USF Libraries
USF Digital Collections

The effects of the presence of a dog on the social interactions of children with developmental disabilities

MISSING IMAGE

Material Information

Title:
The effects of the presence of a dog on the social interactions of children with developmental disabilities
Physical Description:
Book
Language:
English
Creator:
Walters, Stephanie
Publisher:
University of South Florida
Place of Publication:
Tampa, Fla.
Publication Date:

Subjects

Subjects / Keywords:
Developmental disorders
Animals
Communication
Therapy
Classroom
Dissertations, Academic -- Applied Behavior Analysis -- Masters -- USF   ( lcsh )
Genre:
government publication (state, provincial, terriorial, dependent)   ( marcgt )
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Abstract:
ABSTRACT: The effects of the presence of a dog on the social interactions between children with developmental disabilities and their teacher were analyzed in this study. We examined whether the presentation of a dog would improve the social interactions of three children with developmental disabilities. A baseline condition consisting of the child and teacher in the presence of three toys, one of which was a toy dog was followed by an intervention in which a real dog was added to the sessions. A multiple baseline design across participants was employed to assess experimental changes in interactions during the intervention condition. All participants demonstrated an increase in overall positive initiated behaviors (verbal and non-verbal), positive initiated interactions toward the teacher (verbal and non-verbal) and positive initiated interactions toward the dog (verbal and non-verbal).
Thesis:
Thesis (M.A.)--University of South Florida, 2005.
Bibliography:
Includes bibliographical references.
System Details:
System requirements: World Wide Web browser and PDF reader.
System Details:
Mode of access: World Wide Web.
Statement of Responsibility:
by Stephanie Walters.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains 79 pages.

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001680959
oclc - 62412401
usfldc doi - E14-SFE0001162
usfldc handle - e14.1162
System ID:
SFS0025483:00001


This item is only available as the following downloads:


Full Text
xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam Ka
controlfield tag 001 001680959
003 fts
005 20060215071031.0
006 m||||e|||d||||||||
007 cr mnu|||uuuuu
008 051202s2005 flu sbm s000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0001162
035
(OCoLC)62412401
SFE0001162
040
FHM
c FHM
049
FHMM
090
BF176.7
1 100
Walters, Stephanie.
4 245
The effects of the presence of a dog on the social interactions of children with developmental disabilities
h [electronic resource] /
by Stephanie Walters.
260
[Tampa, Fla.] :
b University of South Florida,
2005.
502
Thesis (M.A.)--University of South Florida, 2005.
504
Includes bibliographical references.
516
Text (Electronic thesis) in PDF format.
538
System requirements: World Wide Web browser and PDF reader.
Mode of access: World Wide Web.
500
Title from PDF of title page.
Document formatted into pages; contains 79 pages.
3 520
ABSTRACT: The effects of the presence of a dog on the social interactions between children with developmental disabilities and their teacher were analyzed in this study. We examined whether the presentation of a dog would improve the social interactions of three children with developmental disabilities. A baseline condition consisting of the child and teacher in the presence of three toys, one of which was a toy dog was followed by an intervention in which a real dog was added to the sessions. A multiple baseline design across participants was employed to assess experimental changes in interactions during the intervention condition. All participants demonstrated an increase in overall positive initiated behaviors (verbal and non-verbal), positive initiated interactions toward the teacher (verbal and non-verbal) and positive initiated interactions toward the dog (verbal and non-verbal).
590
Adviser: Trevor Stokes, Ph.D.
653
Developmental disorders.
Animals.
Communication.
Therapy.
Classroom.
0 690
Dissertations, Academic
z USF
x Applied Behavior Analysis
Masters.
773
t USF Electronic Theses and Dissertations.
856
u http://digital.lib.usf.edu/?e14.1162



PAGE 1

The Effects Of The Presence Of A Dog On Th e Social Interactions Of Children With Developmental Disabilities by Stephanie Walters A thesis submitted in partial fulfillment of the requirements for the degree of Master of Arts in Appl ied Behavior Analysis College of Graduate Studies University of South Florida Major Professor: Trevor Stokes, Ph.D. Jennifer Austin, Ph.D. Holly Steele, Ph.D. Date of Approval: March 4, 2005 Keywords: developmental disorders, an imals,communication, therapy, classroom Copyright 2005 Stephanie Walters

PAGE 2

Acknowledgements I wish to thank my committee for thei r support and encouragement throughout this process. Dr. Austin, thank you for your commitment to the students and to seeing us through to the completion of our studies. Your loyalty is to be co mmended. Dr. Steele, you have been remarkable in your encourag ement and support. Thank you for making me laugh and being a calming presence throughout this process. I am honored to have met you. Dr. Stokes, you have been a t eacher, a mentor and a friend throughout my studies and especially in writing this thesis Your encouragement has seen me through many days when I have wanted to give up. Your calming presence and firm responses kept me motivated and committed. You never let me quit. You are an inspiration. Thank you. I would also like to thank Dawn Gonzalez, Jennifer Lotti, Dawn Calder and Arrow for making this thesis a reality.

PAGE 3

i Table of Contents List of Tables iii List of Figures iv Abstract v Chapter 1: Introduction 1 Physiological Effects 2 Psychological Effects 6 Social Effects 9 The Present Study 15 Chapter 2: Method 17 Participants 17 Setting 18 Dependent Variables and Measurement 18 Data Collection and Interobserver Reliability 20 Social Validity 22 Experimental Procedures 22 Teacher Training 22 Baseline 23 Intervention 24 Experimental Design 25 Chapter 3: Results 26 Overall Positive Initiated Interactions 26 Positive Initiated Interactions with the Teacher 28 Positive Initiated Interactions with the Dog 30 Overall Negative Initiated Interactions 33 Negative Initiated Interac tions with the Teacher 35 Negative Initiated Intera ctions with the Dog 36 Classroom Rating for Positive Interactions Following Sessions 42 Classroom Rating for Negative Interactions Following Sessions 46 Interobserver Agreement 47 Social Validity Ratings 49

PAGE 4

ii Chapter 4: Discussion 50 Relation to Literature 52 Limitations 53 Recommendations for Future Research and Practice 55 Conclusion 56 References 58 Appendices 60 Appendix A: Baseline Session Guidelines 61 Appendix B: Intervention Session Guidelines 62 Appendix C: Baseline Session Guidelines Checklist 63 Appendix D: Intervention Session Guidelines Checklist 64 Appendix E: Session Data Collection 66 Appendix F: Observer Proficiency 67 Appendix G: Protocols for Interactions 68 Appendix H: Classroom Data Collection 70 Appendix I: Social Validity 72

PAGE 5

iii List of Tables Table 1 Mean baseline and interven tion session percentages for each dependent variable for each participant. 39 Table 2 Mean number of questions and task suggestions per session for baseline and intervention for each participant. 40 Table 3 Mean number of positive, negative, verbal and nonverbal responses to questions or task suggestions directed toward the teacher and the dog per session. 40 Table 4 Mean rating of each dependent variable during classroom data collection following each sessi on for each participant. 42 Table 5 Mean percentage of interobserver agreement scores for each dependent variable for each participant. 48 Table 6 Post intervention social validity ratings by teacher and teachers assistant using a Likert Scale. 49

PAGE 6

iv List of Figures Figure 1 Multiple baseline acr oss participants of overall positive interactions. 27 Figure 2 Multiple baseline across partic ipants of positive child initiated interactions directed toward the teacher. 29 Figure 3 Multiple baseline across partic ipants of positive child initiated interactions directed toward the dog. 31 Figure 4 Multiple baseline across participants of overall negative interactions. 34 Figure 5 Multiple baseline across partic ipants of negative child initiated interactions directed toward the teacher. 37 Figure 6 Multiple baseline across partic ipants of negative child initiated interactions directed toward the dog. 38 Figure 7 Multiple baseline across participants of intervals completed per session. 43 Figure 8 Multiple baseline across participants of positive classroom interactions following sessions. 44 Figure 9 Multiple baseline across participants of negative classroom interactions following sessions. 45

PAGE 7

v The Effects of the Presence of a Dog on the Social Interactions of Children with Developmental Disabilities Stephanie Walters ABSTRACT The effects of the presence of a dog on the social interactions between children with developmental disabilitie s and their teacher were analyzed in this study. We examined whether the presentation of a dog w ould improve the social interactions of three children with developmental disabiliti es. A baseline condition consisting of the child and teacher in the presence of three toys, one of which was a toy dog was followed by an intervention in which a real dog was added to th e sessions. A multiple baseline design across participants was employed to a ssess experimental changes in interactions during the interven tion condition. All participants demonstrated an increas e in overall positive initiated behaviors (verbal and non-verbal), positive initiated inte ractions toward the teacher (verbal and nonverbal) and positive initiated interactions toward the dog (verbal and non-verbal). The children also showed an overall decrease in negative initiated beha viors (verbal and nonverbal). Two of the three pa rticipants demonstrated a decrease in negative initiated interactions toward their teacher (verbal and non-verbal), while with one participant there was a slight increase in negative non-verbal in teractions toward the teacher. All three children showed slight increases in negative initiated non-ver bal interactions with the dog while negative initiated ve rbal interactions toward the dog remained the same.

PAGE 8

1 Chapter One Introduction Dogs are versatile creatures. They are workers bred to perform many duties such as hunting, herding, protecting livestock and property, and hauling sleds. Through the years, the work that dogs are bred to do has become more refined. Dogs are now trained to sniff out drugs, bombs and other scents fo r the K-9 police force, have been used in search and rescue efforts both at sea and on land, and more recently, have been trained as companion animals for people with disabilities. They have been tr ained as seeing-eye dogs for people with visual impairments (N aderi, Miklosi, Doka & Csanyi, 2001) and signal dogs for people with hearing impairmen ts (Hart, Zasloft & Benfatto, 1996). They have also been trained as seizure-alert dogs and to assist people with physical disabilities with aspects of their daily lives includi ng retrieving items, switching lights on and off, and taking off shoes and socks (Lane, McNicholas & Collis, 1998). For years dog owners have claimed their pets have healing power. Although this is highly subjective, a trend has been devel oping toward using dogs as therapy for the elderly and people with disabi lities. Having people participate in service dog training programs have become an avenue of providing therapy for emotiona lly disabled children and prison inmates (Law & Scott, 1995; Walsh & Mertin, 1994). The effects of the presence of animals as adjuncts to therapy are a growing area of research. Many therapists believe that anim al-assisted therapy (AAT) helps build rapport quickly, especially with children with emo tional difficulties. The media has published

PAGE 9

2 articles concerning the traini ng of service dogs by delinquent children (e.g., Stacy, 2003) as well as incarcerated adults (Walsh & Mertin, 1994). P opulations reported to have experienced the positive effect s of animals in therapeutic situations include children (Hansen, Messinger, Baun & Megel, 1999), i ndividuals with physical (Eddy, Hart & Boltz, 1987), emotional (Kaminski, Pellino & Wish, 2002) or psychiatric impairments (Marr, French, Thompson, Drum, Greening, Mormon, Henderson & Hughes, 2000), individuals with developm ental disabilities (Limond, Bradshaw & Cormack, 1997) or pervasive developmental disorders (Marti n & Farnum, 2002), the elderly (CrowleyRobinson, Fenwick & Blackshaw, 1996), (Fick, 1993), delinquent children(Stacy, 2003) victims of abuse (Williams, 2003), adults with substance abuse (Marr, et. al., 2000) and prisoners (Walsh & Mertin, 1994). In most of these cases, positive effects have been reported. However, there are few quantitative studies to support AAT. The published literature includes case studies that are primarily subjective testimonials with few formal research designs or controls in place. The literature at present has focu sed primarily on the benefits of animals on human health. Physiological Effects Several studies have been conducted to demonstrate the physiological effects of animals on people. Eighteen humans and ei ghteen dogs participated in a study conducted by Odendaal (2000). The purpose of this study was to examine the physiological basis for affiliation behavior (behavior that is mutua lly beneficial) between humans and dogs. Odendaal measured the mean arterial blood pr essure and six neurochemicals related to blood pressure (endorphin, oxytoc in, prolactin, phenylethylami ne, dopamine and cortisol)

