by 3% since the U.S. Census Bureau’s estimate was 10% at the time of th e study, according to Petersen. However, it is important to note that Petersen’s sample consisted of only 247 total characters, a relatively small and ungeneralizable number comp ared to far more cumbersome character totals like 9,131 (Signorielli & Gerbner, 1978) and 16,688 (Gerbner, Gross, Signorielli, & Morgan 1980). Cultivation Theory Cultivation theory follows the cumulative model of media effects, which proposes that repeated exposure can socialize people to adopt the values and norms portrayed most often by the media, including stereotypical de pictions of social groups like people age

PAGE 34

27 65+ (Davis, 1985; Jeffres & Perloff, 1997). The basic premise of cultivation theory, developed by George Gerbner with the help of Larry Gross, is that long-term heavy television viewing causes people to perceive the television world as a representative depiction of what society is rea lly like (Gerbner & Gross, 1976). Gerbner and his colleagues at the Univer sity of Pennsylvania’s Annenberg School of Communications operated under the assumpti on that in terms of its socialization capabilities, television reigns supreme compar ed to other American media. They posited that “the reach, scope, ritualization, orga nic connectedness, and non-selective use of mainstream television makes it different from other media of mass communications” (Gerbner & Gross, 1976, p. 175). Unlike print and film media, television is not limited by the public’s immobility or illiteracy; it directly reaches people in the comfort of their own homes and requires minimal intellectual abil ity (Gerbner & Gross, 1976). Reasoning that television’s omnipresence in American hous eholds grants the medium immense power for socialization, Gerbner, Gross, Morgan, a nd Signorielli (1980) put it best when they wrote: The television set has become a key member of the family, the one who tells most of the stories most of the time… the more time one spends ‘living’ in the world of television, the more likely one is to report perceptions of social reality which ca n be traced to (or are congruent with) television’s most persistent representations of life (p. 14). Employed by the National Commission on the Causes and Prevention of Violence, Gerbner and his research team i nvestigated violence on network television drama in 1967 and 1968. Their research conti nued through 1972 under the sponsorship of the Surgeon General’s Scientific Advisory Committee on Television a nd Social Behavior

PAGE 35

28 (Gerbner & Gross, 1976). In th e spring of 1972, a team of co nsultants to the National Institute of Mental Health suggested that Gerbner’s report to the Surgeon General (Gerbner, 1972) be expanded to account for so cial relationships and television viewer conceptions. Under this recommendation and a grant from the National Institute of Mental Health, Gerbner’s re search team began to pe riodically study television programming content and its effect on child and adult viewers’ c onceptions of social reality (Gerbner & Gross, 1976). The project, called the Cultural Indicators Project involved a three-pronged research strategy. The first prong was an institutional process analysis of the policies that dictated television’s flow of messages. This part of the project proved difficult to fund, and the focus was placed on prongs two and three: message system analysis and cultivation analysis (Gerbner, Gross, Morgan, & Si gnorielli, 1986). Message system analysis involved sampling network televi sion drama and conducti ng rigorous content analyses to identify television norms. These norms were considered to be “potential lessons” that viewers would learn from watching televisi on, and they were used to develop questions for cultivation analysis (p. 22). Gerber and Gross (1976) used cultivation analysis to determine what, if anything, heavy viewers were absorbing about real life from televisi on. Participants were asked various questions about the prevalence of viol ence in the real world. For each question, there was a “real-world answer ”, devised from real life violence estimates, and a “television answer”, formulated from the viol ence patterns that the researchers observed on television programming (p. 182). Controlling for covariates such as sex, age, and

PAGE 36

29 education, Gerbner and Gross then measured the participants’ relative amount of television viewing in comparison to their likelihood of choosing th e television answer. By contrasting the amount of heavy viewers and light vi ewers providing the television answers, Gerbner and Gross uncovered what they termed a “cultivation differential” between the two groups, which hi ghlighted the social reality conceptions that television was supposedly cultivating in heavy viewers (p. 182). They found that participants in their national sample of adu lts who were classified as heavy television viewers (those who watched an average of four of more hours per day) were always more likely than light viewers (those who watched an average of two hours or less per day) to give the television answer. For example, wh en asked to estimate people’s chances of being involved in violen ce during any given week, 52% of the heavy viewers overestimated the odds by giving the televi sion answer while only 39% of the light viewers did so. Gerbner, Gross, Morgan, and Signorielli (1980) defined two additional constructs of the cultivation effect, resonance and mainstreaming The concept of resonance occurs when members of a certain group, such as all the people with a college education, experience comparable levels of a cultivation effect. Hence, the cultivation of some idea “resonates” with the entire group (Ger bner, Gross, Morgan, & Signorielli). Mainstreaming, on the other hand, occurs when heavy television viewing causes a convergence of perceptions of objective rea lity across divergent groups, such as when two heavy viewers, one male and one female hold concurrent opinions about something that is typically an object for disagreement between genders. In this way, cultivation theory suggests that televisi on viewing overrides individual cu ltural, social, and political

PAGE 37

30 differences so that common views are absorbed into society such that heavy television viewing perceptions are even adopted by those who do not watch television at all (Gerbner, Gross, Morg an, & Signorielli). Gerbner, Gross, Morgan, and Signorielli ( 1986) asserted that th eir research on the cultivation effect was not limited to a comp arison of television data and real-world statistics. They said, “Some of the most inte resting and important t opics and issues for cultivation analysis involve the symbolic tr ansformation of message system data into hypotheses about more general issues and assumptions” (p. 28). Hence, statistical information learned from television lends itself to first-order cultivation beliefs, which provide a major source of assumptions that elicit second-order values, ideologies, and personal perspectives. For example, Gerbne r, Gross, Morgan, and Signorielli (1980) discovered that heavy viewers who believe that the real world is just as saturated with violence as the television world experien ce a “mean world syndrome”; in their minds, most people “cannot be truste d” and are “just looking out for themselves” (Gerbner, Gross, Morgan, & Signorielli, 1986). Although cultivation theory was originally devised to describe and predict the effects of violence on television, the theory was broadly-based from the outset; Gerbner and his research team have examined “the ex tent to which televisi on viewing contributes to audience conceptions and actions in such realms as sex and age-role stereotypes, health, science, the family, educational achievement and aspirations, politics, and religion” (p. 22). In terms of age-related cultivation res earch, Gerbner, Gross, Signorielli, and Morgan’s (1980) study was the most significan t. The study took content analysis data

PAGE 38

31 attained through the observation of a tota l 16,688 television programming characters on 1,365 shows broadcast between 1968 and 1978 and compared it with th e results of the National Council on Aging’s “Myth and Real ity of Aging” survey. The survey, conducted by Louis Harris and Associates in 1974, measured both television viewing frequency and perceptions about the age 65+ population. Gerbne r, Gross, Signorielli, and Morgan found a statistically si gnificant positive relationship (a correlation of .10 where p < .001) between heavy television viewing and th e belief that the number of old people in society is diminishing, not grow ing. That relationship was even stronger (a correlation of .20 where p < .001) among participants under age 30, even with the demographic variables of sex, age, income, and education held constant (p. 46). Davis (1985) discussed the implications of cultivation theory on the findings of content analysis related to te levision’s depiction of old-ag e. Positing that the medium’s underrepresentation of elderly people teaches te levision viewers what to think about the significance of the age group is social society, Davis said, “When increasing age equals increasing invisibility on television, the messa ge is clear: To be old is to be without importance” (p. 45). Criticisms of Cultivation Theory Cultivation is a heavily-debated theory of media effects. Many critics disapprove of the theory’s ambiguous definition of what constitutes “heavy” versus “light” television viewing, and they argue that cu ltivation effects are, at best, correlational; they can be reduced or eliminated when control variab les—such as age, sex, income, education,

PAGE 39

32 direct experience, etc.—other than an indivi dual’s relative amount of television viewing are taken into account (Doob & Macdona ld, 1979; Hirsch, 1981a & 1981b). Many critics, the most noted of whic h are Hughes (1980) and Hirsch (1980, 1981a & 1981b), condemn cultivation theory for misattributing spurious relationships as causal ones. They content that by overrati ng the power of television to manipulate viewers’ social reality perceptions, cultivati on scholars fail to recognize that television viewing is but one of many variables that contri bute to an individual’s concept of reality. While these criticisms of cultivation th eory are prevalent, it is the theory’s internal validity that has been called into question the most. Hence, the real bane of cultivation theory’s existence is that it cannot account for the steps that can be expected to mediate the cultivation process (Hughes, 1980). The th eory does not explain the cognitive processes that take place between television viewing and social reality judgment-making. It is almost as if cultivati on takes place within an unobservable “black box” deep inside the mind of the televi sion viewer (Shanahan & Morgan, 1999, p. 172). Cognitive Subprocesses in the Cultivation Effect Cultivation theory posits that television holds the power to influence people’s perceptions of social re ality, but critics (Doob & M acdonald, 1979; Hirsch, 1980, 1981a & 1981b; Hughes, 1980) have argued contend th at no causal relationship can be observed between television viewi ng and judgment-making about the real world. In response to such scrutiny of cultiva tion theory, scholars began developing a process model of cultivation to provide a cl earer understanding of the possible cognitive links that exist between the stimulus (i.e., television consumption) and the response (i.e.,

PAGE 40

33 television-like social reality perceptions) in the cultivation effect. Their goal was to create an empirically verifiable model with testab le links that could provide a theoretical framework for understanding the cognitive conditions that can be expected to either facilitate or inhibit cult ivation. In so doing, they be lieved that they would boost cultivation theory’s credibility by enhancing the case for its inte rnal validity. L. J. Shrum, arguably the most prominent scholar of cultiv ation studies from a cognitive processing perspective, said “In virtually every disc ipline, it is incumbent upon researchers to explain the ‘whys’ and ‘hows’ of an effect, not just establish its existence” (1995, p. 402). Unfortunately, the earliest attempts to achieve this goal were fruitless. Hawkins and Pingree (1980) examined how various social and psychological conditions as well as programming choice could affect cultivati on. Drawing on Hawkins’ (1977) postulation that the degree of trust in television’s cred ibility as an information source varies from person to person, Hawkins and Pingree assumed that increased faith in television’s ability to illustrate a realistic picture of the worl d would help to induce cultivation. However, there was not a significant relationship between perceived reality of television and cultivation. Likewise, there wa s little relation between cu ltivation and other potential variables that the researchers expected w ould mediate the cultiv ation effect. For example, participants who were classified as having high inference-making abilities, those who could presumably better interpre t meaning from information, did not show stronger cultivation effects. Similarly, in Pi ngree’s (1983) study, participants with low inference-making abilities actually exhibited a stronger cultiva tion effect that those with high inference-making abilities.

PAGE 41

34 Perceived Reality of Television Following Hawkins (1977) and Hawkins a nd Pingree’s (1980) seminal work that conceptualized the notion of perceived reality of televisi on, Potter (1986) sought to more explicitly define the construct. In his multid imensional definition of perceived reality, he called the first dimension the magic window which is the degree to which a television viewer thinks that television c ontent is a precise representation of real life. The second dimension, instruction refers to viewers’ perceptions of television’s ability to supplement their real life experiences and broaden th eir limited knowledge of foreign places and cultures. The third dimension is identity ; it is the level of tele vision viewers’ perceived similarity between their lives in the real wo rld and the lives of te levision characters. Five years later, Potter (1986) studied these three perceived reality dimensions of television in relation to average television viewing frequency among a sample of college and high school students. Cont rolling for the variables of sex, age, and race, Potter examined the participant’s estimates of th eir chances of being victimized—by murder, rape, assault, robbery, or fist fight—and of dying from particular causes, including: accidents, cancer, heart disease, homicide, and pneumonia. Potter’s results disproved his assumption that increased perceived reality of television along a ny one of the three dimensions would increase the cultivation effect. Among the partic ipants who scored high on the magic window dimens ion (i.e. those who thought th at television content is a precise representation of real life), incr eased viewing frequency equaled increased estimates of being murdered, robbed, and dying in a car accident. However, the participants who scored low on the identity scale (i.e., those who did not personally identify with television characters) and on th e instruction scale (i.e., those who did not

PAGE 42

35 think that television can serve as an inst ructional aid) exhibi ted stronger cultivation effects than those who scored high. This s upports the supposition that people who tend to think that television provides them with a “magic window” to the world will be more inclined to enculturation th rough television (p. 162). Potter’s magic window dimension was s ubsequently adopted as the standard definition of perceived reality of television, but there was a lot of controversy over how to operationalize the dimension in order to effectively measure a person’s perceived reality of television (Potter, 1988). That is, until Rubin, Perse, and Taylor (1988) developed a five-item perceive d reality scale to be used in their study of perceived realism as a variable in cultivating social re ality perceptions of faith in others, life control, interpersonal connecti on, political efficacy, and safet y. Like the findings reported by Potter (1986), the results of Rubin et al .’s (1988) study suggest ed that perceived realism of television does indeed mediate th e cultivation effect; a sense of political efficacy, faith in others, and concern for pe rsonal safety could be at least partially attributed to a person’s perc eived reality of television. Learning Theory of Cultivation Another wave of scholarly research not only considered viewer’s perceived reality of television as a mediator in the cultivati on effect, but also broke down the cultivation process into four linear stag es, commonly referred to as “learning theory” (Shrum, 2007a, p. 247). Hawkins and Pingree (1982) were the firs t to propose that viewers take what they learn (stage one) about the wo rld from television to construct (stage two) conceptions about social reality. Further, they proposed that the viewer then adopts what Gerbner,

PAGE 43

36 Gross, Morgan, and Signorielli (1980) termed first-order beliefs (stage three), which eventually lead to the formation of generalized second-order attitudes (stage four). Operating under the assumption that “there is considerable psychological distance between what is seen during individual acts of viewing or even th e accumulation of many acts of viewing, and the constr uction of beliefs about social reality similar to those portrayed or implied by television,” Hawkin s, Pingree, and Alder (1987, p. 555) tested two possible subprocesses in the cultivation effect in a two-part study. First, they examined whether perceived reality of televisi on served as an inte rvening step between learning from television and constructing social reality be liefs. Second, they examined whether first-order beliefs based on televisi on content can truly lead to generalized second-order attitudes. A cultivation effect was observed in part one of Hawkins, Pingree, and Alder’s (1987) study; heavy television viewers provide d higher estimates of violence and crime in social reality than light television viewer s. Part two of the study revealed that firstorder social reality beliefs were not conditio nal in the relationship between television viewing and second-order attit udes. Hence, Hawkins et al.’ s (1987) study did not support their proposed four-stage lear ning theory of cultivation. Replicating some of the same conc lusions found by Hawkins, Pingree, and Alder’s (1987) and Potter’s (1991a) study on component subp rocesses in the cultivation effect also refuted the learning theory of cultivation. Potter did find evidence for firstorder cultivation effects; heavy television viewers were more likely to construct television-like perceptions of social reality than light viewers in terms of not only violence and crime, but also sex, affluence, divorce, and health. On the other hand, there

PAGE 44

37 were no significant differences between hea vy and light television viewers’ second-order attitudes, demonstrating that “there is no construction effect with second-order measures” (Potter, 1991a, p. 94). Learning from television was found to lead only to the construction of first-order social reality beliefs. Potter (1991b) conducted another study specifi cally designed to find out if firstand second-order measures of cultivation should be measured simultaneously or independently. The social reality measures used in this study pertained to three frequently-studied cultivation topics, including: women in th e workplace, affluence, and divorce. Potter found that as a whole, first-order beliefs a nd second-order attitudes were only weakly related. Although firs t-order beliefs were more lik ely to predict second-order attitudes than vice versa, in tercorrelations between the tw o measures on any one of the three topics were “very modest ” and suggested “an ability for one of these measures to predict a maximum of less than 5% of the va riance of the other” (Potter, 1991b, p. 107). Hawkins and Pingree (1990) later conclude d that the results of the Hawkins, Pingree, and Alder (1987) study and Potter’s (1 991a; 1991b) studies s uggest that it is very unlikely that viewers form second-order at titudes from first-order beliefs. Therefore, although the learning theory process mode l of cultivation was not supported by the research, these studies did serv e to identify a key idea for futu re research in the area of cultivation processes: first-order beliefs a nd second-order attitudes must be treated as independent of one another and tested as su ch. This major finding shows that “as a first step toward advancing research on the cogni tive subprocesses underl ying cultivation, the contribution of Hawkins and Pingree and others cannot be overemphasized” (Shrum, 2007a, p. 248).

