<%BANNER%>

Child Directed Interaction Training with Kinship Foster Caregivers

MISSING IMAGE

Material Information

Title:
Child Directed Interaction Training with Kinship Foster Caregivers An Evaluation of Parenting Outcomes and Caregiver Benefits
Physical Description:
1 online resource (97 p.)
Language:
english
Creator:
N'zi, Amanda 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, Clinical and Health Psychology
Committee Chair:
Eyberg, Sheila M
Committee Members:
Wiens, Brenda A
Rozensky, Ronald Howard
Diehl, David Christopher

Subjects

Subjects / Keywords:
attachment -- discipline -- fostercare -- parenting -- preschoolers
Clinical and Health 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:
Children in foster care are susceptible to multipleplacements which can result in poor attachment patterns as well as childinternalizing and externalizing problems. Kinship caregivers are the federallymandated preferred placement option for children in the child welfare system.Unfortunately, these caregivers experience greater levels of stress, displaylower levels of warmth, and experience increased difficulties handling childdisruptive behaviors when compared to other caregivers. Child problematic behavioris one of the most prominent stressors among all foster caregivers and aleading risk factor of multiple placements. Parent-Child Interaction Therapy(PCIT), already established as a “best practice” in the field of child abusetreatment, teaches parents to engage their child in nurturing and cooperativeinteractions and manage difficult child behaviors. One component of PCIT, ChildDirected Interaction Training (CDIT), is a scientifically derived therapeuticinteraction that, through behavioral principles consciously applied, changescaregiver and child behaviors, fosters changes in reciprocal emotionalresponses, and results in a stronger attachment relationship. Using a randomized controlled trial design, this studyevaluated the efficacy of CDIT in a community center with 12 kinship caregiversand their children. Results indicated that following CDIT, kinship caregivershad less parenting stress and depression and reported a lower percentage of criticalverbal disciplinary responses than wait-list control caregivers. Changes inparenting stress, depression, and critical verbal discipline practices remainedstable for the CDIT group at 3-month follow-up. Kinship caregivers receiving CDIT also used significantly more positiveattention skills and fewer negative leading behaviors during observedcaregiver-child interactions. No attrition occurred in this study, which isuncommon for parent training interventions. The success of CDIT likely dependedon the collaboration with a community resource center and community agencystaff and should be considered when planning new studies.  It will be important to replicate this studywith a larger sample to determine the stability of these findings. CDIT holdspromising evidence as a short term and cost-effective intervention for thisunderserved population in need of appropriate parent training in order to breakthe generational cycle of abuse and neglect and facilitate the growth ofhealthy reciprocal interactions.
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 Amanda M N'zi.
Thesis:
Thesis (Ph.D.)--University of Florida, 2013.
Local:
Adviser: Eyberg, Sheila M.
Electronic Access:
RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2015-08-31

Record Information

Source Institution:
UFRGP
Rights Management:
Applicable rights reserved.
Classification:
lcc - LD1780 2013
System ID:
UFE0044377:00001

MISSING IMAGE

Material Information

Title:
Child Directed Interaction Training with Kinship Foster Caregivers An Evaluation of Parenting Outcomes and Caregiver Benefits
Physical Description:
1 online resource (97 p.)
Language:
english
Creator:
N'zi, Amanda 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, Clinical and Health Psychology
Committee Chair:
Eyberg, Sheila M
Committee Members:
Wiens, Brenda A
Rozensky, Ronald Howard
Diehl, David Christopher

