xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim
leader nam 22 Ka 4500
controlfield tag 007 cr-bnu---uuuuu
008 s2010 flu s 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0004686
The emotional guardianship of foreign-born and native-born hispanic youth and its effect on violent victimization
h [electronic resource] /
by Amy Eggers.
[Tampa, Fla] :
b University of South Florida,
Title from PDF of title page.
Document formatted into pages; contains X pages.
Thesis (MA)--University of South Florida, 2010.
Includes bibliographical references.
Text (Electronic thesis) in PDF format.
Mode of access: World Wide Web.
System requirements: World Wide Web browser and PDF reader.
ABSTRACT: This study seeks to expand the scope of assimilation theory by integrating it with elements of routine activities theory to better understand what influence assimilation has in regard to violent victimization. Specifically, the purpose of this study is to determine whether or not differences in victimization rates between foreign-born and native-born Hispanic youth are related to variations in emotional guardianship. Emotional guardianship refers to the aspect of relationships (i.e., affection and communication) between Hispanic youth and their parents that serve to protect the youth from being victimized. I hypothesize that foreign-born Hispanics have greater emotional guardianship than native-born Hispanics, and as a result foreign-born Hispanics have lower probabilities of victimization. To test this hypothesis and others, the National Longitudinal Study of Adolescent Health (Add Health) is utilized, as it provides data about the various aspects of assimilation (e.g., country of birth, language spoken at home), routine activities (e.g., sports, clubs, and family outings), and emotional guardianship (e.g., communication of problems, expectations, and satisfaction of parental bond), which are each believed to contribute to the likelihood of being victimized.
Advisor: Ojmarrh Mitchell, Ph.D.
Routine activities theory
t USF Electronic Theses and Dissertations.
The Emotional Guardianship of Foreign Born and Native Born Hispanic Youth and Its Effect on Violent Victimization b y Amy Sheena Eggers A thesis submitted in partial fulfillmen t of the requirements for the degree of Master of Arts Department of Criminology College of Behavioral and Community Science University of South Florida Major Professor: Ojmarrh Mitchell, Ph.D. Shayne Jones, Ph.D Wilson Palacios, Ph.D. Date of Approval: September 1 6, 2010 Keywords: assimilation theory routine activities theory, parent child bonding, parental supervision, Latino youth Copyright 2010, Amy Sheena Eggers
Dedication I would like to dedicate this manuscript to my parents who have provided all the forms of support that a person can imagine. Even during times of doubt and disappointment, I was always encouraged to continue with my studies, all the while being assured that my feelings were simply a part of the overall process. It is for this reason that I consider this manuscript to belong to them just as much as it belongs to me. They are the driving force that has allowed me to have reached this point and accomplish my goals up until now. I thank God for having bles sed me with parents who possess the ability to love, guide, and listen at all times, for these abilities have contributed to my overall success in graduate school, which has been quite a challenging, yet, rewarding experience. Therefore, I feel that it is only appropriate to mention the two individuals whom God provided to serve as my foundation during this part of my life.
Acknowledgements First and foremost, I would like to express my gratitude to my chair, Dr. Ojmarrh Mitchell, who from the beginning, has served as a pillar and a guide in completing my thesis. During this entire process, Dr. Mitchell was very thorough in his revisions and was always available to answer my questions. He served as my mentor and shared hi s wisdom throughout the course of this endeavor. He allowed me to draw from my own intellect ual capacity and provided the finishing touches on those ideas. For these reasons, I will always be thankful for having served under his leadership and direction. I would like to thank my committee members, Dr. Wilson Palacios and Dr. Shayne Jones, who also contributed to the overall development and successful completion of this manuscript. Dr. Palacios assisted me with the merging of criminological and sociological theory, which led to the birth of my thesis idea. As for Dr. Jones, his assistance was provided both in and out of the classroom. Through his courses, I have been able to hone my writing skills, which was necessary and quite helpful throughout this proces s. His honesty and guidance also allowed me to better prepare for the daunting task of completing my thesis, for which I am truly grateful.
i Table of Contents List of Tables ii Abstract iii Chapter I : Introduction 1 Chapter II : Literature Review 8 Assimilation Theory 1 3 Routine Activities Theory 1 5 Target Suitability 16 Guardianship 1 8 Summary 20 Chapter III : Methodology 23 Hypotheses 23 Data 24 Sub Sample 25 Measures 26 Target Suitability 26 Emotional Guardianship 27 Victimization 30 Data Analytic Plan 34 Chapter IV : Results 3 6 Chapter V : Discussion 44 Limitations 47 Implications/Future Research 4 8 References 52
ii List of Tables Table 1. Demographic Characteristics of the Hispanic Sample from Add H ealth (N=595) 2 5 Table 2. Routine Activities Variables from Add Health (N=595) 27 Table 3. Emotional Guardianship Variables from Add Health (N=5 95) 28 Table 4. Reliabilities for Summed RAT and Emotional Guard ianship Variables 29 Table 5. Violent Victimization Variable from Add Health (N=595) 30 Table 6. Correlations of Demographics, Routine Activities, Emotional Guardianship, and Violent Victimization (N=595) 32 Table 7. Correlations of Demographics, Routine Activities, Emotional Guardianship and Violent Victimization (N=595) Cont. 33 Table 8. Logistic Regression Predicting Violent Victimiza tion for U.S. Born Youth, Routine Activities, and Emotional Guardianship 39 Table 9. Summary of Support for Hypotheses 43
iii Abstract This study seeks to expand the scope of assimilation theory by integrating it with elements of routine activities theory to better understand what influence assimilation has in regard to violent victimization. Specifically, the purpose of this study is to determine whether or not differences in victimization rates between foreign born and native born Hispanic youth are related to variations in emotional guardianship. Emotional guardianship refers to the aspect of relationships (i.e., affect ion and communication) between Hispanic youth and their parents that serve to protect the youth from being victimized. I hypothesize that foreign born Hispanics have greater emotional guardianship than native born Hispanics, and as a result foreign born Hi spanics have lower probabilities of victimization. To test this hypothesis and others, the National Longitudinal Study of Adolescent Health (Add Health) is utilized, as it provides data about the various aspects of assimilation (e.g., country of birth, lan guage spoken at home), routine activities (e.g., sports, clubs, and family outings), and emotional guardianship (e.g., communication of problems, expectations, and satisfaction of parental bond), which are each believed to contribute to the likelihood of b eing victimized.
1 Chapter I : Introduction With in the United States, the Hispanic population has more than doubled between 1980 and 2000, from 14.5 million to an estimated 35.3 million, and now equals African Americans as the largest ethnic group (Martinez, 2002, p. 1). This increase in population the Caribbean, [which] has transformed the ethn ic and racial composition of the U.S. Estrada, 2006, p. 66). According to Martinez (2002), due to the size of this population, the immigration debate has mainly focused on Hispanics. For example, man y problems that are attributed to immigration, such as the rise in urban decay, intergroup conflict, and crime have been linked to Hispanics (p. 1). reflected in various way s, such as in immigration policy (i.e., Emergency Immigration Act of 1921 and the Immigration Act of 1924) and the media (i.e., the local and national coverage of the arrival of Mariel Cubans in 1980; Martinez, 2006). One indication of this sentiment remai ning over time is the fact that in 2000, 25 percent of a nationally immigrants cause higher crime rates and another 48 percent answered that it was aut et al., 2006, p. 68). In other words, about three quarters of the sample believed that immigrants cause higher crime rates. Interestingly, this
2 perception continues to cause rifts among people even to this day as evidenced by the current debate regardi ng the new law in Arizona, which allows police to question people making it necessary for Hispanics to carry their legal papers at all times (Reinhard, 2010). Notab (Hannity, 2010). Senator officers ha ve been killed and maimed in the line of duty dealing with illegal immigrants 2010). Clearly, many Americans believe there is an association between immigration and crime; h owever, past research indicates that this fear or belief is unfounded. Prior research, however, demonstrates that Hispanic immigrants (i.e., foreign born first generation Hispanics) are less likely than Hispanic youth born in the United State (i.e., nati ve born youth) to participate in criminal activities. For instance, Morenoff and Astor ( 2006) demonstrated that second and third generation individuals were more likely to engage in the acts of hitting someone, throwing objects at someone, carrying a weapo n, being involved in a gang fight and picking pockets/snatching purses in comparison to first generation individuals. They also determined that among first generation youth, those who arrived to the United States at a younger age were more likely to commit one of these acts during their adolescence (with the exception of pick pocketing) than those who arrived at a later age. Rumbaut et al. ( 2006) also found similar results regarding the incarceration rate for native born individuals, which was four times gr eater than that for foreign born youth. In regard to time in the United States, the more
3 time the immigrants resided in the United States, the higher their incarceration rates were (p. 73). Therefore, the evidence indicates that Hispanic immigrants do not arrive in the United States with elevated criminal tendencies; rather Hispanics seem to become more assimilating to American culture. Past studies have focused almost solely on Hispanic immigrants as perpetrators of crime. For example, several studies examine the relationship between neighborhood immigrant concentration and neighborhood crime rates (Desmond & Kubrin, 2008; Morenoff & Astor, 2006). Likewise, other resear ch assesses the relationship between immigrant concentration and incarceration rat es (Rumbaut et. al., 2006). Much less research, however, has focused on immigrants as crime victims. As noted by Morenoff and Astor (2006), one of the reasons for the lack of research is that many commonly used data sources (e.g., Uniform Crime Reports and the National Victimization Survey) do not include information on the birth status of victims (p. 37). Thus, existing studies only examine one aspect of the relationship betw een immigration and crime. Considering the focus that has been placed on immigrant criminals, especially Hispanic immigrants, it seems only logical that research should also examine victimization among Hispanic immigrants. For reasons that I explain below I believe that it is impossible to fully understand victimization among Hispanic immigrants without taking assimilation and cultural values/beliefs into consideration. To help fill these gaps in the empirical knowledge base, this thesis compares violent victimization rates between native born Hispanic youth and foreign born Hispanic immigrant youth. Specifically, the purpose of this thesis is to determine whether violent victimization rates differ between
4 these two groups of youth, and whether these differences in victimization can be explained by the concepts of assimilation and emotional guardianship. Bearing in mind the results of the above mentioned studies assessing Hispanic immigrants as crime perpetr ators I hypothesize that native born youth have higher rates of victimization than their foreign born counterparts Drawing on concepts from segmented assimilation theory (Portes & Zhou, 1993) and routine activity theory (Cohen & Felson, 1979), I hypothes ize that the higher violent victimization rates of native born youth are due to their greater assimilation into American culture and their reduced emotional guardianship. Assimilation is the product of ethnically diverse individuals moving to another cou ntry and learning to embrace a new culture and behaviors in exchange for or in addition to their old ones. However, the degree to which people begin adapting to this new way of life depends on how much interaction takes place with those of the new and old cultures. S egmented assimilation theory explain s that all immigrant youth will not assimilate uniformly ; instead assimilation depends on various aspects, such as the strength of their family support and co ethnics (Morenoff & Astor, 2006, p. 40). This theo ry argues that immigrant youth who do not possess strong ties with family and co city youth subcultures . and are at risk for experiencing downward assimilation and hence more involvement in cr ). While I believe that segmented assimilation theory contains useful ideas, it has been criticized for not providing a testable set of hypotheses. To derive a set of testable hypotheses, I integrate notions from segment ed assimilation theory with the concept of guardianship from routine activities theory. This
