Evaluations of Placement and Activity Preferences in Foster Care

MISSING IMAGE

Material Information

Title:
Evaluations of Placement and Activity Preferences in Foster Care
Physical Description:
1 online resource (61 p.)
Language:
english
Creator:
Whitehouse, Cristina M
Publisher:
University of Florida
Place of Publication:
Gainesville, Fla.
Publication Date:

Thesis/Dissertation Information

Degree:
Doctorate ( Ph.D.)
Degree Grantor:
University of Florida
Degree Disciplines:
Psychology
Committee Chair:
Vollmer, Timothy R
Committee Members:
Iwata, Brian
Graber, Julia A
Diehl, David C.

Subjects

Subjects / Keywords:
child -- preference -- welfare
Psychology -- Dissertations, Academic -- UF
Genre:
Psychology thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract:
The purpose of these studies is to extend the use of preference assessment methods to child welfare using a web-based preference assessment system. In Study 1, subjects completed a computerized 4-point likert-type questionnaire designed to assess preference for a wide range of activities and community characteristics. Next, stimuli identified as high preferred (HP) and low preferred (LP) in the questionnaire were tested using a computerized paired-stimulus (PS) preference assessment. Results showed complete correspondence between the results of the computerized preference assessment methods for 9 of 15 subjects, 5 subjects showed correspondence for 8 (of 10) stimuli, and one subject showed correspondence for 6 stimuli (of 10). Studies 2 and 3 evaluated whether the stimuli identified as HP and LP in Study 1 could function as reinforcers. Specifically, Study 2 tested response allocation and engagement and Study 3 tested whether stimuli identified in the survey could be used as reinforcers for math problem completion. Results demonstrated that subjects allocated their engagement to stimuli identified as HP in the survey and those HP stimuli functioned as reinforcers for math problem completion. Collectively, these studies demonstrated that the web-based system designed for this study is a feasible approach to quickly identifying preferences among children in the child welfare system. Other features of the computerized system that attempt to match children and foster parents are discussed.
General Note:
In the series University of Florida Digital Collections.
General Note:
Includes vita.
Bibliography:
Includes bibliographical references.
Source of Description:
Description based on online resource; title from PDF title page.
Source of Description:
This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility:
by Cristina M Whitehouse.
Thesis:
Thesis (Ph.D.)--University of Florida, 2012.
Local:
Adviser: Vollmer, Timothy R.
Electronic Access:
RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2013-08-31

Record Information

Source Institution:
UFRGP
Rights Management:
Applicable rights reserved.
Classification:
lcc - LD1780 2012
System ID:
UFE0044504:00001


This item is only available as the following downloads:


Full Text

PAGE 1

1 EVALUATIONS OF PLACEMENT AND ACTIVITY PREFERENCE S IN FOSTER CARE By CRISTINA M. WHITEHOUSE A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2012

PAGE 2

2 2012 Cristina M. Whitehouse

PAGE 3

3 To my Dad, Roni and Colin

PAGE 4

4 ACKNOWLEDGMENTS I would like to express my sincere gratitude to my advisor, Dr. Timothy Vollmer, for all of his immeasurable support and guidance. I would also like to thank th e members of my committee, Drs. David Diehl, Julia Graber, and Brian Iwata for their editorial comments and assistance guiding this research. This project would not have been possible without the support and enthusiasm given by Bennie Colbert and Heartland for Children, especially Eva Horner, Terri Saunders, Kathleen Cowan, Jessica McClemore Corely, Angela Stills, and Vanessa McCottry. Additional thanks are given to Nem Nesic for all his inval uable programming. A special thank you to my undergraduate research volunteers, Danielle Willis and Kelly Benhart, for all their assistance with data analysis. I would also like to thank my mother, Rosalia M. Whitehouse, my sister, Joan M. Whitehouse, for their unwavering love, support, and encouragement. Finally, to Roni, Colin, Tukey, Erica, and Amy, this process would not have been as colorful and enjoyable without you.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 7 LIST OF FIGURES ................................ ................................ ................................ .......... 8 ABSTRACT ................................ ................................ ................................ ..................... 9 CHAPTER 1 GENERAL INTRODUCTION ................................ ................................ .................. 11 2 GENERAL METHOD ................................ ................................ .............................. 18 Subjects and Set ting ................................ ................................ ............................... 18 Placement Preference Assessment (PPA) System ................................ ................ 18 Dependent Variables ................................ ................................ .............................. 19 3 STUDY 1: EVALUATION OF CHILD PREFERENCES ................................ ........... 20 Introduction ................................ ................................ ................................ ............. 20 Method ................................ ................................ ................................ .................... 20 Subjects ................................ ................................ ................................ ............ 20 Procedure ................................ ................................ ................................ ......... 20 Results ................................ ................................ ................................ .................... 23 Discussion ................................ ................................ ................................ .............. 24 4 STUDY 2: AN EVALUATION OF CORRESPONDENCE BETWEEN CHILD SURVEY RESPONSES AND ENGAGEMENT ................................ ....................... 40 Introduction ................................ ................................ ................................ ............. 40 Method ................................ ................................ ................................ .................... 40 Subjects and Settings ................................ ................................ ....................... 40 Data Collection and Analysis ................................ ................................ ............ 40 Dependent Variables ................................ ................................ ........................ 41 Procedure ................................ ................................ ................................ ......... 41 Results ................................ ................................ ................................ .................... 42 Discussion ................................ ................................ ................................ .............. 43 5 STUDY 3: REINFORCER ASSESSMENT ................................ .............................. 47 Introduction ................................ ................................ ................................ ............. 47 Method ................................ ................................ ................................ .................... 47 Subjects and Settings ................................ ................................ ....................... 47

PAGE 6

6 Dependent Variables ................................ ................................ ........................ 47 Data Collection and Analysis ................................ ................................ ............ 47 Procedure ................................ ................................ ................................ ......... 48 Results ................................ ................................ ................................ .................... 49 Discussion ................................ ................................ ................................ .............. 50 6 GENERAL DISCUSSION ................................ ................................ ....................... 54 LIST OF REFERENCES ................................ ................................ ............................... 58 BIOGRAPHICAL SKETCH ................................ ................................ ............................ 61

PAGE 7

7 LIST OF TABLES Table page 3 1 Child Questionnaire ................................ ................................ ............................ 28

PAGE 8

8 LIST OF FIGURES Figure page 3 1 Screenshot of computerized child quest ionnaire presentation ........................... 29 3 2 Screenshot of computerized PS preference assessment presentation .............. 30 3 3 Data for Lauren and Chris ................................ ................................ .................. 31 3 4 Data for Grant and Ana. ................................ ................................ ..................... 32 3 5 Data for Mari and Alex ................................ ................................ ........................ 33 3 6 Data for Kayla and Jack. ................................ ................................ .................... 34 3 7 Data for Lanie ................................ ................................ ................................ .... 35 3 8 Data for Diana and Keith ................................ ................................ .................... 36 3 9 Data for Cameron and Mia ................................ ................................ ................. 37 3 10 Data for Sam ................................ ................................ ................................ ...... 38 3 11 Data for Rex. ................................ ................................ ................................ ...... 39 4 1 Engagement data ................................ ................................ .............................. 46 5 1 Rate of math problem completion ................................ ................................ ....... 53