Subjects

Subjects / Keywords:
attachment -- discipline -- fostercare -- parenting -- preschoolers
Clinical and Health 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:
Children in foster care are susceptible to multipleplacements which can result in poor attachment patterns as well as childinternalizing and externalizing problems. Kinship caregivers are the federallymandated preferred placement option for children in the child welfare system.Unfortunately, these caregivers experience greater levels of stress, displaylower levels of warmth, and experience increased difficulties handling childdisruptive behaviors when compared to other caregivers. Child problematic behavioris one of the most prominent stressors among all foster caregivers and aleading risk factor of multiple placements. Parent-Child Interaction Therapy(PCIT), already established as a “best practice” in the field of child abusetreatment, teaches parents to engage their child in nurturing and cooperativeinteractions and manage difficult child behaviors. One component of PCIT, ChildDirected Interaction Training (CDIT), is a scientifically derived therapeuticinteraction that, through behavioral principles consciously applied, changescaregiver and child behaviors, fosters changes in reciprocal emotionalresponses, and results in a stronger attachment relationship. Using a randomized controlled trial design, this studyevaluated the efficacy of CDIT in a community center with 12 kinship caregiversand their children. Results indicated that following CDIT, kinship caregivershad less parenting stress and depression and reported a lower percentage of criticalverbal disciplinary responses than wait-list control caregivers. Changes inparenting stress, depression, and critical verbal discipline practices remainedstable for the CDIT group at 3-month follow-up. Kinship caregivers receiving CDIT also used significantly more positiveattention skills and fewer negative leading behaviors during observedcaregiver-child interactions. No attrition occurred in this study, which isuncommon for parent training interventions. The success of CDIT likely dependedon the collaboration with a community resource center and community agencystaff and should be considered when planning new studies.  It will be important to replicate this studywith a larger sample to determine the stability of these findings. CDIT holdspromising evidence as a short term and cost-effective intervention for thisunderserved population in need of appropriate parent training in order to breakthe generational cycle of abuse and neglect and facilitate the growth ofhealthy reciprocal interactions.
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 Amanda M N'zi.
Thesis:
Thesis (Ph.D.)--University of Florida, 2013.
Local:
Adviser: Eyberg, Sheila M.
Electronic Access:
RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2015-08-31

Record Information

Source Institution:
UFRGP
Rights Management:
Applicable rights reserved.
Classification:
lcc - LD1780 2013
System ID:
UFE0044377:00001


This item has the following downloads:


Full Text

PAGE 1

1 CHILD DIRECTED INTERACTON TRIANING WITH KINSHIP FOSTER CAREGIVERS: AN EVALUATION OF PARENTING OUTCOMES AND CAREGIVER BENEFITS By AMANDA MAY N ZI 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 Amanda May Nzi

PAGE 3

3 To the dedicated professionals and caregivers that work tirelessly to reduce the frequency of child abuse and negl ect

PAGE 4

4 ACKNOWLEDGMENTS I would like to thank my mentor, Dr. Sheila Eyberg, for her support and guidance throughout graduate school. I would also like to thank my committee members, Drs. David Diehl, Ronald Rozensky, and Brenda Wiens. Finally, I owe many thanks to my husband, Teiva and the rest of my family and friends for their encouragement and unconditional support. I would also like to acknowledge the Center for Family Studies and the College of Public Health and Health Professions for funding this st udy.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS .................................................................................................. 4 LIST OF TABLES ............................................................................................................ 7 LIST OF FIGURES .......................................................................................................... 8 ABSTRACT ..................................................................................................................... 9 CHAPTER 1 INTRODUCTION .................................................................................................... 11 An Overview of Children in the Foster Care System ............................................... 11 A Review of the Foster Care Placements ......................................................... 11 Psychological Functioning of Children in the Foster Care System ................... 12 Special Consideration for Young Children in Foster Care ................................ 13 Kinship Caregivers: The Primary Placement Option ........................................ 13 The Current State of Foster Caregiv er Parent Training Programs and Interventions ........................................................................................................ 16 Attrition and Training Completion ..................................................................... 16 Foster Caregiver Training Programs ................................................................ 16 Multidimensional t reatment f oster care ...................................................... 17 Parent child i nteraction t herapy .................................................................. 20 The Child Directed Interaction Training ................................................................... 23 Current Aims and Study Hypotheses ...................................................................... 25 2 METHOD ................................................................................................................ 28 Participants ............................................................................................................. 28 Trainers ................................................................................................................... 32 Assessors ............................................................................................................... 33 Measures ................................................................................................................ 33 Eyberg Child Behavior Inventory ...................................................................... 33 Child Behavior Checklist ................................................................................... 34 Dyadic Parent Child Interaction Coding System: Third Edition ......................... 35 Parent Daily Report/Daily Discipline Inventory ................................................. 36 Beck Depression Inventory II ........................................................................... 37 Parenting Stress Index Short Form .................................................................. 38 Readiness, Efficacy, Attributions, Defensiveness, and, Importance Scale ....... 39 Procedures ............................................................................................................. 39 Initial Phone Screening .................................................................................... 40 Time 1 Assessment .......................................................................................... 40 Training Procedures ......................................................................................... 42 Training Integrity ............................................................................................... 44