5 between the youth and their family and, serves as the link between assimilat ion theory and routine activities theories influencing criminal opportunity through the convergence of three elements: a motivated offender; a suitable target; and the abs ence of a capable guardian, which they argue are the purpose of this thesis only the latter two elements will be considered. More of an emphasis will be placed on guardianship, since it serve s as the key element in understanding how assimilation and routine activities are related to one another and victimization. According to Cohen and Felson, there are two types of guardianship: formal and informal. Formal guardianship is defined as protectio n provided by the state, such as police officers. Informal guardianship, on the other hand, consists of people familiar with the target (e.g., family, friends, and neighbors). The focus of this study is strictly on informal guardianship, more specifically, on the idea of emotional guardianship, as provided by the parents of these youth. Since the argument being made revolves around the concept of emotional guardianship, parents serve as the best form of guardianship because of the close contact and influenc e they have over their children. Formal guardianship, such as that provided by police officers, does not involve an element of familiarity with the target, which is necessary to understand the potential effects of emotional guardianship. As previously men tioned, research has shown that native born youth are more likely than foreign born youth to engage in criminal activities. Interestingly, part of the explanation for the generational difference in crime participation has been attributed to which refers to
6 (Marin & Gamba, 2003 p. 86). The concept of f amilialism centers on strong identification and attachment with the family (both the immediate family and the extended family) as well as strength, reliability, and unity within relationships between family members. Essentially, the family is viewed as an institution or place of refuge, which can be relied on in times of need. Despite the strong sense of family and unity that seem to exist w ithin Hispanic culture, existing research suggests that familialism is at times susceptible to the effects of assimilation. I use the concept of familialism to develop a form of guardianship that integrates assimilation theory and routine activities theo ry. This type of guardianship is referred to from a strong bond or relationship between the youth and their guardian (i.e., parents). This protection is best provided w hen specific aspects of a relationship are present, such as good communication, trust, and affection. The significance of emotional guardianship stems from the belief that strong relationships will equip guardians with the tools (i.e., information regardin g their whereabouts and who they are going out with) they need to better shield youth from criminals. Essentially, emotional guardianship serves as the key concept in this thesis as it is based on cultural values that are associated with assimilation (i.e ., familialism) and it also potentially aids in understanding wh ether guardianship assists in shielding individuals from criminals. Despite the strong sense of family and family unity that seem to exist within Hispanic culture, existing research suggests that familialism is at times susceptible to the effects of assimilation. For instance, Sabogal, Marin, Otero Sabogal, Vanoss Marin and Perez Stable (1987) reported that familial obligations and the perception of the family
7 appear to diminish in importance with the level of assimilation Assimilation has also been shown to be associated with less family support among Mexican American families (Barrett, Joe, & Simpson, 1991). Thus, existing research supports the notion that the strong family attachment chara cteristic of Hispanic culture is negatively affected by assimilation in American culture. In the present study, the significance of Hispanic familialism is its ties to emotional guardianship provided by the family, which is expected to be negatively impac be the driving force behind emotional guardianship due to the emphasis and imp ortance of family ties. This emphasis is expected to vary between youth born inside and outside of the United States, as a result of assimilation. In other words, the belief is that the attachment/protection of the youth (or the lack thereof) is due, in la rge part, to the emphasis placed on the preservation of familialism. Overall, emotional guardianship links assimilation, familialism, and routine activities together because the level of guardianship will be affected by how assimilated the youth are, which is expected to influence the likelihood of victimization. This thesis proposal is organized as follows. Chapter two discusses segmented assimilation and routine activities theories and the empirical research assessing these theories. Chapter three begi ns by listing the hypotheses and then discusses, in some detail, the AddHealth data set, measures of key concepts, and the analytic strategy used to test the research hypotheses. Chapter four describes the analyses, their results, and their implications. A nd the final chapter discusses my findings and their implications.
8 Chapter II : Literature Review The purpose of this chapter is to examine the theoretical and empirical history of segmented assimilation and routine activities theory as it pertains to this thesis. The present conceptual framework revolve s around the belief that generational differences exist between foreign born and na tive born Hispanic youth in regard to violent victimization rates The following discussion demonstrates how the concept of emotional guardianship (which is based on familialism) links segmented assimilation and routine activities theories. This informatio n will serve as an integral part of understanding differences between foreign born and native born Hispanic youth and whether these differences can be attributed to emotional guardianship and routine activities. Originally, Park and Burgess (1969) devel oped classical a ssimilation theory. These authors described assimilation as consisting of a si ngle path (i.e., all people who assimilate did so in a similar fashion and with like results). T hey argued that assimilation occurs naturally and most rapidly amo ng primary contacts or in relationships that are treasured and intense, such as those found among family members and in intimate social circles. In th fusion in which persons and gr oups acquire the memories, sentiments, and attitudes of other persons or groups, and, by sharing their experience and history, are incorporated ( p. 735). In essence, immigrants are expected to
9 adopt the customs, the lan guage, and way of life of the host country (i.e., the country to which they have immigrated) in a uniform fashion. Segmented assimilation theory on the other hand, was developed by Portes and Zhou (1993) and is based on s too diverse for any one path of assimilation. So, rather than focusing on one path like Park and Burgess, this theory incorporates three irst path is integration into the middle class, which is considered the ideal path and is the one described in classical assimilation theory. The second path leads to permanent poverty and assimilati on into the lower class. For those who do not possess st rong family ties, immigrant youth are turn leads down this path to permanent poverty and lower class status. The third path consists of attaining economic advancement whi le still preserving the immigrant The successful outcome of this path stems from the assistance that is provided by their co ethnic community. Examples of this assistance include employment opportunities, preserving the family and encouraging parental authority over children (Morenoff & Astor, 2006, p. 41). Aside from the different paths that exist, Portes and Rumbaut (2001) suggest that assimilation is affected by the mode or rate of acculturation experienced by youth, whic h can influence the relationship between parents and children. For instance, when parents and children acculturate or do not acculturate together, this is known as consonant acculturation. Selective acculturation refers to when both groups only partially a cculturate; however, when children acculturate faster than their parents, this is referred
10 to as dissonant acculturation. This last mode is said to be the most problematic because it can result in ties being severed, increased conflicts, and loss of commun ication between and provide support for the youth, which is necessary because and community resources available to support children, who c onfront numerous (Xie & Greenman, 2005, p. 3 4). Interestingly, this theory views pre servation of native culture and customs as potentially beneficial with regard to assimilation and upward mobility (i.e. moving into middle class status), u nlike classical assimilation which considers any preservation to be disadvantageous. For the purpose of this thesis, the significance of assimilation is derived from its possible association with familialism. Familia lism was first explored among Mexican American families with the over arching conclusion revealing that the family served as a source of protection, warmth, and solidarity (Becerra, 1988; Grebler, Moore, & Guzman, 1970; Keefe, 1980; Moore, 1970; Schwartz, 2 007). However, according to Portes and Zhou (1997) the lack of family ties can lead to what is referred to as downward assimilation as well as involvement in crime, which has been found to vary among first generation and second generation youth. Since the focus here is on violent victimization, the generational differences that exist regarding crime participation are expected to apply to victimization as well. Considering the manner in which family is esteemed within Hispanic culture, the concept of emotio nal guardianship will be utilized to determine whether variations in family ties are related to an increase or decrease in the likelihood of
11 violent victimization. In order to gauge the likelihood of violent victimization, the theory of routine activities will also be implemented. Cohen and Felson (1979) routine activities theory has three main elements: a motivated offender, a suitable target, and a capable guardian. Essentially, Cohen and Felson theorize that when a motivated offender crosses paths wit h a suitable and unguarded target, a direct contact predatory violation is committed Predatory violations ar p. 589). Specifically, Cohen and Felson describe target suitability as being related to four major components, which are: valuable items (i.e., the desirability of the item or the individual on the part of the offender; in regard to desired items, electronics and automobiles are most coveted by offenders); physical visibility; accessibility; and inertia (i.e., characteristics related to the item or person which may increase or decrease the level of difficulty in regard to their removal by the offender, such as the weight and size of the item or the locks on the item). Target suitability is also related to routine activities in that the availability of illegally removing these items or attacking their owners is affected by encounter with a motivated offender). Essentially, motivated offenders will take these various characteristics (e.g., visibility and accessibility) into consideration to determine the likelihood of successfully completing their attack. Due to the manner in which routine activities theory has been presented in this thesis, it bears a resemblance to soci al bonding theory (1969 ). For example, attachment refers to the
12 importance given to the opinions and feelings of significant others, such as parents, teachers, and peers. This definition is similar to the measure being utilized here fo r guardianship, which focuses on the emotional relationship between the youth and their parents. nvolvement consists of the dedication given to conventional activities, for example, sports and clubs, which are expected to kee p youth too preoccupied to engage in delinquency. Involvement as defined by Hirschi is similar to i ncrease guardianship and reduce target suitability, which in turn can lower the probability of victimization. Thus the concepts of social bonding mirror the definitions and measures that will be utilized in the present study to test whether family ties and participation in different activities are related to violent vi ctimization. However, because r outine activities theory focuses on victimization rather than delinquency, the former theory and its concepts serve the purpose of this thesis more so than social bonding theory. And therefore, this thesis utilizes the routin e activities perspective. The following discussion summarizes past research the measures used in this research, as well as the conclusions that were drawn. These studies demonstrate what remains to be examined in the areas of assimilation (i.e., the eff ect of assimilation on cultural values) and routine activities, especially, guardianship. This understanding of cultural values may provide some insight as to their effect on the likelihood of violent victimization.