PAGE 9

9 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy EVALUATIONS OF PLACEMENT AND ACTIVITY PREFERENCES IN FOSTER CARE By Cristina M. Whitehouse August 2012 Chair : Timothy Vollmer Major: Psychology The purpose of these studies is to extend the use of preference assessment methods to child welfare using a web based preference assessment system. In Study 1, subjects completed a computerized 4 point likert type questionnaire designed to assess preference for a wide r ange of activities and community characteristics. Next, stimuli identified as high preferred (HP) and low preferred (LP) in the questionnaire were tested using a computerized paired stimulus (PS) preference assessment. Results showed complete correspondenc e between the results of the computerized pref erence assessment methods for 9 of 15 subjects, 5 subjects showed correspondence for 8 (of 10) stimuli, and one subject showed corresp ondence for 6 stimuli (of 10). Studies 2 and 3 evaluated whether the stimuli identified as HP and LP in Study 1 could function as reinforcers. Specifically, Study 2 tested response allocation and engagement and Study 3 tested whether stimuli identified in the survey could be used as reinforcer s for math problem completion. Results demonstrated that subjects allocated their engagement to stimuli identified as HP in the survey and those HP stimuli functioned as reinforcers for math problem completion. Collectively, these studies demonstrated that the web based system designed for thi s study is a feasible approach to quickly identifying preferences

PAGE 10

10 among children in the child welfare system. Other features of the computerized system that attempt to match children and foster parents are discussed.

PAGE 11

11 CHAPTER 1 GENERAL INTRODUCTION A critical feature of behavioral interventions is the selection of stimuli that could function as reinforcers. Reinforcer identification is especially difficult if the client has impaired communication skills. Such challenges have led researchers in applied behavior analysis to develop a variety of stimulus preference assessments (e.g., DeLeon & Iwata, 1996; Fisher, Piazza, Bowman, Hagopian, Owens, & Sle vin, 1992; Matson, Bielecki, Mayville, Smalls, Bamburg, & Baglio, 1999; Pace, Ivancic, Edwards, Iwata, & Page, 1985; W indsor, Piche, & Locke, 1994). It is now standard to use one of many such assessments prior to developing an intervention for individuals with intellectual disabilities (ID). In the current series of studies, we sought to develop and partially validate a form of stimulus preference assessment for eventual use in the foster care system. Children enter foster care because of abuse, violence, or neglect (Leslie, Gordon, Lambros, Premji, Peoples, Gist, 2005). Once in care, a child is often placed into a foster home where little is known about the child (Brian, 1963; Lindsay, 1991; R edding, Fried, Britner, 2001). Identifying preferences for children in foster care would serve several potential purposes. One, it could aide behavior analysts, other professionals, and foster parents in identifying reinforcers for person al care and academic programs. Two, it could help to identify activities and stimuli that would make the foster placement more transition and stabilizing the foster placement is a primary goal for the care agency and foster family (Simon & Simon, 1982) Three, it is possible that children at risk for losing placement would be less likely to engage in behavior resulting in a placement change;

PAGE 12

12 research has in fact shown that a child who is content with his or her placement will abstain from engaging in be havior that places them at risk for placement disrupt ion (Sinclair & Wilson, 2003). However, before these potential advantages of stimulus preference assessments can be evaluated, a preference assessment method must be developed, implem ented, and validated Thus, this series of studies will represent a beginning point for examining the utility of preference assessments in child welfare. Before discussing these possibilities further, a brief review of stimulus preference assessments will be provided. Pace e t al., (1985) established one method for assessing preference using a single stimulus (SS) presentation m ethod for individuals with ID. In Study 1 of this seminal paper, 16 stimuli were presented 10 times each, across 8 assessment sessions. Each session co nsisted of 20 trials during which 4 stimuli were presented in isolation 5 times in a counterbalanced order. If the subject approached the stimulus on a given trial within 5 s of presentation, the individual could interact with the item for 5 s. The results showed that subjects demonstrated differential preference between the stimuli, indicating that this manner of assessing preference could be useful in indentify preferred stimuli for individuals with profound ID. However, the results from Pace et al. (1985 ) Study 1 did not necessarily indicate that the stimuli identified as high preferred could function as effective reinforcers for anything other than an approach response. Thus, Pace et al. Study 2 aimed to test whether the stimuli identified in the SS asse ssment could function as reinforcers for a variety of adaptive skills. The authors first defined preferred items as stimuli selected in at least 80% of the trials and non preferred stimuli as stimuli approached in 50% or

PAGE 13

13 fewer of the trails. Using an ABCAB C design, Pace et al tested whether preferred and non preferred stimuli could increase the individually identified adaptive behavior (e.g., loo king, reaching, raising hand). Results indicated that preferred stimuli increased the number of target responses for 5 of the 6 subjects compared to baseline and compared to the non preferred stimulus condition. Overall, the results demonstrated a clear, practical method for assessing preferred stimuli for low functioning individuals. The study also set a course for future re search on stimulus preference. One putative limitation of this study, however, was that relative preference to the array of stimuli presented was not discernible (Fisher et al., 1992). Subsequently, Mason, McGee, Farmer Dougan, and Risley (1989) and Fisher et al. (1992) evaluated relative preference among an array of stimuli by using a c oncurrent operant arrangement. In this arrangement, two stimuli are presented simultaneously and a subject is given the opportunity to select one stimulus. By pair ing each stimulus in this manner, a hierarchy of preference among the stimuli was obtained (Fisher et al.). Concurrent operant arrangements are derived from basic operant experiments that examine choice and preference (e.g., Her rnstein, 1961). In behavior analysis, measuring allocation of responding across two alternatives essentially defines choice of two or more alternative responses, derived from the relative frequenc ies of the responses over an extende Clinically, this is an important conceptualization because individuals must allocate behavior among multiple, concurrently available stimuli (Fisher & Mazur, 1 997). Thus, Fisher et al. com pared SS presentation (e.g., Pace et al., 1985) to a paired stimulus (PS) presentation method.

PAGE 14

14 In the first part of Fisher et al. (1992), subjects completed a SS preference assessment replicating Pace et al. (1985). Next, the subjects were tested using the same stimuli in a PS assessment format. Specifically, the experimenters paired each stimulus once with every other stimulus. A trial began with two stimuli p laced in front of the subject. Approach to one of the stimuli resulted in 5 s access to that stimu lus and t he other stimulus was removed. Similar to Pace et al., the last phase was a reinforcer test. Stimuli tested were those items identified as high preferred in both the SS and PS assessments and stimuli that showed disagreement between the two assess ment methods (i.e., stimuli identified as high preferred in the SS assessment but low preferred in the PS assessment). Overall, the results indicated that the PS method produced greater differentiation among the stimuli tested and the SS method tended to s how more items identified as hig h preferred (false positives). Results also indicated that the PS assessment was more effective at indicating which stimuli could function as reinforcers. Stimuli selected as low as 60% of the presentations were shown to fun ction as effective reinforcers. One major limitation, however, was that the PS format was more time intensive (i.e., required 120 pairings) compared to the SS assessment method, and it did not rule out the possibility that the HP stimuli from the PS would function as effective reinforcers when used in isolation (Roscoe, Iwata, and Kahng, 1999). Since the publications of Pace et al (1985) and Fisher et al (1992), researchers have identified numerous extensions and variations to assess stimulus preference (e .g., DeLeon et al, 1996; Roane et al., 1998). Notably, researchers have demonstrated that stimulus preference assessments identified preferred items more effectively than