PAGE 10

3 before and after conditions of interactions and quiet book reading. Participants were divided into two groups: pet owners and th eir dogs in an experimental group and individuals with unfamiliar dogs in a control group. In the interaction-with-dog condition, all of the neurochemicals increased significan tly (p<.05) except co rtisol, with both the humans and the dog participants. These neur ochemicals are associated with positive affiliation behavior. Therefore the human a nd dog participants increase in affiliation behavior was indicated by the neurochemical changes. According to Odendaals report, both the human and the dog participants showed deceased cortisol. The report does assert a significant decrease for humans, although no significance level was noted, whereas for dogs the decrease was not significant (p=.30). The only significant difference between the experimental and control groups in the baseline and interacting-w ith-a-dog condition was in the level of oxytocin, which was higher in the experimental group (p<.01) in wh ich the participants in teracted with their own dogs. Oxytocin has been used as an indi cator of bsocial attachment in members of the same species. Thus oxytocin can also be associated with measures of social attachment between species as demonstrated by the experimental group interacting with a familiar dog. Increases in endorphin (p<.10), oxytocin (p<.01), and prolactin (p<.01) were higher during the dog interaction condition vers us the reading quietly conditions with no significant changes in other neurochemical levels. Changes in endorphin, oxytocin and prolactin are correlated with social bonding neurochemical changes. Results also indicated that a significant decrease in blood pressure for humans and dogs was achieved

PAGE 11

4 in 5-24 minutes of positive inte ractions with dogs as well as changes in neurochemical plasma levels that are associated with attachment A repeated measures study was conducte d by Wilson (1987), who examined the cardiovascular responses of college students to a dog. Ninety-two undergraduate students ages 18-39 participated in this study. Heart rate, blood pressure (sys tolic and diastolic), and mean arterial pressures were measured be fore or during three te st conditions: reading alone, reading out loud, and petting an unf amiliar dog. Anxiety as assessed as state anxiety (how the participant feels right now) and trait anxiety (how the participant generally feels) were measured followi ng each session using the Spielberger SelfEvaluation Questionnaires. Each session was 10 minutes long and was preceded by a 10 minute adaptation baseline period in which the participant sat quiet ly and did not talk, read or interact. Each part icipant was randomly assigned to one of six orders of the sessions so that treatment order was not th e same for all of the participants. Analysis of variance indicated significant differences in treatment effect (p<.001) in all three conditions. The results of this st udy indicated that intera cting with the dog and reading quietly decreased th e physiological and psychologica l responses of the students from pre-session baseline measures (p< .001). Reading aloud was elevated above baseline for heart rate and blood pressure (systolic and diastolic) (p<.0005). Trait anxiety showed no significant difference between conditions. State anxiety level was significantly different from baseline under al l conditions except petting an unfamiliar dog (p=.937). A possible confounding variable to this st udy includes the following: in the third condition, petting the dog, some participants did talk to the dog which may account for

PAGE 12

5 some elevation in the blood pressure measures However, since all of the participants spoke to the dog in that condition th e differences should be constant. Hansen, et al. (1999) conducted a repeated measures study of 34 two-to-six year old children obtaining ph ysical examinations in a pediatri c outpatient clinic. Children did not share a common diagnosis. The childre n were randomly assigned to one of two groups. The experimental group contained the presence of a dog during the physical examination and the control group did not have a dog present during the physical examination. Physiological symptoms were measured during the physical examination. It was hypothesized that the physical examina tion would be stressful to the children and the presence of the dog would moderate that distress by bein g a distraction to the child because dogs initiate and facilitate interac tions which children may consider to be friendly. Physiological symptoms measured were blood pressure (systolic and diastolic), mean arterial pressures, heart rate and finger tip temperature. The participants were also video taped during the examination and the Observation Scale of Behavioral Distress (OSBD) was used to determine the presence of behaviors indicating distress. These behaviors included information-seeking (ask ing questions), cryi ng, screaming, physical resistance, verbal resistance, verbal pain (reports of actual or anticipated pain or discomfort) and flailing (arms, legs or body). Baseline data were taken prior to the examination and at 2 minute intervals throughout the examination process. No signi ficant differences were found in either demographic variables or in the presence of a dog in the home between the experimental and control groups. There was also no statistically significant diffe rence between groups

PAGE 13

6 in measures of systolic, diastolic or mean ar terial blood pressures, heart rate or finger tip temperature at baseline or during the examinations. Behavioral indicators of distress were measured at baseline and throughout the examination and were apparent in both groups. These behaviors increased in frequency over time in both groups. Participants in the groups in which the dog was present exhibited fewer behaviors indicating distress overall, scoring lower on the OSBD (p=0.034). There was no statistically si gnificant difference between the groups at baseline. Therefore, although the presen ce of the dog did not significantly impact physiological indicators of distress, behavioral indicators of distre ss were lower in the presence of the dog. This study indicates that the presence of a dog in medical/clinical settings may alleviate dist ress in children, allowing mo re thorough examinations and more accurate diagnoses. Psychological Effects How play therapy and pet therapy affect hospitalized children were examined by Kaminski et al. (2002). Seventy children with a mean age of 9.86 and who were diagnosed with chronic medical disorders participated in this study. The children were divided into two groups: play therapy gr oup and a pet therapy group. Play therapy consisted of developmentally appropriate pl ay opportunities in th e hospital playroom. The pet therapy consisted of th e presence of a visiti ng dog in the sessions to interact with the child. The dependent variables included ratings on a mood rating scale completed by the patient as well as a separate mood rating scale completed by the parent/caregiver of the child. These mood rating scales included questions about feeli ng happy, lonely, sad, worried, bored, (feel) like crying and (feel) like playing with ot her kids (Kaminski et al.,

PAGE 14

7 2002, p. 325). The children were also asked to identify what they were feeling using a faces chart of facial expressions. The mood rating scale done by the parent/caregiver contained the following four items: happy, scared lonely and relaxed. On all three mood rating scales, the higher the score, the more pos itive the mood or condition of the child. At the beginning of each session, the child ren were videotaped for approximately 2 minutes and again at 10 minutes and 20 minut es into the session. The videotapes were reviewed and the children were assessed for positive affect, negative affect, anxiousfearful affect, neutral affect, touch-physical contact and persistence-on task. All of these behaviors were operationally defined a nd a coding system was developed. The percentage of the videotaped time a child engaged in one of the above behaviors (indicators of affect were beha viorally defined) was analyzed Physiological indicators were also measured. These included salivary cortisol measures (a steroid associated with increased adrenocorticol responses and stress), heart rate a nd blood pressure, all of which showed positive improvements. According to parent/caregiver ratings, children were reported to be happier after pet and play therapy. The pet therapy group was reported to be happier than the play group after therapy (p<.001). Children in the pet therapy group also displayed significantly more positive affect and touchi ng than the play therapy group (p<.05). Heart rate was higher in the pet therapy group after therapy. The salivary cortisol levels were similar in both groups prior to thera py and decreased in both groups after therapy. The results are from only a portion of the samples of saliva taken due to the evaporation of some samples. The results of the salivary cortisol measure were not statistically significant.

PAGE 15

8 Unfortunately, this study was not tightly co ntrolled in some areas. For example, the children did not always begin their sessions immediately following the initial measurements. Therefore, other factors pr eceding the sessions could account for the changes observed. Overall, the play and pet therapy were reported as a positiv e experience for the child by the parents/caregivers. Introducing play and/or pets into a treatment setting such as a hospital will helped promote normalcy for the children as well as provide them with opportunities to particip ate in activities that were likely to decrease boredom and loneliness associated with l ong stays in the hospital. Marr et al. (2000) conducted a study on the effects of animal assisted therapy on the pro-social behaviors of 69 adults diagnos ed with a mental illness and at risk for substance abuse in an inpatient psychiatric f acility. A repeated measures analysis of variance design was used to evaluate mean w eekly scores on the Social Behavior Scale (SBS, Perelle & Granville, 1993). Participan ts were randomly assigned to two groups. Group therapy consisted of substance abuse education. The content of the training was identical for both groups with the exception of the presence of animals in the room for the AAT group. The Social Behavi or Scale was used for baselin e and ongoing measurement. Some of the items measured were socializat ion, helpfulness, cooperativeness, activeness, response to surroundings, sociab le with others, likely to in teract with other patients, smiling and other indications of pleasure. By the fourth week of group sessions, the AAT group was found to interact more with other patients (p=0.022), to be more active (p=0.04), responsive to their surroundings (p=0.0 3), more sociable with others (p=0.05),

PAGE 16

9 more helpful (p=0.04), likely to interact w ith other patients (p=0.008) and were more likely to smile or indicate pleasur e (p=0.003) than the control group. Social Effects The beneficial effects of animals on individuals with developmental disabilities/mental retardation and physical disabilities also have been noted in the literature. Specifically noted were the effect s of social interactions between individuals with disabilities and a therap ist, as well as the interac tions of the public toward individuals with disabilitie s (Eddy et al., 1987; Hart et al., 1987; Lane et al., 1998; Mader, Hart & Bergin, 1989; Martin & Farnum, 2002). Eddy et al. (1987) published a study in which the frequency of social acknowledgement including smiles, conversatio n, and touch toward individuals with physical impairments (participants using wheelchairs) and toward people who were ambulatory was examined. The person in the wheelchair was followed by an observer who then recorded the following behaviors of the passersby: smiles, touch, conversation, gaze aversion, path avoidance, or no response. The passerbys responses to the dog were recorded separately than those directed towa rd the individual in the wheelchair. The behaviors of the passerby toward the part icipants with physical impairments were recorded for those participants with a se rvice dog and those without a service dog. The results indicated that smiles and conversations from passersby increased in the presence of the service dog. This study suggests th at social recognition and acceptance for individuals with physical impairments can be improved with the pr esence of service animals in social settings.

PAGE 17

10 Social interactions among strangers have also been examined to determine the effects of animals on increasing social inte ractions. The role of small animals on increasing social interactions among stra ngers was examined by Hunt, Hart and Gomulkiewicz (1991). A woman serving as a confederate sa t in a park reading and taking notes. There were f our conditions to the study: the woman with one of the following for each condition: a rabbit, a tur tle, a small portable television set that was playing and a bottle of bubbles. The occurren ce of smiles, conversation or touching by individuals approaching within 1.5m was recorded for each condition. The individuals were categorized as either adults or child ren based on appearance. Observations were recorded for a total of 6 hours in e ach condition (1 hour sessions). The results indicated that the adults appr oached the rabbit significantly more than any other stimulus (p<.004) and children were significantly more lik ely to approach the bubbles (p=.002) than any other stimuli. Child ren were also signifi cantly more likely to touch the stimuli (p=.016) than the adults and the adults were more likely to smile (p<.001) and converse (p<.001) than the children. A similar study was conducted by McNicholas and Collis (2000) to determine if the presence of a dog increased social intera ctions between strangers. The study included a baseline condition of interactions without the presence of a dog and two conditions, one in which a neatly dressed male walked a dog in a public area and one in which a scruffily dressed male walked a dog in a public area. The results of the study indicated that there were more interactions between strangers in the presence of a dog than without a dog. The results also indicated that the interactions were higher with the neatly dressed male than with the scruffily dressed male and the interactions in both conditions were