PAGE 6

6 Time 2 Assessment .......................................................................................... 45 Time 3 Assessm ent .......................................................................................... 45 Time 4 Assessment .......................................................................................... 45 Statistical Analyses ................................................................................................. 46 3 RESULTS ............................................................................................................... 55 Aim 1: Analyses of Change Following CDIT ........................................................... 55 Caregiver Self Report of Depression and Parenting Stress ............................. 55 Observed Parent Child Interaction ................................................................... 55 Caregiver Report of Daily Discipline Practices ................................................. 56 Aim 2: Predictors of Positive Attention Skills Acquisition ........................................ 56 Aim 3: Moderators of Positive Attention Skills Acquisition ...................................... 57 Aim 4: Attrition Analyses ......................................................................................... 58 Aim 5: Follow up Analyses ...................................................................................... 58 4 DISCUSSION ......................................................................................................... 71 Limitations ............................................................................................................... 78 Future Directions .................................................................................................... 81 APPENDIX A DAILY DISCIPLINE INTERVIEW CODING DEFINITIONS ..................................... 84 B READI ..................................................................................................................... 85 LIST OF REFERENCES ............................................................................................... 87 BIOGRAPHICAL SKETCH ............................................................................................ 97

PAGE 7

7 LIST OF TABLES Table page 2 1 Demographic Characteristics of Immediate Training and Wait List Groups ............ 49 2 2 Measures and Administration Time Points .............................................................. 50 2 3 Inter Coder Agreement Statistics for the DPICS III Categories at the Time 1 Assessment ............................................................................................................ 51 3 1 Mean Scores for Caregiver Self Report Measures at Time 1 and 3 Assessments 60 3 2 Time 1 and 3 Scores for Caregivers during 5Minutes of Observed ChildLed Play ........................................................................................................................ 61 3 3 Time 1 and 3 Percentage of Reported Discipline Practices .................................... 62 3 4 Equivalency Testing Analyses between Time 3 and Time 4 Caregiver Report Variables ................................................................................................................ 63

PAGE 8

8 LIST OF FIGURES Figure page 2 1 Sampling and Flow of Participants throughout Child Directed Interaction Training. ................................................................................................................. 52 2 2 Child Directed Interaction Training Assessment and Training Flow Chart for Participants. ............................................................................................................ 53 2 3 Mean Frequency Rating of Child Behavior Problems on the ECBI throughout the course of CDIT for the Immediate T raining Group ................................................... 54 3 1 Caregiver Depression at the Time 3 Assessment. .................................................. 64 3 2 Childrearing Stress at the Time 3 Assessment ....................................................... 65 3 3 Percentage of O bserved Caregiver V erbalizations with their C hild by C ategory at the Time 3 Assessment ...................................................................................... 66 3 4 Percentage of disc ipline used by category at the Time 3 Assessment .................... 67 3 5 Maintenance of Caregiver Depression from Post Training to the 3month Follow up A ssessment ........................................................................................... 68 3 6 Maintenance of Childrearing Stress from Post Training to the 3month Follow up Assessment. ........................................................................................................... 69 3 7 Maintenance of Parent Reported Discipline Practices from Post Training to the 3 month Follow up Assessment. ............................................................................ 70