PAGE 15

15 caregiver nominated stimuli (Green, Reid, Canipe, Gardener, 1991 ). Additionally, Nort hup, George, Jones, Broussard, and Vollmer (1996) extended the use of preference assessments with children diagnosed with attention deficit hyperactivity disorder (ADHD) whose verbal communication was not impaired. Often, correspondence elf report and subsequent allocation of behavior is poor (e.g., Risley & Hart, 1969); thus, Northup et al. examined the use of verbal and pictorial stimulus assessments after the subjects self reported their preferences using a Likert type survey. They als o tested whether the stimuli identified as high preferred would function as reinforcers. Overall, results demonstrated that the verbal and pictorial PS assessment were more likely than the survey to identify high and low preferences. Results demonstrated a lso that stimuli scored as highly preferred by the child on the survey did not always function as reinforcers, suggesting that the survey may lead to false positives. A comparison of the two formats of the PS assessment indicated that the pictorial PS asse ssment enhanced identification of stimuli that functioned as reinforcers over the verbal PS assessment. the subjects did not select individual stimuli; instead, they select ed categories of stimuli (e.g., activities, edibles, etc.). Second, subjects did not actually receive the stimuli throughout the prefe rence assessments, which is not the standard method in the assessment methods discussed previously. Third, the procedures were used to identify stimulus preference among children who were not ID. Despite the demonstration of the usefulness of preference assessment use among children who are not ID, little research has emerged in this area.

PAGE 16

16 Given the advancements of reinforcem ent based procedures using stimulus preference assessments in the area of ID and the extensions identified in Northup et al. (1996), parallel assessment methods could be a useful addition in child welfare. As of 2010 there were 509,000 children in foster c are in the US and 19,156 in foster care in Florida (US Department of Health and Human Services, 2011). Of those children, approximately 70% were typically developing (US Department of Health and Human Services). As discussed previously, when a child enters care for the first time very little is known about the child and, to date, there is no systematic procedure in place to a child moves from one foster home to an preferences is not systematically shared with the new foster parent. Preference assessment results could provide foster parents with a list of preferred and non preferred stimuli or activities that could function as rei nforcers and be used to enrich the the home. There is a need for the development and validation of a preference assessment method that would be applicable on a wide scale for this population. To date, the extension of preference assessmen t methods into child welfare has not been evaluated. Overall behavior analytic research in child welfare has been relatively sparse ( Stoutimore, Williams, Neff, & Foster 2008). How ever, research has emerged in the area of token economies (Phillip s 1968), parent training (Van Camp, Vollmer, Goh, Whitehouse, Reyes, Montgomery, Borrero, 2008) staff training ( Crosland, Dunlap, Sager, Neff, Wilcox, Blanco, & Giddings, 2008) and runaway assessments ( Witherup, Vollmer, Van Camp, Goh Borrero & Mayfield, 2008 ) Behavior

PAGE 17

17 analysis can make meaningful contributi ons in child welfare (Vancamp, B orrero ,& Vollmer, 2003 ) and t he followin g studies extend behavior analytic research in this a rea with the use of p reference assessment s The use of preference assessment method s and the web based program designed for this r esearch is a novel approach that may provide t is identified. Thus, the overall purpose of the following studies was to evaluate the use of a computerized web based preference assessmen t method with foster children. Specifically, in Study 1, subjects completed a 4 point Likert type questionnaire designed to assess preference for different activities and environmental characteristics. Following the administration of the questionnaire, the subjects completed a PS preference assessment as a way of providing initial empirical validation of the Likert type survey responses. Study 2 was designed to provide further validation of the survey responses by observing the subjects interact with (HP versus LP) stimuli identified during Study 1 in a concurrent operant arrangement. Study 3 aimed to test whether t he stimuli identified in Study 1 could be used as reinforcers for completing math problems. Combined, the findings from this research have the potential to advance routine child welfare prac tices by quickly identifying preferred stimuli and empirically val idated reinforcers.

PAGE 18

18 CHAPTER 2 GENERAL METHOD Subjects and Setting At the outset of the study, there were 18 subjects. Subjects were children and adolescents ages 6 to 17. All subjects were in foster care in Florida Circuit 10, which includes Polk, Highla nds, and Hardee counties. Sessions took place either in the foster home, group home, or at the Heartland for Children (Heartland) offices. Heartland is a private agency contracted by the state of Florida to provide dependent care services in Circuit 10. P lacement Preference Assessment (PPA) System management, and a professional programmer designed a web based computer program for the following series of studies. Heartland funded the development of the computer program and it was designed for routine staff use. Staff members, foster parents, and foster children accessed the system individually with a unique username and password. Foster parents and children using the system were only able to complete preference assessments. Staff members were able to add foster parent and foster child users to the system, set up preference sessions, view and print preference assessment reports, and identify potential child placements. A ser ver contracted by Heartland housed and maintained the system. The PPA system was incorporated into the daily job responsibilities for all the placement, licensing, retention, and service counselors. The primary investigator trained Heartland staff members and created short (2 4 minutes) training videos to provide instructional support for staff users.

PAGE 19

19 To date, approximately 30 Heartland staff members, 254 foster parents, and 70 foster children have used the system. However, to participate in the validation procedures described herein, a subject must have had informed consent signed by their biological p arent. We anticipate being able to conduct subsequent larger scale retrospective studies with other existing data. Dependent Variables The web based computer program recorded responses for Study 1. Measures the preference assessments, the latency to respond to each question, the time spent answering each question, and the number and duration of any breaks taken. Additional dependent variables will be specified for each study.

PAGE 20

20 CHAPTER 3 STUDY 1: EVAL UATION OF CHILD PREF ERENCES Introduction The purpose of this study was to assess child preferences for a variety of activities and environmental characteristics. To this end, a 57 item computerized child questionnaire (Table 3. 1) was created. In the first phase of the study, subjects completed the 4 point Likert type questionnaire. In the second phase, subjects completed a computerized PS preference assessment as a way of providing initial empirical validation of the survey resp onses. Method S ubjects Eighteen sub jects ages 6 17 began Study 1. All subjects were foster children placed in foster care Three children were dropped from the study because they inadvertently were allowed to discuss their answers during the computerize d avoid separate placement (which was of course in no way a possible outcome of this assessme nt per se). Thus, a total of 15 subjects completed Study 1. Procedure In order to complete the computerized preference assessments, each subject was assigned a unique username and password by the primary researcher or foster care counselor to log into the PPA system. The subjects completed two preference assessments on the web based system. The first was a computerized questionnaire and the second was a paired stimulus (PS) assessment. The computerized questionnaire will be described first.