PAGE 18

11 significantly higher than without the presence of a dog. Thes e results suggest that the presence of a dog may increase soci al interactions in the scruff ily and the neatly dressed. The effect animals have on the social lives of humans is perhaps one of the more widely researched areas of animal assisted therapy. Many pet owners acknowledge changes in their social lives resulting from owning a pet. These changes can include increased social contact with strangers, increased activity re sulting from taking their pets out and an extension in their network of suppor t from other pet owners. Pet owners often congregate in dog parks, dog beaches, pet shops and at specialized doggie events. In a retrospective study of 19 people with service dogs utilizing wheelchairs, Hart et al., (1987) found that the part icipants reported that when the service dog was present, social greetings from strangers (adult and chil d) increased. Hart et al. (1987) found that the participants were approached significan tly more often when the dog was present than when the dog was not present on a typical trip downtown (p< 0.01). The social behavior of the participants also increased with 11 of 19 participan ts reporting more frequent outings (without an attendant) into the community since obtaining a service dog. The role of service dogs for people with disabilities often has a combined benefit of assisting in social inte gration, acknowledgement or acceptance among society. Lane et al., (1998) studied the benefits of service dogs for 57 subjects receiving service dogs from the Dogs for the Disabled organization. Part icipants answered a questionnaire assessing the perceived changes in thei r lifestyle and well-being follo wing receiving a service dog. The dogs role as a social f acilitator, provider of an a ffectionate relationship, an emotional and esteem supporter and as an influence on perceived health was assessed through the questionnaire. The re sults indicated that 92% of th e participants reported that

PAGE 19

12 people frequently stopped to talk with them when they were out with their dogs. Seventy-five percent of the pa rticipants reported that sin ce obtaining a service dog they had made new friends, and over one third of the participants reported having an overall improvement in their social lives as a result of the service dog. Children with disabilities are not often th e recipients of service dogs and often receive less social acknowledgement than a dults (Mader et al., 1989). Mader et al. (1989) conducted a study of five children in the California school system with service dogs. The childrens age ranged from 10-15 y ears old. A control group of 5 children of similar age, race and degree of disability was selected. The frequency of social acknowledgement, defined as friendly glances smiles and conversations was noted in a school setting and in a shopping mall. The resu lts of this study indicated that children with service dogs received significantly more looks and conversations from passersby and conversations were longer in duration th an the children without the service dogs in the school setting and in the public setting. These results suppor t the hypothesis that service dogs facilitate social acknowledgement for children wi th disabilities (Mader et al., 1989). Martin and Farnum (2002) used a within participants repeated measures design to measure pro-social interactions (behaviors th at indicated interest and engagement with the environment) and nonsocial interactions (behaviors such as hand flapping and ignoring questions) during three conditions: with a ball, with a stuffed dog toy and in the presence of a live dog. Therapy sessions o ccurred three times a week and lasted 15 minutes each. During the therapy sessions, the child was presented with the toy ball, the stuffed dog or the real dog and the therapis t. The therapist followed a predetermined

PAGE 20

13 protocol designed to elicit pr o-social behaviors. The pr otocol was based on questions having to do with the indepe ndent variable. Ten childre n diagnosed with pervasive developmental disorder ranging from the ag e of 3 to 13 participated in the study. According to this study, the children who were exposed to a real dog were more focused (duration of looking p<.017) and aw are (looking at object therapist or dog) (p<.001) of their environments and displaye d a more playful mood (indicated by laughing p<.025, giving treats p<.001, and hand-flappi ng p<.005) when in the presence of a therapy dog. Children were also more likely to talk to the dog (p< .001) or about the dog (p<.001) when in the presence of the dog. One area of contradiction to the hypothesis within this study is the evid ence that children re sponded with less deta iled explanations (p<.001) and were less likely to initiate c onversations about themselves (p<.017) or the therapist (p<.001) in the presence of the dog (Martin and Farnum, 2002). Also, handflapping was included in Martin and Farnum s description of a non-social behavior however in their discussion of the results of their study they refer to it as a pro-social behavior. Most significant to the cu rrent study is the research by Limond et al., (1997) in which eight children ranging in age from 7 to 12 and diagnosed with Down syndrome participated in a study to dete rmine the effects of the presen ce of a dog on the behavior of children with developmental disabilities. B ecause of the lack of adequate quantitative methodological studies, this st udy also sought to develop pr ocedures for assessing the effects of an animals presence on an individu als behavior with controls in place in the environment.

PAGE 21

14 A repeated measures design with two conditions was utilized. The conditions consisted of the handler and an imitation dog sim ilar in size, color and texture to the live dog and two toys in the first condition and the handler and a real dog (a 7 year old male black Labrador Retriever) in the second condition. Each session contained both conditions and was a total of 14 minutes each. The children were ini tially exposed to one condition for 7 minutes, followed by the second condition for another 7 minutes. In each condition the handler encouraged the child to perform activit ies involving the test dog (real or imitation) but the child was free to in teract in any way with the dog, the toys or the handler. The conditions were alternated weekly to control for effects of order and habituation to or fear of dogs. The behaviors of interest in this study were the duration of time spent in looking toward the therapist, dog, toys or other obj ects; the frequency of verbal and non-verbal initiations and the frequency of verbal and non-verbal responses to ward the therapist, dog, toys or other objects. The results of this study i ndicated that the children dire cted their gaze at the real dog for a significantly longer duration than th ey did to the therap ist, imitation dog and toys, or other objects in the room. The child ren did not respond to th e therapist as often in the imitation condition as th e real dog condition, with a statistical difference of p<0.05. Non-verbal responses to sugge stions concerning the dog were more frequent in the real dog condition (p<0.05), however the frequenc y of nonverbal responses to suggestions about items other than the dog were similar in both conditions. The frequency of verbal responses concerning the dog (real or imitati on) was similar across both conditions but

PAGE 22

15 there was a statistically significant increase in the verbal responses to items other than the dog in the imitation dog condition. Verbal responses were categorized as eith er positive or negative. Positive verbal responses were defined as those that were appropriate to the situa tion and/or expressed interest or enthusiasm, whereas negative verbal responses were defined as being inappropriate and/or expressi ng disinterest or a lack of enthusiasm. The children responded positively with significantly greater frequency to questions concerning the test dog in the real dog condition and negative verb al responses concerni ng the test dog were significantly more frequent in the imitation condition. These results indicated that the children responded non-verbally more often and more pos itively in the real dog condition. Initiations were similar in frequency in both conditions but toys and other objects (room fixtures such as light switches, or items found in the room other than those selected for the study) elicited significantly more non verbal initiations than the test dog or therapist in the imitation dog condition. There were significantly more verbal initiations to the test dog in the real dog condition and significantly more verbal initiations to other objects in the imitation condition. While the Limond et al. study noted the effects of the presence of a dog on the frequency and quality of inter actions, the quality of those interactions was considered to be either positive or negative based on the verbal behavior of the child. Non-verbal responses were not subdivided into eith er positive or negative interactions. The Present Study In summary, research regarding the eff ects of dogs on people shows that social interactions, psychological well being, and phys iological reactions are affected by the

PAGE 23

16 presence of dogs and therapeutic interactions with dogs. Most research is characterized as anecdotal and qualitative methodologies devoid of obj ective data within nonexperimentally controlled designs. The purpose of the present study was to mo re objectively assess the effects of the presence of a dog on the positive and negative social responsiveness (both verbal and non-verbal) of children with developmental di sabilities. In additi on, the research was conducted within a systematic single case experimental design with replicated effects across participants.

PAGE 24

17 Chapter Two Method Participants Three children between the ages of 5-9 in an Exceptional Student Education (ESE) kindergarten through second grade classr oom of a public elementary school were identified for this study. The children in this group were all diagnosed as having mental retardation. The participants consisted of two males and one female. Two of the three children, Kirsten and Georgie were also dia gnosed with Down syndrome. The other boy, Owen was diagnosed as hearing impaired. Each child displayed the ability to communicate using one word verbalizations. Two children displayed the ability to use short 2-3 word sentences. All three children either own a dog or have a relative or friend with a dog that they see regularly. Criteria for inclusion were: a diagnosis of mental retardation, placement in a special education class, between the ages of 5-9 years old, and the ability to communicate using a minimum of one word utterances. Criteria for exclusion in this study incl uded allergies to or fear of animals/dogs. After the participants were nominated by their teach er, written informed consent was obtained from the parents/guardians of the participants prior to c onducting this study consistent with approval from the Institutional Revi ew Board and the Pinellas County Schools. Parents also were asked whether their childr en have a dog at home and the extent to which the children had experien ce interacting with dogs.

PAGE 25

18 Setting This study was conducted in a public elementary school. The sessions were located in a room adjoining the participant s ESE classroom. The children were allowed to have access to that room prior to the study beginning to control for possible confounds due to a novel environment. The door was left open for the children to wander through throughout the school day. Children also used the room for one to one academic sessions with the teacher as well as to use the cot when ill. The room was approximately 6 x 2.75 meters with three side by side windows. The room contained a cot, two file cabinets, two storage cabinets, shelves c ontaining books and educational materials on three of the walls, a small refrigerator, a microwave, a counter that contained drawers and had a computer and printer on it, one office chair, one childrens chair and various toys located in the far corner of the room. Items were placed out of arms r each of the child when possible and items the children might find interesting or distracti ng were removed or covered with white paper during the sessions, when possible. A video camera was set up on one of the book shelves facing the child and teacher. It was turned on before the child en tered the room and turned off after the child left the room. The child and teacher sat on th e floor, across from each other, and with the child facing the camera. The camera was c oncealed amongst other items on the shelf and had a cloth draped over it to decrea se reactivity to being videotaped. Dependent Variables and Measurement The social behaviors of the participants served as the dependent measures in this study. Behaviors were categorized as positive or negative, verbal or non-verbal. Positive verbal statements were defined as those utte rances indicating pleasure or interest in the

PAGE 26

19 situation (e.g., that was fun, I like it, more, can I stay?, I love the dog. et cetera) or requests for help (e.g., help opening bag containing dog trea t). Negative verbal statements were defined as those utterances indicating displeasure or disinterest in the situation (e.g., This sucks, I hate this, Can I leave?, I hate the dog, Get me out of here et cetera). Verbal refusals to pa rticipate in the session also were scored as negative verbal statements (e.g., I dont want to, No, no more, et cetera). Positive nonverbal behaviors were defined as those behaviors indicating pleasure or interest in the situation (e.g., smili ng, laughing, touching the dog by petting, hugging or kissing, et cetera.), cl apping hands, nodding head, co mplying with a request nonverbally, blowing kisses and social agreement uh-huh, sharing or handing things to the teacher, throwing/handing treats to the dog, holding the leash or walking the dog, et cetera. Negative nonverbal behaviors were defined as those behaviors indicating displeasure or disinterest in the situa tion (e.g., turning body or face away from the teacher, crying, frowning (corners of lips turned down), hiding face, attempts to leave the room, physical aggression, property destruc tion (throwing things, knocking things off shelves/table, playing with computer if these actions would cause damage if uninterrupted) or no response to dog-related questions or task suggestions). Interactions were further assessed as eith er child-initiated responses toward the teacher or the dog (e.g., child responded inde pendently without prompting) or teacherprompted responses toward the teacher or the dog (e.g., child responded to a request to perform a task with the dog or answered a question when asked).

PAGE 27

20 Data Collection and Inter observer Reliability Each child participated in 8-minute sessions 5 days per week w ith the teacher. The sessions were scheduled to occur upon ar rival at school followi ng breakfast and the times of the sessions remained constant, about 9 am. During experimental sessions, partial interval record ing (ten seconds for observation, fi ve seconds for recording) was used to measure the dependent variables (A ppendix E). Each session was video taped. The observer, who was a guidance counselor wi th a Masters degree, was experienced in the behavior of children w ith developmental disabilities, and was blind to the experimental predictions, was present at the sessions when possible and sat in the far corner of the room and remained as unobt rusive as possible. The observer was instructed not to speak or make eye contact during the sessions. When it was not possible for the observer to be present, the video tape s were reviewed and subsequently scored by the observer. The primary observer was present for all but three sessions. The reliability observer was present for all but four sessions. A data sheet was designed for use in m easuring the dependent variables during sessions (Appendix E). Observers were cued at the end of each interval using a cassette tape that signaled the elapsed time. Session data were reported as the percentage of intervals in which each target ed behavior occurred (number of intervals in which the behavior was scored divided by the total number of intervals x 100). Measures of dependent variables also we re conducted for half an hour following each intervention session (i.e. when the child is reintegrated into the classroom) (Appendix I). During reintegration sessions, data were collected by the teacher or instructional assistant using a rating scale (Appendix I).