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 CHILD DIRECTED INTERACTON TRIANING WITH KINSHIP FOSTER CAREGIVERS: AN EVALUATION OF PARENTING OUTCOMES AND CAREGIVER BENEFITS By Amanda May Nzi August 2013 Chair: Sheila Eyberg Major: Psychology Children in foster care are susceptible to multiple placements which can result in poor attachment patterns as well as child internalizing and externalizing problems. Kinship caregivers are the federally mandated preferred placement option for children in the child welfare system. Unfortunately, these caregivers experience greater levels of stress, display lower levels of warmth, and experience increased difficulties handling child disruptive behaviors when compared to other caregivers Child problematic behavior is one of the most prominent stressors among all foster caregivers and a leading risk factor of multiple placements. Parent Child Interaction Therapy (PCIT), already established as a best practice in the field of child abuse treatment, teaches par ents to engage their child in nurturing and cooperative interactions and manage difficult child behaviors. One component of PCIT, Child Directed Interaction Training (CDIT) is a scientifically derived therapeutic interaction that, through behavioral princ iples consciously applied, changes caregiver and child behaviors fosters changes in reciprocal emotional responses and results in a stronger attachment relationship.

PAGE 10

10 Using a randomized controlled trial design, this study evaluated the efficacy of CDIT in a community center with 12 kinship caregivers and their children. Results indicated that following CDIT, kinship caregivers had less parenting stress and depression and reported a lower percentage of critical verbal disciplinary r esponses than wait list control caregivers. Changes in parenting stress, depression, and critical verbal discipline practices remained stable for the CDIT group at 3month follow up. Kinship c aregivers receiving CDIT also used significantly more positive attention skills and fewer negative leading behaviors during observed caregiver child interactions. No attrition occurred in this study, which is uncommon for parent training interventions. The success of CDIT likely depended on the collaboration with a community resource center and community agency staff and should be considered when planning new studies It will be important to replicate this study with a larger sample to determine t he stabil ity of these findings. CDIT holds promising evidence as a short term and cost effective intervention for this underserved population in need of appropriate parent training in order to break the generational cycle of abuse and neglect and facil itate the gro wth of healthy reciprocal interactions.

PAGE 11

11 CHAPTER 1 INTRODUCTION An Overview of Children in the Foster Care System A Review of the Foster Care Placements According to recent statistics, the number of children in the foster care system has nearly doubled since the 1980s, with current estimates at more than 400,000 children (USDHHS 2011). Children in foster care are at a higher risk than other children in community samples for developing internalizing and externalizing psychological disorders, social skills deficits, and adaptive functioning delays (Clausen, Landsverk, Ganger, Chadwick, & Litrownik, 1998). This quickly growing population of vulner able children creates a need for caregivers who are skilled in addressing these psychological difficulties. T he federally mandated primary placement option for children in the foster care system is with kinship caregivers ; however, kinship caregivers still do not receive equivalent training to traditional foster parents (Grimm, 2003; Hu r lburt, Leslie, Barth, & Landsverk, in press ). In addition, caregiver training that is provided to kinship and nonkinship foster caregivers often lacks instruction on handli ng the problematic behaviors commonly associated children with maltreatment histories (Chamberlain, Price, Leve, Heidemarie, Landsverk, & Reid, 2008; Denby, Rindfleisch, & Bean 1999). Unprepared foster caregivers likely contribute to placement instability. Placement disruption rates are reported to be between 25 and 70% and children experience an average of 4 to 7 different placements while in the foster care system (Berrick, 1998; Smith, Stormshak, Chamberlain, & Whaley, 2001). To further understand what might protect against multiple placements, it is important to explore the child and caregiver factors that influence the stability of a foster placement.