PAGE 21

21 After a subject signed into the system, brief and standard instructions appeared specifi c to the questionnaire. The instructions informed the subject that the questions way for any of their answers. Following the instructions, each question was presented on t he screen one at a time. Each question appeared at the top of the screen and answer options appeared toward the bottom of the screen ( Figure 3 1) The answering The top of the screen showed the questionnaire number (e.g., 1 out of 53), a link to sign out and a link to pause the questionnaire. In addition to the survey question, a submenu field appeared on the screen for some questions testing a larger stimulu s nu field listed a variety of specific sports (e.g., baseball, soccer, football). Subjects could select as many specific stimuli as they wanted to select. A free form text field was also available to allow the subjects to type their own item(s). Once a ques tion was answered, the subject could not change his or her answer. The first 49 questions of the questionnaire were presented randomly. Questions 50 52 were three questions that required the child to type a response and were always presented in the same o rder and at the end of the questionnaire. Following ques tion 52, the subject was given five opportunities to add items not asked on the questionnaire. If the subject did not add anything else, the questionnaire ended. Assistance reading was provided if a s ubject requested help

PAGE 22

22 during the preference assessments. Only one subject, Mia, required any reading assistance. Mia was the youngest subject (age 7). Throughout the answering session, the subject could pause, resume, and sign out. If a subject paused and logged off the system, the questionnaire resumed at the last question the subject answered. The results of the preference questionnaire were immediately available to the counselor to download and to share with the subject, the both. Following completion of the questionnaire, a PS preference assessment was generated by the computer. Analogous to the procedures described by Fisher et al each other ( Figure 3 2). Due to the large num ber of stimuli assessed on the child questionnaire, and the potential for a large number of pairings the maximum number of items tested was set at ten, resulting in 45 pairings. All subjects had 45 pairing. The program was designed to randomly select five Like stimuli added in by the subjects at the end of the questionnaire (i.e., stimuli the subject had the option of adding in the last five questions of the questionnaire) also c ould be selected into the PS. The primary investigator and counselors could view the PS stimuli the program had randomly selected. The purpose of viewing these stimuli before the assessment was to fix any spelling errors made for stimuli added in by the su bject and

PAGE 23

23 to make sure any stimulus added by the subject was not inappropriate (e.g., drug use). No changes had to be made to the PS stimuli generated by the computer. The computer program was tested bi weekly to check for any program errors. No program re cording errors were identified. It should be noted that subjects not in foster care who participated in this study completed preference assessments on a separate but functionally equivalent version of the system originally used for staff training purposes. Results Figures 3 3 through 3 11 depict data for the fifteen s ubjects who completed Study 1. In each of the figures, the percentage of selections during the PS is depicted in descending rank order along the Y axis and the stimuli tested are listed along the X available upon request. For all subjects, a differentiated hierarchy of selections was observed throughout the PS. In total, nine subjects showed complete correspondence between how the stimulus was rated by the subject on the survey and the percent of se lections on the ranked those stimuli in the top 5 of 10 possible rankings. Likewise, stimuli subjects in the bottom 5 of 10 possible rankings during the PS. Figures 3 3 through 3 7 depict the data for the

PAGE 24

24 subjects for whom complete correspondence was observed. It should be not ed that 7 ) showed a tie between her 5th and 6th rankings. Five subjects (Figures 3 8 and 3 10 ) had correspondence for 8 stimuli (note that if one LP stimulus moves into the top five, one HP stimulus necessarily moves to the bottom five, so these subjects represent cases when one LP stimulus moved into the top fi the survey but ranked in the bottom five rankings in the PS assessment One subject, Rex, (Figure 3 11 ) showed correspondence for six stimuli (two LP stimuli moved into the top five). A Spearman rank order correlation (rho) showed an overall statistically significant relationship between the survey rating and the PS ranking (rs [ .563], p < .001). Individually, 15 subjects showed a statistically significant relationship (p <.05). The tw o subjects who did not show statistically significant results were Keith (rs [0.59], p = .07 [bottom panel of Figure 3 8 ]) and Rex (rs [0.522], p=0.12 [Figure 3 11 ]). Mean response time for the questionnaire was 14 min and 26s; the range was 5 min 24 s an d 27 min 47 s. Mean response time for the PS was 4 min an d 31 s; the range was 2 min and 31 s and 9 min and 57 s. Discussion In summary, 9 subjects showed complete correspondence for all 10 stimuli, 5 subjects showed correspondence for 8 stimuli, and 1 sub ject only showed correspondence for 6 stimuli. Overall, the findings from Study 1 suggest the computerized assessment is a viable method of quickly assessing preference for

PAGE 25

25 children in foster care. Results also imply that asking the foster child to rate h is preferences in a survey format resulted in good correspondence with rankings in the PS preference. show correspondence between the questionnaire and the PS. A closer examina tion of his data did not reveal obvious reasons for the disparities. It does not appear that Rex simply clicked one side of the screen or switched back and forth between stimulus options. Additionally, his response latencies throughout the PS did not indic ate that he clicked through the assessment at a rapid pace. For comparison, Rex completed the PS in 11min and 48s and Diana had the fastest completion PS time (2min and 31s) and her data showed correspondence for 8/10 stimuli. Assessments were all complete d on the same day for all subjects, which would minimize potential fluctuations in preference. It is possible, however, that Rex originally endorsed the stimulus incorrectly on the survey. For example, he may have incorrectly made during the PS. Conversely, given that the assessments were administered on the same day with a short break in between, it is possible that the subjects became fatigued. However, Lauren, whose data demonstrate complete correspondence, took the longest time to complete both the survey and PS (approximately 38 min). One limitation of the study is that we did not test all the stimuli in a PS format. If we had done so, there was a potential for la rge number of pairings and the time necessary to complete the assessment may have become impractical (DeLeon, Fisher, Rodriguez Catter, Maglieri, Herman, & Marhefka, 2001). Future investigations will attempt to

PAGE 26

26 evaluate all the stimuli using a PS format; h owever, this study constrained the maximum to 45 pairings by randomly selecting stimuli to test. By randomly selecting stimuli, we did not attempt to control for testing similar types of stimuli during the PS. For example, Lauren (top panel of Figure 3 3 ) had 3 stimuli included in the PS that were all related to visiting and seeing certain people. Additionally, the children did not receive access to the stimuli, which is the standard method in preference assessment procedur es. Additionally, despite the re latively small number of pairings, the subjects often instructions presented before the PS clearly stated that each pairing was unique, it was a commonly stated complaint by the subjects PS rankings corresponded with how each initially rated the stimuli on the survey. subjects were siblings that were being moved to a new foster placement. They were all seated in the same office space and they were asked to complete the assessments at the same time using different computers. As the subjects completed the assessments, they discussed the idea of res ponding similarly in order to be placed into the same foster home and thereafter answered the survey questions similarly. Although the instructions presented at the start of the computerized assessments do not mention anything about placement, the subjects formed this erroneous rule. Although the data were not included, the problem exemplifies the immediate need and importance of staff training on how to instruct, present, and describe the computerized preference assessments.