13 Assimilation Theory S everal studies have been conducted to test the relationship between assimilation and illicit behavior. The results of these studies generally indicate that assimilation is positively related to involvement in illicit behavior. For instance, Caetano (1987) examin ed the relationship between alcohol use and assimilation as measured by language use, English or Spanish media preference, among other variables in a nationally representative sample of Hispanic immigrants. Caetano found that youth categorized into the hig h assimilation group reported more frequent episodes of heavy drinking and increased opportunities for drinking in comparison to the low acculturation group. Similarly, Myers et al. (2009) found that greater English proficiency (an indication of assimilat ion) was associated with a greater likelihood of using illicit drugs These researchers suggested that English proficiency may make Hispanic youth more these youth in contact with those who have pro substance abuse attitudes, since they have the ability to interact with a heterogeneous group of youth. In another study, Allen et al. (2008) discovered that Spanish language use was associated with less substance abuse. However, when parental monitoring and social network characteristics (i.e., the percentage of extended family members and substance abusers in the network) were controlled for the association was no longer significant (p. 376). As a result, Allen et al. co ncluded that the importance of language use stems from who m the youth interact with and how they interact. This is an important finding as it suggests English proficiency may tained
14 cultural understanding faster than their parents have, which can affect the effectiveness of their parenting. Another focal point of the assimilation/crime literature is assessing the relationship between length of residence in the United States a nd involvement in illicit activities. The main hypothesis tested in these studies is that the more time immigrants spend in the United States, the greater the likelihood o f committing a crime when compared to their foreign born counterparts So in theory, second and third generation immigrants are expected to have higher rates of crime in comparison to first generation immigrant s Support for this hypothesis was found in Butcher and Morrison Piehl (1998) in that these researchers found that immigrants ha d lower rates of institutionalization (i.e., confinement to correctional facilities, mental hospitals, or other institutions) than native born U.S. residents, even after controlling for other variables. Further, these authors found that more recent immigra nts were less likely to be institutionalized than earlier immigrants. Other research also supports the hypothesis of assimilation leading to illicit behavior. Bui (2009), for example, found that first generation students were less likely to report substanc e abuse, property delinquency, and violent delinquency than second generation and third plus generation students. Bui interprets this finding as a negative aspect of assimilation in that assimilation caused a decrease in the strength of family relationship s as measured by parent child conflicts. Collectively, these studies clearly indicate that greater assimilation into American society leads to greater involvement in illicit behaviors. Further, these findings suggest that assimilation leads to crime becaus e assimilation leads to a break down in parent authority and family functioning.
15 Routine Activities Theory When the theory of routine activities was introduced many researchers implemented and tested its propositions as a macro level theory, which is how it was first presented by Cohen and Felson. For example, Andresen (2006) combined routine activity theory and social disorganization theory and found that the presence of a young population served as a strong predictor of crime. Rice and Smith (2002) also combined the above two theories to study how and why auto theft occurs. Other macro level tests of the theory have consisted of the location of leisure time and rates of serious crime (Messner & Blau, 1987); urban stratification and routine activities on suburban crime rates (Stahura & Sloan, 1988); and juvenile arrest rates for burglaries across cities (Pollock, Joo, & Lawton, 2010). Within each of these studies, the focus was on cities, neighborhoods, and rates of crime for which target suitability was d escribed as possessing certain objects that are desired by motivated offenders. As for guardianship, measures consisted of the percentage of racial heterogeneity, participation in activities away from home, and family households. However, over time, new st udies began emerging that incorporated lifestyle theories (i.e., micro level or individual) into its examination and was first introduced by Hindelang, Gottfredson, & Garofalo ( 1978 ), in which they argued that activity patterns and criminal victimization a re interrelated. For instance, Sampson and Lauritsen (1990) explored the effect of individual lifestyle and ecological proximity in relation to personal violence using both individual and community and Johnston (1996) focused on how unstructured (individual) activities are related to deviant behavior. Essentially, these studies implement a theoretical foundation that revolves on
16 the convergence of the three elements of routine activities but within an individual perspective. In other words, different lifestyles exist for individuals with varying demographic information (e.g., gender, age, and income; Miethe, Stafford, & Long, 1987), which may increase or decrease their likelihood of victimization. T he following discussion consists of various individual level studies and the different measures that were used to test the elements of target suitability and guardianship. Over the years, many studies have been conducted in an attempt to test and better understand the link between victimization and the three elements that comprise routine activities theory. Researchers have applied this theory to different types of victimization from minor crimes (Mustaine & Tewksbury, 1998) to serious ones like sexual as sault (Cass, 2007; Jackson, Gilliland, & Veneziano, 2006; Tewksbury & Mustaine, 2001). Researchers have also used a variety of measures of the target suitability and guardianship concepts. Perhaps because of the various measures used in these studies, the results of these studies have been mixed with some studies finding support for routine activities, while others do not. This research is reviewed below. Note, however, existing studies have not really attempted to understand whether and how birth status an d culture are related to the elements of routine activities theory. Yet, it is entirely possible that birth status and culture have important influences on the elements of routine activities theory. Target Suitability Target suitability plays a major role Cohen and Felson argue that in order for a direct contact predatory victimization to occur a motivated offender must meet in time and space with a suitable target. Suitable targets
17 are vulnerable targets in t hat they lack protection from others (i.e., capable guardians). This suggests that the suitability of a target is affected by where one hangs out as some places offer more guardianship, with whom one hangs out as some companions may be motivated offenders (or capable guardians), and when one hangs out as the guardianship available at one location may vary with time of day. In other words, the suitability of a target is affected by the choices one makes. Existing research has used various measures of targe t suitability, despite these different measures; most studies have supported the importance of target suitability in explaining victimization. Target suitability has been measured by: time kids spend unsupervised (Osgood et al. 1996); frequency of going to sporting events and bars (Kennedy & Forde, 1990); eating out frequently (Mustaine & Tewksbury, 1998); time spent watching television (Plass & Carmody, 2005); among others. These measures vary substantially but they all were found to be predictive of victi mization. For example, in a recent study conducted by Plass and Carmody (2005), they measured target suitability using time spent watching television rather than going out. They found that frequent television watching decreased the odds of victimization f or all youth. Another important finding from this body of research is that engaging in delinquent acts consistently increases the likelihood of victimization. For example, Plass and Carmody (2005) found that high school students who engaged in drug use were more susceptible to victimization. Similarly, research conducted by Cass (2007) and Tewksbury and Mustaine (2001) discovered that drug use was also related to a greater risk for sexual as sault. Further, Mustaine and Tewksbury (1998) found that college students who committed illegal acts (e.g., smoking marijuana, threatening other people)