PAGE 28

21 The principle investigator provided a training session for the observers. Training included direct instructions on session procedures, operational definitions of the targeted behaviors and data collection procedures, demonstration of session procedures and data collection procedures, role plays demonstrating two examples of each behavior (positive verbal and nonverbal behavior s, negative verbal and nonverbal behaviors and initiations and responses), guided feedback on session pr ocedures, operational definitions of the targeted behaviors and data collection pro cedures, and corrected role play on session procedures. Each observer participated in one training session each. The teachers assistant scored 95% proficiency and the gui dance counselor (prima ry observer) scored 100% proficiency following the training. Proficiency was calculated using the following equation to obtain the percentage of agr eement: Proficiency = number of correct number of correct plus incorrect x 100. Inter-observer agreement was assessed in 65% of the sessions for Kirsten, in 68% of Georgies, and 44% of Owens sessions di stributed across all conditions. The sessions or videotapes were viewed simultaneously by two observers who were seated at least 1 meter from each other with the video camcord er set up between them and did not confer regarding what they were viewing to insure independence of observations. Interobserver agreement was calculated by dividi ng the number of agreement intervals by the number of agreement intervals plus disagreement intervals and multiplying by 100. The primary observer was the school guidance c ounselor with a masters degree and the reliability observer was a graduate student in applied behavior analysis.

PAGE 29

22 Social Validity Social validity was assessed using a ques tionnaire which was administered to the teacher and instructional assistant followi ng the completion of the study (Appendix J: Social Validity). This was intended to meas ure the appropriateness of the procedures, the social importance of the goals and the social importance of the effects. Experimental Procedures Teacher training. Prior to data collection, traini ng was conducted to ensure the teacher conducted experimental sessions according to prescribed protocols. A certified Exceptional Student Education (ESE) teacher w ith a Masters degree in Special Education conducted all experimental sessions. The principle investigator provided a training session with the teacher and a session guideline was given to the teacher to follow each session (Appendix A: Baseline Session Guidelines; Appendix B: Intervention Session Guidelines). A Protocol for Interactions (Appendix H), specifying the conten t of the interactions with the child was given to the teacher to follow along with the Baseline Session Guidelines and Intervention Session Guidelines (A ppendices A and B) each session. Training included direct instructions on baseline and intervention session procedures and the operational definitions of the targeted behaviors; demonstration of baseline and intervention session procedures; role plays demonstrating two examples of each behavior (positive verbal and nonverbal behaviors, negative verbal and nonverbal behaviors and initiations and responses); guided feedback on baseline and intervention session procedures; and corrected role play on baseline and intervention session procedures. The teacher scored 100% profic iency following the training. Proficiency

PAGE 30

23 was scored using the attached Baseline Se ssion Guideline Checklist (Appendix C) or Intervention Session Guideline Checklist (A ppendix D). Proficiency percentage was calculated using the following equation: Prof iciency = number of correct number of correct plus incorrect 100. Baseline. Baseline sessions consisted of the presence of the teacher in the room and a choice of three toys, one of which was a toy dog. Prior to beginning each baseline session, the toy dog and two other toys were brought to the session room. The teacher gathered the following items and placed them on the floor for the session: a toy Koosh ball, a toy car, a stuffed dog, a dog leash, dog bi scuits in a bag, a brush, and a dog toy. The teacher then went to the classroom and walked the child back to the session room. Once in the room, the teacher asked the chil d to be seated in the designated area and the child was prompted: Lets play with the dog today. The teacher waited 10 seconds for the child to initiat e interactions with the toys or with the teacher. A predetermined guideline for interactions was us ed in the sessions (Appendix H: Protocols for Interactions), which include d questions relating to the dog such as What color is the dog?, Do you remember the dogs name? a nd tasks related to the dog such as Give the dog a treat and Brush the dog. If no interactions with the toys or with the teacher were initiated within ten seconds, the teacher asked the child a dog related question from the protocol. The teacher waited 10 seconds for a response. If no response was made, the teacher asked the child to do a task from the protocol. If ther e was still no response, the teacher asked the child the next dog-related que stion from the protocol. Questions and tasks were alternated throughout the session. Once a question or task from the protocol was asked or

PAGE 31

24 offered, the question or task was checked off to ensure the teacher did not ask the same question or task suggestion more than once per session. Criteria to discontinue sessions were changed during the 10 th session for Georgie and during the 4 th session of baseline for Kirsten and Owen. The criteria to terminate a session due to a no response to three consecuti ve dog related questions or task suggestions were dropped. All further sessions were discontinued only if the child engaged in an attempt to leave the room, physical aggression toward the teacher or dog such as grabbing, hitting or kicking, or property destruction. At the end of each sessi on (regardless of the reason for termination), the teacher led the child back to the classroom. Intervention. Following the stabilization of baseline data, the second condition was introduced. The procedures for these sessi ons were identical to the procedures for the baseline condition, with the exception of the presence of a real dog. The dog chosen for this study was a one and a half year old male German Shepherd/Labrador Retriever mix named Arrow. Arrow was obedience traine d and currently enrolled in therapy dog training. He had experience inte racting with children in a sp ecial education classroom for over a year. Prior to beginning each session, the dog was brought to the session room while the children were out of the classroom to avoid disruptions. Inte rvention sessions ended (regardless of the reason for termination) w ith the teacher saying The dog is tired, its time to say goodbye. The teacher then led the child back to the classroom and returned for the dog. The camera was turned off after the child left the room.

PAGE 32

25 When sessions occurred consecutively the dog was given a 10 minute break every 30 minutes to go outside for a drink of water and a short walk. Arrow was not hurt in the conduct of this study. Experimental Design A multiple baseline design across participants was utilized to demonstrate the effects of the presence of a dog on the social interactions of the participants as measured by the dependent variables. Baseline data were taken on all th ree participants. Intervention with participant one began with the stabilizati on of baseline data. Upon the stabilization of intervention for participant one and baseline for participants two and three, intervention was then applied to pa rticipant two. Again following the stabilization of all data, intervention was applied to pa rticipant three (Kazdin, 1982, Parsonson, 2003).

PAGE 33

26 Chapter Three Results Overall Positive Initiated Interactions Figure 1 presents a multiple baseline desi gn across participan ts of overall child initiated positive interactions during baseline and intervention. Positive initiated verbal interactions were low but stable in all three participants during the baseline condition. Positive initiated verbal interactions had a m ean of 1% and a range of 0%-6% for Kirsten, a mean and range of 0% for Georgie and a mean of 4% and a range of 0%-13% for Owen. With Kirsten, positive initiated verbal interactions (mean 4%) did not show an initial increase when the dog was introduced however, as the intervention progressed an increase was noted with little variabil ity for Kirsten (range 0%-16%). Georgie demonstrated an immediate increase in posit ive initiated verbal interactions (mean 24%) also with moderate variability (range 3%-50%). Owens positive initiated verbal interactions (mean 14%) showed an increas e when the dog was introduced as well with some variability (range 3%-28%). Positive initiated non-verbal interactions (mean 11%) were low with some initial variability (range 0%-57%) for Kirsten the baseline condition. Positive initiated nonverbal interactions (mean 33%) were initia lly at a moderate level of occurrence but displaying a downward trend in the baseline condition for Georgie (r ange 0%-65%) and

PAGE 34

F igure 1. Multiple baseline across participants of overall positive interactions -20%0%20%40%60%80%100% BaselineInterventionKirsten -20%0%20%40%60%80%100%Percentage of intervals of positive interactions Georgie -20%0%20%40%60%80%100%12345678910111213141516171819202122Sessions Positive Initiated Verbal Interactions Positive Initiated Non-verbal Interactions Owen 27

PAGE 35

28 positive initiated non-verbal interactions (mean 63%) for Owen were moderate and level (range 38%-75%) during the baseline condition. When the intervention was introduced, pos itive initiated non-verbal interactions increased for all three participan ts with an upward trend noted. Kirstens positive initiated non-verbal interactions increased to a m ean of 72% and with a range of 41%-97%, Georgies positive initiated non-verbal interact ions increased to a mean of 99% and a range of 91%-100% and Owens positive initiated non-verbal interactions increased to a mean of 76% and a range of 59%-88%. Positive Initiated Interactions With the Teacher Figure 2 shows the percentage of positive initiated verbal or nonverbal interactions that occurred between the child and teacher fo r all three participants during baseline and intervention within the sessions. Kirsten demonstrated a low (mean 1%), stable (range 0%-6%) occurrence of positive initiated verbal interactions with the teacher during baseline. During intervention there was a slight increase in the occurrence of positive initiated verbal interactions with the teach er (mean 2%). The range was 0%-13%. She showed a more moderate (mean 11%) occurrence of positive initiated non-verbal interactions with the t eacher in baseline, although there was slightly more variability to the data (range 0%-57%) and during intervention pos itive initiated non-verbal interactions with the teacher increased (mean 22%) with a range of 7%-97%.

PAGE 36

Figure 2. Multiple baseline across participants of positive child initiated interactions directed toward the teacher. -20%0%20%40%60%80%100% BaselineInterventionKirsten -20%0%20%40%60%80%100%Percentage of intervals of positive interactions with the teacher Georgie -20%0%20%40%60%80%100%12345678910111213141516171819202122Sessions Positive Initiated Verbal Interactions with Teacher Positive Initiated Non-verbal Interactions with Teacher Owen 29

PAGE 37

30 Georgie showed a stable baseline condition with a mean and range of 0% for positive initiated verbal interactions with the teacher and a mean of 11% for positive initiated nonverbal interactions with the teacher with a range of 0%-34%. When the intervention condition was introduced at the 10 th session, there was an immediate and sustained increase in positive non -verbal interactions with the teacher (mean 48%). The range was 19%-81%. There was an initial increase in positive initiated verbal interactions with the teacher (mean 4%) as well however, th e occurrence of that behavior declined over the course of the study (range 0%-16%). Owen demonstrated a stable baseline w ith the highest occurrence of positive initiated non-verbal interactions with the teacher (mean 54%) during baseline compared with Kirsten and Georgie. The range was 38%-68%. Owen also demonstrated a stable baseline for positive initiated verbal interactio ns with the teacher (range 0%-9%) and had a mean of 4%. Intervention was introduced at the 12 th session for Owen. Owen demonstrated a level, though slightly lower occurrence of positive initiated nonverbal interactions with the teacher (mean 50%)with little variability (range 38%-59%) and a slightly higher occurrence of positive initiated verbal interact ions with the teacher (mean 11%) with little variability (3%-19%). Positive Initiated Interactions with the Dog Figure 3 shows the percentage of positive verbal or nonverbal interactions that occurred between the child and dog for all three participants during baseline and intervention within intervention sessions.

PAGE 38

Figure 3. Multiple baseline across participants for positive child initiated interactionsdirected toward the dog. -20%0%20%40%60%80%100% BaselineInterventionKirsten -20%0%20%40%60%80%100%Percentage of intervals of positive initiated interactions with the dog -20%0%20%40%60%80%100%12345678910111213141516171819202122Sessions Positive Initiated Verbal Interactions with Dog Positive Initiated Non-verbal Interactions with Dog Owen Georgie 31

PAGE 39

32 Kirsten demonstrated a mean and range of 0% fo r positive initiated verb al interactions with the dog and positive initiated non-verbal intera ctions with the dog during baseline. During intervention there was a slight increase in the occurrence of positive initiated verbal interacti ons with the dog (mean 1%) with little variability (range 0%-6%). Positive initiated non-verbal interactions w ith the dog increased immediately (mean 62%) and although variable (range 7%-97%), cont inued to show an upward trend during the intervention sessions. For Georgie, positive initiated non-verba l interactions with the dog (mean 22%) showed a downward trend in occurrence with va riability (range 0%-48%) during baseline. Positive initiated verbal interactions with the dog had a mean and a range of 0% during baseline. With the introduction of intervention there was an immediate and sustained increase in positive initiated non-verbal interactions with the dog (mean 94%) with little variability (range 56%-100%). There was also an increase in positive initiated verbal interactions with the dog (mean 20%) with a higher level of variability (range 0%-44%) but showing a slight upward trend. Owen demonstrated a downward trend in ba seline with a mean of 12% and a range of 0%-34% for positive initiated non-verbal interactions with the dog and a mean of 0% for positive initiated verbal interactions with the dog with a range of 0%-3%. When the intervention was applied Owen demons trated an increase in positive initiated non-verbal interactions with the dog (mean 47 %) with an upward tr end and a high degree of variability (range 22%-72%). Positive initiated verbal interactions with the dog (mean 3%) slightly increased and remain ed stable (range 0%-6%).