PAGE 27

27 Another limitation of this stud y is that these data do not indicate whether the children would engage with stimuli identified as preferred if given an opportunity and whether those stimuli selected as preferred could function as reinforcers. Studies 2 and 3 were design ed to address thi s limitation

PAGE 28

28 Table 3 1. Child questionnaire Questions I like doing arts and crafts I like talking walks in the neighborhood I like going to amusement parks I like spending time with my friends I like going to restaurants I like having a Facebook, Twitter, MySpace page I like playing with toys and video games I like swimming I like pets and animals I like cooking I like gift cards I like having my own room I like going to community parks I like going to school I like eating junk f oods I like would like to go to college I like sports activities I like living in the city or town of my choice I like board games and indoor games I like going to the school of my choice l like shopping I like visiting certain people I like talking on the telephone I like staying up last ** I like shopping I like doing certain chores ** I like listening to the radio, MP3 players, cds Ipod I like having my friends come over and hang out I like eating fruit and healthy foods I like going to the salon, beauty parlor, or barber shop I like using a computer and surfing the internet I like getting a massage, manicure, or pedicure I like playing musical instruments or signing I like having a job I like outdoor adventure activit ies I like getting an allowance I like dancing, gymnastics, and fitness activities I like going to church, temple, mosque I like reading books and magazines I like going to school I like cosmetics and hair care products I don't like [open field] I like photography and taking pictures What get's you upset? I like watching TV and renting movies When upset what calms you down? I like hanging out with other kids in the neighborhood Is there anything you wish you were asked? [open field] I like having my own cell phone Is there anything you wish you were asked? [open field] I like earning money for following the rules Is there anything you wish you were asked? [open field] I like decorating my own room Is there anything you wish you were asked? [open field] I like riding a bicycle and/or skateboarding Is there anything you wish you were asked? [open field] *Indicates questions with a submenu field

PAGE 29

29 Figure 3 1. Screenshot of computerized child questionnaire presentation

PAGE 30

30 Figure 3 2 Screenshot of computerized PS preference assessment presentation

PAGE 31

31 Figure 3 3 Data for Lauren and Chris. Percent of selections during the PS is depicted in descending order along the Y axis and the stimuli tested are depicted along the X axis completion times are also depicted on the figures. Both subject show correspondence for all 10 stimuli denotes stimuli added in to the survey by the subject

PAGE 32

32 Figure 3 4 Data for Grant and Ana. Both subject s show correspondence for all 10 stimuli

PAGE 33

33 Figure 3 5 Data for Mari and Alex. Both subject s show correspondence for all 10 stimuli

PAGE 34

34 Figure 3 6 Data for Kayla and Jack. Both subject show correspondence for all 10 stimuli. denotes stimuli added in to the survey by the subject.

PAGE 35

35 Figure 3 7 Data for Lan th and 6 th rankings are a tie.

PAGE 36

36 Figure 3 8 Data for Diana and Keith Both subjects show correspondence for 8 stimuli

PAGE 37

37 Figure 3 9 Data for Cameron and Mia. Both subjects show correspondence for 8 stimuli

PAGE 38

38 Figure 3 10 Data for Sam. 8 stimuli

PAGE 39

39 Figure 3 11 6 stimuli

PAGE 40

40 CHAPTER 4 STUDY 2: AN EVALUATION OF CORRES PONDENCE BETWEEN CHI LD SURVEY RESPONSES AND ENGAGE MENT Introduction This experiment was designed to test the correspondence between stimuli ranked as HP on the survey and PS and the allocation of time engaging with those items and activ ities. Using the questionnaire items amenable to this type of experimentation (e.g., video games, TV, computer, arts and crafts), the subjects were given an opportunity to se items was measured. Method Subjects and Settings Four subject ages 7 17 who completed Study 1 participated in this experiment. These four subjects were selected because they had informed consents signed and their schedules allowed them to participate home, foster home, shelter facility, or group home. Data Collection and Analysis Data were collected using handheld devices (hp iPAQ) using Instant Data, a behavioral recording program. Measures of IOA were obtai ned by having a second observer score session vi deo tapes for 30% of sessions. IOA was calculated using the proportional method by breaking each session into consecutive 10 s intervals and by dividing the smaller duration by the larger duration in each int erval, and then multipl ying by 100 for each interval. Agreement was then summarized by averaging the agreement scores obtained for each interval. IOA averaged 98% ; the range for individual observation intervals was 80 100%.

PAGE 41

41 Dependent Variables The depen or activities presented. Engagement was defined as manipulating, holding, touching, or orienting towards an item or activity. A subject could be engaged with both items or none of the it ems presented. There was a 3 s pause allowed; specifically, if a subject stopped or paused for less than 3s then engagement was still scored. If the pause was longer than 3 s then engagement was no longer scored. Procedure Two stimuli selected from the computerized assessments (survey and PC) were placed on two different sides of a session room (approximately 91 cm apart), if space permitted. If not, the two stimuli were presented side by side (approximately 25 cm apart) on a table. For each session, one high preference (HP) and one low preference item (LP) were presented. A HP item was defined as any item scored above 60% of the questionnaire. An L P item was defined as an item scored below 40% of the PC HP item was paired with one LP item. Stimuli that were scored above 60% on the PC questionnaire were not tested. pla pointed to and named the two stimuli presented. The placement of HP items and LP items were switched between sessions so that the HP items and LP items were not

PAGE 42

42 associate d with one side of the room or table. It was possible that the stimulus tested could not be placed on the table (e.g., riding a bike or doing chores.) When this occurred, the stimulus word or phrase was written on a white sheet of paper along with some phy sical stimulus representation (e.g., paper towels and dish detergent representing chores). If the subject selected the stimulus with the printed name, then the subject was allowed to get up from the table to engage with the stimulus (e.g., go outside to ri de a bike or do chores). min session was recorded. The subjects were allowed to switch their choice throughout the interval. At the end of the trial, the subject was told that the time had elapsed and was asked to stop playing with items. There was a short 2 min break between sessions. Sessions were conducted until stability of the data paths were observed via visual inspection of the data. Additionally, s, Target, or Walmart for participating. Results Figure 4 1 depicts data for all six participants. Engagement (percent of session) is depicted along the Y axis and sessio ns are depicted on the X axis. Closed squares represent HP stimuli and open circle represent LP stimuli. The top panel, left to right, show s data for Diana and Alex. Both subjects engaged exclusively with the HP stimuli throughout all sessions and did not engage with the LP stimuli. Di ana spent 100% of all 100% of sessions. The bottom panel, left to right, of Figure 4 1 depicts data for Maria and Jack. Maria engaged with the HP stimuli throughout most the five se engagement ranged from 88 100% of session. Maria was the only subject who engaged

PAGE 43

43 with the LP stimulus, which occurred during the first session for 3% of the session. Her engagement with the HP dropped to 88% during the fourth session. Jack with the HP stimuli ranged from 94 100% of session. Discussion All subjects engaged either exclusively or almost exclusively with the HP stimuli. Maria was the only subject who allocated any engagement to the LP stimulus (a fashion magazine ). Collectively, these data provide further empirical support for the subsequent allocation of behavior. This demonstrates that for the subjects tested, the stimuli initially id given the opportunity. Furthermore, the subjects engaged with these stimuli more often than with the usefulness of the initial survey for its efficiency and predictive value in identi fying response allocation. This finding has potentially important implications for foster car e agencies and foster parents. First, the procedure and survey results provide a fast way of simply identifying preferred and non preferred items and activities i n a systematic manner. Currently, in child welfare, there is no systematic way of obtaining suc h information about the child. Second, using stimuli identified on the survey could be used to enrich s may make the foster reinforcing. Numerous investigations in the area of ID have shown the usefulness of EE