18 had higher victimization rates in comparison to college students who did not participate in illicit beh avior. These findings, taken together, indicate that participation in different activities place individuals in vulnerable positions and contribute to their victimization risk. Vulnerability comes about when the target is no longer being guarded or protec ted, which propels motivated offenders to attack because the likelihood of success has increased. E ngaging in illicit behavior also increases vulnerability for victimization. These findings suggest that birth status and culture may affect target suitability in two ways: 1) cultural differences like familialism may reduce target suitability by limiting unsupervised time with other kids, frequency of going out, and so forth; and, 2) cultural differences lead to reduced participation in illicit behaviors, which are strongly related to victimization risk. Despite the clear possibility of culture affecting target suitability no research (to my knowledge) has tested these effects. Gua rdianship The element of greatest interest to this thesis is that of guardianship. The idea of a capable guardian arch is that Cohen and Felson never truly developed ways to measure capable guardianship; instead, the theory proposes that two types of capable guardians are in existence (i.e., formal and informal), with formal guardians considered police officers and in formal guardians being ordinary citizens, family, or friends (Cass, 2007). The routine activity theory posits that a crime will not be committed if a capable guardian is present to either protect the individual or
19 their property. However, since no real mea sure of guardianship exists, researchers simply have created their own measures. For example, some researchers proposed that teenage drug use would decrease if their routine activities were supervised by adults (Bratt, 2008). In order to test this predicti on, social workers and volunteers were placed in areas where local thirteen and fifteen year olds were known to gather to dissuade drug use. Bratt found that drug use decreased but the observed decrease in drug use was not attributed to the guardians, bec ause many of the teenagers who were not present at the gatherings also reported a decrease. Other measures of capable guardianship that have been tested include institutional guardianship (i.e., university; Cass, 2007), which was measure d as the availabi lity of a self defense class, the availability of off campus escort services during the day and in the evening (p. 357). Most research on guardianship, however has revolved around individual guardianship Examples of individual levels of guardianship in clude marital status, employment status, and daily activities (i.e., going to the movies, attending parties; Belknap, 1987; Cass, 2007). Tewksbury and Mustaine (2003) used a different approach by measuring guardianship using possession of a gun or mace. Finally, and most important to the current research is the study of various aspects of family life, from which measures of guardianship have been derived, such as parental monitoring. Interestingly, parental monitoring has been found to be effective in no t only preventing behavior problems, but also in reducing poor outcomes among those youth with behavioral problems (Robertson, Baird Thomas, & Stein, 2008). Another family
20 measure of guardianship comes from Schreck and Fisher (2004), who used the emotional relationship between family members as a measure of guardianship. These authors found that family attachment promotes effective guardianship by making children less attractive targets and reducing contact with motivated offenders. This finding is supporte d by other research that concludes measures of family function and parental monitoring are important predictors of youth victimization (see e.g., Benedict & Zautra, 1993; Reid & Sullivan, 2009; Turner, Finkelhor, & Ormrod, 2007). Again, guardianship, part icularly measures involving family life may be influenced by immigrant culture and assimilation but the existing research largely has not examined this possibility. In particular, the close family relations of Hispanic immigrants may provide them with grea ter levels of guardianship than more assimilated Hispanics. Summary Overall, previous measures of assimilation have consisted mainly of generational differences (e.g., first and second) and the age of entry into the United States (Morenoff & Astor, 2006, p. 43). Generational differences are an indication of whether the youth was born inside or outside of the United States, which is believed to be associated with the amount of time that they have to assimilate to American culture. Essentially, the belief is that the more time that is spent in the United States the more likely the individual is to participate in crime and other types of risky behavior (Morenoff & Astor, 2006, p. 38). Though the results indicated significant differences in crime (as perpetrato rs) based on the above characteristics, this study seeks to determine how elements of assimilation (i.e., language spoken in the home, whether the parents were born in the United States,
21 whether the youth was born in the United States) affect the likelihoo d of being victimized. Specifically, the purpose behind the measures of assimilation is to determine child, which may increase or decrease the likelihood of being a v ictim of a violent crime. In the past, routine activities theory has been applied to a number of criminological aspects through the use of surveys (Jackson, Gilliland, & Veneziano, 2006; Tewksbury & Mustaine, 2001; Tewksbury & Mustaine, 2003). In the pre sent stu dy, though routine activities are analyzed, emotional guardianship serves as the main focus. The weakest element of routine activities is that of capable guardianship because there is no real definition of capability or of who can be considered a guardian; however, for this study, the parents of the youth will serve as guardians. Their capability as guardians will be based on the emotional connection (i.e., discussing problems, youth receiving encouragement from parents, the youth feeling that his/ her parents care) that has been established (or the lack thereof) between parents and children. Low levels of affection/communication will indicate poor guardianship (i.e., incapability), which is expected to correlate with higher levels of victimization. The rationale for this measure of guardianship is that well established emotional connections between the youth and their whereabouts when out of the house. It is this aspect of their relationship that is expected to decrease the likelihood of victimization because the parents will be better able to dangers that may be lurking in such areas. The se well established relationships are expected to be found more so between less assimilated youth (i.e., foreign born youth)
22 and their parents because they will be more dependent on their family for support since they have not acclimated to American cultur e. The activities that the youth take part in will also be considered in regard to the likelihood of victimization, for example, participation in school functions (e.g., clubs and sports), youth groups, and parties. These activities may influence the indiv
23 Chapter III : Methodology The purpose of this chapter is to present the hypotheses that have been derived fro m the integration between assimilation and routine activities and describe the methodology used to test these hypotheses. The measures that represent each of the main constructs (i.e., assimilation, emotional guardianship, and routine activities) are defin ed and outlined. Last the analyses are described and descriptive statistics are provided for each of the measures. Hypotheses (1) Foreign born Hispanic youth have lower levels of violent victimization than native born Hispanic youth, even after controlling for demographic factors. (2) Participation in various activities, such as religious gatherings and family dinners, serve to protect youth after controlling for competing factors. (3) Emotional guardianship is related to the likelihood of be ing victimized after controlling for competing factors. (4) Foreign comparison to native born Hispanic youth is explained by differences in various routine activities and emotional guardiansh ip.
24 Data This study tests the above hypotheses by using data provided by the National Longitudinal Study of Adolescent Health (Add Health) project. This study provides data regarding adolescents from across the United States ranging from grades 7 12 (Wa ve I), which was collected during the 1994 1995 school year. The purpose for the Add Health study was to gather data related to fields of study connected to social and behavioral sciences. As such, participants provided data regarding their social, economi c, psychological and physical well being as well as other aspects of their lives, such as their family, neighborhood, school, friendships, and peer groups. The participants were derived from 80 high schools and 52 middle schools from all over the United St ates. Systematic sampling and stratification methods were implemented to ensure a representative sample of US schools. The variables included in the Add Health dataset suit the purposes of this study in regard to providing data relating to aspects of life pertaining to assimilation, the routine activities of Hispanic youth and the emotional relationship with parents. With this knowledge, comparisons will be made to determine whether emotional guardianship and certain activities are related to victimization, and, if so, how they are related (i.e., serves as a protective factor). Despite the dataset being longitudinal, only the data from Wave I will be utilized for this thesis. Thus, this study focuses on the cross sectional relationship between assimilation, emotional guardianship and routine activities and their effects on violent victimization.