PAGE 12

12 Psychological Functioning of Children in the Foster Care System Predictors of severe psychological def icits in children with maltreatment histories include a history of trauma that was severe, prolonged, and/ repeated (Carr, 2004; Pine & Cohen 2002). P sychological deficits can be further exacerbated after entering the foster care system if children experience a significant decrease in social support, immediate and severe post incident anxiety, and/or cognitive or neurobiological deficits resulting from maltreatment (Carr 2004; Lansford et al., 2002; Pine & Cohen, 2002). Children in foster care resemble a cl inical population more than a community population in terms of their psychological functioning (Landsverk, Litrownik, Newton, Ganger, & Remmer, 1996). They are twice as likely as their normally developing peers to experience mental health difficulties (Clausen et al., 1998, Fisher, Burraston, & Pears, 2005). Multiple placements in the child welfare system result in the need for additional mental health services (Rubin et al., 2004) and impact the childs ability to form a strong attachment relationship w ith a primary caregiver (American Academy of Pediatrics, 2000; Dozier et al. 2006). Strong attachment relationships promote the development of regulatory capabilities and are characterized by high levels of warmth, and emotional availability (Dozier et al. 2006; Schofield & Beek, 2005). Children who have experienced maltreatment and multiple placements fail to develop strong attachments to primary caregivers resulting in the lack of development of appropriate self regulatory mechanisms at the biobehavioral level (Dozier et al., 2006; Howe & Fearnley, 2003; Pears & Fisher, 2005; Schofield & Beek, 2005). However, promising research indicates that when maltreated children experience a secure attachment relationship with a foster caregiver that is characterized by warm and positive

PAGE 13

13 interactions, they experience fewer psychological disturbances (Marcus, 1991). The development of an effective foster caregiver intervention that increases the responsiveness of the caregiver to the child may increase the stability of the foster care placement and decrease the risk of psychological insult. Special Consideration for Young Children in Foster Care Young children under the age of 7 represent over half of the children in child welfare system, and their overall representat ion in the system is disproportionately high ( U S DHHS, 2010 ). A review of the literature indicates the most commonly experienced psychological deficits of preschoolers in the foster care system are behavior problems, developmental delay, and/or difficulties with emotion regulation (Patterson, Reid, & Dishion, 1992). Likely related to the presence of these psychological deficits, foster caregivers of young children tend to demonstrate lower levels of emotional availability and more negative discipline strateg ies than foster caregivers of older children (Lawler, 2008; Linares, Maltanto, Rosbruch, & Li, 2006). To create stable placements for preschoolers in the foster care system that foster psychological growth, it is important to develop training programs that teach developmentally sensitive parenting strategies that promote a warm and nurturing environment. Kinship Caregivers: The Primary Placement Option The role of the kinship caregiver, like traditional foster parents, is to provide a stable placement for t he child in order to reduce negative psychological sequelae associated with multiple placements as well as further abuse or neglect. Kinship caregiver placements are believed to decrease the number of placements a child experiences when compared to placeme nts with a traditional foster caregiver (Barth,

PAGE 14

14 Courtney, Berrick, & Albert, 2004; Price et al., 2008); however, kinship caregiver placements remain controversial. Advocates of kinship caregiver placements believe that these caregivers provide continuity of family identity and knowledge, access to relatives and pre existing relationships, as well as residence within the ethnic, cultural, and religious practices of their biological parents (Hegar, 1999). Research supporting this argument confirms kinship ca regivers investment in maintaining connections to the childrens biological family and culture of origin (LeProhn, 1994). However, these studies have not explored whether or not remaining in the childs community of origin improves and/or protects against the difficult psychological sequelae experienced in foster care. Recent research has indicated that kinship caregivers are a very heterogeneous group of caregivers and it is difficult to make overgeneralizations about this population as a whole (Zinn, 2012). Zinn (2012) did find that, overall, kinship caregivers are more likely to adopt children in their care and to reduce the rate of displacement and older kinship caregivers have a higher rate of disruption than younger kinship caregiver. Due to the heter ogeneous nature of this population and the lack of studies investigating the psychological benefits of a kinship caregiver placement it is difficult to determine the extent of benefits from a kinship caregiver placements, but these placements may assist in reducing multiple placements under some circumstances (Zinn, 2012). Those opposing kinship placements argue that connections to the original social network where the child maltreatment occurred may be detrimental to the childs well being. Kinship caregi vers tend to engage in greater conflict with the child (Harden, Clyman, Kreibel, & Lyons, 2004) and endorse more depression, less marital support,