PAGE 44

44 in decreasing problem behavior such as stereotypy (e.g., Hor ner, 1980; Rapp, 2006). Additionally, research has shown that EE is enhanced by using highly preferred activities over less preferred activities (Vollmer, Marcus, & Leblanc, 1994). However, to date, there have not been any investigations of the potential e ffects of EE using stimulus preference assessment results in foster homes. The use of this computerized preference assessment method sets the stage for an evaluation of the use of EE with empirically validated reinforcers in attempting to reduce placement disruptions. There are a number of limitations to this study. First, only a small sample of the stimuli from the questionnaire was tested. Second, only stimuli that met the definition of HP and LP were tested. For example, if a stimulus was scored as LP on the survey PS that stimulus was not tested. Future studies will specifically test engagement with those items that did not show correspondence between preference as sessments. Third, some stimuli from the questionnaire were not amenable to testing (e.g., visiting certain people, going to a certain school, having a job). Fourth, the array of stimuli tested was further constrained by group and shelter home rules. For ex ample, access to internet, riding a bike, talking on the telephone are a few examples of stimuli prohibited by some group home rules. Additionally, the measurement of approach and engagement with a stimulus may not be predictive of rein forcer effectiveness in other contexts (Pace et al. 1985). That is, although the subjects allocated responding to the stimuli identified as HP on the survey and PS, there remains the possibility that those stimuli would not be effective reinforcers

PAGE 45

45 to incr ease meaningful behavior su ch as academic work or chores. This limitati on will be addressed in Study 3.

PAGE 46

46 Figure 4 1 Engagement (percent of session) is depicted along the Y axis and sessions are depicted on the X axis. Closed squares represent engagement with the HP stimuli and open circles represent engagement with LP stimuli.

PAGE 47

47 CHAPTER 5 STUDY 3: REINFORCER ASSESSMENT Introduction The purpose of this experiment was to further test whether the stimuli identified in Study 1 could serve as effective reinforcers for an academic task. Math problems were selected because they represent an activity that all subjects were requi red to do while living at the foster home or group home (i.e., it simulated a homework session). Method Method Subjects and Settings Mia (age 7), Ana (age 11), and Kayla (age 14) served as subjects in this study. All three subjects previously completed Study 1 and were in foster care. Assessments were Dependent Variables The primary dependent variable was the rate of math problem completion reported as responses per minute. Problems did not hav e to be completed correctly to be counted, however accuracy measures were scored and will be reported in the results. Data Collection and Analysis Data were collected during the sessions using handheld devices. All sessions were videotaped. Measures of (I OA) were collected by having a second observer score session vi deo tapes for 30% of sessions. IOA was calculated in the same manner as in Study 2. IOA average agreem ent across subjects was 96%, the range for individual observations ranged from 91 % 100%.

PAGE 48

48 Pr ocedure An assessment began with the subject sitting across from the experimenter. In front of the subject were three stacks of math problems approximately 10 cm apart. Each pile contained the same twelve pages of math problems. The math problems were basi multiplication, and division. Above each stack of math worksheets was a white sheet of paper with the written name of the stimuli associated with each stack. The stimuli were an HP item, an LP item, and a control, in which work on those math problem sheets produced tokens but did not produce any item after the session. HP and LP stimuli were defined the same as in Study 2. Different HP and LP stimuli were tested per assess ment (stimuli are in dicated in Figure 5 1). The stimuli were reviewed with the subject prior to the assessment. The reinforcement schedule used was a fixed ratio (FR) 3 schedule, in which every 3 problems completed resulted in a token. Math problems comple ted did not have to be correct, though this contingency was not specified in the instructions. Tokens were exchangeable for 1 min access to the stimuli associated with the math stack under which the token was earned. The placement of the HP, LP and control stacks switched between assessments. The instructions given to no work at all. Depending on which stack you choose, you can earn tokens toward that/those items. For eve ry three math problems you complete, you will earn one token, (e.g., gift card), each token was worth $.25. All tokens earned were white poker chips. If the subject re quested assistance, the use of a calculator was permitted. Two subjects, Ana and Kayla, used a calculator for some math problems.

PAGE 49

49 A subject could switch their allocation of work completion throughout the math time. The first assessment was always for 4 mi n of math work time. Math time ended if the subject stated that he or she did not want to do any more math problems, said that he or she was finished, or worked for four consecutive minutes. The tokens were exchangeable after the math time. The first asses sment session consisted of 4 min of math completion time plus time earned with the stimuli. Total assessment duration was variable because it depended upon math time plus reinforcer access time the subject earned. The average assessment duration was 28 min and the range was 12 50 min. All assessments were completed on the same day. Due to time constraints, math nts, Kayla participated in two assessments. As in Study 2, subjects received a $5.00 gift card to Results All subjects worked exclusively on one stack th roughout all sessions. Figure 5 1 shows data for all three subjects and for all of the assessments. The rate of math problem completion is depicted along the Y axis and minutes are depicted on the X axis. Closed squares represent responding for the HP stack, open circles represent respon ding on the LP stack, and the closed triangles represent responding on the LP pairings identified on the figures. For 7 of 8 assessments, subjects completed math problems exclusively to the stack associated with the HP item identified from the survey and PS exclusively to the stack of math problems associated with the LP stimulus. Although the

PAGE 50

50 contingency was pla ced on math problem completion, rather than accuracy, the average percentage correct was 84% (range, 22 more errors. Discussion surve y and PS in Study 1 could serve as effective reinforcers for math problem completion. For 7 of the assessments, subjects completed math problems exclusively to the stack associated with the HP item identified from the survey. Additionally, testing whether the stimuli could function as reinforcers for math problem completion addressed a limitation of Study 2. Notably, Ana was the only subject who allocated exclusive responding to the stack associated with the LP stimulus (board games) during her second ass essment. There are a number of potentially important variations throughout that assessment that merit discussion. First, the HP stimulus tested (a gift card to Target, which was specifically identified by the subject on the questionnaire) was the only stim ulus tested for any of the subjects that had a dollar amount rather than an immediate, tangible reinforcer. Although a gift card with the amount earned would have been given to the subject following math problem completion, it could represent a more delaye d, conditioned reinforcer in that the subject would need to be driven to the store to pick out a tangible item. By comparison, the LP stimulus paired with the gift card was playing board games, an immediate, tangible consequence. This difference may have c ontributed to the obtained results. time (identified as HP). She earned a total of 18 min and while playing video games she