PAGE 45

38 First-Order Cultivation Effects Models Operating under the assumption that they should focus their research efforts on either first-order beliefs or second-order attitudes, scholars who continued to search for a process model of cultivation effects chose th e former. This is because the research has consistently supported the presen ce of first order cultivation e ffects, but it has been much less definitive about the existence of cult ivated second-order attitudes (Shrum, 1995). First-order beliefs (e.g., estimates of the fre quency of people being mugged) also have a television answer, meaning that they can be clearly defined and observed in order to support the observation of a cultivation effect. Conversely, it is harder to find support for a cultivation effect with second-order attitu des (e.g., fear of walking alone at night), which are not directly comparable to some quantifiable reference in television content (Hawkins & Pingree, 1990). The two most prominent social cognition m odels of first-order cultivation effects are the weighing and balancing model (Sha piro & Lang, 1991) and the accessibility model (Shrum, 1999b; 2002). Both of these models examine cultivation subprocesses from the perspective that social reality judgments are memory-based and not made online In other words, people rely on their r ecollections of their previous exposure to things—in this case, through television view ing—instead of basing their judgments on the circumstantial context of those things in their present everyday lives (Shrum, 2007a). Weighing and Balancing Model of Cultivation Shapiro and Lang (1991) proposed a weighi ng and balancing mode l to explain the cultivation effect. According to the model, the earliest stag e of the cultivation process

PAGE 46

39 occurs when a television view er processes television inform ation and stores it as “event memories” in his or her mind (Shapiro & Lang, 1991, p. 686). Viewers create memories of the event, its context, and the source of its memory (i.e., television). Then, in the stage where social reality judgments need to be ma de about something, viewers retrieve all of the relevant memories they have of that particular thing and “weigh and balance” the significance that each one b ears on the decision at hand. To test the weighing and ba lancing model, Shapiro (1991 ) compared participant’s perceptions of the like lihood of crime and the proportion of occupational or social groups in the population to their usage of various communication media, in cluding: television, newspapers, magazines, and books. He also asked each particip ant to freely recall examples of information stored his or her memory about the likelihood of crime and the proportion of occupational or so cial groups; the participants then self-reported the source of each memory. Shapiro found that memories in a part icular domain (e.g., crime) predicted participants’ worldviews in that domain and that participants truly did associate the perceived source of a memory with the memory itself. The results also at least partially supported Shapiro’s hypothesis that memories categorized by source would better predict a participant’s worldview than event memories alone. Finally although the participants in Shapiro’s study attributed more examples of recalled memories to sources of communication that they used most ofte n, memories categorized by source better predicted. According to Shapir o, this showed that “the im portant process is not the accumulation of event memories but the weighing and balancing of those memories” to form judgments about so cial reality (p. 8).

PAGE 47

40 According to the weighing and balancing model, if a viewer realized that television was the source of a particular me mory, he or she would likely chose to deem that memory an unreliable means for cons tructing judgments about the real world. Therefore, careful weighing and balancing of information retrieved from a memory, especially the source of that memory, coul d mediate the cultivation effect. However, Johnson, Hashtroudi, and Lindsay’s (1993) work on what they call source monitoring suggests that people cannot alwa ys remember the source of their memories; they might mistakenly attribute a memory to some other source instead of recogni zing that television was the memory’s originator. Such errors in source attribution could be caused by an individual’s lack of motivati on to closely examine the root of a memory or by a cognitive deficiency that disables him or her from correctly attributing the source. Accessibility Model of Cultivation While the weighing and balancing model assumes that people make laborious attempts to verify information retrieved from memory and that they will discount information if they think the original s ource lacks credibility, the accessibility model “does not assume that people (necessarily) make source discounting erro rs, but rather that they usually do not make the effort to sour ce discount at all” (S hrum, 2007a, p. 259). The most basic premise of Shrum’s accessibility model (1999b, 2002) is that people simplify decision-making about social reality by usi ng cognitive shortcuts, which are known as “heuristics” (Sherman & Corty, 1984). Shrum’s model rests on two fundamental social cognitive concepts, Wyer and Srull’s construct accessibility and Tversky and Kahneman’s availability heuristic. Wyer

PAGE 48

41 and Srull (1986) conceptualized an individual’s long-term me mory as an assortment of mental storage bins separated by subject, each containing vertical stacks of information. According to Wyer and Srull, when an in formed decision needs to be made, people search the appropriate bin by starting from the top and moving down so that the most recent information is what is called upon. Wyer and Srull proposed that as new information is obtained and old information is reinforced, that information is automatically stacked in the top slot of a me ntal bin. They also posited that sometimes the information at the top of a storage bin can be overlooked. “This means that the more often a representation appears ne ar the top of the bin, the gr eater the likelihood of it being retrieved” (Shrum & O’Guinn, p. 442). Theref ore, Wyer and Srull’s storage bin model suggests that both the recency and frequency of information on a certain topic affect how accessible that subject matter is in a person’ s mind, and hence, what information they might recall with the greatest ease. Tversky and Khaneman’s (1973) availability heuristic assumes that the ease with which something comes to mind creates a bias toward that thing. When asked to make judgments about the probability of someth ing—whether it is a person, place, or occurrence—in real life, people will likely overe stimate or underestimate that thing based on how extensively they have to search their minds for releva nt examples of it. Tversky and Kahneman reasoned that “a vailability is an ecologically valid clue for the judgment of frequency because, in general, frequent ev ents are easier to r ecall or imagine than infrequent ones” (p. 209). Shrum (1995) applied the concepts of cons truct accessibility and the availability heuristic to cultivation theory and reasoned that it is quite possible that television, as a

PAGE 49

42 potential information source, hol ds the power to influence pe ople’s perceptions of social reality by offering recent and frequent illustrati ons of a particular subj ect. In other words, an individual will perceive that the “prevale nce of a construct” (e.g., the percentage of people age 65+ in the U.S. population) or the “trait characteristics of individuals” (e.g., the persona of people age 65+) correspond with the relative ease with which he or she can recall examples of that particular th ing from memory (Shrum, 2007a, p. 254). The resulting accessibility bias is the basis for the first two propositions in Shrum’s accessibility model of cultivation, which is made up of five testable propositions including the following: 1) Tele vision viewing influences acc essibility; 2) Accessibility mediates the cultivation effect; 3) Televisi on exemplars are not discounted; 4) Motivation to process information mediates the cultiv ation effect; and 5) Ability to process information mediates the cultiv ation effect (2002, p. 80-86). As a whole, these propositions posit th at television viewing makes television exemplars more available in people’s minds. Th is creates an accessibility bias that can be expected to promote or discourage cultiva tion, depending on whether people process the television information they retrieve from me mory heuristically or systematically, the latter of which requires a laborious effort to scrutinize the re trieved information. According to the model, if people are mo tivated and able to process television information, they will engage in systema tic processing. In tur n, they will recognize television as the source of the informati on and discount the information because they have deemed television an unreliable sour ce. On the other hand, if people are not motivated to process a message or if they ar e motivated but unable to process a message,

PAGE 50

43 they will engage in heuristic processing. Si nce they do not realize that information is coming from something they saw on television, they do not see the need to di scount it. Proposition three, which assumes that te levision exemplars are not discounted, then acts as the determining factor in Shrum’s accessibility model (1999b, 2002). If people who engaged in heuristic processing due to their lack of motiv ation or ability are primed to recognize that television is the s ource of information retrieved from memory, they will discount the source such that no cu ltivation effect will occur. However, if people are not primed to recognize television as the information source, they will not discount the source and a cultivat ion effect will be observed. Therefore, Shrum’s accessibility model of cultivation posits that when either motivation or ability is absent, people will engage in heuristic processing of a message, and the presence or absence of source-priming will determine if the source (i.e., television) will be identified and discounted or simply left undefined. The model has been tested and many studies have shown support for each of its five propositions (1999b, 2002). Several of these studies are outlined be low. Proposition number three, which says that television exemplars are not discounted, is discussed last because it serves as the final stage in the accessibility model’s flow from television exposur e to judgment-making about social reality. Television Viewing Infl uences Accessibility In an unprecedented attempt to determine whether the accessibility of information in memory can help explain cultivation e ffects, Shrum and O’Guinn (1993) conducted an experiment using reaction time testing, which measures the amount of time participants

PAGE 51

44 use to generate responses to questions. If a certain construct has been activated in a person’s mind recently and/or frequently, th at information becomes more accessible; thus, the more accessible a construct is to a pe rson, the faster he or she will be able to respond (Fazio, 1990). In Shrum and O’Guinn’s (1993) study, the participants were asked to provide percentage estimates to questions given to th em via a computer screen that pertained to constructs that are overrep resented on television (e.g., crime, substance abuse, prostitution). Consistent with their hypothe ses, the research te am found that heavy television viewers gave higher es timates than light television viewers of the constructs’ frequency in social reality. What’s more, the heavy viewers were able to generate their responses faster than the light viewers, even when controlling for other variables such as grade point average and use of other media. In an extension of the 1993 study conducted by Shrum and O’Guinn, Shrum (1996) used reaction time tes ting to measure participant re sponses to questions about three construct that appear very often on televi sion: crime, marital di scord, and particular occupations. Shrum’s findings replicated those found by Shrum and O’Guinn (1993); a cultivation effect was observed as heavy view ers provided higher fr equency estimates of the three constructs than light viewers, and an accessibility effect was noted as heavy viewers were also able to provi de their answers faster than light viewers. Similar results were uncovered in another study conducte d by O’Guinn and Shrum (1997), which examined the effect of television exposure on consumer socialization. Heavy soap opera viewers gave higher prevalence estimates of products and behavior s associated with affluence than light viewers, and they also constructed their answers much faster.

PAGE 52

45 Accessibility Mediates Cultivation Shrum and O’Guinn (1993) concluded that “enhanced accessibility of relevant information for heavier viewers can at least partially account for th e cultivation effect” (p. 436). They based this positi on on indirect evidence uncover ed in their research study. When the research team controlled for acce ssibility, which was measured by speed of response, the cultivation eff ect was significantly reduced. Still, the indirect evidence obtained from Shrum and O’Guinn’s study did not seem to be enough; just because televisi on viewing influences accessibility does not mean that accessibility acts a mediator in the cultivation effect. To show that both the former and the later were true, Shrum (1996) us ed path analyses. His results showed that there was a significant relationship between the variables when moving from television viewing frequency to accessibility, and from accessibility to social reality judgments. Busselle (2001) also found support for accessibility as a mediator in the cultivation effect. In his study, participants were exposed to two different experimental conditions. In the first condition, participants estimated the prevalence of three constructs that are consistently overrepresented on television—Black doct ors, shootings, and extramarital affairs—before they were as ked to recall a specif ic example of each construct. In the second condition, participan ts recalled a specific example of each construct before providing thei r prevalence estimates. Intere stingly, a cultivation effect was observed in the first condition but not in the second. Heavy tele vision viewers gave higher prevalence estimates than light televisi on viewers only when th e participants were not asked to recall an example th e constructs first. In other words, the accessibility bias

PAGE 53

46 for heavy viewers was eliminated in the s econd condition due to a leveling out of the accessibility of the constructs in the minds of the participants. Motivation to Process Information Moderates Cultivation The theoretical framework for this pr oposition comes from Petty and Cacioppo’s elaboration likelihood mode l (Cacioppo & Petty, 1982; Pe tty & Cacioppo, 1986). The model suggests that the amount of cognitive elab oration, or “the extent to which a person carefully thinks about issue-re levant information” to make a decision, depends on his or her motivation and ability (Petty & C acioppo, 1986, p. 7). When conditions promote motivation and ability, the “ likelihood” of cognitive “el aboration” is high; when conditions discourage motivation and ability, the “likelihood” of cognitive “elaboration” is low. Both the message and the message recipi ent can serve as motivational variables (Petty & Cacioppo, 1986). In terms of the message itself, an i ndividual is unlikely to be motivated to systematically process informati on that has little or no personal relevance. Individuals also have a varying need for cognition which can be defined as “a need to understand and make reasonable the experienti al world” (Cohen, Stotland, & Wolfe, 1955). A message recipient with a high need for cognition is more likely to process information systematically while a message recipient with a low need for cognition will probably engage in heuristic pr ocessing (Cacioppo & Petty, 1982). Shrum (2001) conducted and experiment to test the proposition that people who are not motivated to process a message will engage in heuristic processing while people who are motivated to process the same messa ge will engage in systematic processing.