25 Sub Sample The sample used in thesis consists of only Hispanic youth. There are 595 Hispanic youth in this dataset with valid data on all variables of interest, which represents 9.14% of the entire Add Health sample. The Hispanic groups represented in this sample include Mexicans, Chicanos, Cubans, Puerto Ricans, and Central /South Americans. Table 1 provides basic demographic information about this sample. Approximately 70% (69.8) of these youth were born in the United States, and the other 30% were not. A little less than half of the sample is male (46.6%) and the larger por tion of the sample is female (53.4%). The average age of the youth is 16.06 years. Among this sample, 3.2% had no mother and 32.4% of the youth had no father. Table 1. Demographic Characteristics of the Hispanic Sample from Add Health (N=595) Variable N % Mean SD Range Born in the U.S. No 180 30.3 Yes 415 69.7 Gender Male 277 46.6 Female 318 53.4 Age 595 16.06 1.77 12 20 Mother No 19 3.2 Yes 576 96.8 Father No 193 32.4 Yes 402 67.6
26 Measures Target Suitability Routine a ctivities are those activities of Hispanic youth daily lives, which increase or decrease their suitability as target s. Table 2 includes those variables that represent various routine activities. To test the above hypotheses I utilize seven measures total number of clubs, organizations, and teams each youth reported participating in. The Add Health data lists fourteen activities. Theref ore, the highest possible value for this variable is fourteen and the lowest is zero; the mean number of activities is 1.54. The also summed composite variables. Specific ally, youth were asked a series of questions regarding if they attend religious services, church related event s, go to movie s play s museum s concert s or sports event s with their dad. Then the same series of questions were asked about activities with their mom. Responses to these questions are summed to create the activities variables. The average number of activities with each parent is .7647 (Dad) and 1.49 (Mom). Th e fourth measure is frequency of eating dinner with at least one parent present in the room each week. Participants reported having dinner with at least one parent present in the room an average of 4 times over the course of a week. The fifth measure conce rns household chores, participants were asked how often they worked around the house during the course of one week, their choices were: not at all; once or twice; three or four times; or 5 or more times. The mean was 1.91 indicating that the youth worked b etween one and four times a week. The sixth and seventh measures concern frequency of attending youth religious groups and religious services during the
27 once a month/le ss than once a week; (3) once a week or more. The mean for both activities signified that the number of time the youth attended these events fell between once a mo nth and less than once a week. Table 2. Routine Activities Variables from Add Health (N=595) Variable Mean SD Range Sum of Participation 1.15 1.87 0 14 Sum of activities with Dad .7647 1.07 0 5 Sum of Activities with Mom 1.49 1.09 0 5 Frequency of Dinner with Parents 4.55 2.61 0 7 Chores 1.91 .939 (0) Not at all 5 or more times(3) Past Year Attend Youth Groups .9361 1.24 (0)Never Once a week or more(3) Past Year Attend Religious Services 1.75 1.20 (0)Never Once a week or more(3) Emotional Guardianship Emotional g uardianship is an aspect of relationships (i.e., affection and communication) betwe en the youth and their parents that serve to protect the youth from being victimized. Table 3 provides details regarding the six measures of emotional guardianship used in this study. The first two measures concern communication with
28 series of questions co ncerning whether the youth talked to each parent about: personal problem ; someone they ar e dating or a party they went to; school work or grades; and other things they are doing at school. On average, the youth discussed about two of the above issues with their mom and one issue with their dad. The third and fourth measures concern how close the youth feel to each parent with possible scores ranging from 0 (no father/mother) to 5 (very close). On average, respondents reported feeling close to their mom than their dad. The last two measures concern closeness of relationship with dad/mom. Participants were asked how much they agreed or disagreed with the statement (strongly ag ree). The mean scores on these two variables were 2.69 for dad and 4.12 for mom. Table 3. Emotional Guardianship Variables from Add Health (N=595) Variable Mean SD Range Sum of Communication with Dad .8538 1.12 0 4 Sum of Communication with Mom 1.83 1.32 0 4 Close to Dad 2.84 2.14 (0) No father Very much(5) Close to Mom 4.41 1.16 (0) No mother Very much(5) Dad Good Relationship 2.69 2.05 (0) No father Strongly agree (5) Mom Good Relationship 4.12 1.20 (0) No mother Strongly agree (5)
29 To assess the reliability of the summed measures of routine activities and Table 4 includes the reliabilities for each of these measures. The results of this analysis were interpreted using a relatively lenient standard of reliability of .60. The two summed routine activities produced low reliabilities. The two emotional guardianship measures, however, exhibited adequate reliabilities. Despite the relatively low reliabi lities of the summed routine activities measures, I used them for parsimony. It is important to note that as a result of the low reliabilities, the absolute value of the regression coefficients are affected in that they are smaller than they would be if th e measures had greater reliability making it more difficult to obtain statistically significant regression coefficients. An exploratory factor analysis was also conducted for these measures; however, it was difficult to conduct because the variables were m easured differently. For example, some of the variables were measured on a Likert scale, whereas others were dichotomous (i.e., yes or no). Table 4. Reliabilities for Summed RAT and Emotional Guardianship Variables Variable Alpha No. of Items Total number of activities with Dad .559 5 Sum of activities with Mom .462 5 Total communication with Dad .654 4 Sum of communication with Mom .612 4
30 Victimization The dependent variable, v ictimization consists of delinquent acts suffered by the Hispanic youth utilized in this study The measure of violent victimization provided by the Add Health data asks youth if s/he had ever: had s omeone pull a knife or gun on them; been shot; been cut or stabbed; or been jumped. Within this sub sample, 24.4% of the youth indicated that they had suffered at least one form of violent victimization. Table 5. Violent Victimization Variable from Add Health (N=595) Variable N % Violent Victimization Yes 145 24.4 No 450 75.6 I also conducted correlations for all of the variables mentioned above. Table 6 indicates how each of these measures correlates with one another. However, due to the large number of measures for demographics, routine activities, emotional guardianship, and violent victimization, I used two separate tables to report the correlations. Overall, table 6 reveals that none of the variables are highly correlated with violent victimization. The highest correlations existed between close with dad and not having a father ( r = .922, p <.01 ) and having a good relationship with dad and not having a father (r = .912, p <.01); this is clearly evidence of multicollinearity Table 7 indicated that the strongest correlations appeared between the sum of activities with dad and close to dad ( r = 560, p <.01 ) good relationship with dad ( r = .568, p <.01 ) and the sum of communication with dad ( r = .555, p <.01 ) Lastly, table 7 also provided evidence of multicollinearity between close to dad and good relationship with dad (r = .949, p <.01). Ther e was also a strong relationship between good relationship with mom and close to mom (r = .738, p <.01) and sum of communication
31 with dad and close to dad (r = .554, p <.01) and good relationship with dad (r = .559, p <.01).
32 Table 6. Correlations of Demographics, Routine Activities, Emotional Guardianship, and Violent Victimization (N=595) 1 2 3 4 5 6 7 8 9 1 2 .084 + 3 .169* .016 4 .044 .175* .054 5 .030 .026 .060 .054 6 .075 .066 .064 .038 .126* 7 .115 + .060 .188* .077 .043 .039 8 .120 + .048 .009 .249* .068 .098 + .158* 9 .114* .090 + .115* .075 .030 .052 .061 .053 10 .062 .031 .058 .096 + .032 .054 .099 + .078 .056 11 .106* .053 .121* .095 + .034 .115* .127* .155* .077 12 .095 + .052 .098 + .152* .248* .060 .128* .232* .165* 13 .054 .012 .139 + .128* .102 .495* .030 .225* .111* 14 .063 .047 .060 .015 .690* .095 + .044 .192* .018 15 .116* .045 .105 + .067 .130* .922* .029 .161* .048 16 .095 + .040 .094 + .062 .097 + .912* .029 .145* .055 17 .052 .028 .074 .056 .624* .034 .046 .250* .015 18 .062 .027 .126* .003 .252* .061 .081 + .072 .114* 19 .076 .066 .071 .037 .101 + .528* .034 .137* .077 Note: 1= Violent Victimization; 2=US Born; 3=Male; 4=Age; 5=No Mother; 6=No Father; 7=Chores; 8=Dinner with Parents; 9=Sum of Participation; 10=Youth Group; 11=Religious Group; 12= Sum of Activities Mom; 13= Sum of Activities Dad; 14= Close to Mom; 15=Close to D ad; 16=Good Relationship with Dad; 17=Good Relationship with Mom; 18= Sum of Communication Mom; 19=Sum of Communication Dad. +Correlation is significant at the 0.05 level (2 tailed). Correlation is significant at the 0.01 level (2 tailed).
33 Table 7 Correlations of Demographics, Routine Activities, Emotional Guardianship, and Violent Victimization (N=595) Cont. 10 11 12 13 14 15 16 17 18 1 2 3 4 5 6 7 8 9 10 11 .491* 12 .134* .214* 13 .156* .203* .305* 14 .016 .057 .285* .001 15 .064 .132* .030 .560* .024 16 .067 .130* .007 .568* .004 .949* 17 .002 .072 .312* .088 + .738* .045 .083 + 18 .019 .004 .309* .070 .282* .071 .065 .318* 19 .084 + .117* .104 + .555* .027 .554* .559* .053 .270* Note: 1= Violent Victimization; 2=US Born; 3=Male; 4=Age; 5=No Mother; 6=No Father; 7=Chores; 8=Dinner with Parents; 9=Sum of Participation; 10=Youth Group; 11=Religious Group; 12= Sum of Activities Mom; 13= Sum of Activities Dad; 14= Close to Mom; 15=Close to Dad; 16=Good Relationship with Dad; 17=Good Relationship with Mom; 18= Sum of Communication Mom; 19=Sum of Communication Dad. +Correlation is significant at the 0.05 level (2 tailed). Correlation is significant at the 0.01 level (2 tailed).
34 Data Analytic Plan In order to gain a better understanding of the relationship between assimilation, emotional guardianship and violent victimization, a logistic regression was used to test each of the hypotheses. Since the dependent variable is dichotomous, a logistic regr ession must be implemented rather than an ordinary least squares regression (OLS), in order to ensure that the assumptions for the error term are not violated and that the statistical significance of the estimates is not affected. Violent victimization se rves as the dependent variable and the measures for emotional guardianship; routine activities; assimilation, and demographic variables serve as the independent variables. In particular, three nested logistic regression models are estimated. To test hypot hesis number 1 (i.e., foreign born Hispanics have lower victimization than native born Hispanics) victimization is regressed on birth status and demographic variables. This model reveals whether birth status is a predictor of victimization, above and beyon d demographic variables known to predict victimization. The second model adds the seven routine activities to the variables in model 1 to test the predictive power of the routine activity variables. If hypothesis number 2 is correct, then model 2 should pr edict victimization more accurately than model 1. In other words, model 2 should fit the data better than model 1. The third and final model tests hypotheses number 3 (emotional guardianship predicts victimization) and 4 (differences in guardianship and ro utine activities explain the difference in victimization between foreign and native born Hispanic), by adding the measures of emotional guardianship to the variables included in model 2. If hypothesis number 3 is correct, then model 3 should fit the data better than model 2. If hypothesis number 4 is correct, then after taking into account emotional
35 guardianship and routine activities, the effect of birth status (i.e., foreign born or native born) should disappear. In summary, this methodology assists in determining whether generational differences exist in violent victimization and whether these differences are explained by measures, and analytic strategy. The results detailed in the next chapter tests the validity of my hypotheses.