PAGE 15

15 and poorer health conditions than nonkinship caregivers (Cuddeback, 2004). Kinship foster caregivers also have fewer financial resources, less social support, less education, less marital support, and poorer health conditions than traditional foster parents (Harden et al. 2004), and it is unclear whether differences in parenting styles and caregiver mental heal th between kinship and nonkinship caregiver population are due to their social and economic disadvantages or other personality and cultural factors (Cuddeback, 2004). Specific research on how kinship caregiver characteristics, including caregiver depressi on and stress, affect the acquisition of training skills and knowledge has not been explored. Often overlooked when considering the controversy surrounding kinship foster caregiver placements is the fact that effective kinship foster training in behavior management and relationship building skills could result in placement success, lowered re abuse rates, foster caregiver continuation, and improvem ents in child psychopathology, as found in nonkin foster parent training programs (Chamberlain et al., 2008; Price et al., 2008; Timmer, Sedlar, & U r quiza, 2004). Despite evidence of positive outcomes following foster parent training, the unfortunate reality is that kinship caregivers seek fewer mental health resources than other foster caregivers (Leslie et al., 2000) and that children who have ever been placed with a kinship caregiver receive 41% fewer services than those in traditional foster placements (James, Landsverk, Slymen, & Leslie, 2004). Further, kinship caregivers are generally excluded from research on foster parent training. The need to study the effects of parent training within the kinship caregiver population in future research is clear.

PAGE 16

16 The Current State of Foster Caregiver Parent Training Programs and Interventions Attrition and Training Comple tion The factors that influence treatment completion are understudied in the foster caregiver population; most of the research on attrition has addressed foster caregiver termination from caregiving rather than from specific training or treatment programs. Inadequate training in managing child behavior problems and little support from agency personnel are the two most commonly reported reasons for termination of care of foster children (Brown & Calder, 2000; Buehler Cox, & Cuddeback 2003; Denby et al. 1999). However, parent management training increases support provided for foster caregivers as well as tools to manage child behavior, so the factors predicting attrition from caregiving may not apply to training programs. One possible influence on treatment completion in the child welfare population is an improvement in motivation following referral to parenting interventions and trainings. Chaffin et al. (2009) used a motivational intervention before beginning behavioral parent training, which improved attrition rates in low to moderately motivated caregivers in the child welfare system. This finding suggests that an increase in motivation may improve treatment attendance and resulting skill acquisitio n in the kinship foster caregiver population. Foster Caregiver Training Programs Child welfare agencies that evidence greater placement stability and improved child behavior during placement are successful because of two factors: high levels of foster car egiver involvement with the agency and foster caregiver training in positive reinforcement and behavior management techniques (Fisher, Ellis, & Chamberlain, 1999; Fisher & Chamberlain 2000; Grimm 2003). However, the two most commonly

PAGE 17

17 used programs to train foster parents, T he Model Approach to Partnership in Parenting (MAPP) and Parent Resources for Information Development and Education (The PRIDE Program), have little empirical support to suggest that they improve foster caregivers ability to man a ge diffi cult behaviors or increase involvement with the agency. MAPP is described as a program that prepares foster parents for the demands associated with managing child behavior problems, but this program does not succeed in implementing significant changes in caregiver empathy or use of physical punishment, nor does it increase knowledge of behavior principles, developmental expectations, or parenting roles (Puddy & Jackson 2003; Lee & Holland, 1991). The PRIDE Program also claims to prepare foster parents with the knowledge and ability to care for children in the child welfare system, but it has not been thoroughly experimentally evaluated for effectiveness (Grimm 2003). Understudied and nonempirically supported training programs for foster parents are a signi ficant problem in the child welfare system because caregivers are not adequately prepared to serve this at risk and vulnerable population of children. Multidimensional Treatment Foster Care Chamberlain and colleagues (2008) have developed evidencebased training programs that yield positive results for therapeutic foster care and enhance service delivery to maltreated children and youth. One of these programs, Multidimensional Treatment Foster Care (MTFC) has been classified as an evidencebased treatment for youth ages 9 to 18 in foster care (Fisher & Chamberlain, 2000). MTFC typically lasts for 6 to 9 months, and it has been shown to decrease child behavior problems, decrease foster parent attrition and limit disrupted placements (Chaffin & Friedrich 20 04; Chamberlain 2003; Eyberg, Nelson, & Boggs, 2008). In this program, foster parents,