PAGE 54

47 Participants in the study were randomly assigne d to three experimental conditions. In the heuristic processing condition, participants were asked to provide prevalence estimates of constructs overrepresented on television—crime, certain o ccupations, affluence, and marital discord—quickly and spontaneously. Par ticipants in the systematic processing condition were informed that accuracy was im perative for the important study they were taking part in, encouraging th em to think carefully before providing their prevalence estimates. The other participants were part of the control grou p, who were asked to simply answer the questions. The results of Shrum’s (2001) experiment suggested that “processing strategy moderated the cultivation e ffect such that cultivation effects were noted in the heuristic and c ontrol groups but not in the systematic group” (p. 94). Ability to Process Information Moderates Cultivation Petty and Cacioppo’s elaboration lik elihood model (Cacioppo & Petty, 1982; Petty & Cacioppo, 1986) also serves as the th eoretic framework for this proposition. As previously mentioned, cognitive elaborati on depends on both motivation and ability to process a message. To discover whether people who are unabl e to process a message will engage in heuristic processing while able people will use systematic processing, Shrum (1999a) conducted an experiment that used time pressu re to manipulate participants’ ability to engage in systematic processing to answer questions. Randomly sampled participants were selected for a mail survey, which was c onsidered low time pressure, or a telephone survey, which was considered high time pressu re. Shrum hypothesized that if cultivation was a result of heuristic pro cessing of television informati on retrieved from memory, the

PAGE 55

48 cultivation effect should be stronger among participants taki ng the phone survey than the mail survey. His predictions were supporte d; for four of the five constructs overrepresented on television—including societal crime, societal vice, marital discord, affluence, and certain occupations—, partic ipants who took the telephone survey exhibited a greater cultivation e ffect. These results were replicated in a later experiment conducted by Shrum (2007). Television Exemplars are Not Discounted Several scholars have posited that if people recall an example of a construct retrieved from memory and deem it irrele vant to the judgment at hand, they will disregard that example and employ other in formation as the basis for decision-making (Higgins & Brendl, 1995; Shapiro & La ng, 1991). Rationalizing “the somewhat counterintuitive notion that people would use information from nonveridical sources (e.g., fictional programs) to form judgments about the real world,” Shrum (2007) said that because they lack either motivation or ability, “people will likely not attend to source features in constructing their judgments” ( p. 256). Unmotivated people might engage in low involvement processing (Pet ty & Cacioppo, 1982), or people might lack the ability to remember the source of information and ma ke source monitoring errors (Mares, 1996; Shrum, 1997). A study conducted by Shrum, Wyer, and O’Gu inn (1998) provided support for the proposition that people do not discount info rmation they learn from television because they do not identify the source of retrieved information unless prompted to do so. To manipulate participants’ awareness that they were basing their judgments on television

PAGE 56

49 information, Shrum et al. (1998) used three experimental priming conditions. Priming is an experimental technique that employs a stimulus to sensitize an individual to subsequent exposure to that person, place, thing, or idea (Iyengar & Kinder, 1987). Scholars like Iyengar and Kinde r (1987) and Roskos-Ewoldsen, D., Klinger, and RoskosEwoldsen, B. (2002) have also applied the c oncept of priming to the media’s ability to sensitize the public to violence and even to particular polit ical candidates. In the first condition of Shrum, Wyer, and O’Guinn’s (1998) experiment, the nopriming condition, participants were asked to provide prevalence estimates of crime and certain occupations before they were questio ned about their televisi on viewing habits. In the source priming condition, the order of the questions was reversed such that participants answered questions about their television viewing habits before providing prevalence estimates of crime and certain occupations. Participants in the relationpriming condition answered questions in the same order as the nopriming condition, but they were first exposed to an instructional st atement intended to prime them to think that television might be the source of the inform ation they were calling upon to answer the prevalence estimate questions. The statement wa s as follows: “In order to answer these questions, you will use information from a va riety of sources. You should be aware that the subjects of these questions are often depicted on televisio n, more so than occurs in real life. Consequently, people often use this information to formulate answers” (Shrum, Wyer, & O’Guinn, 1998, p. 450). As was predicted by Shrum, Wyer, and O’ Guinn, priming did have an effect on cultivation. Only in the no-priming conditi on, when participants gave prevalence estimates of crime and certain occupations before they were questioned about their

PAGE 57

50 television viewing habits, did heavy televi sion viewers provide more television-like estimates than light viewers. In both the source-priming and relation-priming conditions, no cultivation effect was observed. Hypotheses Cultivation Within Priming Conditions Shrum, Wyer, and O’Guinn’s (1998) resear ch on the role of source discounting in cultivation revealed that priming conditions m oderate the cultivation effect; a cultivation effect was observable only when television wa s not primed as the source of information. Specifically, in the no-priming condition, heavy television viewers’ perceptions of crime and occupations were more in line with what is portrayed on television than were the perceptions of light television viewers. If Shrum et al.’s ( 1998) findings are generalizable and not limited to television’s cultivation of perceptions about crime and occupations, similar results should be obtained when employing the same methodology to a study of the relationship that exists between televi sion viewing and social reality perceptions about people 65 and older. As illustrated by Table 1, six priming groups were examined in this study. The vertical axis represents a pre-existing condi tion, participants’ televi sion viewing level. The horizontal axis represents the experimental condition that participants were exposed to, including: no-priming, sourcepriming, and relation-priming.

PAGE 58

51 Table 1: Priming Groups Heavy TV Viewers (H) Light TV Viewers (L) No Priming (P1) P1H: Heavy TV viewers exposed to the nopriming experimental condition P1L: Light TV viewers exposed to the nopriming experimental condition Sourcepriming (P2) P2H: Heavy TV viewers exposed to the source-priming experimental condition P2L: Light TV viewers exposed to the source-priming experimental condition Relationpriming (P3) P3H: Heavy TV viewers exposed to the relation-priming experimental condition P3L: Light TV viewers exposed to the relation-priming experimental condition Most of the research conducted on the availability heuristic, including Tversky and Khanemnan’s (1973) seminal work, focu sed on how increased ease in recall of something from memory leads to increased es timates of the occurrence of that thing in real life. According to Schw arz, Bless, Strack, Klumpp, R ittenauer-Schatka, and Simons (1991), who studied the availability heuris tic from the opposite end of the continuum, “difficulty in recall may decrease judgment s of frequency, probability, or typicality, much as ease of recall has been assumed to increase these judgments” (p. 201) Therefore, the following hypothesis was proposed: H1: In the no-priming condition, hea vy television viewers (P1H) will report lower estimates of the percentage of people age 65+ in the U.S. population than light television viewers (P1L). While the participants in Shrum, Wyer, and O’Guinn’s (1998) no-priming condition exhibited a cultivation effect, par ticipants in the sour ce-priming and relationpriming conditions did not. This suggests that when television was primed as a source of information, either through source-priming or relation-priming, heavy viewers and light viewers provided similar responses to ques tions about social r eality. Hence, it was expected that in both the source-priming and relation-priming conditions in this

PAGE 59

52 experiment, there would not be a significant relationship between pa rticipants’ television viewing level and their percep tions of people 65 and older. The following two hypotheses were proposed: H2: In the source-priming condition, hea vy television viewers (P2H) and light television viewers (P2L) will report comparable estimates of the percentage of people age 65+ in the U.S. population. H3: In the relation-priming condition, hea vy television viewers (P3H) and light television viewers (P3L) will report comparable estimates of the percentage of people age 65+ in the U.S. population. Shrum, Wyer, and O’Guinn’s (1998) me thodology can be used to study not only the relationship that exists between televi sion viewing and social reality perceptions about age 65+ people’s demographic set size, but also the relations hip between viewing and perceptions about old-age trait characterist ics. Given that an ex tensive review of the literature on television’s depi ction of people 65 and older s upported the notion that the age group is consistently portrayed as sexles s, insignificant, and comical, the following hypothesis and propositions were proposed: H4: In the no-priming condition, hea vy television viewers (P1H) will report less favorable perceptions of people ag e 65+ than light television viewers (P1L).

PAGE 60

53 P1: A greater proportion of the hea vy television viewers (P1H) than the light television viewers (P1L) will perceive people age 65+ to be sexless. P2: A greater proportion of the hea vy television viewers (P1H) than the light television viewers (P1L) will perceive people age 65+ to be insignificant. P3: A greater proportion of the hea vy television viewers (P1H) than the light television viewers (P1L) will perceive people age 65+ to be comical. Again, since Shrum, Wyer, and O’Guinn ( 1998) did not find participants in the source-priming and relation-priming conditions to exhibit a cultivation effect, it was assumed that heavy viewers and light viewer s in these two conditi ons would provide not only similar estimates of the age 65+ population in the U.S., but also similar perceptions of people 65 and older. The following two hypotheses were proposed: H5: In the source-priming condition, hea vy television viewers (P2H) and light television viewers (P2L) will report comparable perceptions of people age 65+. H6: In the relation-priming condition, he avy television viewers (P3H) and light television viewers (P3L) will report comparable perceptions of people age 65+.

PAGE 61

54 Cultivation Across Other Pre-Existing Conditions In addition to television viewing level, tw o other pre-existing variables expected to influence cultivation were considered in this study: perceived re ality of television and direct experience with people 65 and older in real life. These two variables were expected to work in an opposing manner; while incr eased perceived realit y of television was expected to enhance television-like perceptions of people 65 and older, increased direct experience with people 65+ was expected to diminish television-like perceptions of people 65+. Since a cultivation effect was only ex pected to be observed when participants were not prompted to recognize television as an information source, these two variables were only used to compare the responses of heavy television viewers in the no-priming condition. Table 2: Perceived Reality Groups Heavy TV Viewers (H) High Perceived Reality (HPR) HPRH: Heavy TV viewers who have a high level of perceived reality Low Perceived Reality (LPR) LPRH: Heavy TV viewers who have a low level of perceived reality As illustrated by Table 2, heavy viewers with high perceived reality were compared to heavy viewers with low perceived reality. Viewers’ perceived reality of television as a potential window to the world affects the degree to which they apply television’s values and norms to real life (Hawkins, 1977). Therefore, heavy viewers with high perceived reality of television were assumed to be more likely to report televisionlike beliefs about people 65 and older than h eavy viewers with low perceived reality of television. The following two hypotheses were proposed:

PAGE 62

55 H7: In the no-priming condition, heavy te levision viewers with high perceived reality (HPRH) will report lower estimat es of the percentage of people age 65+ in the U.S. population than heavy viewers with low perceived reality (LPRH). H8: In the no-priming condition, heavy te levision viewers with high perceived reality (HPRH) will report less favorable perceptions of people age 65+ than heavy viewers with low perceived reality (LPRH). P1: A greater proportion of hea vy television viewers with high perceived reality (HPRH) than he avy viewers with low perceived reality (LPRH) will perceive pe ople age 65+ to be sexless. P2: A greater proportion of hea vy television viewers with high perceived reality (HPRH) than he avy viewers with low perceived reality (LPRH) will perceive people age 65+ to be insignificant. P3: A greater proportion of hea vy television viewers with high perceived reality (HPRH) than he avy viewers with low perceived reality (LPRH) will perceive pe ople age 65+ to be comical. Table 3: Direct Experience Groups Heavy TV Viewers (H) High Direct Experience (HDE) HDEH: Heavy TV viewers who have a high level of direct experience with people 65 and older Low Direct Experience (LDE) LDEH: Heavy TV viewers who have a low level of direct experience with people 65 and older

PAGE 63

56 As illustrated by Table 3, heavy viewers with a high level of direct experience with people 65 and older were compared to heavy viewers with a low level of direct experience with people 65 and older. Life expe riences can serve to disprove television’s picture of the world (Hawkins & Pingree, 1980) This means that the presence or absence of people 65 and older in the lives of heavy television viewers coul d have an effect on their beliefs about the number and nature of people age 65+. Hence, it was assumed that heavy television viewers with high levels of direct experience w ith people 65 and older would have less television-like perceptions about people 65 and older than heavy television viewers with low levels of dire ct experience with people 65 and older. Two final hypotheses were proposed in this study: H9: In the no-priming condition, heavy television viewers with high direct experience (HDEH) will report higher estimates of the percentage of people age 65+ in the U.S. population than heavy viewers with low direct experience (LDEH). H10: In the no-priming condition, heavy television viewers with high direct experience (HDEH) will report more favorable perceptions of people age 65+ than heavy viewers with lo w direct experience (LDEH). P1: A greater proportion of heavy te levision viewers with low direct experience (LDEH) than heavy viewers with high direct experience (HDEH) will perceive people age 65+ to be sexless.

PAGE 64

57 P2: A greater proportion of heavy te levision viewers with low direct experience (LDEH) than heavy viewers with high direct experience (HDEH) will perceive people age 65+ to be insignificant. P3: A greater proportion of heavy te levision viewers with low direct experience (LDEH) than heavy viewers with high direct experience (HDEH) will perceive people age 65+ to be comical.

PAGE 65

58 METHODOLOGY Sample The sample for this study was dr awn from an undergraduate mass communications course at a large, southeaste rn university. College undergraduates were sampled because research shows that cultivat ion of social reality perceptions could be especially prominent among young adults; pe ople between the ages of 18 and 29 are particularly susceptible to television’s pow er to cultivate their perceptions of people 65 and older, perhaps because they are so “distant” from old age (Gerbner, Gross, Signorielli, & Morgan, 1980b, p. 46). What’s mo re, Peterson and Ross (1997) asserted that the elderly tend to be depicted in a less favorable light when younger individuals make up the target audience. It is logical, then, to study the extent to which televised portrayals of people age 65+ are cultivating corresponding views of old-age in the psyche of young adults. Procedure Modeling Shrum, Wyer, and O’Gui nn’s (1998) methodological approach, participants in this experiment were random ly assigned to one of three experimental conditions, one control condition (no-primi ng) and two priming conditions (sourcepriming and relation-priming). Participants in the three conditions completed a two-page questionnaire that in cluded the following:

PAGE 66

59 Dependent measures to gauge participant’s percep tions about the demographic set size (number) and trait characteristic s (nature) of people 65 and older Demographic measures to ascertain the age, sex, and grade point average of each participant Television measures to estimate total weekly viewing time and perceived reality of television Direct experience measures to determine whether participants’ real life experiences corroborate television content The order of the first three sections in the three questionnaires differed according to the priming condition. The no-priming questionnaire included dependent measures first, followed by demographic measures and television measures. In the source-priming questionnaire, the order of the sections was reversed; demographic measures and television measures came first, a nd dependent measures followed. Finally, the relation-priming questionnaire prompted participants to read an introductory statement about how television typically portrays people age 65+ before completing the questionnaire, which was orde red in the same manner as the no-priming questionnaire. The introductory statemen t, which was based on Shrum, Wyer, and O’Guinn’s (1998) introductory statement, wa s as follows: “In order to answer these questions, you will use information from a va riety of sources, including television. You should be aware that people age 65 and older are underrepresented on television compared to their actual percentage in th e population. The age group is also negatively

PAGE 67

60 portrayed on television. Conse quently, people often use this information to formulate answers.” In all three experimental conditions —no-priming, source-priming, and relationpriming—direct experience measures were listed last so as to avoid any interference with the intended priming condition. Dependent Measures The three questionnaires contained items to measure participants’ social reality perceptions of both the number and nature of people 65+. In terms of the relative number of people in the age group, one open-ended it em was employed to measure participant’s perceived size of the national population of people 65 and older. Specifically, the item was as follows: What percent (between 0% and 100%) of the current United States population is age 65+? Twelve Likert-type scale items measured pa rticipants’ perceptions of age 65+ trait characteristics. Participants provided a numer ical response to each given statement with a number between one (strongly di sagree) and five (strongly ag ree). The content of these items was based on a careful review of c ontent analyses conduced on television’s portrayal of people 65 and older, which revealed that they ar e consistently depicted as sexless, insignificant, and comical. To measure perceptions of people age 65+ as sexless, the following statements were used: 1) People 65+ are not sexually activ e; 2) People 65+ are sexually aroused; 3) People 65+ are not sexually passionate; and 4) Peopl e 65+ are not seductive.