36 Chapter IV: Re sults This chapter provides a discussion of the models, results, and findings from the analyses that were computed in order to test the validity of the proposed hypotheses pertaining to this thesis. The main hypotheses center on foreign born Hispanic youth having lower victimization rates than native born Hispanic youth. The lower victimization rates are explained by foreign born Hispanic youth participating in different routine activities and having greater emotional guardianship than native born Hisp anic youth. In order to determine the validity of these statements, a series of logistic regressions were conducted. Recall hypothesis 1 predicts foreign born Hispanic youth have lower levels of violent victimization than native born Hispanic youth. To t est this prediction, violent victimization was regressed on birth status (i.e., foreign born vs. native born), while controlling for important demographic variables. Age and gender were included because prior research suggests these variables are related t o victimization. The measures for no mother and no father were included to control for between group differences on these variables and to assist with the missing data that emerged in the later analyses due to the absence of one or both parents. The resu lts of this first model indicated that violent victimization was collectively predicted by birth status and demographics 2 (5 ) =27.744; p < 001). Table 8 points out
37 that native born Hispanic youth were significantly more likely to be violently victimized i n comparison to their foreign born counterparts, after controlling for other demographic factors. The regression coefficient indicates that native born Hispanics have 65% greater odds of violent victimization than foreign born Hispanics (Odds Ratio= 1.653) Males were also significantly more likely to be victims of violent crime when compared to females. Specifically, this model suggests the odds of violent victimization is 117% greater for males than females (OR= 2.271). Those youth without fathers were al so more likely to be violently victimized in comparison to those youth who had fathers. The results of this analysis indicate that the hypothesis proposing that foreign born Hispanic youth have lower levels of violent victimization than native born Hispani c youth was supported. In other words, there is a difference in violent victimization that exists between the two groups that warrants an explanation. Hypothesis 2 predicts that participation in various activities, such as religious gatherings and family dinners, serve to protect youth after controlling for competing factors. To determine the validity of this statement, a second logistic regression was conducted. The first and second models are nested, meaning that the variables in the first model were i ncluded in the second model to determine whether routine activities predicts violent victimization above and beyond birth status and demographics. In essence, the first model served as the baseline model or the constrained model and the two models that fol low serve as the unconstrained models in which the variables measuring routine activities and emotional guardianship were included. Model 2 regressed violent victimization on demographic and routine activity variables. The activities ranged from sports, cl ubs, and organizations in school to a variety of activities in which the youth
38 participated with their parents (e.g., seeing a movie, attending a religious service, and working on a school project). Since model 1 is nested in model 2, the collective expl anatory power of the routine activity variables was tested by the way of the likelihood ratio test, which was used to compare the fit of two models where one model is nested within the other. In SPSS, the model fit statistics include the 2 log likelihood ( 2LL) for which smaller numbers are an indication of the model serving as a better fit for the data. For the baseline model, the 2LL equaled 633.069 and once the routine activity variables were included in the second model, the 2LL decreased to 616.711, an indication of a better fit. The likelihood ratio test indicates that model 2 fits the data significantly better than model 1 2 (7) = 16.358; p=. 022). Table 8 reveals that of the various activ ities included in model 2, only the sum of participation wa s individually significant. Specifically, the model results indicate that the odds of violent victimization decreased by 13% for every additional club, sport, or organization that the youth participated in (OR=.870). Though the other activities were not in dividually significant, they were all negatively associated with violent victimization, with the exception of the sum of activities with dad. The lack of additional significant routine activity measures came as a surprise, especially since none of the meas ures were highly correlated with one another (i.e., none of the variance inflation factors were above 4). The addition of the routine activities predicted victimization, but these measures do not predict victimization very strongly.
39 Interestingly, in mod el 2, birth status was still a statistically significant predictor of violent victimization According to hypothesis 4, differences in routine activities would help account for the differences in victimization between groups. Had this been the case, the re gression coefficient for birth status would have decreased in comparison to the baseline model; however, rather than decrease, the coefficient actually increased. Essentially, this means that partici pation in routine activities did not aid in explaining th e difference in violent victimization rates between foreign born and native born Hispanic youth as expected. Therefore, there is no preliminary support for hypothesis 4. Table 8 Logistic Regression Predicting Violent Victimization for U.S. Born Youth, Routine Activities, and Emotional Guardianship Model 1 Model 2 Model 3 Predictor b(SE) b(SE) b(SE) US Born .502* .527* .569* (.229) (.235) (.239) Male .820** .699** .781** (.199) (.209) (.218) Age .068 .010 .006 (.056) (.060) (.061) No Mother .457 .093 .049 (.523) (.565) (.841) No Father .418* .335 1.149 (.207) (.251) (.615) Chores .164 .169 (.109) (.110) Dinner with Parents .069 .050 (.041) (.042) Sum of Participation .139* .143* (.067) (.069) Youth Group .022 .030 (.095) (.096) Religious Services .072 .066 (.096) (.098) Sum of Activities Mom .090 .101 (.111) (.116) Sum of Activities Dad .015 .120 (126) (.142)
40 Model 1 Model 2 Model 3 Predictor b(SE) b(SE) b(SE) Close to Mom .009 (.151) Close to Dad .512** (.186) Good Relation with Dad .130 (.178) Good Relation with Mom .012 (.135) Sum of Communication Mom .029 (.094) Sum of Communication Dad .019 (.131) Cox and Snell R 2 .046 .071 .091 2LL 633.069 616.711 604.289 N 595 595 595 *p <.05. **p <.01. Lastly, model 3 tests hypotheses 3 and four. Hypothesis 3 predicts emotional guardianship is related to the likelihood of being victimized after controlling for competing factors. T he third model examined the relationship between emotional guardianship and likelihood of violent victimization according to various measures of emotional guardianship while also including the previous demographic and routine activity variables The 2LL for the l ast model decreased to 604.289, indicating a somewhat better fit to the data with the inclusion of the emotional guardianship variables. Yet, the additional variables fell slightly short of significance as determined by the likelihood ratio test 2 (6 ) =12. 421 ; p=. 053), which by statistical standards means that the discussion of these results should end here. However, since the concept of emotional guardianship is being introduced here and because the measures of emotional
41 guardianship had low reliabilities, which attenuate their predictive power (see chapter 3), the results of this model are discussed largely for exploring purposes. The results in table 8 reveal that of the six measures of emotional guardianship only one, feeling close to dad, was individua lly and significantly related to victimization. Specifically, feeling close to dad was associated with a 40% decrease in the odds of violent victimization (OR=.599). The rest of the measures were all negatively related to violent victimization with the exc eption of having a good relationship with dad, but not statistically significant. The reason for the lack of significance among the other measures of emotional guardianship is partially affected by multicollinearity as several of the variables in model 3 e xhibited high variance inflation factors. In particular, being close to father, having a good relationship with father, and not having a father, all had large variance inflation factors, an indication of multicollinearity. Consequently, additional analyses were conducted that consisted of removing some of these highly correlated variables from the models to see if the results changed substantively. These alternative models (not shown in Table 8) produced substantively similar results to those shown in Table 8. Emotional guardianship was introduced as the main focus of this thesis in that this concept was expected to be significantly related to the outcome. Surprisingly, the majority of these measures were not significant. It seems that the variables measuri ng communication and parent child attachment did not serve to protect the youth from being victimized. Overall, based on the results in the last model, the addition of the emotional guardianship measures did not significantly predict violent victimization as a whole (i.e., the model was not significant) nor were the individual measures related to the outcome.