PAGE 40

33 Overall Negative Initiated Interactions Figure 4 shows the percentage of overall ne gative verbal or nonverbal interactions that occurred for all three participants duri ng baseline and intervention within intervention sessions. At baseline Kirsten displa yed a high mean percentage of negative initiated nonverbal interactions (mean 86% and range 35 %-100%) and a low (mean 2%), stable (range 0%-12%) baseline for negative initiated verbal interactions. During intervention negative initiated non-ve rbal interactions declined (mean 16%) and showed a downward trend with some initial variability (range 0%-53%). Negative initiated verbal interactions remained low with a mean and range of 0%. Baseline and intervention conditions for Georgie showed similarly low to zero occurrence of both negative init iated verbal interactions (0% baseline and intervention means with a 0% range for each) and negative initiated non-verbal in teractions (baseline mean 1% range 0%-3%, interventi on mean 0% and range 0%-6%). Owen demonstrated a 0% mean and range in baseline for both negative initiated verbal interactions and negative initiated non-verbal interactions. At the introduction of the intervention, there was an initial increase in both negative initiated verbal interactions (mean 1%, ra nge 0%-9%) and negativ e initiated non-verbal interactions (mean 15% and range 6%-25%) Negative initiated verbal interactions did not maintain that increase and declined back to 0% for the majority of the intervention sessions.

PAGE 41

Figure 4. Multiple baseline across participants of overall negative interactions. -20%0%20%40%60%80%100% BaselineInterventionKirsten -20%0%20%40%60%80%100%Percentage of intervals of negative interactions Georgie -20%0%20%40%60%80%100%12345678910111213141516171819202122Sessions Negative Initiated Verbal Interactions Negative Initiated Non-verbal Interactions Owen 34

PAGE 42

35 Negative Initiated Interactions with the Teacher Figure 5 shows the percentage of negative verbal or nonverbal interactions that occurred between the child and teacher for all three participants during baseline and intervention within intervention sessions. Kirsten showed a high and stable occu rrence of negative initiated non-verbal interactions with the teache r during baseline with a mean of 86% and a range of 35%100%. Negative initiated verbal interactions with the teach er had a mean of 2% and a range of 0%-12%. There was little variability to the baseline data. At the point of intervention negative init iated non-verbal interactions with the teacher (mean 11%) showed an immediate decline with some initial variability but becoming level at near 0% during the last fi ve sessions of the intervention condition. The range was from 0-53% for negative initiated non -verbal interactions with the teacher. Negative initiated verbal interact ions with the teacher declined to a mean of 0%. Negative initiated verbal interactions with the teacher had a range of 0%. Georgie showed very little to no negative in teractions toward his teacher in either baseline or intervention. In baseline, negative initiated non-v erbal interactions with the teacher had a mean of 1% and a range of 0%-3 % and negative initiated verbal interactions with the teacher had a mean of 3% and a range of 0%-3%. In the intervention c ondition, negative initia ted non-verbal inte ractions with the teacher and negative initiated verbal interacti ons with the teacher re mained at a mean of 0% and a range of 0%.

PAGE 43

36 In the baseline condition for Owen, negative initiated non-verbal interactions with the teacher and negative initiate d verbal interactions with the teacher each had a mean of 0% and a range of 0%. When intervention was introduced, there was a slight increase in negative initiated non-verbal interactions with the teacher (mean of 2%) with a range of 0%-6%. Negative initiated verbal interactions with the teacher remained at a mean of 0% with a range of 0%3%. Negative Initiated Interactions with the Dog Figure 6 shows the percentage of negative verbal or nonverbal interactions that occurred between the child and dog for all three participants during baseline and intervention within intervention sessions. Kirs ten had a mean of 0% and a range of 0% for negative initiated non-verbal interactions with the dog a nd negative initiated verbal interactions with the dog during baseline. At the point of interventi on negative initiated non-verba l interactions with the dog increased slightly to a mean of 1% with a range of 0%-33%. Nega tive initiated verbal interactions with the dog remained at a mean and range of 0%. In the baseline condition for Georgie, nega tive initiated non -verbal intera ctions with the dog and negative initiated verbal interacti ons with the dog each had a mean and range of 0%. When intervention was introduced the mean remained at 0% for negative initiated non-verbal interactions with the dog and negati ve initiated verbal in teractions with the dog.

PAGE 44

Figure 5. Multiple baseline across participants of negative child initiated interactions directed toward the teacher. -20%0%20%40%60%80%100% KirstenBaselineIntervention -20%0%20%40%60%80%100%Percentage of intervals of negative initiated interactions with the teacher Georgie -20%0%20%40%60%80%100%12345678910111213141516171819202122Sessions Negative Initiated Verbal Interactions with Teacher Negative Initiated Non-verbal Interactions with Teacher Owen 37

PAGE 45

F igure 6. Multiple baseline across participants of child inititated negative interactions -20%0%20%40%60%80%100%BaselineIntervention Kirsten -20%0%20%40%60%80%100%Percentage of intervals of negative initiated interactions with the dog Georgie -20%0%20%40%60%80%100%12345678910111213141516171819202122Sessions Negative Initiated Verbal Interactions with Dog Negative Initiated Non-verbal Interactions with Dog Owen 38

PAGE 46

39 Negative initiated non-verbal interactions with th e dog had a range of 0%-6% and negative initiated verbal interactions with the dog had a range of 0%. In the baseline condition for Owen, negative initiated non-verbal interactions with the dog and negative initiated verbal interacti ons with the dog each had a mean and range of 0%. When intervention was introduced, there was a slight increase in negative initiated verbal interactions with th e dog with a mean of 1% and a range of 0%-6%. Negative initiated non-verbal interactions with the dog increased to a mean of 13% and a range of 6%-22%. The mean session percentages of each depende nt variable for all three participants in baseline and intervention condi tions are presented in Table 1. Table 1 Mean baseline and intervention session percentages of each dependent variable for each participant. Kirsten Georgie Owen Dependent Variable BL Interv. BL Interv. BL Interv. Positive Initiated Verbal Interactions with Teacher 1% 2% 0% 4% 4% 11% Positive Initiated Non-verbal Interactions with Teacher 11% 22% 11% 48% 54% 50% Positive Initiated Verbal Interactions with Dog 0% 1% 0% 20% 0% 3% Positive Initiated Non-verbal Interactions with Dog 0% 62% 22% 94% 12% 47% N egative Initiated Verbal Interactions with Teacher 2% 0% 0% 0% 0% 0% Negative Initiated Non-verbal Interactions with Teacher 86% 11% 1% 0% 0% 2% N egative Initiated Verbal Interactions with Dog 0% 0% 0% 0% 0% 1% Negative Initiated Non-verbal Interactions with Dog 0% 5% 0% 0% 0% 13%

PAGE 47

40 Figure 7 presents the total number of in tervals (10 seconds each) completed for each session for all three participants. Kirsten did not complete a full session of 32 intervals during baseline. Sessions were terminated for the following reasons: leav ing the room or attempts to leave the room (5 sessions) and property destruction (1 se ssion). The mean length of time Kirsten remained in the room during baseline was 8.83 intervals. During the intervention condition, Kirsten remained in the room for all 32 intervals 6 times. The four sessions that were terminated early for the following reasons: leaving the room or attempts to leave the room (4 sessions). The mean length of time Kirsten remained in the room during intervention was 24.27 intervals. Therefore when the dog was present, Kirstens time spent in the room increased. Georgie remained in the room for all 32 intervals for all sessions except session 5 where he left after 28 intervals for no respons e to three consecutive dog related questions or task suggestions (prior to the change in cr iteria to terminate sessions). Owen remained in the room for all 32 interv als for all sessions. During the sessions, questions and task s uggestions were offered only when there was a 10 second period of no child interacti on with the teacher, dog or toys. Table 2 presents the mean number of questions or task suggestions per session and the mean number of responses to questions or task s uggestions for baseline and intervention for each participant. For Kirsten and Georgie, less que stions or task suggestions were needed to encourage the child to interact when in the presence of the dog. With Owen, more questions or task suggestions were required to enc ourage interaction.

PAGE 48

41 Table 2 Mean number of questions or task suggestions per session for baseline and intervention for each participant. Kirsten Georgie Owen Questions Or Task Suggestions And Responses BL Interv. BL Interv. BL Interv. Mean Number Of Questions Or Task Suggestions Per Session 4.2 3.1 9.1 0.4 0.0 2.3 Mean Number Of Responses To Questions Or Task Suggestions Per Session 0.5 1.3 3.0 0.3 0.0 1.0 Table 3 presents the mean number of positive, negative, verbal and non-verbal responses to questions or task suggestions di rected toward the teacher and dog per session. Table 3 Mean number of positive, negative, ver bal and non-verbal responses to questions or task suggestions directed toward the teacher and dog per session. Kirsten Georgie Owen Verbal and Non-verbal Responses BL Interv BL e Interv. n BL e Interv. n Positive Verbal Responses with Teacher 0.0 0.8 1.2 0.2 N/A 0.7 Positive Non-verbal Responses with Teacher 0.0 0.1 0.7 .08 N/A 0.1 Negative Verbal Responses with Teacher 0.5 0.0 0.1 0.0 N/A 0.1 Negative Non-verbal Responses with Teacher 0.0 0.2 0.1 0.0 N/A 0.0 Positive Verbal Responses with Dog 0.0 0.0 0.1 0.0 N/A 0.0 Positive Non-verbal Responses with Dog 0.0 0.0 0.8 0.0 N/A 0.0 Negative Verbal Responses with Dog 0.0 0.0 0.0 0.0 N/A 0.0 Negative Non-verbal Responses with Dog 0.0 0.2 0.0 0.0 N/A 0.0 Figures 8 and 9 present the teacher assistan ts ratings in the classroom for 30 minutes immediately following each session.