PAGE 68

61 Perceptions of people age 65+ as insigni ficant where measured by the following statements: 1) People 65+ are insignificant in society; 2) People 65+ are not successful in society; 3) People 65+ are thriving members of society; and 4) Pe ople 65+ are forgotten by society. Finally, the following statements measur ed perceptions of people age 65+ as comical; 1) People 65+ are comical; 2) People 65+ are laughed at; 3) People 65+ are not amusing; and 4) People 65+ are unintentionally funny. Demographic Measures The three conditional questionnaires also included items to measure variables including the following: sex, age, and grade point average. These covariates were used to identify trends among groups in this e xperiment, yielding information about the prevalence of the cultivation effect across the two sexes, divergent ages, and varying levels of intelligence, which was measured by grade point average. Television Measures Since Shrum, Wyer, and O’Guinn (1998) did not find television program category to be a significant predictor of social reality perceptions, and the research team posited that “total television viewi ng is the preferred predictor variable” (p. 451), television viewing level was measured in terms of total average weekly viewing. Hence, the type of television programming watched wa s not included in this anal ysis, and for simplification purposes, it was assumed that “televisi on viewing” included viewing of both programming and commercials.

PAGE 69

62 Following Shrum, Wyer, and O’Guinn’s ( 1998) procedure, six open-ended items asked participants to estimate the number of hours they watch televi sion during particular time periods of an average week. The following items were used: 1) On an average weekday morning, how many hours of televi sion do you watch; 2) On an average weekday afternoon, how many hours of televisi on do you watch; 3) During prime time on an average weekday, how many hours of television do you watc h; 4) During late night on an average weekday, how many hours of television do you watch; 5) On an average Saturday, how many hours of television do you watch; and 6) On an average Sunday, how many hours of television do you watch. The four weekday measures were multiplied by five and added to the viewing estimates for Saturday and Sunday in order to identify an averag e weekly television viewing estimate for each participant. The sums from this calculation yielded a wide range of weekly television vi ewing estimates. The range of estimates was divided into thirds such that participants who gave viewi ng estimates that ranked in the top third were labeled heavy viewers, and participants who gave viewing estimates that ranked in the bottom third were labeled light viewers. Participants whose estimates ranked in the middle third of all responses, then, were labe led moderate viewers; their data was thrown out because the objective of this research study, and cultivations studies as a whole, is to compare heavy and light television viewers. To access the extent to which each partic ipant believed that television provides accurate representations of real life, the que stionnaires also included Rubin, Perse, and Taylor’s (1988) five-item per ceived reality scale. Rubin et al.’s (1988) perceived reality scale is frequently relied upon as a means for accessing peop le’s perceptions about the

PAGE 70

63 realism of television content and making comp arisons between television viewing level and the adoption of social reality beliefs. Th e five items were measured using a Likerttype scale where participants provided a numerical response to each given statement with a number between one (strongly disagree) and five (strongly agree).The following statements were used for this measure: 1) Television shows life as it really is; 2) Television presents things as they really are in life; 3) If I see something on television I can be sure it really is that way; 4) Televi sion lets me see how other people live; and 4) Television lets me see what happens in ot her places as if I’m really there. Direct Experience Measures Since “even if television messages affect the construction of social reality, it is not done in a vacuum,” it was imperative that this research study examine the effect that direct experiences have on cultivation (Weimann, 2000, p. 74). For this reason, the questionnaires also included an item to m easure participant’s di rect experience with people 65+. Using a Likert-type scale where pa rticipants provided a numerical response to each given statement with a number between one (strongly disagree) and five (strongly agree), participants responded to five items. The following statements were used for this measure: 1) I live or have lived with at least one person who is 65+; 2) People 65+ are present in my life; and 3) I freque ntly interact with people 65+. Data Analysis Independent-Samples t-Tests, which are in terval-level tests, were employed to analyze the data collected in this study. This particular statistical method was chosen

PAGE 71

64 because of its ability to comp are the statistical significan ce of a difference between the means of two independent groups on some meas ure, which is applicable to all of the hypotheses posed by this study. All t-Tests were analyzed at the 95% confidence level, which is typical in soci al science research. Although some researchers denigrate the use of interval-level testing to analyze data obtained through Likert-type scales, th e practice is common in social science research, especially when the scale has at least five items and one middle item. Jaccard and Wan (1996) said, “For many statistical tests, rather severe departures (from intervalness) do not seem to affect Type I a nd Type II errors dramatically” (p 4). The Likert-type scale used in this study’s survey instrument not only contained five items with one middle item (i.e., strongly disagr ee, disagree, undecided, agree, and strongly agree), but also clearly implied a symmetry of response levels in its wording. Hence, it was assumed that respondents perceived each le vel in the Likert-type scale as equidistant from the others, and interval-level testing was deemed appropriate for data analysis. Further, exploratory analysis empl oyed a univariate ANOVA to analyze the participants’ experimental condition, thei r television viewing level, and the two categories’ interaction effect for the dependent variable “What percent of the total current U.S. population is 65+. The experimental condition—no-priming, source-priming, or relation-priming—was entered as a fixed fact or, and the television viewing level, heavy viewing or light viewing, was entered as a random factor.

PAGE 72

65 RESULTS Sample A total sample of 247 undergraduate stude nts enrolled in a mass communications course at a large, southeastern university completed questionnaires representing one of three experimental conditions, one contro l condition (no-priming) and two priming conditions (source-priming and relation-primi ng). Before any other data was analyzed, the range of total weekly television viewing es timates was divided into thirds. Eighty-two participants whose week ly television viewing equaled 0 to 17.5 hours were listed in the bottom third and labeled light viewers; 82 pa rticipants whose week ly television viewing equaled 35.5 to 128 hours were listed in th e top third and labeled heavy viewers. The remaining 83 participants whose week ly television viewing equaled 18 to 34.5 hours were listed in the middle third and la beled moderate viewers. This finding is consistent with a 2006 study by Nielsen Medi a Research, which found that the average time an individual spent watching televi sion during the 2005-2006 television year was about 32 hours per week. Moderate viewers’ da ta was not included in this study, which is designed to analyze the differences between heavy and light television viewers. The remaining sample included 82 heavy view ers and 82 light viewers (N = 164). An Independent-Samples t-Test revealed a statistically significant difference in the mean of television viewing hours watched per week by heavy television viewers and

PAGE 73

66 light television viewers (t = -22.99, p = .00, p < .05). As illustrated by Table 4, light viewers’ mean number of viewing hours per week was 8.48; heavy viewers’ mean of viewing hours per week was 50.78. Table 4: Television Viewing LevelStatistics TV Viewing Level N Mean Number of Viewing Hours per Week Std. Deviation Std. Error Mean Total Weekly TV Viewing Light TV Viewing 82 8.476 5.6617 .6252 Heavy TV Viewing 82 50.780 15.6746 1.7310 The median age of the participants wa s 19 years with a nine-year age range between 18 and 27 years. There were 44 male participants and 120 female participants. Since moderate viewers’ data was thrown out, there was an unequal number of usable questionnaires in each of the three experimental conditions. As illustrated by Table 5, there were 53 no-priming questionnaires, 66 source-priming questionnaires, and 45 relation-priming questionnaires. Table 5: Frequency of Experimental Conditions Frequency Percent Valid Percent Cumulative Percent Valid No-Priming 53 32.3 32.3 32.3 Source-Priming 66 40.2 40.2 72.6 Relation-Priming 45 27.4 27.4 100.0 Total 164 100.0 100.0

PAGE 74

67 Cultivation Within Priming Conditions Hypothesis one (H1), hypothe sis two (H2), and hypothe sis three (H3) were examined through an analysis of the responses to one ope n-ended item that measured participant’s perceived size of the nationa l population of people 65+. Specifically, the item was as follows: What percent (between 0% and 100%) of the total current United States population is age 65+? Table 6: H1, H2, & H3 Statistics Dependent Variable: What percent (between 0% and 100%) of the total current population is 65+? Experimental Condition TV Viewing Level Mean Std. Deviation N No-Priming Light TV Viewing 34.31 10.641 29 Heavy TV Viewing 36.42 18.252 24 Total 35.26 14.472 53 Source-Priming Light TV Viewing 31.24 11.478 33 Heavy TV Viewing 36.48 13.300 33 Total 33.86 12.606 66 Relation-Priming Light TV Viewing 30.00 12.994 20 Heavy TV Viewing 33.75 13.957 24 Total 32.05 13.504 44 Total Light TV Viewing 32.02 11.573 82 Heavy TV Viewing 35.65 14.969 81 Total 33.83 13.451 163 H1 predicted that in the no-priming condition, heavy televi sion viewers would report lower estimates of the percentage of people age 65+ in the U.S. population than light television viewers. As illustrated by Ta ble 6, in the no-priming condition, the mean of light television viewers’ estimates was 34% and the mean of heavy television viewers’ estimates was 36%. An Independent Samples t-Test did not reveal a statistically

PAGE 75

68 significant difference in the mean of light and heavy viewers’ estimates of the percentage of people age 65+ in the U.S. population (t = -.50, p = .62, p > .05). H1 was not supported. H2 and H3 predicted that heavy televi sion viewers and light television viewers would report comparable estimates of the pe rcentage of people age 65+ in the U.S. population in the source-priming and relation-priming conditions, respectively. As illustrated by Table 6, in the source-primi ng condition, the mean of light television viewers’ estimates was 31%, and the mean of heavy television vi ewers’ estimates was 36%. An Independent Samples t-Test did not re veal a statistically significant difference in the mean of light and heavy viewers’ estimat es of the percentage of people age 65+ in the U.S. population (t = -1.71, p = .09, p > .05) Since the difference between the means of heavy viewers’ estimates a nd light viewers’ estimates was not statistically significant, the two groups’ estimates were comparable in the source-priming condition. H2 was supported. As illustrated by Table 6, in the relati on-priming condition, the mean of light television viewers’ estimates was 30%, and the mean of heavy television viewers’ estimates was 34%. An Independent Samples t-Test did not reveal a statistically significant difference in the mean of light and heavy viewers’ estimates of the percentage of people age 65+ in the U.S. population (t = -.92, p = .37, p > .05). Since the difference between the means of heavy viewers’ estimates and light viewers’ estimates was not statistically significant, the two groups’ estimates were comparable in the relationpriming condition. H3 was supported.

PAGE 76

69 Hypothesis four (H4), hypothesis five (H5), and hypothesis six (H6) were examined through an analysis of the responses to 12 Li kert-type scale items that measured participants’ perceptions of people 65+ as sexless, insignificant, and comical. Four items were created to measure these three constructs for the purpose of creating three indexes: a sexless index, an insigni ficant index, and a comical index. However, Cronbach’s Alpha testing revealed low inte ritem reliability among the items used to measure perceptions of people 65+ as sexless ( = .63, M = 2.95, SD = .70), insignificant ( = .40, M = 2.21, SD = .55), and comical ( = .27, M = 3.36, SD = .52). These numbers were lower than the lowest acceptable valu e for Cronbach’s Alpha in social science research, .70; therefore, the three indexes were not created because the alphas showed that the three constructs—sexless, insigni ficant, and comical—were multidimensional, not one-dimensional. Instead, ea ch of the four items that were created to measure the three constructs were tested individually. H4, which predicted that in the no-priming condition, heavy television viewers would report less favorable perceptions of pe ople age 65+ than light television viewers, was based on three propositions. The first proposition (P1) of H4 predicted that a greater proportion of the heavy television viewers than the light television viewers would perceive people age 65+ to be sexless. Th e following items were used: 1) People 65+ are not sexually active; 2) People 65+ are sexually aroused; 3) People 65+ are not sexually passionate; and 4) People 65+ are not seductive. An Independent Samples t-Test did not re veal a statistically significant difference in the mean of heavy viewers’ and light vi ewers’ perceptions of people 65+ as “not sexually active” (t = -.14, p = .89, p > .05), “ not sexually passionate” (t = .81, p = .51, p >

PAGE 77

70 .05), and “not seductive” (t = 1.87, p = .07, p > .05). However, as illustrated by Table 7, the test did reveal a statistically significant difference in heavy viewers’ mean perception (4.50) and light viewers’ mean percepti on (3.86) of people 65+ as “not sexually attractive” (t = -2.35, p = .03, p < .05). Since th e difference between the means of heavy viewers’ and light viewers’ pe rceptions was only statisticall y significant for one of the four items used to measure the sexless constr uct, P1 of H4 was onl y partially supported. Table 7: H4-P1 t-Testfor No-Priming t-Test for Equality of Means t df Sig. (2tailed) Mean Difference Std. Error Difference 95% Confidence Interval of the Difference Lower Upper People 65+ are not sexually attractive. -2.346 48 .023 -.638 .272 -1.184 -.091 The second proposition (P2) of H4 pred icted that a greate r proportion of the heavy television viewers than the light television viewers would perc eive people age 65+ to be insignificant. The following items were used: 1) People 65+ are insignificant in society; 2) People 65+ are not successful in society; 3) People 65+ are thriving members of society; and 4) People 65+ are forgotten by society. An Independent Samples t-Test did not re veal a statistically significant difference in the mean of heavy viewers’ and light viewers’ perceptions of people 65+ as “insignificant in society” (t = -.19, p = .85, p > .05), “not successful in society” (t = 1.14, p = .26, p > .05), “not thriving members of society” (t = .12, p = .91, p > .05), or “forgotten by society” (t = .48, p = .63, p > .05). Since the difference between the means

PAGE 78

71 of heavy viewers’ and light viewers’ percep tions was not statistica lly significant for any of the four items used to measure the insigni ficant construct, P2 of H4 was not supported. The third proposition (P3) of H4 predicted that a greater proportion of the heavy television viewers than the light television vi ewers would perceive pe ople age 65+ to be comical. The following items were used: 1) People 65+ are comical; 2) People 65+ are laughed at; 3) People 65+ are no t amusing; and 4) People 65 + are unintentionally funny. An Independent Samples t-Test did not re veal a statistically significant difference in the mean of heavy viewers’ and light view ers’ perceptions of people 65+ as “comical” (t = -.14, p = .89, p > .05), “la ughed at” (t = .81, p = .51, p > .05), “amusing” (t = 1.87, p = .07, p > .05), and “unintentionally funny” (t = -2.26, p = .03, p < .05). Since the difference between the means of heavy viewer s’ and light viewers’ perceptions was not statistically significant for any of the four items used to m easure the comical construct, P3 of H4 was not supported. P1 of H4 was only partially supported; P2 of H4 was not supported; P3 of H4 was not supported. Overall, H4 was only partially supported for the sexles s construct, and it was not supported at all for the insi gnificant and comical constructs. H5 predicted that in the source-primi ng condition, heavy television viewers and light television viewers would report compar able perceptions of people age 65+. An Independent Samples t-Test did not reveal a statistically significant difference in the mean of heavy viewers’ and light viewers’ perceptions of people 65+ for 10 of the 12 items used to measure the three constructs—s exless, insignificant, and comical. Thus, as illustrated by Table 8, the test did reveal a statistically significant difference between heavy viewers’ mean perception (2.79) and light viewers’ mean perception (3.45) of