PAGE 51

51 asked the experimenter to play games with her. Thus, prior to the gift card vs board game pairing (the second assessment), the experimenter had played video games with Ana for approximately 15 minutes. This video game time with the experimenter may have served as an establishing operation, defined as an ant ecedent event that momentarily increased the effectiveness of, in this case, social interaction, thus establishing board games as a reinforcer and making completing math problems more probable. Essentially, playing video games prior to the gift card board game pairing may have become a conditioned reinforcer by virtue of pairing with the video game. Another possibility is that because, as noted in the procedures, all the s ubjects earned a gift card for participating, the gift card reinforcer was de valuated. All of these variables may have contributed to the unexpected and different results in the one assessment for Ana. Future studies should attempt to control for these ef fects and examine whether changes in preference can be manipulated given pre session exposure. There are a number of limitations of this study. First, math time tested was limited to only 3 or 4 minutes. This was due to the variable assessment length that depended upon the time the subject earned and time constraints imposed by group home visitation rules and foster family engagements. Longer math completion times would have tested the effectiveness of the reinforcer more rigorously. In retrospect, we could have reduced the earned access time or increased the ratio schedule to ensure that the session s fit into a reasonable time period. Second, a social component to some of the stimuli tested was not controlled. For example, stimuli such as board games, manicures, certain video games are typically accompanied by other naturally occurring social

PAGE 52

52 intera ctions. Thus, it is unclear whether it is the tangible stimulus property, the social stimulus property, or some combination that makes the stimulus preferred. Third, and perhaps most importantly, although the HP stimuli were clearly shown to be reinforcer s, this study did not show that the LP stimuli would not also serve as reinforcers if presented in isolation (Roscoe, Iwata, Kahng, 1999). For example, it is possible that LP items such as reading books and magazines, playing board game, and doing fitness activities could be used as effective reinforcer s when presented in isolation. We plan to address this limitation in future studies.

PAGE 53

53 Figure 5 1 Rate of math problem completion is depicted along the Y axis and minute is depicted on the X axis. Close d squares represent responding for the HP stack, open circles represent responding on the LP stack, and the closed triangles represent responding on the control stack. Subjects HP LP pairings are shown by session in each column

PAGE 54

54 CHAPTER 6 GENERAL DISCUSS ION Collectively, these experiments suggest that the computerized preference assessment method is a feasible way to assess preferences for typically developing foster children. For the 17 subjects in Study 1, 11 showed complete correspondence between child questionnaire ratin gs and PS assessment outcomes. Five subjects showed near complete correspondence and 1 subject showed partial correspondence (6 stimuli). The results from Studies 2 and 3 further validate the computerized preference assessment method b and math problem completion. Prior to this series of studies, the extension of preference assessment methods into c hild welfare had not been evaluated. Additionally, a computerized preference assessment method for verbal children had not been reported in the pre ference assessment literature. Preference assessment method and the web based program designed for this resea rch is a novel approach and may provide some useful Providing foster parents with a list of potential reinforcers may also assist the foster parents when implementing contingency management strategies for increasing appropriate behavior and reducing problem behavior. To date, there is no systematic manner of asking foster children their general preferences and information about the child is not systemically shared with the foster parents. Such a system could be an important contribution to child welfare. In fact, the funding agency for this project has

PAGE 55

55 now adopted the program and has used it to assess preferences for 70 foster children and 234 foster parents. A major l imitation of the studies presented relate to the stimuli tested. For Studies 2 and 3, the array of stimuli testable were constrained due to procedural nuances and home rules often prohibited access to the internet. Additionally, activities such as going to amusement parks were activities that were not easily scheduled. Future studies should look into testing these stimuli, perhaps as back up reinforcers in a token system ; however, for some stimuli there may still be extenuating circumstances that may prevent evaluation (e.g., experimenters would have to comply with a court mandated family visitation schedule). Additional limitations of the computerized method emerged with the possibility of limitation highlights that simply asking a child what they like has an implied demand c haracteristic that may influence responses. A child may not respond consistently with actual preference but may respond in a manner he or she may think is more the uncert ainty of the type of home he or she will experience. It is imperative that staff is adequately trained in how to present, administer, and discuss the computerized preference assessments. Any research on preference assessments must recognize that individua l preference is not static; preferences can shift over time, and will be influenced by states of satiation and deprivation (e.g., DeLeon et al., 2001; Kennedy and Haring, 1993).

PAGE 56

56 Thus, frequent preference assessments may have to be administered as foster ch ildren contact novel experiences living in new foster homes. A benefit of the computerized system, then, may prove to be its overall brevity and ease of completion. Overall, there may be some benefits to foster children, foster parents, and child welfare a gencies by incorporating the use of preference assessments. Ultimately, increasing placement stability would be the over arching aim of the incorporation of preference assessment methods into child welfare. Placement stability is an important issue for fos ter care agencies and placement disruptions are highly undesirable for a number of reasons. Placement disruptions are associated with increases in reported child behavior problems (Newton, Litrownik, and Landsverk, 2000), are associated with poor school pe rformance (Johnson Reid & Barth, 2000), and have an emotional cost for both foster children and foster parents (Fanshel, Finch, & Grun dy, 1990; van der Kolk, 1987). N ational goals have stated also that foster children should have two or fewer moves during their entire time in care (U.S. Department of Health and Human Services, 2003). Additionally, placement disruptions have a financial impact on the child welfare system. It is estimated that a placement disruption results in an average of over 25 h of casew orker time to find and place a child in a new home, and document changes (Chamberlain, Price, Reid, Landsverk, Fisher, Stoolmiller, 2006). Although there are a myriad of variables that influence placement decisions (e.g., availability of a bed in a partic ular foster home, foster home licensing regulations), there is a need for identifying placements that are more likely to be successful (i.e., homes in which the child will likely remain and, thus, avoid a placement change). Currently, no such system to att empt to identify placements for foster children exists. Heartland, the

PAGE 57

57 funding agency for the current project, plans to initiate a large scale study in which the parents in order to evaluate effects on placement duration. They have named the Overall, the studies described herein represent a small portion of current and potential future investigations of the PPA syst em and child welfare represents an area in critical need of systematic evaluations. Whereas preference assessments in ID have hope that the PPA system has given childre n in foster care a voice that can be heard. Too often, foster children are not given much of a chance to express their wants and needs.

PAGE 58

58 LIST OF REFERENCES Adoption and Foster Care Analysis and Reporting System (2010) U.S. Department of Health and Human Services, Administration for Children and Families, Administration on Children, Youth and Families, Children's Bureau, www.acf.hhs.gov/programs/cb Bojak, S. L., & Carr, J. E. (1999). On the displacement of leisure items by food during multiple stimulus preference assessments. Journal of Applied Behavior Analysis, 32, 515 518. Brian, S. (1963). Clinical judgment i n foster care placement. Child Welfare,42, 161 169. Carr, J. E., Nicolson, A. C., & Higbee, T. S. (2000). Evaluation of a brief multiple stimulus preference assessment in a naturalistic context. Journal of Applied Behavior Analysis, 33 353 357. Crosland, K. A., Dunlap, G., Sager, W., Neff, B., Wilcox, C., Blanco, A., & Giddings, T. (2008). The effects of staff training on the types of interactions observed at two group homes for foster care children. Research on Social Work Practice, 18 (5), 410 420. DeLeon I. G., Fisher, W. W., Rodriguez Catter, V., Maglieri, K., Herman, K., & Marhefka, J. M. (2001). Examination of relative reinforcement effects of stimuli identified through pretreatment and daily brief preference assessments. Journal of Applied Behavior A nalysis, 34 463 473. DeLeon, I. G., & Iwata, B. A. (1996). Evaluation of a multiple stimulus presentation format for assessing reinforcer preferences. Journal of Applied Behavior Analysis, 29, 519 533. DeLeon, I. G., Iwata, B. A., & Roscoe, E. M. (1997). Displacement of leisure reinforcers by food during preference assessments. Journal of Applied Behavior Analysis, 30, 475 484. DeLeon, I. G., Fisher, W. W., Rodriguez Catter, V.,Maglieri, K., Herman, K., & Marhefka, J. M. (2001). Examination of relative rei nforcement effects of stimuli and daily brief preference assessments. Journal of Applied Behavior Analysis, 34, 463 473. Egel, A. L. (1981). Reinforcer variation: Implications for motivating developmentally disabled children. Journal of Applied Behavior Analysis, 14 345 350. Fanshel, D., Finch, S., & Grundy, J. (1990). Foster children in a life course perspective. New York: Columbia University Press. Favell, J. E., & Cannon, P. R. (1976). Evaluation of entertainment materials for severely etarded persons. American Journal of Mental Deficiency, 81(4) 357 361.