PAGE 49

Table 4 presents the condition means for each dependent variable as rated by the teachers assistant. Table 4 Mean rating of each dependent variable during classroom data collection following each session for each participant. Kirsten Georgie Owen Dependent Variable 1=not at all 3=sometimes 5=most of the time BL Interv. BL Interv. BL Interv. Positive Initiated Verbal Interactions 1.67 2.18 1.56 2.85 1.82 4.00 Positive Initiated Non-verbal Interactions 1.83 2.27 2.89 3.23 2.45 2.71 Positive Verbal Responses 2.83 2.36 3.56 4.31 2.64 3.86 Positive Non-verbal Responses 2.83 3.20 3.00 3.62 2.91 3.71 Negative Initiated Verbal Interactions 3.17 1.91 1.22 1.00 1.73 1.71 Negative Initiated Non-verba l Interactions 3.17 3.45 1.56 1.38 3.00 2.86 Negative Verbal Responses 3.17 2.64 1.11 1.38 2.55 1.86 Negative Non-verbal Responses 3.33 2.40 1.56 1.31 3.45 3.00 Classroom Rating For Positive Interactions Following Sessions Kirsten demonstrated a mean of 1.67 for positive initiated verbal interactions, 1.83 for positive initiated non-verbal interactions, 2.83 for positive verbal responses and 2.83 for positive non-verbal responses in baseline. In the intervention condition, positive initiated verbal interactions increased to a mean of 2.18, positive initiated non-verbal interactions increased to a mean of 2.27, positive verbal responses decreased to a mean of 2.36 and positive non-verbal responses increased to a mean of 3.20. Georgie demonstrated a mean of 1.56 for positive initiated verbal interactions, 2.89 for positive initiated non-verbal interactions, 3.56 for positive verbal responses and 3.00 for positive non-verbal responses in baseline. In the intervention condition, positive initiated 42

PAGE 50

F igure 7. Multiple baseline across participants of intervals completed per session 048121620242832 BaselineInterventionKirsten 048121620242832Number of intervals completed per session Georgie 04812162024283212345678910111213141516171819202122Sessions Owen 43

PAGE 51

Figure 8. Multiple baseline across participants of positive classroom interactions 12345 N ot at allSometimesMost of the time BaselineInterventionKirsten 12345Likert Scale of positive classroom interactions following sessions N ot at allSometimesMost of the time Georgie 1234512345678910111213141516171819202122Sessions Positive Initiated Verbal Interactions Positive Initiated Non-verbal Interactions Positive Verbal Responses Positive Non-verbal Responses N ot at allSometimesMost of the ti me Owen 44

PAGE 52

Figure 9. Multiple baseline across participants of negative classroom interactions following sessions. 12345 N ot at allSometimesMost of the time BaselineInterventionKirsten 12345Likert Scale of negative classroom interactions following sessions N ot at allSometimesMost of the time Georgie 1234512345678910111213141516171819202122Sessions Negative Initiated Verbal Interactions Negative Initiated Non-verbal Interactions Negative Verbal Responses Negative Non-verbal Responses N ot at allSometimesMost of the time Owen 45

PAGE 53

46 verbal interactions increased to a mean of 2.85, positive initi ated non-verbal interactions increased to a mean of 3.23, positive verbal responses increased to a mean of 4.31 and positive non-verbal responses increased to a mean of 3.62. Owen demonstrated a mean of 1.82 for posit ive initiated verbal interactions, 2.45 for positive initiated non-verbal interactions, 2.64 for positive verbal responses and 2.91 for positive non-verbal responses in baseline. In the intervention condition, positive initiated verbal interactions increased to a mean of 4.00, positive initi ated non-verbal interactions increased to a mean of 2.71, positive verbal responses increased to a mean of 3.86 and positive non-verbal responses increased to a mean of 3.71. Classroom Rating For Negative In teractions Following Sessions Kirsten demonstrated a mean of 3.17 for negative initiated verb al interactions, 3.17 for negative initiated non-verba l interactions, 3.17 for negativ e verbal responses and 3.33 for negative non-verbal responses in baseli ne. In the interven tion condition, negative initiated verbal interactions decreased to a mean of 1.91, negativ e initiated non-verbal interactions increased to a m ean of 3.45, negative verbal responses decreased to a mean of 2.64 and negative non-verbal response s decreased to a mean of 2.40. Georgie demonstrated a mean of 1.22 for negative initiated verb al interactions, 1.56 for negative initiated non-verba l interactions, 1.11 for negativ e verbal responses and 1.56 for negative non-verbal responses in baseli ne. In the interven tion condition, negative initiated verbal interactions decreased to a mean of 1.00, negativ e initiated non-verbal interactions decreased to a m ean of 1.38, negative verbal resp onses increased to a mean of 1.38 and negative non-verbal response s decreased to a mean of 1.31.

PAGE 54

47 Owen demonstrated a mean of 1.73 for nega tive initiated verb al interactions, 3.00 for negative initiated non-verba l interactions, 2.55 for negativ e verbal responses and 3.45 for negative non-verbal responses in baseli ne. In the interven tion condition, negative initiated verbal interactions decreased to a mean of 1.71, negativ e initiated non-verbal interactions decreased to a m ean of 2.86, negative verbal resp onses decreased to a mean of 1.86 and negative non-verbal response s decreased to a mean of 3.00. Interobserver Agreement The interobserver agreement for each depe ndent variable was assessed in 65% of the sessions for Kirsten, in 68% of Georgies, and 44% of Ow ens sessions. The mean percent interobserver agreement score for the measured dependent variables for each of the participants ranged from 77% to 100%. Table 5 presents the mean percent observer agreement scores by dependent variable and child. The range of the interobserver agreement sc ores were as follows: positive initiated verbal interactions with the teacher wa s 94%-100% for Kirsten and Georgie and 88%100% for Owen, positive initiated verbal inte ractions with the dog was 94%-100% for Kirsten and Georgie and 97%-100% for Owen, po sitive initiated non-verbal interactions with the teacher was 91%-100% for Kirste n, 22%-100% for Georgie and 41%-100% for Owen, positive initiated non-verbal interactio ns with the dog was 97%-100% for Kirsten, 50%-100% for Georgie and 84%-100% for Owen, positive verbal responses with the teacher was 94%-100% for Kirsten and Georgie and 100% for Owen, positive verbal responses with the dog was 100% for Kirs ten and Owen and 97%-100% for Georgie, positive non-verbal responses with the teach er was 100% for Kirsten and Owen and 97%100% for Georgie, positive non-verbal respon ses with the dog was 100% for Kirsten and

PAGE 55

48 Owen and 88%-97% for Georgie, negative initia ted verbal interactions with the teacher and negative initiated verbal interactions with the dog had a range of 100% for all three participants. Negative initiated non-verbal in teractions with the teacher had a range of 71%-100% for Kirsten and 100% for Georgie and Owen, negative in itiated non-verbal interactions with the dog ha d a range of 100% for Kirsten and Georgie and 94%-100% for Owen, negative verbal responses with the te acher had a range of 88%-100% for Kirsten and 100% for Georgie and Owen, negative verb al responses with the dog, negative nonverbal responses with the teacher and negativ e non-verbal responses with the dog each had a range of 100% for all three pa rticipants. The interobserver ra tings show a wide range with a high mean. The wide range usually reflected one low point in reliability. Table 5 Mean percentage of interobserver agreement scores for each dependent variable for each participant. Dependent Variable Kirsten Georgie Owen Positive Initiated Verbal Interactions wit h Teacher 99% 100% 97% Positive Initiated Non-verbal Interaction s with Teacher 98% 77% 86% Positive Verbal Responses with Teacher 99% 99% 100% Positive Non-verbal Responses with Teacher 100% 99% 100% Positive Initiated Verbal Interactions with Dog 99% 99% 99% Positive Initiated Non-verbal Interaction s with Dog 99% 91% 95% Positive Verbal Responses with Dog 100% 99% 100% Positive Non-verbal Responses with Dog 100% 93% 100% Negative Initiated Verbal Interactions with Teacher 100% 100% 100% Negative Initiated Non-verbal Interactions with Teacher 97% 100% 100% Negative Verbal Responses with Teacher 99% 100% 100% Negative Non-verbal Responses with Teacher 100% 100% 100% N egative Initiated Verbal Interactions wit h Dog 100% 100% 100% Negative Initiated Non-verbal Interaction s with Dog 100% 100% 99% Negative Verbal Responses with Dog 100% 100% 100% Negative Non-verbal Responses with Dog 100% 100% 100%

PAGE 56

49 Social Validity Ratings Table 6 displays the results of the post intervention soci al validity ratings. The social validity data showed that both the teacher and teachers assistant found the intervention to be appropriate, easy to use, and socially signif icant. They also strongly agreed that the intervention was effective a nd led to improved interactions with their teacher. Table 6 Post intervention social validity ratings by the teacher and teachers assistant using a Likert Scale. 1=strongly disagree 2=disagree 3=slightly disagree 4=slightly agree 5=agree 6=strongly agree Teacher Teachers Assistant Appropriateness of Procedures This intervention was easy to use. 5 5 I would recommend this intervention to other educators and parents. 5 6 I liked the procedures used in this intervention. 5 6 Social Significance of the Goals It is important to increase the social responsiveness of students with their teacher. 6 6 It is important to learn new interventions to change the behavior of children with mental retardation. 6 6 It is useful to examine how a childs interactions with a dog can lead to positive outcomes. 5 6 Social Importance of the Effects I would use this intervention in the classroom setting again because it is effective. 6 6 The presence of a dog led to an improvem ent in the social interactions of the children with their teacher. 6 6 This intervention was valuable for the child 6 6

PAGE 57

50 Chapter Four Discussion The purpose of this study was to assess the effects of the presence of a dog on the positive and negative interactions (both ve rbal and non-verbal) of children with developmental disabilities toward their teach er. The study showed that the presence of the dog during sessions increased positive initi ated verbal and non-verbal interactions with the teacher in a ll three participants. The presence of a dog also contributed to an increase in participation in the sessions by one participant who was not participating fully. In addition, when there was a high rate of occurrence of negativ e interactions, those decreased with the intervention. Furthermore, most of the mean ratings within the classroom following the intervention session showed consistent improvement in positive interactions and decrease in negative interact ions within the classroom. In addition, social validity assessment established positive ratings of procedures, goals, and effects in this research. The multiple baseline design across three participants was used to demonstrate the effects and generalization of the treatment in an experimentally controlled manner. The controlled effects were determined by systematically introducing the intervention to different participants, at diffe rent points in time, and showing the changes in behavior demonstrated after intervention. Controlled e ffects were demonstrated with the dependent variables in the intervention setting as we ll as improvements within the generalization setting, the childrens classroom.

PAGE 58

51 The data show that with regard to inter action with the teacher, when interactions were low in frequency during baseline, there wa s an increase in interactions, yet when there was a higher level of interaction (Owen) there was not a large change in positive initiated interactions. With Georgie, there was also a declining trend in positive initiations. It is probable that the level of positive interactions would settle into a stable level of about 2040% although this would need to be docume nted over a longer period of time than was assessed in this study. Although there was an increase in verbal interactions among the participants of this study it should be noted that t hose interactions consisted most ly of one word utterances rather than full sentences. This limited improvement may reflect the verbal deficits associated with the diagnosi s of mental retardation. Negative initiated interactions seen in this study primarily were turning away or moving away from the teacher or dog. In addi tion, Kirsten attempted or actually left the room. During baseline when Kirsten left the room she did not turn back. However, during intervention on one occasion Kirs ten attempted to leave the room, stating mom as she was leaving with the dog and when told the dog could not go with her she left on her own. With regard to overall negative behavior s, Kirsten showed a decrease, Georgie showed no change, and Owen showed an in crease in negatives. Kirstens negative interactions were primarily toward her teacher whereas Owens negative interactions were directed at the dog. Owen was initially both excited and intimidated by the dog. The teacher needed to modify how he interacted with the dog by holding the dog leash and keeping the dog from climbing on or licking Owe n. By the end of the treatment sessions

PAGE 59

52 Owen was very comfortable with Arrow, shown by his holding of the leash and his brushing of the dog. During baseline sessions Owen and Georgie interacted with all the toys while Kirsten did not interact with a ny toys. Baseline data reflects only those interactions directed toward the toy dog, not the other toys. Relation to Literature In comparison with the previous litera ture, this study supports the research by Martin and Farnum (2002) in which pros ocial behaviors were measured under three conditions, one of which was the presence of a real dog. This study differs in design, diagnosis of participants, procedures and operational definitions. Although operational definitions of positive behaviors vary betw een both studies, laughing, giving treats and talking to the dog were included in this study as well. As with Martin and Farnums study, this study demonstrated an increase in those beha viors as well for all participants when in the presence of the dog. Although different designs were used this study was procedurally more similar to the study done by Limond et al.(1997). Limond et. al used a repeated measures design with two conditions, one of which was the presen ce of a real dog. The therapist followed a predetermined guideline for interactions as we ll and the participants in both studies were similar in age and diagnosis. The behaviors of interest in Limond et. als study had some similarities to those in this study including, initiations and re sponses that were rated as positive and negative verbal and non-verbal. Operational definitions varied slightly between both studies. Limond et als study found that the children re sponded non-verbally more often and more positively, they responded to the therapist more frequently and they