PAGE 79

72 people 65+ as “not seductive” (t = 2.56, p = .01, p < .05) and a statistically significant difference in heavy viewers’ mean perception (3.45) and light view ers’ mean perception (3.00) of people 65+ as “unintentionally f unny” (t = -2.01, p = .05, p = .05). Since the difference between the means of heavy viewers’ and light viewers’ perceptions of people 65+ was not statistically significant for only 10 of the 12 items, the two groups’ estimates were not entirely comparable in the sour ce-priming condition. H5 was only partially supported. Table 8: H5 t-Testfor Source-Priming t-Test for Equality of Means t df Sig. (2tailed) Mean Difference Std. Error Difference 95% Confidence Interval of the Difference Lower Upper People 65+ are not seductive. 2.559 64 .013 .667 .261 .146 1.187 People 65+ are unintentionally funny. -2.007 64 .049 -.455 .227 -.907 -.002 H6 predicted that in the relation-prim ing condition, heavy television viewers and light television viewers would report compar able perceptions of people age 65+. An Independent Samples t-Test did not reveal a statistically significant difference in the mean of heavy viewers’ and light viewers’ perceptions of people 65+ for 11 of the 12 items used to measure the three constructs—s exless, insignificant, and comical. Thus, as illustrated by Table 9, the test did reveal a statistically significant difference in heavy viewers’ mean perception (2.52) and light vi ewers’ mean perception (1.85) of people 65+ as “not sexually passionate” (t = -2.35, p = .02, p < .05). Sinc e the difference between the means of heavy viewers’ and light view ers’ perceptions of people 65+ was not

PAGE 80

73 statistically significant for only 11 of the 12 items, the two groups’ estimates were not entirely comparable in the relation-primi ng condition. H6 was only partially supported. Table 9: H6 t-Testfor Relation-Priming t-Test for Equality of Means t df Sig. (2tailed) Mean Difference Std. Error Difference 95% Confidence Interval of the Difference Lower Upper People 65+ are not sexually passionate. -2.351 43 .023 -.670 .285 -1.245 -.095 Cultivation Across Other Pre-Existing Conditions Hypothesis seven (H7) and hypothesis eigh t (H8) analyzed heavy television viewers in the no-priming condition (N = 24) to determine the effect that perceived reality of television would have on their estimat es of the percentage of people age 65+ in the U.S. population and their perceptions of people 65+ as sexless, insignificant, and comical. Five Likert-type scale items measured part icipants’ perceived r eality of television. The following items were used: 1) Television shows life as it really is; 2) Television presents things as they really are in life; 3) If I see something on television I can be sure it really is that way; 4) Televi sion lets me see how other people live; and 4) Television lets me see what happens in other places as if I’m really there. Cronbach’s Alpha testing revealed acceptable interitem reliability among the five items ( = .71, M = 2.18, SD = .59). Therefore, a perceived reality i ndex for the five items was created. A median split (Mdn = 2.40) was used to differentiate high versus low perceived reality. Any case with an i ndex score in the range of 1.00-2.40 was assigned to the low

PAGE 81

74 perceived reality category (N = 14); any case with an index score in the range of 2.415.00 was assigned to the high percei ved reality category (N = 10). An Independent-Samples t-Test revealed a statistically significant difference in the index score means of the two perceived reality groups (t = -5.36, p = .00, p < .05). As illustrated by Table 10, participants with lo w perceived reality had a mean score of 2.14 on the perceived reality index; participants with high perceived reality had a mean score of 2.94 on the perceived reality index. Table 10: Perceived Reality Level Statistics Perceived Reality Level N Mean Std. Deviation Std. Error Mean Perceived Reality Index Low Perceived Reality 14 2.1429 .37970 .10148 High Perceived Reality 10 2.9400 .32728 .10349 H7 predicted that in the no-priming c ondition, heavy televisi on viewers with high perceived reality would report lower estimates of the percentage of people age 65+ in the U.S. population than heavy viewers with low perceived reality. Part icipants with low perceived reality had a mean estimate of 34% and participants with high perceived reality had a mean estimate of 40%. An Inde pendent Samples t-Test did not reveal a statistically significant difference in the two perceived reality groups’ mean estimates of the percentage of people age 65+ in the U.S. population (t = -.85, p = .40, p > .05). H7 was not supported. H8, which predicted that in the no-priming condition, heavy television viewers with high perceived reality would report le ss favorable perceptions of people age 65+ than heavy viewers with low perceived rea lity, was based on three propositions. The first

PAGE 82

75 proposition (P1) of H8 predic ted that a greater proportion of heavy television viewers with high perceived reality than heavy viewer s with low perceived re ality would perceive people age 65+ to be sexless. An Independent Samples t-Test did not re veal a statistically significant difference in the mean of the two perceived reality groups’ perceptions of people 65+ as “not sexually active” (t = -.15, p = .88, p > .05), “ not sexually attractive” (t = -.52, p = .61, p > .05), “not sexually passionate” (t = -.09, p = .93, p > .05), and “not seductive” (t = .66, p = .52, p > .05). Since the difference between the means of the two perceived reality groups’ perceptions was not sta tistically significant for any of the four items used to measure the sexless construct, P1 of H8 was not supported. The second proposition (P2) of H8 predic ted that a greater proportion of heavy television viewers with high perceived reality than heavy viewers with low perceived reality would perceive people ag e 65+ to be insignificant. An Independent Samples t-Test did not re veal a statistically significant difference in the mean of the two perceived rea lity groups’ perceptions of people 65+ as “insignificant in society” (t = .51, p = .61, p > .05), “not successful in society” (t = .21, p = .84, p > .05), “not thriving members of society” (t = -.74, p = .47, p > .05), or “forgotten by society” (t = -.39, p = .70, p > .05). Since the difference between the means of the two perceived reality groups’ perceptions was not statistically significant for any of the four items used to measure the insignifi cant construct, P2 of H8 was not supported. The third proposition (P3) of H8 predicte d that a greater proportion of heavy television viewers with high perceived reality than heavy viewers with low perceived reality would perceive people age 65+ to be comical.

PAGE 83

76 An Independent Samples t-Test did not re veal a statistically significant difference in the two perceived reality groups’ perceptions of people 65+ as “laughed at” (t = -.29, p = .77, p > .05), “amusing” (t = -.45, p = .67, p > .05), and “unintenti onally funny” (t = .66, p = .52, p > .05). However, as illustra ted by Table 11, the test did reveal a statistically significant difference in high perc eived reality participants’ mean perception (4.00) and low perceived reality participants ’ mean perception (2.71) of people 65+ as “comical” (t = -3.55, p = .00, p < .05), Since the difference between the means of the two perceived reality groups’ percepti ons was only statistically sign ificant for one of the four items used to measure the comical construct, P3 of H8 was only partially supported. Table 11: H8-P3 t-Test t-Test for Equality of Means t df Sig. (2tailed) Mean Difference Std. Error Difference 95% Confidence Interval of the Difference Lower Upper People 65+ are comical. -3.547 22 .002 -1.286 .362 -2.037 -.534 P1 of H8 was not supported; P2 of H8 was not supported; P3 of H8 was only partially supported. Overall, H8 was only pa rtially supported for th e comical construct, and it was not supported at all for the sexless and insignificant constructs. Hypothesis nine (H9) and hypothesis 10 (H 10) analyzed heavy television viewers in the no-priming condition (N = 24) to determine the effect that direct experience with people 65+ would have on their estimates of th e percentage of people age 65+ in the U.S. population and their perceptions of people 65+ as sexless, insignificant, and comical.

PAGE 84

77 Three Likert-type scale it ems measured participants’ direct experience with people 65+. The following items were used: 1) I live or have lived with at least one person who is 65+; 2) People 65+ are present in my life; and 3) I frequently interact with people 65+. Cronbach’s Alpha testing revealed low interitem reliability among the three items ( = .50, M = 3.26, SD = 1). However, deleti ng the first item (“I live or have lived with at least one person who is 65+”) brought the interitem reliability among the remaining two items (“People 65+ are present in my life” and “I frequently interact with people 65+”) up to an acceptable level ( = .70) Therefore, a dire ct experience index for the remaining two items was created. A median split (Mdn = 4.00) was used to differentiate high versus low direct experience. Any case with an index score in the range of 1.00-4.00 was assigned to the low direct experience category (N = 13); any case with an index score in the range of 4.01-5.00 was assigned to the high dir ect experience category (N = 11). An Independent-Samples t-Test revealed a statistically significant difference in the index score means of the two direct e xperience groups (t = -7.12, p = .00, p < .05). As illustrated by Table 12, participants with lo w direct experience had a mean score of 2.54 on the direct experience index; participants w ith high direct experience had a mean score of 4.68 on the direct experience index. Table 12: Direct Experience Level Statistics Direct Experience Level N Mean Std. Deviation Std. Error Mean Direct Experience Index Low Direct Ex perience 13 2.5385 .96742 .26831 High Direct Experience 11 4.6818 .25226 .07606

PAGE 85

78 H9 predicted that in the no-priming c ondition, heavy televisi on viewers with high direct experience would report higher estimate s of the percentage of people age 65+ in the U.S. population than heavy viewers with lo w direct experience. Pa rticipants with low direct experience had a mean estimate of 31%, and participants with high direct experience had a mean estimate of 43%. An I ndependent Samples t-Test did not reveal a statistically significant difference in the two direct experience groups’ mean estimates of the percentage of people age 65+ in the U.S. population (t = -1.61, p = .12, p > .05). H9 was not supported. H10, which predicted that in the no-pr iming condition, heavy television viewers with high direct experience would report more favorable perceptions of people age 65+ than heavy viewers with low direct experi ence, was based on three propositions. The first proposition predicted that a gr eater proportion of heavy televi sion viewers with low direct experience than heavy viewers with high di rect experience would perceive people age 65+ to be sexless. An Independent Samples t-Test did not re veal a statistically significant difference in the mean of the two direct experience groups’ perceptions of people 65+ as “not sexually active” (t = 1.91, p = .07, p > .05), “not sexually attractive” (t = .26, p = .80, p > .05), “not sexually passionate” (t = 1.59, p = .13, p > .05), and “not seductive” (t = 1.44, p = .16, p > .05). Since the difference between the means of the tw o direct experience groups’ perceptions was not sta tistically significant for any of the four items used to measure the sexless construct, P1 of H10 was not supported.

PAGE 86

79 The second proposition predicted that a greater proportion of heavy television viewers with low direct experience than hea vy viewers with high di rect experience would perceive people age 65+ to be insignificant. An Independent Samples t-Test did not re veal a statistically significant difference in the mean of the two direct experience groups’ perceptions of people 65+ as “not successful in society” (t = 1.61, p = .12, p > .05) “not thriving members of society” (t = .79, p = .44, p > .05), or “forgotten by society” (t = -.10, p = .92, p > .05). However, as illustrated by Table 13, the test did reveal a statistically significant difference in high direct experience participants’ mean per ception (1.45) and low direct experience participants’ mean perception ( 2.31) of people 65+ as “insigni ficant in society” (t = 2.48, p = .02, p < .05). Since the difference between the means of the two direct experience groups’ perceptions was only sta tistically significant for one of the four items used to measure the insignificant construct, P2 of H10 was only partially supported. Table 13: H10-P2 t-Test t-Test for Equality of Means t df Sig. (2tailed) Mean Difference Std. Error Difference 95% Confidence Interval of the Difference Lower Upper People 65+ are insignificant in society. 2.481 22 .021 .853 .344 .140 1.566 The third proposition predicted that a gr eater proportion of heavy television viewers with low direct experience than hea vy viewers with high di rect experience would perceive people age 65+ to be comical.

PAGE 87

80 An Independent Samples t-Test did not re veal a statistically significant difference in the mean of heavy viewers’ and light view ers’ perceptions of people 65+ as “comical” (t = -1.26, p = .22, p > .05), “laughed at” (t = .73, p = .47, p > .05), “amusing” (t = .22, p = .83, p > .05), and “unintentionally funny” (t = -.44, p = .67, p > .05). Since the difference between the means of the two pe rceived reality groups’ perceptions was not statistically significant for any of the four items used to m easure the comical construct, P3 of H10 was not supported. P1 of H10 was not supported ; P2 of H10 was only partially supported; P3 of H10 was not supported. Overall, H10 was only partially supported for the insignificant construct, and it was not supported at all for the sexless and co mical constructs. Exploratory Analysis Exploratory analysis of the data usin g univariate ANOVA tes ting revealed one important statistically significant relationship that can be applied to the better understand the results of H1, H2, and H 3, which are the three hypotheses concerning estimates of the percentage of people 65+ in the U.S. popul ation. Although the expe rimental condition variable was not statistically significant (F = 4.03, p = .20, p > .05), the television viewing level variable was st atistically significant (F = 14.19, p = .05). These results are illustrated in Table 14.