PAGE 59

59 Fisher, W., & Mazur, J. E. (1997). Basic and applied research on choice responding. Journal of Applied Behavior Analysis, 30, 387 410. Fisher, W., Piazza, C. C., Bowman, L. G., Hagopian, L. P., Owens, J. C., & Slevin, I. (1992). A comparison of two approaches for identifying reinforcers for persons with severe and profound disabilities. Journal of Applied Behavior Analysis, 25 491 498. Hagopian, L. P., Rush, K. S., Lewin, A. B., & Long, E. S. (2001). Evaluating the predictive validity of a single stimulus engagement preference assessment. Journal of Applied Behavior Analysis, 34 475 485. Hanley, G. P., Cammilleri, A. P., Tiger, J.H., & I ngvarsson E. T. (2007) A method for Journal of Applied Behavior Analysis, 40 603 618. Homer, R. D. (1980). The effects of an environmental enrichment" program on the behavior of institutionalized profoundly retarded children. Journal of Applied Behavior Analysis, 13, 473 491. Jonson Reid M ., Barth, R. (2000). From maltreatment to juvenile incarceration: uncovering the role of child welfare ser vices. Child Abuse Neglect, 24 505 520. Kennedy, C. H., & Haring, T. G. (1993). Teaching choice making during social interactions to students with profound multiple disabilities. Journal of Applied Behavior Analysis, 26 63 76. Matson, J. L., Bielecki J., Mayville, E. A., Smalls, Y., Bamburg, J. W., & Baglio, C. S. (1999). The development of a reinforcer choice assessment scale for persons with severe and profound mental retardation. Research in Developmental Disabilities, 20 379 384. Northup, J., Ge orge, T., Jones, K., Broussard, C., & Vollmer, T. R. (1996). A comparison of reinforcer assessment methods: The utility of verbal and pictorial choice procedures. Journal of Applied Behavior Analysis, 29, 201 212. Pace, G. M., Ivancic, M. T., Edwards, G. L ., Iwata, B. A., & Page, T. J. (1985). Assessment of stimulus preference and reinforcer value with profoundly retarded individuals. Journal of Applied Behavior Analysis, 18 249 255. Phillips, E. L. (1968). Achievement place: Token reinforcement procedures in a home Journal of Applied Behavior Analysis, 1 213 223. Phillips, M., Shyne, A., Sherman, E., Haring, B. (1971). Factors associataed with placement decisions in child welfare. New York: Child Welfare League of America, Inc.

PAGE 60

60 Rapp, J. T. (2006). Toward an empirical method for identifying matched stimulation: A preliminary investigation. Journal of Applied Behavior Analysis, 39, 137 140. Roanne, H. S., Vollmer, T. R., Ringdahl, J. E., & Marcus, E. A. (1998). Evaluation of a brief stimulus preference assessment. Journal of Applied Behavior Analysis, 3, 605 620. Roscoe, E. M., Iwata, B. A., & Kahng, S.W. (1999). Relative versus absolute reinforcement effects: Implications for preference assessments. Journal of Applied Behavior Analysis, 32 479 493. Simon, R. D., & Simon, D. K. (1982). The effect of foster parent selection and training on service delivery. Child Welfare : Journal of Policy, Practice, a nd Program, 61 (8), 515 524. Sinclair, I., & Wilson K. (2003). Matches and Mismatches: The Contribution of Carers and Children to the Success of Foster Placements. British Journal of Social Work, 33 (7), 871 884. Stoutimore, M. R., Willia ms, C. E., Neff, B., & Foster, M (2008). The Florida Child Welfare Behavior Analysis Services Program. Research o n Social Work Practice, 18 (5), 367 376. Van Camp, C. M., Vollmer, T. R., Goh, H., Whitehouse, C. M., Reyes, J., Montgomery, J. L., & Borrero, J. C. (2008). Behavioral parent training in child welfare: Evaluat ions o f skills acquisition. Research on Social Work Practice, 18 (5), 377 391. Van Camp, C. M., Borrero, J. C., & Vollmer, T.R. (2003). The family safety/applied behavior analysis initiative: An introduction and overview The Behavior Analyst Today, 3 (4), 389 404. van der Kolk, B. A. (1987). The separation cry and the trauma response: Developmental issues in the psychobiology of attachment and separation. In B. A. van der Kolk (Ed.), Psychological trauma (pp. 31 62). Washington, DC: American Psychological Association. Vollmer, T. R., Marcus, B. A., & LeBlanc, L. (1994). Treatment of self injury and hand mouthing following inconclusive functional analyses. Journal of Applied Behavior Analysis, 27 331 344. Witherup, L. R., Vollmer, T. R., Van Camp, C. M., Goh, H., Borrero, J. C., & Mayfield, K. (2008). Baseline measurement of running away among youth in foster care. Journal of Applied Behavior Analysis, 41 (3), 305 318. Windsor, J., Piche, L. M., & Locke, P. A. (1994). Preference testing: A comparison of tw o presentation methods. Research in Developmental Disabilities, 15 439 455.

PAGE 61

61 BIOGRAPHICAL SKETCH Cristina Whitehouse graduated from Rollins College in 1995 with a B.A. in psychology. Cristina became interested in Behavior Analysis while at Rollins and com pleted her senior thesis on the Personalized System of Instruction with her undergraduate advisor, Dr. Maria Ruiz. Upon graduation Cristina began working at Threshold Inc. providing early intervention behavioral services for children with autism and developmental delays. Cristina then spent two years working as a Primary Therapist providing behavioral services at Quest Kids These opportunities lead to work at Community Services for Autistic Adults and Children in Rockville, MD where she was responsib le for in home behavioral programming and staff training. After moving back to Florida, Cristina spent 3 years working as a Behavior Analyst providing behavioral services to foster children and teaching foster parents behavioral parenting skills under the supervision of Dr. Timothy Vollmer. This opportunity motivated Cristina goal to pursue a graduate degree in applied behavior analysis, and she began her graduate studies at the University of Florida in 200 4 Since beginning graduate school, Cristina has had the opportunity to conduct research in the areas of assessment of preference and large scale program evaluation. Upon co mpletion of her graduate degree, Cristina will begin a post doctoral position at the University of Florida