42 Birth status wa s significantly related to an increased likelihood of violent victimization, even after controlling for routine activities and emot ion al guardianship. Being male wa s also significantl y associated with victimization. As for the routine activity measures, the sum of participation remained significant. However, the regression coefficient for birth status increased with the inclusion of emot ional guardianship, which indicates that neither the routine activities nor emotional guardianship are accounting for the difference in victimization between foreign born and native born youth. Ultimately, hypothesis 4 was not supported. Overall, each of these models allowed for the testing of the hypotheses proposed in this thesis. First, it was determined that victimization rates differed between foreign born and native born youth. It is clear that a difference in victimization rates exists. However, the explanations provided in this thesis did not account for those differences as expected. First, the routine activity variables were statistically significant but weak predictors of violent victimization. In regard to emotional guardianship, though being cl ose to dad was significant, the model itself fell short of statistical significance, so it appears that the differences in the likelihood of violent victimization between foreign born and native born youth could not be attributed to emotional guardianship as predicted. Aside from the lack of a significant model, the regression coefficient for birth status increased in comparison to the previous models, despite controlling for routine activities and emotional guardianship. Therefore, neither routine activiti es nor emotional guardianship are accounting for the difference in the likelihood of victimization. In conclusion native born youth were significantly more likely to be violently victimized in comparison to foreign born youth, holding all other variable s constant in all
43 of the models This finding is consistent with that of other researchers who have studied generational differences among Hispanics youth regarding participation in crime (Morenoff & Astor, 2006; Rumbaut et al., 2006) However, the explana tion that routine activities and emotional guardianship account for the differences in the likelihood of violent victimization which was proposed in this thesis was not supported. Table 9 provides a summary of the support that was shown or not shown for ea ch of the hypotheses proposed in this thesis. Table 9 Summary of Support for Hypotheses Support Shown Hypothesis Yes No 1 2 3 4
44 Chapter V: Discussion Overall, the purpose of this thesis was to determine whether differences in violent victimization existed between foreign born and native born Hispanic youth and whether these differences could be attributed to differences in routine activities and emotion al guardianship. Specifically, participation in various activities was expected to place youth out of the reach of motivated offenders thus lowering their suitability as targets; especially, if the youth were to be engaging in activities with their parents (i.e., hypothesis 2). The relationship of most interest, however, was that of emotional guardianship and violent victimization. Emotional guardianship is defined as the aspect of a parent child relationship that fosters communication, affection, and trust These elements were expected to better equip guardians because vital information could be transferred from the youth to the parent, such as the whereabouts of the youth and any problems that the youth may be facing. Another element pertaining to emotio nal guardianship was birth status. Since the sample utilized in the analyses consisted of Hispanic youth who were born inside and outside of the United States, assimilation was believed to affect the existence and strength of the parent child bond. In part icular, assimilation was expected to affect a cultural value known as familialism, which is described as strong identification with and attachment to the family Essentially, familialism served as the link between assimilation and routine activities theory. For this reason, foreign born youth were hypothesized to be victimized
45 less in comparison to native born youth (i.e., hypothesis 1). Considering the emphasis that is placed on family, foreign born youth were expected to manifest this cultural value through their relationships with their parents more so than native born youth. The presentation of these hypotheses led to analyses that were conducted to determine the validity of these statements. Based on the results of the logistic regressions suppor t was provided for some of the hypotheses, but not for all. Overall, the first hypothesis was clearly supported because in each of the models native born Hispanic youth were significantly more likely to be violently victimized in comparison to foreign born Hispanic youth. Though this thesis focused on the opposite side of the spectrum by examining violent victimization, the finding regarding generational differences was consistent with that of other researchers. In their study, Morenoff and Astor (2006) sug gest that differences in crime between immigrant youth and native youth may be the result of selection bias in that those who choose to migrate may be less prone to participate in criminal activities than the general population from their country of origin (p. 53). On that same note, it may be possible that immigrant families are less likely to be victimized because they make it a point to remain united and protect one another. In other words, those families that choose to make the United States their home may make more of a conscience effort to guard one another in comparison to those from their country of origin or those who have always lived in the United States. Throughout all three models, the need for an explanation of this difference in victimizatio n was established. One of the theoretical explanations introduced in this thesis consisted of participation in routine activities serving as a protective shield to the
46 youth. Specifically, differences in participation were expected to account for differenc es in the likelihood of violent victimization between foreign born and native born Hispanic youth. However, as a whole, the model containing the demographic and routine activity variables demonstrated that the difference in the likelihood of violent victim ization between the two groups increased in comparison to the baseline model. This increase indicated that the routine activities did not serve as an explanation for the dissimilarity in the likelihood of victimization between youth. For had it been the re ason, the regression coefficient would have decreased in magnitude. As for the specific measures of routine activities, only the sum of participation in clubs, sports and organizations was significant. It seems that these extracurricular activities serve as a protective shield for those who participate in comparison to those who do not. The other activities in the model had negative relationships with the outcome with the exception of the total number of activities with dad. In general, the activities cor responded with the theoretical argument and with the conclusions of past studies (Mustaine & Tewksbury, 2002; Plass & Carmondy, 2005) in regard to certain activities being negatively associated with victimization. Despite the negative relations with the ou tcome, each of these variables was surprisingly insignificant because the activities that were chosen as measures were believed to be age appropriate for the sample utilized in the analyses. For example, chores, youth groups and specific activities with pa rents were all expected to predict violent victimization; however this was not the case. Therefore, no support was provided for hypothesis 2. The other theoretical explanation provided involved emotional guardianship. Specifically, the parent child relati onship was expected to differ between foreign born
47 and native born youth and thus mark the difference between the two groups of youth in relation to their likelihood of victimization. However, the third model also failed to provide an explanation for the d ifference between foreign born and native born youth as a whole. The model fit statistics demonstrated that the inclusion of the emotional guardianship did not significantly predict violent victimization. Once again, the regression coefficient for birth status increased in magnitude with the inclusion of the additional variables in the last model. This served as an indication that emotional guardianship did not provide an explanation for the difference in victimization rates between groups. However, despi te the different variables that were included in the model, being close to dad was the only one significantly related to the outcome. Basically, the measures for emotional guardianship did not provide the protection that was predicted. Therefore, despite t he significance of being close to dad, the overall importance of emotional guardianship in explaining violent victimization was not supported by the results in the model. Limitations In spite of the theoretical foundations on which these hypotheses were based, the results demonstrated that there was not much support for many of the hypotheses. This lack of support of the hypotheses may be due to the limitations of this research. First, the data utilized for this study was secondary data, which restricted the choice of demographic and emotional guardianship measures available to be included in the models. Perhaps if better, more reliable, measures were available, more support for the hypotheses would have been found. In other words, the low internal consist ency of the
48 variables reduced absolute values of the regression coefficients and as a consequence reduced the predictive power of these measures. Secondly the data included in the analyses were limited to Wave I, which leads to question of proper temporal order. In addition, the dependent variable was limited to violent victimization, which may have been related to the lack of significant relationships among measures of emotional guardianship and routine activities. It is possible that a more concrete or physical type of guardianship is necessary to ward off offenders who are looking to violently attack their targets. In other words, emotional guardianship may not be a valid measure for capable guardianship when measuring violent victimization. Though the emotional stability/relationship between parent and child is important, the emotional aspect of this relationship may not act as an efficient form of guardianship. Implications /Future Research In conclusion, in spite of these drawbacks, the findings pre sented here indicate that there is something to be learned from studying immigration and crime, especially between foreign born and native born Hispanics. Though generational differences have been studied among criminals, this thesis demonstrates that this finding is applicable to victimization as well. However, it is not clear as to why this difference exists. Also, as evinced by the magnitude of the birth status, routine activities and emotional guardianship did not provide an explanation for the differe nce in violent victimization between groups. Nonetheless, some of the individual variables were significant, such as participation in various clubs, sports and organizations, which showed that target suitability seems to decrease among the Hispanic youth i n this sample. In regard to
49 emotional guardianship, being close with dad was the only significant variable, which coincides with the argument presented in this thesis. It seems that the element of closeness between father and child serves as a marker for d etermining what characteristics are associated with being a capable guardian. Overall, future studies need to focus on identifying what measures best capture the true nature of guardianship. Guardians are those who are able to keep individuals from being victimized. However, this protection may be the result of prevention or intervention. Since Cohen and Felson (1979) did not specify what characteristics could be used to identify capability in guardians, it is necessary to examine those individuals with w hom youth tend to establish relationships. Though it is possible for some relationships whom youth interact and are shielded from harm. The latter relationships are the ones that need to be examined to detect those aspects of these relationships that are most related to minimizing the risk of victimization. Future studies may also want to further examine the emotional relationship between parents and children to determi ne whether this relationship serves to shape or define what a capable guardian truly is. Though emotional guardianship was the focus of this thesis, future studies should focus on determining whether different forms of guardianship exist and whether they provide similar forms of protection. Though communication and affection within a parent child relationship did not significantly predict violent victimization in the present analyses, future analyses may want to study the effect of different forms of guard ianship on various outcomes. It may also help to determine what type of guardianship (i.e., emotional or physical) best suits different forms of victimization (e.g., violent
50 victimization versus non violent victimization). Also, as mentioned above, examini ng the effect of physical and emotional guardianship among different individuals who may potentially serve as capable guardians (e.g., coaches, teachers, and youth leaders) may produce a different set of findings. For some youth may feel more comfortable s haring certain pieces of information with people other than their parents. In general, using the relationship between father and child as a measure of guardianship coincides with one of the types of guardianship (i.e., informal guardianship) that Cohen a nd Felson (1979) discussed in their introduction of routine activities theory. Aside from providing protection, fathers also offer an element of familiarity with the target (i.e., youth), which differs from formal guardianship (e.g., police officers). In e ssence, it is assumed that fathers have already established a relationship with their children (i.e., suitable targets), which differs from t he protection provided by formal guardians. Formal guardians usually provide protection from behind a wall of propr iety and professionalism and though certain guardians, such as police officers, may be better equipped to cope and react in certain situations, the aid that is offered would likely be in the form of intervention rather than prevention; unless the problem w as large enough to have been brought to the attention of the police (i.e., gang activity or drug activity). In other words, though police officers may be able to better protect people because of their training, fathers may be better able to guard their chi ldren because of the knowledge they possess as a parent. Therefore, it would be interesting to examine whether familiarity or training serve to better equip people as capable guardians. In conclusion, there is much to be learned from studying the element of guardianship in relation to routine activity theory. Though past studies have indicated that
51 certain activities and lifestyles are associated with victimization, it is still somewhat unclear what capabilities are required for an individual to be classif ied as a guardian. However, aside from determining who is a capable guardian, it is necessary to determine in what form that guardianship is provided or what skills, abilities, or knowledge the individual possesses that allows the individual to provide pro tection to those who need it. This latter aspect may lead to information that can be disseminated to others who wish to also successfully defend their loved ones, such as parents, teachers, coaches, and other family members.