PAGE 88

81 Table 14: H1, H2, & H3 Tests of Between-Subjects Effects Dependent Variable: What percent (between 0% and 100%) of the total current population is 65+? Source Type III Sum of Squares df Mean Square F Sig. Partial Eta Squared Intercept Hypothesis 178995.654 1 178995.654 331.920 .035 .997 Error 539.274 1 539.274a EXPERIMENTCOND Hypothesis 290.34 2 2 145.171 4.032 .199 .801 Error 72.002 2 36.001b TVVIEWLEVEL Hypothesis 539.274 1 539.274 14.188 .051 .861 Error 87.136 2.292 38.010c EXPERIMENTCOND TVVIEWLEVEL Hypothesis 72.002 2 36.0 01 .199 .820 .003 Error 28396.843 157 180.872d a. MS(TVVIEWLEVEL) b. MS(EXPERIMENTCOND TVVIEWLEVEL) c. .986 MS(EXPERIMENTCOND TVVIEWLEVEL) + .014 MS(Error) d. MS(Error) Hence, participants’ television viewing le vel had a significant influence on the participants’ estimates of the 65+ population while their assi gnment to one of the three experimental conditions did not More significantly, there wa s no significant interaction effect between the experimental condition and the television viewing level of the participants. This finding compliments the results of the t-Test s for H1, H2, and H3, which did not find a statistically significant difference between heavy viewers’ and light viewers’ mean estimates of the 65+ populat ion in any of the three experimental conditions but did reveal that heavy viewers’ estimates were slightly higher than light viewers’ estimates in all th ree experimental conditions. Exploratory analysis of the demographic information obtained in this study—sex, age, and grade point average—also revealed an important finding about the relationship between GPA and perceived reality of tele vision. The high GPA participants’ (with a

PAGE 89

82 GPA of 3.5 to 4.0) mean score on the perceive d reality index (2.27) was higher than the low GPA participants’ (with a GPA of 1.8 to 2.7) mean score on the perceived reality index (1.95). As illustrated by Table 15, an Inde pendent Samples t-Test revealed that the difference in the two groups’ mean scores on the perceived reality index approaches significance (t = -1.94, p = .056). Table 15: GPA and Perceived Reality t-Test t-Test for Equality of Means t df Sig. (2tailed) Mean Difference Std. Error Difference 95% Confidence Interval of the Difference Lower Upper Perceived Reality Index -1.940 69 .056 -.315 45 .16258 -.63979 .00888

PAGE 90

83 DISCUSSION Cultivation Effects Gerbner and Gross’ (1976) cultivation theory posits that people’s understanding of the real world results from the cumulati ve effects of heavy television viewing. The literature on television’s depict ion of people 65 and older supp orts the notion that the age group is consistently underrepresented and por trayed as sexless, insignificant, and comical. Hence, heavy viewers were expected to posses these television-like perceptions about the number and nature of people 65+. Cultivation effects were expected to be observed in the no-priming condition. In this condition, however, heavy television viewers did not repo rt lower estimates of the percentage of people 65+ in the U.S. population than light television viewers. In fact, all of the participants overestimated the 65+ population (M = 33%), but heavy viewers actually reported slightly higher estimates of the 65+ population than light viewers. The difference between the two group’s estimates wa s not statistically significant, but the results imply that heavy viewers are at leas t somewhat more likely to overestimate the 65+ population than light viewers. Hence, the results of this study suggest that when people are not prompted to recognize television as an information source, heavy television viewing does not cultivate them to underestimate the 65+ population in the real world, despite television’s gross

PAGE 91

84 underrepresentation of the age group (Gerbne r, Gross, Signorielli, & Morgan, 1980; Signorielli, 2004). It was also expected that in the no -priming condition, heavy television viewers would report less favorable perceptions of th e age group than light television viewers, being more likely to identify them as sexle ss, insignificant, and comical. However, heavy viewers were not significantly more likely th an light viewers to pe rceive people 65+ as “not sexually active,” “not sexually passionate,” “not se ductive,” “insignificant in society,” “not successful in society,” “not thriving members of society,” “forgotten by society,” “comical,” “laughed at,” “amusing,” or “unintentionally funny.” This means that heavy viewers’ perceptions of people 65+ were not significantly different from light viewers’ perceptions of people 65+ for three of the four items used to measure the sexless construct, all four items used to measure th e insignificant construct, and all four items used to measure the comical construct. Consequently, the results of this study s uggest that when people are not prompted to recognize television as an informati on source, heavy televi sion viewing does not cultivate them to perceive people 65+ as insignificant or comical in the real world, despite television’s depiction of the age group as such (C assata & Irwin, 1997; Greenberg, Graef, Fernandez-Collado, Ko rzenny, & Atkin, 1980; Harwood & Giles, 1992). This is not entirely the case when it comes to the sexless construct, however, because when heavy television viewers are no t prompted to recogni ze television as an information source, heavy viewers were signif icantly more likely than light viewers to perceive people 65+ as “not sexually attractive.” Hence, th e results of this study also suggest that when people are not prompted to recognize te levision as an information

PAGE 92

85 source, heavy television viewing cultivates them to perceive people 65+ as sexually unattractive. Priming Effects Shrum, Wyer, and O’Guinn (1998) posite d that when making a judgment, people do not discount information learned from te levision because they are unmotivated or unable to identify the source of retrieved information unless prompted to do so. Participants in the source-priming and rela tion-priming conditions were prompted to realize that they were calli ng on television information to answer the questionnaire items about people 65+. This realization was supposed to discourage heur istic processing and encourage systematic processing, which would lead participants to disc ount television as a viable source of information such that no cultivation effect would be observed (i.e. heavy viewers’ perceptions of people 65+ would be comparable to light viewers’ perceptions of people 65+). At first glance, it appears that priming cau sed this study’s participants to discount television information when estimating the pe rcentage of people age 65+ in the U.S. population; heavy television vi ewers and light television vi ewers reported comparable estimates of the 65+ population in the s ource-priming and relation-priming conditions. However, since heavy and light viewers in the no-priming condition also provided comparable estimates of the 65+ population, the priming conditions did not actually have a significant effect on part icipants’ thought processing fo r this particular item. It seems that participants in all three conditions, not just in the two priming conditions, employed systematic processing to decide what percentage of the U.S. population is 65+. Therefore, the results of this study suggest that whether people are

PAGE 93

86 prompted to recognize television as an inform ation source, they mi ght discount television information when making judgments about the number of people 65+. It was also expected that in the sour ce-priming and relation-priming conditions, heavy television viewers and light tele vision viewers would report comparable perceptions of people age 65+ Indeed, in the source-pri ming condition, heavy viewers’ perceptions of people 65+ were not significantly diffe rent from light viewers’ perceptions of people 65+ for three of the four items used to measure the sexle ss construct, all four items used to measure the insignificant constr uct, and three of the four items used to measure the comical construct. However, li ght viewers were significantly more likely than heavy viewers to perceive people 65+ as “not seductive,” and heavy viewers were significantly more likely than light viewers to perceive pe ople 65+ as “unintentionally funny.” Likewise, in the relation-priming cond ition, heavy viewers’ perceptions of people 65+ were not significantly different from li ght viewers’ perceptions of people 65+ for three of the four items used to measure the sexless construct, all four items used to measure the insignificant construct, and all four items used to measure the comical construct. However, heavy viewers were signi ficantly more likely than light viewers to perceive people 65+ as “not sexually passionate.” It seems that participants in all three conditions, not just the two priming conditions, employed heuristic processing to ma ke decisions about people 65+ as sexless, insignificant, and comical. Therefore, the result s of this study suggest that whether people are prompted to recognize television as an information source, they might count television information when making judgmen ts about the nature of people 65+.

PAGE 94

87 Effects of Other Pre-Existing Conditions Hawkins (1977) posited that viewers’ percei ved reality of televi sion as a potential window to the world affects the degree to whic h they apply television’s values and norms to real life. Therefore, high perceived reality of television was expected to enhance the cultivation effect among h eavy viewers in the no-priming condition in this study. Viewers with high perceived real ity of television were expected to be more likely to report television-like perceptio ns of people 65 and older than viewers with low perceived reality of television.Participants with high perceived realit y did not report lower estimates of the percentage of people age 65+ in the U.S. popul ation than participants with low perceived reality. In fact, participants with high percei ved reality actually re ported slightly higher estimates of the 65+ population than part icipants with low perceived reality. The difference between the two group’s estimates wa s not statistically significant, but the results imply that heavy viewers with high pe rceived reality are at least somewhat more likely to overestimate the 65+ population than heavy viewers with low perceived reality. Hence, the results of this study suggest that perceived reality of television does not have a significant effect on heavy viewers’ estimates of the percentage of people age 65+ in the U.S. population. It was also expected that participants with high percei ved reality would report less favorable perceptions of people 65+ than part icipants with low perceived reality, being more likely to identify them as sexless, insignificant, and comical. However, high perceived reality participants’ perceptions of people 65+ were not significantly different from low perceived reality participants’ percep tions of people 65+ for all four items used

PAGE 95

88 to measure the sexless constr uct, all four items used to measure the insignificant construct, and three of the four items used to measure the comical construct. Hence, the results of this study suggest that perceived reality of television does not have a significant effect on heavy viewers’ perceptions of people 65+ as sexless or insignificant. This is not entirely the cas e when it comes to the comical construct, however, because heavy viewers with high pe rceived reality were significantly more likely than heavy viewers with low perceived reality to perceive people 65+ as “comical.” Hence, the results of this study also suggest that high pe rceived reality of television encourages heavy viewers to pe rceive people 65+ as comical. Hawkins and Pingree (1980) posited that di rect experience can serve to disprove television’s picture of the world. In this st udy direct experience with people 65 and older, like perceived reality of television, was expe cted to influence the cultivation effect among heavy viewers in the no-priming c ondition in this stu dy. Unlike increased perceived reality, which was expected to e nhance television-like perceptions of the number and nature of people 65+, increased di rect experience was expected to diminish television-like perceptions of the age group. It was expected that participants with high direct experience with people 65+ would be le ss likely to have television-like perceptions about the age group than partic ipants with low direct e xperience with people 65+. Participants with high direct experience reported slightly higher estimates of the percentage of people age 65+ in the U.S. popul ation than participants with low direct experience, implying that heavy viewers with high direct experience are at least somewhat more likely to pr ovide higher estimates of the 65+ population than heavy viewers with low direct e xperience. However, the diffe rence between the two group’s

PAGE 96

89 estimates was not statistically significant. Hence, the results of this study suggest that high direct experience with people 65+ doe s not have a significant effect on heavy viewers’ estimates of the percentage of people age 65+ in the U.S. population. It was also expected that participants with high di rect experience would report more favorable perceptions of people 65+ than participants with low direct experience, being less likely to identify them as sexless, insignificant, and comical. However, high direct experience participants’ perceptions of people 65+ were not significantly different from low direct experience participants’ percep tions of people 65+ for all four items used to measure the sexless construct, three of the four items used to measure the insignificant construct, and all four items used to measure the comical construct. Hence, the results of this study suggest that direct experience with people 65+ does not have a significant effect on heavy vi ewers’ perceptions of people 65+ as sexless or comical. This is not entirely the case when it comes to the insignificant construct, however, because heavy viewers with low di rect experience were significantly more likely than heavy viewers with high direct experience to percei ve people 65+ as “insignificant in society.” Hence, the result s of this study also suggest that high direct experience with people 65+ discourages hea vy viewers to perceive people 65+ as insignificant in society. Limitations of the Study The most intriguing discovery of this resear ch study also happens to allude to one of the study’s limitations: nearly all of the participants overestimated the percentage of people age 65+ in the U.S. population. The mean estimate was 33%, and the estimates

PAGE 97

90 ranged from 10% to 80%. The U.S. Census Bureau (2006) estimates that in 2006, 12.4% of the U.S. population was 65 and older. Only three out of the 164 pa rticipants in this study underestimated the percentage of people 65+ in the U.S. population; their estimate was 10%. The U.S. Census Bureau’s (2006) estimate for the 65+ population in the state of Florida, however, is 16.8%, which means that the percentage of people 65+ in Florida is 4.4% higher than the national estimate. Still, only 10 participants estimated the U.S. population of people 65+ to be betwee n 12.4% (U.S. estimate of 65+ population percentage) and 16.8% (Florida estimate of 65+ population percentage); two participants estimated it to be 13%, and eight part icipants estimated it to be 15%. It is possible that this study’s participants had a Floridian bias that caused them to overestimate the 65+ population, especially in light of the fact that the questionnaires were administered during a winter month when many people 65+ temporarily reside in Florida due to the warm climate. This potential bias could explain why the heavy television viewers in this study provided higher estimates of the percentage of people 65+ in the U.S. population than the light televisi on viewers. It seems reasonable to assume that television content reflects the demographics of its audience, and since Florida has a larger 65+ population, more people 65+ might appear on television in Florida than in other states. This is especi ally plausible when it comes to television commercials featuring people 65+ in advertisements for health-related products and services, retirement communities, nursing homes, etc. Since heavy viewers watch more television than light viewers, they might also watch these types of commercials more than light viewers and be more likely to overestimat e the 65+ population in the real world.

PAGE 98

91 This study’s participants had somethi ng else in common that could have influenced the results—they were all enro lled in a mass communications course where they were learning about various forms of ma ss media, including televi sion. It is possible that the participants were already sensitized to recognize that tele vision might influence their perceptions about the real world. Th is potential sensitization could explain why participants in this study discounted television information when making judgments about the number of people 65+. Further, this study’s sample had far more female participants (120) than male participants (44), which could have influenced the results. What’s more, throwing out the data of all the participants who were deemed moderate tele vision viewers caused there to be an unequal number of usable questionna ires in each of the three experimental conditions; there were 53 no-priming questi onnaires, 66 source-priming questionnaires, and 45 relation-priming questionnaires. The results of this stu dy might have been different if quota sampling had been used in th e data collection proce ss (i.e., if the study required a total of 50 of each type of questionnaire). Finally, another limitation to the study was that the 12 Likert-type scale items used to measure participants’ perceptions of people 65+ were not founded on previous research on this topic because an extensive review of the literature did not uncover any survey items that could be used to measur e the sexless, insignificant, and comical constructs. The 12 Likert-type items were enti rely original, and as a result, suffered low interitem reliability. Hence, an index could not be created for any of the three constructs because the Cronbach’s Alpha numbers were too low, revealing that the three constructs were multidimensional, not one-dimensional. Hence, each of the four items that were

PAGE 99

92 created to measure the three constructs ha d to be tested indi vidually, which did not provide rich data concerning the participant’s overall perceptions of people 65+ as sexless, insignificant, and comical.

PAGE 100

93 CONCLUSION The results of this study suggest that when people ar e not prompted to recognize television as an information source, heavy television viewing does not cultivate them to underestimate the 65+ population or to perceive people 65+ as insignificant or comical in the real world; under this circumstance, how ever, heavy television viewing does cultivate them to perceive people 65+ as “not se xually attractive.” Hence, heavy exposure to television’s underrepresentation and negativ e portrayal of people 65+ as sexless, insignificant, and comical does not appear to cause people to assume such television-like perceptions of the age group in the real world. Cultivation theory (Gerbner and Gross, 1976) has been supported by studies dedica ted to many television-related topics— violence, gender roles, divorce rates, etc.—but in this study, it was only supported in the no-priming condition by the fact that heavy vi ewers were significantly more likely than light viewers to perceive people 65+ as “not sexually attractive.” It is interesting to note that cultivation effects were also observed in the priming conditions. Heavy viewers reported more tele vision-like perception of people 65+ for the “sexless” and “comical” constructs, but not for the “insignificant” construct. In the source-priming condition, heavy viewers were significantly more likely than light viewers to perceive people 65+ as “unint entionally funny,” and in the relation-priming condition, heavy viewers were significantly mo re likely than light viewers to perceive people 65+ as “not sexually passionate.” T hus, although cultivation was not expected in either of the priming conditions, the presence of these two effects’ presence provides

PAGE 101

94 further support for the proposition that televi sion might be cultivating heavy television viewers to perceive people 65+ as sexless and comical. It appears that Shrum, Wyer, and O’Gui nn’s (1998) findings might be limited to television’s cultivation of perceptions about crime and occupations. Applying their methodology to this study’s analysis of percep tions of people 65+ did not reveal similar results. Shrum et al.’s (1998) research found that priming conditions moderated the extent to which viewers would report television-like perceptions of the real world; in the nopriming condition, heavy televisi on viewers’ perceptions of crime and occupations were more television-like than th e light television viewers’ pe rceptions, and in the sourcepriming and relation-priming condition, a cult ivation effect was not observed. In this study, however, cultivation effects were not obs erved in any of the three conditions in terms of participants’ perceptions of th e size of the 65+ popula tion, but they were observed in all three conditions in terms of participants’ perceptions of people 65+. This study’s results also sugge st that perceived reality of television does not have a significant effect on heavy viewers’ estima tes of the percentage of people age 65+ in the U.S. population or their pe rceptions of people 65+ as sexless or insignificant; however, high perceived reality of televisi on encourages heavy viewers to perceive people 65+ as “comical,” which implies that when heavy viewers believe that television paints a realistic picture of the world, they ar e more likely to be cultivated by television’s characterization of people 65+ as a subject of amusement because of their irrational and eccentric behavior. Further, the results of this study suggest that high direct experience with people 65+ does not have a significant effect on hea vy viewers’ estimates of the percentage of

PAGE 102

95 people age 65+ in the U.S. population or thei r perceptions of people 65+ as sexless or comical; however, high direct experience with people 65+ discourages heavy viewers to perceive people 65+ as “ins ignificant in society.” Hence, it appears that when heavy viewers have real life experiences with peopl e 65+, they perceive them to be more significant members of our society. In conclusion, the results of this study revealed six major findings concerning the relationship between the indepe ndent variables in this st udy—television viewing level, priming, perceived reality of television, and direct experience with people 65+—and the dependent variables, viewer perceptions of the number and nature of people 65+. First, heavy television viewing does not cu ltivate viewers to underestimate the 65+ population in the U.S.; second, heavy televisi on viewing cultivates vi ewers to perceive people 65+ as sexless (specifically, “not sexually attractive” and “not sexually passionate”) and comical (specifically, “uni ntentionally funny”); th ird, priming is not necessary to induce source di scounting of television information for judgment-making about the number and nature of the elderly in the real worl d; fourth, whether people are prompted to recognize television as an information source, they will discount television information when making judgments about th e number of people 65+, and they will count television information when making j udgments about the nature of people 65+; fifth, high perceived reality of television encourages heavy viewers to perceive people 65+ as “comical”; sixth, high direct experi ence with people 65+ discourages heavy viewers to perceive people 65+ as “insignificant in society.”