PAGE 60

53 initiated verbal interactions more frequently when in the presence of the real dog, which is supported by the current study. Limitations Some limitations became evid ent during this study. Th e data collection system used in this study does not re liably indicate the number of questions/task suggestions offered or responses; however, the teachers pr ocedural checklist was used to obtain data on the number of questions and task suggesti ons offered per session. Some responses and questions occurred during recording interval s therefore although the teacher may have indicated that a question was asked or a ta sk suggestion was offered, the data may not reflect the participants response when that response occurred during the recording interval. A better measure of responsiveness would ha ve been the number of responses per opportunity which might have shown a larger effect. It may be noted, however, that Table 2 shows that the mean of the participants re sponding during interventi on was higher than the responsiveness during baseline. That is, duri ng intervention, fewer que stions were asked, while more questions were answered, whic h is a higher hit rate for questions. Although the teacher did not respond to ch ild-initiated intera ctions to avoid a confounding variable, it is not recommended that the teacher not respond in a real setting as this is not natural and does not reinforce initiated communication. Future studies could control for this potential conf ounding variable by introducing one praise statement for each positive initiated interaction across participan ts and sessions. By reinforcing initiated communication we may have seen a greater effect over time. Furthermore, non-verbal behaviors were more difficult to score. Also, it is recommended that future studies conduct observations of the childre n prior to conducting

PAGE 61

54 the study in order to develop more comprehens ive and inclusive operational definitions of the target behavior. Procedurally, the teacher di d not ask a dog related questi on or task suggestion when the child was playing with any of the toys offered. However, in retrospect, having the teacher ask the questions when the child was in teracting with the toys and neither the toy dog, the real dog nor the therapis t would have provided more of an opportunity to observe positive and negative responses in baseline and intervention. Treatment concerns arising in this study include the carryover e ffects demonstrated with Kirsten. Anecdotally, it appeared as t hough Kirstens time spent in the sessions was influenced by events occurring either before or during her sessions. For example, if Kirsten was reprimanded or went to time out pr ior to a session or heard a preferred activity such as circle time occurring during her session she was more likely to leave the sessions early. Future studies may consider examini ng these context variab les and running sessions in a room further from the classroom or during a free period. The present study also examined the gene ralization of effects across settings (Stokes & Baer, 1977). Ratings in the classr oom showed improvements, although it is important to recognize that systematic obser vations of behavior were not completed. Although unanalyzed in terms of the factor s controlling generalization, there was an important common salient stimulus present in both the intervention setting and the classroom, the teacher. Further evaluation of this discriminative stimulus variable controlling the occurrence of ge neralization is warranted. As a single case experimental design, the generalizability of th ese data would be established by further replication. It is interesting to note that two of the three participants

PAGE 62

55 were children with Down syndrome and the th ird was a child with a hearing impairment, although all were children with retardation. It is possible that the differences between the children relate to the characteristics of the participants but repli cation is important to establish generalizabilit y, as in any one study. Recommendations for Future Research and Practice One parents anecdotal report was that he r child had begun talking much more at home during the intervention condition of this study. Future studies should look at the effects of the intervention and generalization of the behaviors across various settings and times of day. One suggestion is that clas sroom data be evaluated 30 minutes following each session as done in this study and again at the end of the school day to determine how long lasting the effects are following the sessions. Interactions with peers were not studied in this research. An ecdotally, the teacher and teachers assistant reported that they had observed more positive interactions among the children in this study a nd their peers in the classroom including more sharing, talking and positive statements. Another area of recommended future study is to look at interactions with the teacher regarding specific tasks i.e. academics, to determine if the childs academic tasks improve either as a result of the sessions or as a re sult of the increased positive communication with the teacher resulting from this study. The dog can be used to establish stimulus control with the teacher over sessions because the teach er is so actively involved in the therapeutic sessions with the dog. The teacher occasions th e presentation of the dog and the positive consequences of its presence. This may serve to improve interactions between the teacher and child, thus aiding in teaching academic skills. In this manner, generalization from the

PAGE 63

56 treatment session to the classroom can be faci litated as the teacher, as a common salient stimulus, moves with the child from one se tting to another (St okes & Baer, 1977). In a similar manner, it may be valuable to consider how the increase in the interactions with the dog may lead to increased opportunities for th e teacher to provide positive consequences, exemplifying the value of coming into contact with natural communities of reinforcement (Stokes & Baer, 1977). When working with children with developm ental disabilities it is important to discover various ways of teaching them. This study is significant in demonstrating that the presence of a dog can increase communication between a teacher and a child with developmental disabilities. This increas ed communication can then be focused on educational tasks and training. It would be beneficial to use dogs in schools as assistants to the school counselor, psychologist or speech and physical therapists to assist in increasing communication, speech or motor skills. Dogs can also be used as assistants in the classroom in teaching specific tasks such as daily living skills, or as part of a curriculum such as reading, writing, story time, circle time, etc. A dog can act as the subject for creative writing, for reading stories about dogs or can partic ipate with children in group activities with th e dog being counted as a member of the group. This may increase participation for the children in some activities. It may not be beneficial to have a do g present throughout the school day as this would be exhausting for the dog a nd disruptive to the children. Conclusion There are very few studies to date on th e effects of dogs on so cial interactions. Much of the current literature is anecdotal in nature. This study supports prev ious findings

PAGE 64

57 that the presence of a dog can increase co mmunication and positive non-verbal behaviors which will enable children with developmental disabilities to recruit reinforcement from their natural environments. This study suggests that children with developmental disabilities may greatly benefit from the use of dogs as teaching assistants and adjuncts to therapy.

PAGE 65

58 References Eddy, J., Hart, L. & Boltz, R. (1987). Th e effects of service dogs on social acknowledgement of people in wheelchairs. Journal of Psychology, 122(1) 3945. Crowley-Robinson, P., Fenwick, D.C. & Black shaw, J.K. (1996). A long-term study of elderly people in nursing homes with visiting and resident dogs. Applied Animal Behaviour Science 47, 137-148. Fick, M. (1993). Influence of an animal on so cial interactions of nursing home residents in a group setting. American Journal of Occupational Therapy 47 (1), 529-534. Hanson, K.M., Messinger, C.J., Baun, M. & Me gel, M. (1999). Companion animals alleviating distress in children. Anthrozoos 12(3), 142-148. Hart, L.A., Hart, B.L. & Bergin, B (1987). So cializing effects of service dogs for people with disabilities. Anthrozoos 1(1), 41-44. Hart, L.A., Zasloff, R.L. & Benfatto, A.M. ( 1996). The socializing ro le of hearing dogs. Applied Animal Behaviour Science, 47 7-15. Hunt, S., Hart, L. & Gomulkiewicz, R. ( 1991). Role of small animals in social interactions with strangers. The Journal of Social Psychology 132(2), 245-256. Kaminski, M., Pellino, T. & Wish, J. (2002). Play and pets: the physical and emotional impact of child-life and pet th erapy on hospitalized children. Childrens Health Care 31(4), 321-335. Kazdin, A.E. (1982). Single-case research desi gns. New York: Oxford University Press. Lane, D.R., McNicholas, J. & Collis, G.M. (1998). Dogs for the disabled: benefits to recipients and welfare of the dog. Applied Animal Behaviour Science 59, 49-60. Law, S., & Scott, S. (1995). Tips for practitioners. Focus on Autistic Behavior 10(2), 1718. Limond, J.A., Bradshaw, J.W.S. & Cormack, K.F. M. (1997). Behavior of children with learning disabilities interact ing with a therapy dog. Anthrozoos, 10 (2/3), 84-89. Mader, B., Hart, L.A. & Bergin, B. (1989). Social acknowledgement for children with disabilities: effects of service dogs. Child Development, 60, 1529-1534.

PAGE 66

59 Marr, C.A, French, L., Thompson, D., Drum, L., Greening, G., Morman, J., Henderson, I. & Hughes, C.W. (2000). Animal assisted therapy in psychiatric rehabilitation. Anthrozoos, 13(1), 43-47. Martin, F. & Farnum, J. (2002). Animal a ssisted therapy for children with pervasive developmental disorders. Western Journal of Nursing Research 24(6), 657-670. McNicholas, J. & Collis, G. (2000). Dogs as catalysts for social in teractions: robustness of the effect. British Journal of Psychology 91, 61-70. Naderi, S., Miklosi, A., Doka, A. & Csanyi V. (2001). Co-operative interactions between blind persons and their dogs. Applied Animal Behaviour Science 74, 5980. Odendaal, J.S.J. (2000). Animal assi sted therapy-magic or medicine? Journal of Psychosomatic Research 49; 275-280. Parsonson, B.S. (2003). Visual analysis of graphs: seeing is believing. In Budd, K. and Stokes, T. (Eds.). A Small Matter of Proof: The Legacy of Donald M. Baer (pp. 35-51). Reno, Nevada: Context Press. Perelle, I.B. & Granville, D.A, (1993). Assessment of the effectiveness of a pet facilitated therapy program in a nursing home setting. Society and Animals 1(1), 91-100. Stacy, M. Dogs offer academic motivation to students. (2003, December 25). Tampa Tribune pp. Metro 3. Stokes, T., and Baer, D.M. (1977). An implicit technology of generalization. Journal of Applied Behavior Analysis, 10 349-367. Walsh, P.G. & Mertin, P.G. (1994). The training of pets as therapy in a womens prison: A pilot study. Anthrozoos, 7(2), 124-128. Wilson, C. (1987). Physiological respons es of college students to a pet. The Journal of Mental and Nervous Disease, 175(10), 606-612. Wolf, M.M (1978). Social Validity: The case for subjective measurement or how applied behavior analysis is finding its heart. Journal of Applied Behavior Analysis 11 203-214.

PAGE 67

60 Appendices

PAGE 68

61 Appendix A: Baseline Session Guidelines Prior to Beginning Each Session 1. The toy dog and two other toys will be brought to the session room. 2. The teacher will gather the following items and place them on the floor for the session: a toy dog leash, bisc uits, brush, and dog toy. 3. The teacher will then go to the classroom and walk the child back to the session room. During the Session 1. Once in the room the teacher will ask the child to be seated in the designated area and the child will be prompted: L ets play with the dog today. 2. The teacher will wait 10 seconds for the child to initiate interac tions with the toy dog or with the teacher. 3. If no interactions with th e toy dog or with the teache r are initiate d within ten seconds, the teacher will ask the child a dog related question from the Protocols for Interactions (Appendix F). 4. The teacher will wait 10 seconds for a response. 5. If no response is made the teacher w ill ask the child to do a task from the Protocols for Interactions (Appendix F) with the dog. 6. If there is still no response the teacher will ask the child the next dog-related question from the protocol. 7. If the child does not respond the session will be terminated. 8. Questions and tasks will be alternated throughout the session. Once a question or task from the Protocol for Interactions (Appendix F) has been asked or offered, the question or task will be checked off to ensure the teacher does not ask the same question or task suggestion mo re than once per session. Ending the Session 1. Sessions will be discontinued if the child engages in any inappropriate behaviors such as attempts to leave the room, phys ical aggression toward the teacher or dog such as grabbing, hitting or kicking, propert y destruction or if the child does not respond to three consecutive dog related questions or task suggestions. 2. The teacher will end the session by saying The dog is tired, its time to say goodbye. 3. The teacher will then lead the child back to the classroom. 4. The camera will be turned off after the child has left the room.