52 References Allen, M. L., Elliott, M. N., Fuligini, A. J., Morales, L. S., Hambarsoomian, K., & Schuster, M. A. (2008). The relationship between Spanish language use and substance abuse behaviors among Latino youth: A social network approach. Journal of Adolescent Health, 43, 372 379. Andresen, M.A. (2006). A spatial analysis of crime in Vancouver, British Columbia: A synthesis of social disorganization and routine activity theory. The Canadian Geographer, 50(4), 487 502. Barrett, M.E, Joe, G.W., & Simpson, D.D. (1991). Acculturation influences on inhalant use. Hispanic Journal of Behavioral Sciences, 13, 276 296. Becerra, R. M. (1988). The Mexican American family. In C. H. Mindel, R. W. Habenstein, R.Wright, Jr. (Ed.), Ethnic families in America (141 159). New York: Elsevier. Belknap, J. (1987). Routine activity theory and the risk of rape: Analyzing ten years of national crime survey data. Criminal Justice Policy Review, 2(4), 337 356. Benedict, L. W., & Zautra, A. J. (1993). Family environmental characteristics as risk factors for childhood sexual abuse. Journal of Clinical Child Psychology, 22(3), 365 374. Bratt, C. (2008). Guardians to counter adolescent drug use? Limitatio ns of a routine activities approach. Youth and Society, 39, 385 405. Bui, H.N. (2009). Parent child conflicts, school troubles, and differences in delinquency across immigration generations. Crime & Delinquency, 55(3), 412 441.
53 Butcher, K.F. & Morrison Piehl, A. (1998). Recent immigrants: Unexpected implications for crime and incarceration. Industrial and Labor Relations, 52(4), 654 679. Caetano, R. (1987). Acculturation, drinking, and social settings among U.S. Hispanic s. Drug and Alcohol Dependence, 19, 215 226. Cass, A. I. (2007). Routine activities and sexual assault: An analysis of individual and school level factors. Violence and Victims, 22, 350 366. Cohen, L. E., & Felson, M. (1979). Social change and crime rate trends: A routine activity approach. American Sociological Review, 44, 588 608. Desmond, S. A. & Kubrin, C.E. (2008). Immigrant Communities and Adolescent Violence. Forthcoming in Sociological Quar terly. Grebler, L., Moore, J. L., & Guzman, R. C. (1970). The Mexican American people. New York: The Free Press. Hannity, S. (Interviewer) & McCain, J. (Interviewee). (2010). Senator McCain: Arizona Is Under Siege. [Interview Transcript]. Re trieved from Fox News Web site: http://www.foxnews.com/story/0,2933,592767,00.html Hindelang, M. J., Gottfredson, M. R. & Garofalo J ( 1978 ). Victims of personal crime: An empirical foundation for a theory of personal victimization. Cambridge, MA: Ballinger. Hirschi, T. (1969). Causes of delinquency Berkley, CA: University of California Press. Jackson, A., Gilliland, K., & Veneziano, L. (2006). Routine activity theor y and sexual deviance among male college students. Journal of Family Violence, 21, 449 460. Keefe, S. M. (1980). Acculturation and the extended family among urban Mexican Americans. In A. M. Padilla (Ed.), Acculturation: Theory, models, and some new findings (85 110). Boulder: Westview.
54 Kennedy, L.W. & Forde D.R ( 1990 ). Routine activities and crime: An analy sis of victimization in Canada. Criminology, 28, 137 151. Marin, G. & Gamba, R. J. (2003). Acculturation and changes in cultural values. In K. Chun, P. Balls Organista, & G. Marin (Ed.), Acculturation: Advances in theory, measurement, and applied research (83 93). Washington, D.C.: American Psychological Association. Martinez Jr., R. (2002). Latino homicide: Immigration, violence, and community New York, NY: Routledge Taylor and Francis Group. Martinez Jr., R. (2006). Coming to America: The impact of the new immigration on crime. In R. Mar tinez Jr. & A.Valenzuela Jr. (Ed.), Immigration and crime: Race, ethnicity, and violence (1 19). New York, NY: New York University. Messner, S.F. & Blau, J.R. (1987). Routine leisure activities and rates of crime: A macro level analysis. Social Forces, 65(4), 1035 1052. Miethe, T.D. & Stafford, M.C. & Long, J.S. (1987). Social differentiation in criminal victimization: A test of routine activities/lifestyle theories. American Sociological Review, 52(2), 184 194. Moore, J. W. (1970). Mexican A mericans New Jersey: Prentice Hall. Morenoff, J.D. & Astor, Avraham. (2006). Immigrant assimilation and crime. generational differences in youth violence in Chicago. In R. Martinez Jr. & A. Valenzuela Jr. (Ed.), Immigrat ion and crime: Race, ethnicity, and violence (36 63). New York, NY: New York University. Mustaine, E.E. & Tewksbury R ( 1998 ). Predicting risks of larceny theft victimization: A routine activity analysis us ing refined lifestyle measures. Criminology, 36, 829 858.
55 Mustaine, E.E. & Tewksbury, R. (2002). Sexual assault of college women: A feminist interpretation of a routine activities analysis. Criminal Justice Review, 27(1), 89 123. Myers, R., Chou, C., Sussman, S., Baezcond e Garbanati, L., Pachon, H., & Valente, T. W. (2009). Acculturation and substance use: Social influence as a mediator among Hispanic alternative high school youth. Journal of Health and Social Behavior, 50, 164 179. Time Retrieved from http://www.time.com/time/nation/article/0,8599,1982268,00.html Osgood, D.W., Wilson, J.K., O'Malley, P.M., Bachman, J. G. & Johnston L.D ( 1996 ). Routine activities a nd individual deviant behavior. American Sociological Review, 61 635 655. Park, R E. & Burgess, E W. (1969). Introduction to the scien ce of s ociology. Chicago: University of Chicago Press. Plass, P. S., & Carmody, D. C. (2005). Routine activities of delinquent and non delinquent victims of crime. American Journal of Criminal Justice, 29, 235 247. Pollock, W., Joo, H.J., & Lawton, B. (2010). Juvenile arrest rates for burglary: A routine activities approach. Journal of Criminal Justice, 38, 572 579. Portes, A. & Zhou, M. (1993). The new second generation: Segmented assimilation and its variants. Annals of the American Academy of Political and Social Science, 530, 74 96. Portes, A., & Rumbaut, R. (2001). Legacies: The story of the immigrant second generation New York, NY: Russell Sage Foundation.
56 Reid, J. A. & Sullivan, C. J. (2009). A latent class typology of juvenile victims and exploration of risk factors and outcomes of victimization. Criminal Justice and Behavior, 36(10), 1001 1024. Reinhard, B. (2010, May 14). McCollum flips on Arizona immigration. St. Pete Tim es pp. 1B 8B. Rice, K.J. & Smith, W.R. (2002). Socioecological models of automotive theft: Integrating routine activity and social disorganization approaches. Journal of Research in Crime and Delinquency, 39, 304 336. Robertson, A. A., Baird Thomas, C., & Stein J. A. (2008). Child victimization and parental monitoring as mediators of youth problem behaviors. Criminal Justice and Behaviors, 35, 7 55 771. Rumbaut, R. G., Gonzales, R.G., Komaie, G., Morgan, C.V., & Tafoya Estrada, R. (2006). Immigrant and incarceration: Patterns and predictors of imprisonment among first and second generation young adults. In R. Martinez Jr. & A.Valenzuela Jr. (Ed.), Immigration and crime: Race, ethnicity, and violence (64 89). New York, NY: New York University. Sabogal, F., Marin, G., Otero Sabogal, R., Vanoss Marin, B., & Perez Stable, E. J. Hispanic Journal of Behavioral Sciences, 9, 397 412. Sampson, R.J. & Lauritsen, J.L. (1990). Deviant lifestyles, proximity to crime, and the offender victim link in personal violence. Journal of Research in Crime and Delinquency, 27, 110 1 39.
57 Schreck, C. J., & Fisher, B.S. (2004). Specifying the influence of family and peers on violent victimization: Extending routine activities and lifestyle theories. Journal of Interpersonal Violence, 19, 1021 1041. Schwartz, S. J. (2007). The applicability of familism to diverse ethnic groups: A preliminary study. The Journal of Social Psychology, 147(2), 101 118. Stahura, J.M. & Sloan, J.J.(1988). Urban stratification of places, routine activities, and suburban crime rates. Social Forces, 66(4), 1102 1118. T ewksbury, R., & Mustaine, E. E. (2001). Lifestyle factors associated with sexual assault of men: A routine activity theory analysis. 153 182. T ewksbury, R., & Mustaine, E. E. (2003 ). protective behaviors: Further considerations of the guardianship concept in routine activity theory Criminal Justice and Behavior, 30(3) 302 327 Turne r, H. A., Finkelhor, D., & Ormrod, R. (2007). Family structure variations in patterns and predictors of child victimization. American Journal of Orthopsychiatry, 77(2), 282 295. Xie, Y. & Greenman, E. (2005). Segmented assimilation theory: A reformation and empirical test. Population Studies Center Research Report. Zhou, M. (1997). Segmented assimilation: issues, controversies, and recent research on the second generation. International Migration Review, 31, 975 1008.