PAGE 103

96 Directions for Future Research The results of this study s uggest that future research on television’s cu ltivation of perceptions of people 65+ in the real world should focus on perceptions of the age group as sexless and comical. However, future studi es concerning direct e xperience with people 65+ should focus on perceptions of the age gr oup as insignificant in society, and future studies concerning perceived re ality of television should fo cus on perceptions of people 65+ as comical. Since explorator y analysis revealed that in this study, participants with higher grade point averages actually had higher perceived reality levels, future studies that consider television view ers’ perceived reality of the medium should attempt to discover whether increased GPA is truly indicative of incr eased perceived reality of television, which counters the assumption that people with higher GPA’s are smarter and more likely to realize that te levision does not always depict the world as it really is. It is not suggested that Shrum, Wyer, and O’Guinn’s (1998) priming conditions be used to study the topic of old-age on tele vision. Priming conditions did not moderate the cultivation effect in this study, and explor atory analysis revealed that there was not a significant interaction effect between the experimental condition and participants’ television viewing level. Fu rther research should be c onducted, however, to determine whether people will generally discount te levision information when making judgments about the set size of a group (e.g. in this st udy, “the total current percentage of people 65+ in the U.S. population”) and count tele vision information when making judgments about the characteristics of a group (e.g. in this study, “people 65+ as sexless, insignificant, and comical”).

PAGE 104

97 Finally, future research is needed for th e creation of a set of internally-reliable items that can be indexed and used to meas ure television viewers’ perceptions of people 65+ as sexless, insignifican t, and comical in order to provide one-dimensional data concerning the concerning people’s overa ll perceptions of the age group.

PAGE 105

98 LIST OF REFERENCES Aronoff, C. (1974). Old age in prime time. Journal of Communication, 24(4) 86-87. Bell, J. (1992). In search of a discour se on aging: The el derly on television. The Gerontologist, 32(3) 305-311. Bishop, J. M. & Krause, D. R. (1984). Depi ctions of aging and old age on Saturday morning television. The Gerontologist, 24(1) 91-94. Cacioppo, J. T. & Petty, R. E. (1982). The Need for Cognition. Journal of Personality and Social Psychology, 42(1) 116-131. Cassata, M. B., Anderson, P. A., & Skill, T. D. (1983). Images of old age on daytime. In M. Cassata and T. Skill (Eds.), Life on daytime television: Tuning-in American serial drama (pp. 37-44). Norwood, NJ: Ablex Publishing Corp. Cassata, M. B. & Irwin, B. J. (1997). Young by day: The older person on daytime serial drama. In H. S. Noor Al-Deen (Ed.), Cross-Cultural Communication and Aging in the United States (pp. 215-230). Mahwah, NJ: La wrence Erlbaum Associates. Cohen, A. R., Stotland, E. & Wolfe, D. M. ( 1955). An experimental investigation of need for cognition. Journal of Abnormal and Social Psychology, 51 291-294. Dail, P. W. (1988). Prime-time television portr ayals of older adults in the context of family life. The Gerontologist, 28(5) 700-706. Davis, R. H. (1985). TV’s Image of the El derly: A Practical Gu ide for Change. United States of America: D. C. Heath and Company. Davis, R. H. & Kubey, R. W. (1982). Growi ng old on television and with television. In D. Pearl, L. Bouthilet, & J. Lazar (Eds.), Television and behavior: Ten years of scientific progress and imp lications for the eighties (pp. 201-208). Washington, D.C.: Government Printing Office. Doob, A. N. & Macdonald, G. E. (1979). Televisi on viewing and fear of victimization: Is the relationship causal? Journal of Personality and Social Psychology, 37(2) 170179. Elliott, J. (1984). The daytime televisi on drama portrayal of older adults. The Gerontologist, 24(6) 628-633.

PAGE 106

99 Francher, J. S. (1973). “It’s the Pepsi Gene ration…” accelerated ag ing and the television commercial. International Journal of Aging and Human Development, 4(3) 245255. Gerbner, G. (1972). Violence in television dram a: Trends and symbo lic functions. In G. A. Comstock and E. A. Rubinstein (Eds.), Television and Social Behavior, Vol. I, Media Content and Control (pp. 28-187). Washington, D.C.: Government Printing Office. Gerbner, G. (1997). Gender and age in primetime television. In S. Kirschner and D. A. Kirschner (Eds.), Perspectives on Psychology and the Media (pp. 69-94). Washington, D.C.: American Psychological Association. Gerbner, G. & Gross, L. (1976). Living with television: Th e violence profile. Journal of Communication, 26(2) 173-199. Gerbner, G. & Signorielli, N. (1982). The world according to television. American Demographics, 4(9) 14-17. Gerbner, G., Gross, L., Morgan, M., & Signorielli, N. (1980). The "mainstreaming" of America: Violence profile no. 11. Journal of Communication, 30(3) 10-29. Gerbner, G., Gross, L., Morgan, M., & Signor ielli, N. (1986). Living with Television: The Dynamics of the Cultivation Process. In J. Bryant and D. Zillmann (Eds.), Perspectives on media effects (pp. 17-40). Hillsdale, NJ: Lawrence Erlbaum Associates, Inc. Gerbner, G., Gross, L., Signorielli, N., & Morgan, M. (1980). Aging with television: Images on television drama and c onceptions of social reality. Journal of Communication, 30 37-47. Greenberg, B. S. (1980). Three seasons of te levision characters: A demographic analysis. Journal of Broadcasting, 24(1) 49-60. Greenberg, B. S., Edison, N., Korzenny, F ., Fernandez-Collado, C., & Atkin, C. K. (1980). Antisocial and prosocial behaviors on television. In B. S. Greenberg (Ed.), Life on television (pp. 99-128). Norwood, NJ: Ablex Publishing Corp. Greenberg, B. S, Graef, D., Fernandez-Coll ado, C., Korzenny, F., & Atkin, C. K. (1980). Sexual intimacy on commercial television dur ing prime time. In B. S. Greenberg (Ed.), Life on television (pp. 129-136). Norwood, NJ: Ablex Publishing Corp. Greenberg, B. S., Korzenny, F., & Atkin, C. K. (1980). Trends in the portrayal of the elderly. In B. S. Greenberg (Ed.), Life on television (pp. 23-33). Norwood, NJ: Ablex Publishing Corp.

PAGE 107

100 Hajjar, W. J. (1997). The image of aging in television commercials: An update for the 1990s. In H. S. Noor Al-Deen (Ed.), Cross-Cultural Communication and Aging in the United States (pp. 231-244). Mahwah, NJ: La wrence Erlbaum Associates. Harris, A. J., and Feinberg, J. F. (1977). Te levision and Aging: Is What You See What You Get? The Gerontologist, 17 464-468. Harris, R. J. (2004). A cognitive psychology of mass communication (pp. 53-91). Mahwah, NJ: Lawrence Erlbaum Associates, Inc. Harwood, J. (2000). “SHARP!” Lurking incoheren ce in a television portrayal of an older adult. Journal of Language and Social Psychology, 19(1) 110-140. Harwood, J. & Giles, H. (1992). ‘Don’t ma ke me laugh’: Age re presentations in a humorous context. Discourse & Society, 3(3) 403-436. Hawkins, R. (1977). The dimensional structur e of children’s perceptions of television reality. Communication Research, 4 299-320. Hawkins, R. P. & Hoch, S. J. (1992). Lo w-involvement learning: Memory without evaluation. Journal of Consumer Research, 19 212-225. Hawkins, R. P. & Pingree, S. (1980). So me processes in the cultivation effect. Communication Research, 7 193-226. Hawkins, R. P. & Pingree, S. (1982). Televisi on’s influence on constructions of social reality. In D. Pearl, L Bout hilet, & J. Lazar (Eds.), Television and behavior: Ten years of scientific progress and implications for the eighties (Vol. 2, pp. 224247). Washington, DC: U.S. Government Printing Office. Hawkins, R. P. & Pingree, S. (1990). Diverg ent psychological proce sses in constructing social reality from mass media content. In N. Signorielli & M. Morgan (Eds.), Cultivation analysis: New directi ons in media effects research (pp. 35-50). Newbury Park, CA: Sage. Hawkins, R. P., Pingree, S., & Adler, I. ( 1987). Searching for cognitive processes in the cultivation effect. Human Communication Research, 13 553-577. Hiemstra, R., Goodman, M., Middlemiss, M. A., Vosco, R., & Ziegler, N. (1983). How older persons are portr ayed in television advertising: Implications for educators. Educational Gerontology, 9 111-122. Higgins, E. T., Brendl, C. M. (1995). Accessi bility and Applicability: Some Activation Rules Influencing Judgment. Journal of Experimental Social Psychology, 31(3) 218-243.

PAGE 108

101 Hirsch, P. M. (1980). The “scary world” of the nonviewer and other anomalies: A reanalysis of Gerbner et al.’s finding of cultivation analysis, part 1. Communication Research, 7 403-456. Hirsch, P. M. (1981a). Distinguishing good speculation from bad theory: Rejoinder to Gerbner et al. Communication Research, 8 73-95. Hirsch, P. M. (1981b). On not learning from one’s own mistakes: A reanalysis of Gerbner et al.’s findings on cultivation analysis. Communication Research, 8 3-37. Hughes, M. (1980). The fruits of cultivation an alysis: A reexamination of some effects of television watching. The Public Opinion Quarterly, 44(3) 287-302. Hummert, M. L., Garstka, T. A., Shaner, J. L., & Strahm, S. (1994). Stereotypes of the elderly held by young, middle-ag ed, and elderly adults. Journal of Gerontology: Psychological Sciences, 49 240-249. Iyengar, S. & Kinder, D. R. (1987). News That Matters : Television and American Opinion. The University of Chicago Press. Jaccard, James and Choi K. Wan (1996). LISR EL approaches to interaction effects in multiple regression. Thousand Oaks, CA: Sage Publications. Jeffres, L. W. & Perloff, R. M. (1997). Ma ss Media Effects. Prospect Heights, IL : Waveland Press, Inc. Johnson, M. K., Hashtroudi, S., & Linds ay, D. S. (1993). Source Monitoring. Psychological Bulletin, 114 3-28. Kite, M. E., Deaux, K., & Miele, M. (1991) Stereotypes of old and young: Does age outweigh gender? Psychology and Aging, 6 19-27. Kubey, R. W. (1980). Television and ag ing: Past, present, and future. The Gerontologist, 20(1) 16-35. Miller, D., Leyell, T. S., & Mazachek, J. (2004). Stereotypes of the elderly in U.S. television commercials from the 1950s to the 1990s. International Journal of Aging and Human Development, 58 315 -340. Northcott, H. C. (1975). Too young, too old – Age in the world of television. The Gerontologist, 15(2) 184-186. O'Guinn, T. C. & Shrum, L. J. (1997). The role of television in the construction of consumer reality. The Journal of Consumer Research, 23(4) 278-294. Petersen, M. (1973). The visibility an d image of old people on television. Journalism Quarterly, 50(3) 569-573.

PAGE 109

102 Peterson, R. T. & Ross, D. T. (1997). A Cont ent Analysis of the Portrayal of Mature Individuals in Televi sion Commercials. Journal of Business Ethics, 16(4) 425433. Petty, R. E. & Cacioppo, J. T. (1986). Communication and Persuasion. New York: Springer-Verlag New York Inc. Pingree, S. (1983). Children’s cognitive pr ocesses in constructing social reality. Journalism Quarterly, 60 415-422. Potter, W. J. (1986). Perceived Reality and the Cultivation Hypothesis. Journal of Broadcasting & Electronic Media, 30(2) 159-174. Potter, W. J. (1988). Perceived Real ity in Television Effects Research. Journal of Broadcasting & Electronic Media, 32(1) 23-41. Potter, W. J. (1991a). Examining cultivat ion from a psychological perspective: Component subprocesses. Communication Research, 18(1) 77-102. Potter, W. J. (1991b). The Relationships Betw een Firstand Second-Order Measures of Cultivation. Human Communication Research, 18(1) 92-113. Robinson, T. & Anderson, C. (2006). Older char acters in children’s animated television programs: A content analys is of their portrayal. Journal of Broadcasting and Electronic Media, 50(2) 287-304. Robinson, J. D., & Skill, T. (1995). The invisibl e generation: Portrayals of the elderly on prime-time television. Communication Reports, 8(2) 111-119. Roskos-Ewoldsen, D. R., Klinger, M. R., & Roskos-Ewoldsen, B. (2007). Media priming: An meta-analysis. In R. W. Prei ss, B. M. Gayle, N. Burrell, M. Allen, & J. Bryant (Eds.), Mass media effects research: Advances through meta-analysis (pp. 53-80). Mahwah, NJ: Lawrence Erlbaum Associates, Inc. Roy, A., & Harwood, J. (1997). Underrepresente d, positively portrayed: Older adults in television commercials. Journal of Applied Communication Research, 25 39-56. Rubin, A. M., Perse, E. M., & Taylor, D. S. (1988). A Methodological Examination of Cultivation. Communication Research, 15(2) 107-134. Schmidt, D. F., & Boland, S. M. (1986). The structure of impressi ons of older adults: Evidence for multiple stereotypes. Psychology and Aging, 1 255-260. Schwarz, N., Strack, F., Bless, H., Klum pp, G., Rittenauer-Schatka, H., & Simons, A. (1991). Ease of recall as information: Anot her look at the avai lability heuristic. Journal of Personality and Social Psychology, 61(2), 195-202.