PAGE 69

62 Appendix B: Intervention Session Guidelines Prior to Beginning Each Session 1. The dog will be brought to the session r oom, while the children are out of the classroom to avoid disruptions. 2. The teacher will gather the following ite ms and place them on the floor for the session: a dog leash, biscuits, brush, and dog toy. 3. The teacher will then go to the classroom and walk the child back to the session room. During the Session 1. Once in the room the teacher will ask the ch ild to be seated in the designated area and the child will be prompted: L ets play with the dog today. 2. The teacher will wait 10 seconds for the child to initiate interac tions with the dog or with the teacher. 3. If no interactions with the dog or with the teacher are initiated, the teacher will ask the child a dog related question from the Protocols for Interactions (Appendix F). 4. The teacher will wait 10 seconds for a response. 5. If there is still no response the teacher w ill ask the child the ne xt dog-related question from the protocol. 6. If the child does not respond the session will be terminated. 7. Questions and tasks will be alternated throughout the session. Once a question or task from the Protocol for Interactions (Appendix F) has been asked or offered, the question or task will be checked off to ensure the teacher does not ask the same question or task suggestion more than once per session. Ending the Session 1. Sessions will be discontinued if the child engages in any inappropr iate behaviors such as attempts to leave the room, physical aggression toward the teacher or dog such as grabbing, hitting or kicking, property destruction or if the child does not respond to three consecutive dog related ques tions or task suggestions. 2. The teacher will end the session by saying The dog is tired, its time to say goodbye. 3. The teacher will then lead the child back to the classroom and return for the dog. The camera will be turned off after the child has left the room. When sessions will occur consecutively the dog will be given a 10 minute break every 30 minutes to go outside for a drink of water a nd a short walk. When sessions are completed for the day, the dog will be brought into the clas sroom for all of the children to play with for 10 minutes. He will then be brought outside for a walk and some water and returned to an area with no children (such as the session room) to rest.

PAGE 70

63 Appendix C: Baseline Session Guidelines Checklist Date:______________ Session #:_______________ Rater:________________ + correct incorrect Prior to Beginning Each Session 1. The toy dog and two other toys will be brought to the session room. 2. The teacher will gather the following items and place them on the floor for the session: a toy dog leash, bisc uits, brush, and dog toy. 3. The teacher will then go to the classroom and walk the child back to the session room. During the Session 1. Once in the room the teacher will ask the child to be seated in the designated area and the child will be prompted: L ets play with the dog today. 2. The teacher will wait 10 seconds for the child to initiate interac tions with the toy dog or with the teacher. 3. If no interactions with the toy dog or with the teacher ar e initiated, the teacher will ask the child a dog related question from th e Protocols for Interactions (Appendix F). 4. The teacher will wait 10 seconds for a response. 5. If there is still no response the teacher will ask the child the next dog-related question from the protocol. 6. If the child does not respond the session will be terminated. 7. Questions and tasks will be alternated throughout the session. Once a question or task from the Protocol for Interactions (Appendix F) has been asked or offered, the question or task will be checked off to ensure the teacher does not ask the same question or task suggestion mo re than once per session. Ending the Session 1. Sessions will be discontinued if the child engages in any inappropriate behaviors such as attempts to leave the room, phys ical aggression toward the teacher or dog such as grabbing, hitting or kicking, propert y destruction or if the child does not respond to three consecutive dog related questions or task suggestions. 2. The teacher will end the session by saying The dog is tired, its time to say goodbye. 3. The teacher will then lead the child back to the classroom. 4. The camera will be turned off after the child has left the room. ________% Proficiency Proficiency = number of correct number of correct plus incorrect 100 = _____%

PAGE 71

64 Appendix D: Intervention Session Guidelines Checklist Date:______________ Session #:_______________ Rater:________________ + correct incorrect Prior to Beginning Each Session 1. The dog will be brought to the sessio n room, while the children are out of the classroom to avoid disruptions. 2. The teacher will gather the following items and place them on the floor for the session: a dog leash, bi scuits, brush, and dog toy. 3. The teacher will then go to the clas sroom and walk the child back to the session room. During the Session 1. Once in the room the teacher will as k the child to be seated in the designated area and the child will be prompted: Lets play with the dog today. 2. The teacher will wait 10 seconds for the child to initiate interactions with the dog or with the teacher. 3. If no interactions with the dog or with the teacher are initiated, the teacher will ask the child a dog related question from the Protocols for Interactions (Appendix F). 4. The teacher will wait 10 seconds for a response. 5. If there is still no response the t eacher will ask the child the next dogrelated question from the protocol. 6. If the child does not respond the session will be terminated. 7. Questions and tasks will be alternated throughout the session. Once a question or task from the Protocol for Interactions (Appendix F) has been asked or offered, the question or task will be checked off to ensure the teacher does not ask the same question or task suggestion more than once per session. Ending the Session 1. Sessions will be discontinued if the child engages in any inappropriate behaviors such as attempts to leave the room, physical aggression toward the teacher or dog such as grabbing, hitting or kicking, property destruction or if the ch ild does not respond to three consecutive dog related questions or task suggestions. 2. The teacher will end the session by saying The dog is tired, its time to say goodbye. 3. The teacher will then lead the child back to the classroom and return for the dog. The camera will be turned off after the child has left the room.

PAGE 72

65 Appendix D continued 4. When sessions will occur consecu tively the dog will be given a 10 minute break every 30 minutes to go out side for a drink of water and a short walk. Appendix D: Treatment Session Guidelines Checklist continued 5. When sessions are completed for the day, the dog will be brought into the classroom for all of the children to play with for 10 minutes. He will then be brought outside for a walk and some water and returned to an area with no children (such as the session room) to rest. ________% Proficiency Proficiency = number of correct number of correct plus incorrect 100 = _____%

PAGE 73

66 Appendix E: Session Data Collection Date:_________________ Participant:______________________________ Sessi on #:____________________ Rater:__________________ ____ = Child-initiated Responses X= Teacher-prompted Responses D= Child-initiated behaviors toward the dog Xd= Teacher-prompted responses toward dog Record if the behavior occurred at any time durin g the interval. Interval size is 10 seconds. Interval# 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Positive Verbal Statements Negative Verbal Statements Positive Nonverbal Behaviors Negative Nonverbal Behaviors Interval # 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 Positive Verbal Statements Negative Verbal Statements Positive Nonverbal Behaviors Negative Nonverbal Behaviors 1. Positive verbal statements = statements indicating enjoyment or interest in the situation ex ample: that was fun, I like it, more, can I stay?, and I love the dog or requests for help. 2. Negative verbal statements = statements indicating disinterest of a lack of enjoyment in th e situation example: This sucks, I hate this, Can I leave? , I hate the dog, Get me out of here or refusals I dont want to, No, no more. 3. Positive nonverbal behaviors =smiling, laughing, touching the dog (i.e. petting, hugging or kissing, et cetera.), clapping hands nodding head, complying wit h a request nonverbally and social agreement uh-huh, sharing or handing things to teacher, holding the leash or walking the dog. Negative nonverbal behaviors = turning body or face away from the teacher, crying, frowning (corners of lips turned down, hiding face, attempts to leave the r oom, physical aggression, prope rty destruction (throwing things, knocking things off shelves/table, playing with computer if these actions would caus e damage if uninterrupted) or threats or no response to dog-re lated questions or task suggestions.

PAGE 74

67 Appendix F. Observer Proficiency Observer name:___________________ Date:_______________ + correct incorrect Observer correctly identified examples of the following during role play: 1. ____ positive verbal behavior 2. ____ positive verbal behavior 3. ____positive non-verbal behaviors 4. ____positive non-verbal behaviors 5. ____negative verbal behaviors 6. ____negative verbal behaviors 7. ____negative nonverbal behaviors 8. ____negative nonverbal behaviors 9. ____initiations 10. ____initiations 11. ____responses 12. ____responses Proficiency = number of correct number of correct plus incorrect 100 = _____%

PAGE 75

68 Appendix G: Protocols for Interactions Check off each item as it is used. Do not repeat items in the same session. Dog Related Questions 1. ____Do you know the dogs name? 2. ____What color is the dog? 3. ____Is the dog a boy or girl? 4. ____How many legs does the dog have? 5. ____What color is the dogs collar? 6. ____What color is the dogs leash? 7. ____What color eyes does the dog have? 8. ____How does the dogs hair feel? 9. ____How old is the dog? 10. ____Does the dog look happy? 11. ____Is the dog big or small? 12. ____What would you call the dog if he were yours? 13. ____What does the dogs toy look like? 14. ____Does the dog want to play? 15. ____Does the dog want to eat? 16. ____Was the dog good today? 17. ____Do you like the dog?

PAGE 76

69 Appendix G: Protocols for Interactions continued Dog Related Task Suggestions 1. ____Can you call the dog? 2. ____Can you please give the dog a treat? 3. ____Can you shake hands with the dog? 4. ____Can you put the dogs leash on him? 5. ____Can you give the dog his toy? 6. ____Can you play with the dog? 7. ____Can you brush the dogs hair? 8. ____Can you tell the dog to sit? 9. ____Can you pet the dog? 10. ____Can you give the dog a hug?

PAGE 77

Appendix H: Classroom Data Collection Date:____________________________ Participant:______________________________ Session #:________________________ Rater:__________________________________ Time began:______________________ Time ended:_____________________________ Please answer the following questions 30 minutes after the child has completed a session and has reintegrated into the classroom. 1. Did the child initiate positive verbal statements? 1 2 3 4 5 not sometimes most at all of the time 2. Did the child initiate negative verbal statements? 1 2 3 4 5 not sometimes most at all of the time 3. Did the child initiate positive nonverbal behaviors? 1 2 3 4 5 not sometimes most at all of the time 4. Did the child initiate negative nonverbal behaviors? 1 2 3 4 5 not sometimes most at all of the time 5. Did the child respond to a request or answer a question when asked using positive verbal statements? 1 2 3 4 5 not sometimes most at all of the time 70

PAGE 78

Appendix H: Classroom Data Collection continued Date:____________________________ Participant:______________________________ Session #:________________________ Rater:__________________________________ Time began:______________________ Time ended:_____________________________ Please answer the following questions 30 minutes after the child has completed a session and has reintegrated into the classroom. 6. Did the child respond to a request or answer a question when asked using negative verbal statements? 1 2 3 4 5 not sometimes most at all of the time 7. Did the child respond to a request or answer a question when asked using positive nonverbal behaviors? 1 2 3 4 5 not sometimes most at all of the time 8. Did the child respond to a request or answer a question when asked using negative nonverbal behaviors? 1 2 3 4 5 not sometimes most at all of the time Positive Verbal Statements =statements indicating enjoyment or interest in the situation example: that was fun, I like it, more, can I stay?, and I love the dog. Negative Verbal Statements= statements indicating disinterest of a lack of enjoyment in the situation example: This sucks, I hate this, Can I leave?, I hate the dog, Get me out of here or refusals I dont want to, No, no more. Positive Nonverbal Behaviors=smiling, laughing, touching the dog (i.e. petting, hugging or kissing, et cetera.), clapping hands, nodding head, complying with a request non-verbally and social agreement uh-huh Negative Nonverbal Behaviors=turning body or face away from the teacher, crying, frowning (corners of lips turned down), hiding face, attempts to leave the room, physical aggression, property destruction or threats or refusals to respond to dog-related questions or task suggestions. 71

PAGE 79

72 Appendix I: Social Validity Name:_____________________ Date:______________________ 1=strongly disagree 2=disagree 3=slightly disagree 4=slightly agree 5=agree 6=strongly agree Appropriateness of Procedures 1. This intervention was easy to use. 1 2 3 4 5 6 2. I would recommend this intervention to other educators and parents. 1 2 3 4 5 6 3. I liked the procedures used in this intervention. 1 2 3 4 5 6 Social Significance of the Goals 4. It is important to increase the social re sponsiveness of students with their teacher. 1 2 3 4 5 6 5. It is important to learn new interventions to change the behavi or of children with mental retardation. 1 2 3 4 5 6 6. It is useful to examine how a childs in teractions with a dog can lead to positive outcomes. 1 2 3 4 5 6 Social Importance of the Effects 7. I would use this intervention in the classr oom setting again because it is effective. 1 2 3 4 5 6 8. The presence of a dog led to an improveme nt in the social interactions of the children with their teacher. 1 2 3 4 5 6 9. This intervention was valuable for the child. 1 2 3 4 5