PAGE 110

103 Shanahan, J. & Morgan, M. (1999). Television and its Viewers: Cultivation theory and Research (pp. 172-197). Cambridge University Press. Shapiro, M. A. (1991). Memory and decision processes in the construction of social reality. Communication Research, 18(1) 3-24. Shapiro, M. A., & Lang, A. (1991). Making tele vision reality: Unconscious processes in the construction of social reality. Communication Research, 18(5) 685-705. Sherman, S. J. & Corty, E. (1984). Cognitive he uristics. In R. S. Wyer & T. K. Srull (Eds.), Handbook of social cognition (Vol. 1, pp. 189-286). Hillsdale, NJ: Lawrence Erlbaum. Shrum, L. J. (1995). Assessing the social in fluence of television: A social cognition perspective on cultivation effects. Communication Research, 22(4) 402-429. Shrum, L. J., Wyer, R. S. & O’Guinn, T. C. (1998). The effects of television consumption on social perceptions: The use of priming procedures to investigate psychological processes. Journal of Consumer Research, 24 447-458. Shrum, L. J. (1999b). Television and Persua sion: Effects of the Programs between the Ads. Psychology and Marketing, 16(2) 119-140. Shrum, L. J. & O’Guinn, T. C. (1993). Processe s and effects in the c onstruction of social reality: Construct accessibility as an exploratory variable. Communication Research, 20(3) 436-471. Shrum, L. J. (2002). Media Consumption and Pe rceptions of Social Reality: Effects and Underlying Processes. In J. Bryant & D. Zillmann (Eds.), Media Effects: Advances in Theory and Research (pp. 69-95). Hillsdale, NJ: Lawrence Erlbaum Associates, Inc. Shrum, L. J. (2007a). Social cognition and cu ltivation. In D. R. Roskos-Ewoldsen & J. L. Monahan (Eds.), Communication and Social Cognition: Theories and Methods (pp. 245-272). Mahwah, NJ: Lawrence Erlbaum Associates, Inc. Signorielli, N. (1983). Health, prevention, a nd television: Images of the elderly and perceptions of social reality. Prevention in Human Services, 3(1) 97-117. Signorielli, N. (1985). Role portrayal and stereotyping on televi sion: An annotated bibliography of studies relating to women, minorities, aging, sexual behavior, health, and handicaps (pp. 142-165). We stport, CN: Greenwood Press. Signorielli, N. (2004). Aging on television: Messages Rela ting to Gender, Race, and Occupation in Prime Time. Journal of Broadcasting & Electronic Media 48(2) 297-301.

PAGE 111

104 Signorielli, N. & Gerbner, G. (1978). The im age of the elderly in prime-time network television drama. Generations, 3(2) 10-11. Swayne, L. E., & Greco, A. J. (1987). The por trayal of older Americans in television commercials. Journal of Advertising, 16 47-54. Tversky, A. & Kahneman, D. (1973). Availabil ity: A heuristic for judging frequency and probability. Cognitive Psychology, 5 207-232. U.S. Census Bureau (2004). “U.S. Interim Projections by Age, Sex, Race, and Hispanic Origin,” Retrieved November 3, 2007. http://www.census.gov/ipc/www/isinterimproj/ U.S. Census Bureau (2006). “State and C ountry Quick Facts,” Retrieved March 10, 2008. http://quickfacts.census.gov/qfd/states/12000.html Weimann, G. (2000). Communicating unreality: Modern media and the reconstruction of reality (pp. 39-76). Thousand Oaks, CA: Sage Publications, Inc. Wyer, R. S. & Srull, T. K. (1986). Hu man cognition in its social context. Psychological Review, 93 322-359.

PAGE 112

105 BIBLIOGRAPHY Ferreira, M.B., Garcia-Marques, L., Sher man, J.W., & Sherman, S.J (2006). Automatic and controlled components of judgment and decision making. Journal of Personality and Social Psychology, 91(5) 797-813. Gerbner, G., Gross, L., Morgan, M., Signorie lli, N., & Shanahan, J. (2002). Growing up with television: Cultivation Processes. In J. Bryant & D. Zillmann (Eds.), Media Effects: Advances in Theory and Research (pp. 43-67). Hillsdale, NJ: Lawrence Erlbaum Associates, Inc. Higgins, E. T., Rholes, W. S., & Jones, C. R. (1977). Category accessibility and impression formation. Journal of Experimental Social Psychology, 13 141-154. Lichter, S. R., Lichter, L. S., Roth man, S., & Amundson, D. (1987). Prime-time prejudice: TV’s images of blacks and Hispanics. Public Opinion 13-16. Manis, M., Shedler, J., Jonides, J., & Nels on, T. E. (1993). Availability heuristic in judgments of set size and frequency of occurrence. Journal of Personality and Social Psychology, 65(3) 448-457. Passuth, P. M. & Cook, F. L. (1985). Eff ects of television viewing on knowledge and attitudes about older adults : A critical reexamination. The Gerontologist, 25(1) 69-77. Perse, E. M. (2001). Media Effects and Society (pp. 164-178). Mahwah, NJ: Lawrence Erlbaum Associates, Inc. Powell, L. & Williamson, J. (1985). The mass media and the aged. Social Policy, 16 3849. Robinson, T. (1998). Portraying older pe ople in advertising New York: Garland. Rubin, A. M. (1982). Directions in television and aging research. Journal of Broadcasting, 26(2) 537-551. Shrum, L. J. (2007b). The implications of survey method for measuring cultivation effects. Human Communication Research, 33 64-80.

PAGE 113

106 Srull, T. K., & Wyer, R. S. (1979). The role of category accessibility in the interpretation of information about persons: Some determinants and implications. Journal of Personality and Social Psychology, 37(10) 1660-1672. Srull, T. K., & Wyer, R. S. (1980). Categor y accessibility and social perception: Some implications for the study of person memory and interpersonal judgments. Journal of Personality and Social Psychology, 38(6) 841-856. Wyer, R. S. (1980). The acquisition and use of social knowledge: Basic postulates and representative research. Personality and Social Psychology Bulletin, 6 558-573. Tversky, A. & Kahneman, D. (1982). Judgment under uncertainty: Heuristics and biases. In D. Kahneman, P. Slovic & A. Tversky (Eds.), Judgment under uncertainty (pp. 3-20). New York, NY: Cambridge University Press.

PAGE 114

107 Appendices

PAGE 115

108 Appendix A: No-Priming Questionnaire QUESTIONNAIRE This questionnaire has been composed by a grad uate student at the University of South Florida as part of a research study on per ceptions of United States residents age 65 and older (65+). Your answers are guaranteed to remain absolutely confidential and the questionnaire will only take about 10 minutes to complete. SECTION I: Please write your response on the given line. 1. What percent (between 0% and 100%) of the current U.S. population is age 65+? __________ SECTION II: Please write the corresponding number for your response to each statement on the given line. Use the following rating scale: 1 = Strongly Disagree 2 = Disa gree 3 = Undecided 4 = Agree 5 = Strongly Agree _______ 2. People 65+ are not sexually active. _______ 3. People 65+ are in significant in society. _______ 4. People 65+ are comical. _______ 5. People 65+ are sexually attractive. _______ 6. People 65+ are not successful in society. _______ 7. People 65+ are laughed at. _______ 8. People 65+ are not sexually passionate. _______ 9. People 65+ are thri ving members of society. _______ 10. People 65+ are not amusing. _______ 11. People 65+ are not seductive. _______ 12. People 65+ are forgotten by society. _______ 13. People 65+ are unintentionally funny. SECTION III: Please write your response on the given line or place a check mark on the appropriate line. 14. What is your sex? _______ Male _______ Female 15. What is your age? _______ 16. What is your current GPA to the nearest tenth? _______ 17. On an average weekday morning, how many hours of television do you watch? _______ 18. On an average weekday afternoon, how many hours of television do you watch? _______

PAGE 116

109 Appendix A: No-Priming Qu estionnaire (continued) 19. During prime time on an average w eekday, how many hours of television do you watch? _______ 20. During late night on an average w eekday, how many hours of television do you watch? _______ 21. On an average Saturday, how many hours of television do you watch? _______ 22. On an average Sunday, how many hours of television do you watch? _______ SECTION IV: Please write the corresponding number for your response to each statement on the given line. Use the following rating scale: 1 = Strongly Disagree 2 = Disa gree 3 = Undecided 4 = Agree 5 = Strongly Agree _______ 23. Television shows life as it really is. _______ 24. Television presents things as they really are in life. _______ 25. If I see something on television I can be sure it really is that way. _______ 26. Television lets me see how other people live. _______ 27. Television lets me see what happens in other places as if I’m really there. SECTION V: Please write the corresponding number for your response to each statement on the given line. Use the following rating scale: 1 = Strongly Disagree 2 = Disa gree 3 = Undecided 4 = Agree 5 = Strongly Agree _______ 28. I live or have lived with at least one person who is 65+. _______ 29. People 65+ are present in my life. _______ 30. I frequently interact with people 65+. Thank you. Have a nice day.

PAGE 117

110 Appendix B: Source-Priming Questionnaire QUESTIONNAIRE This questionnaire has been composed by a grad uate student at the University of South Florida as part of a research study on per ceptions of United States residents age 65 and older (65+). Your answers are guaranteed to remain absolutely confidential and the questionnaire will only take about 10 minutes to complete. SECTION I: Please write your response on the given line or place a check mark on the appropriate line. 1. What is your sex? _______ Male _______ Female 2. What is your age? _______ 3. What is your current GPA to the nearest tenth? _______ 4. On an average weekday morning, how many hours of television do you watch? _______ 5. On an average weekday afternoon, how many hours of television do you watch? _______ 6. During prime time on an average weekday, how many hours of television do you watch? _______ 7. During late night on an average weekday, how many hours of television do you watch? _______ 8. On an average Saturday, how many hours of television do you watch? _______ 9. On an average Sunday, how ma ny hours of television do you watch? ______ SECTION II: Please write the corresponding number for your response to each statement on the given line. Use the following rating scale: 1 = Strongly Disagree 2 = Disa gree 3 = Undecided 4 = Agree 5 = Strongly Agree _______ 10. Television shows life as it really is. _______ 11. Television presents things as they really are in life. _______ 12. If I see something on television I can be sure it really is that way. _______ 13. Television lets me see how other people live. _______ 14. Television lets me see what happens in other places as if I’m really there. SECTION III: Please write your response on the given line. 15. What percent (between 0% and 100%) of the current U.S. population is age 65+? _________

PAGE 118

111 Appendix B: Source-Priming Questionnaire (continued) SECTION IV: Please write the corresponding number for your response to each statement on the given line. Use the following rating scale: 1 = Strongly Disagree 2 = Disa gree 3 = Undecided 4 = Agree 5 = Strongly Agree _______ 16. People 65+ are not sexually active. _______ 17. People 65+ are in significant in society. _______ 18. People 65+ are comical. _______ 19. People 65+ are sexually attractive. _______ 20. People 65+ are not successful in society. _______ 21. People 65+ are laughed at. _______ 22. People 65+ are not sexually passionate. _______ 23. People 65+ are th riving members of society. _______ 24. People 65+ are not amusing. _______ 25. People 65+ are not seductive. _______ 26. People 65+ are forgotten by society. _______ 27. People 65+ are unintentionally funny. SECTION V: Please write the corresponding number for your response to each statement on the given line. Use the following rating scale: 1 = Strongly Disagree 2 = Disa gree 3 = Undecided 4 = Agree 5 = Strongly Agree _______ 28. I live or have lived with at least one person who is 65+. _______ 29. People 65+ are present in my life. _______ 30. I frequently interact with people 65+. Thank you. Have a nice day.

PAGE 119

112 Appendix C: Relation-Priming Questionnaire ***READ THIS BEFORE YOU CONTINUE*** In order to answer these ques tions, you will use information from a variety of sources, including television. You shoul d be aware that people age 65 and older (65+) are underrepresented on television compared to thei r actual percentage in the population. The age group is also negatively portrayed on tele vision. Consequently, people often use this information to formulate answers.

PAGE 120

113 Appendix C: Relation-Primi ng Questionnaire (continued) QUESTIONNAIRE This questionnaire has been composed by a grad uate student at the University of South Florida as part of a research study on per ceptions of United States residents age 65 and older (65+). Your answers are guaranteed to remain absolutely confidential and the questionnaire will only take about 10 minutes to complete. SECTION I: Please write your response on the given line. 1. What percent (between 0% and 100%) of the current U.S. population is age 65+? __________ SECTION II: Please write the corresponding number for your response to each statement on the given line. Use the following rating scale: 1 = Strongly Disagree 2 = Disa gree 3 = Undecided 4 = Agree 5 = Strongly Agree _______ 2. People 65+ are not sexually active. _______ 3. People 65+ are in significant in society. _______ 4. People 65+ are comical. _______ 5. People 65+ are sexually attractive. _______ 6. People 65+ are not successful in society. _______ 7. People 65+ are laughed at. _______ 8. People 65+ are not sexually passionate. _______ 9. People 65+ are thri ving members of society. _______ 10. People 65+ are not amusing. _______ 11. People 65+ are not seductive. _______ 12. People 65+ are forgotten by society. _______ 13. People 65+ are unintentionally funny. SECTION III: Please write your response on the given line or place a check mark on the appropriate line. 14. What is your sex? _______ Male _______ Female 15. What is your age? _______ 16. What is your current GPA to the nearest tenth? _______ 17. On an average weekday morning, how many hours of television do you watch? _______ 18. On an average weekday afternoon, how many hours of television do you watch? _______

PAGE 121

114 Appendix C: Relation-Primi ng Questionnaire (continued) 19. During prime time on an average w eekday, how many hours of television do you watch? _______ 20. During late night on an average w eekday, how many hours of television do you watch? _______ 21. On an average Saturday, how many hours of television do you watch? _______ 22. On an average Sunday, how many hours of television do you watch? _______ SECTION IV: Please write the corresponding number for your response to each statement on the given line. Use the following rating scale: 1 = Strongly Disagree 2 = Disa gree 3 = Undecided 4 = Agree 5 = Strongly Agree _______ 23. Television shows life as it really is. _______ 24. Television presents things as they really are in life. _______ 25. If I see something on television I can be sure it really is that way. _______ 26. Television lets me see how other people live. _______ 27. Television lets me see what happens in other places as if I’m really there. SECTION V: Please write the corresponding number for your response to each statement on the given line. Use the following rating scale: 1 = Strongly Disagree 2 = Disa gree 3 = Undecided 4 = Agree 5 = Strongly Agree _______ 28. I live or have lived with at least one person who is 65+. _______ 29. People 65+ are present in my life. _______ 30. I frequently interact with people 65+. Thank you. Have a nice day.