PAGE 18

18 biological parents, and the youth work closely together with a treatment team to provide positive community experiences, manage problem behavior to prepare for permanent placements for the youth. The treatment team consists of a behavior support specialist who intervenes in the school system, an individual youth therapist, a family therapist who teaches behavior management to biological parents, a psychiatrist, a phone caller who assess daily behavior problems, and a case manager or team supervisor. Each of these members has a unique and well defined role on the team and they interact closely to ensure proper delivery of services and management of child behavior problem. MTFC has been adapted for application with younger children in foster care (Fisher et al., 1999) The adaptation, Multidimensional Treatment Foster Care for Preschoolers (MTFC P) focuses predominately on developing emotion regulation strategies for the children and behavioral management techniques for the biological parents and foster caregivers such as consistency, structure, clear expectations, rewards, redirection, and timeout (Fisher et al., 1999). MTFC P has been empirically evaluated with childre n ages 3 to 6 years old and typically the duration is 9 to 12 months, which includes a period of transition into a permanent placement (Fisher, Kim & Pears, 2009). MTFC P has been show to increase placement stability (Fisher et al., 2009), decrease foster parent self reported parenting stress (Fisher & Stoolmiller, 2008), and increase child attachment to foster parents ( Fisher & Kim, 2007). In addition to a daily phone caller, family therapist, behavior specialist, case manager, and a psychiatrist, the MT FC P program also includes weekly socialization peer groups.

PAGE 19

19 MTFC P is not yet designated an empirically supported treatment, but initial evidence has been promising. To execute MTFC or MTFC P with fidelity, agencies need to have the fiscal and personnel resources required to support social service workers with a case load no greater than 10 children for the duration of the childs placement in foster care (Chamberlain, 2003). Unfortunately, most child welfare agencies do not have the resources necessary to implement either of these programs (Chaffin & Freidrich, 2004). While these services may be beneficial for severely disturbed youth and children, it is not a feasible or cost ef fective model for use in agency wide foster parent training programs Furth er, these time and resource intensive programs may not be feasible for the kinship caregiver population because of the sheer number of existing kinship placements in the child welfare system. To address some of the resource difficulties with the MTFC and MTFC P treatment program, the principles of MFTC and parent management training were recalibrated into an efficacious training program, the Keeping Foster Parents Trained and Supported program (KEEP; Price at al., 2008). KEEP focuses on teaching positive reinforcement and non harsh discipline, and has been shown to decrease behavior problems and increase the stability of child placements (Price et al., 2008). One positive aspect of this training program is that it can easily include kinship caregivers because the intervention is conducted in community settings and in a group format. This program last s for 16 weeks, and each participant was paid $15 per session, home visits were made when a session was missed, and refreshments were provided at each session ( Price et al., 2008). The incentives for each caregiver and the length of the treatment

PAGE 20

20 program may be difficult to implement on an agency wide level. Furthermore, this intervention has not been researched with the most prominent population of foster childr en, children under the age of 5 years. T o develop cost effective and feasible training programs for kinship caregivers, it is important to select and design an intervention that takes into account the deconstructed elements of effective foster parent trai ning models that have been beneficial to establishing change in parenting practices. Effective foster caregiver training programs include live behavioral training with the child in session and support the development of caregiver child attachment and nurturance (Grimm 2003; Puddy & Jackson, 2003). In general, programs that yield higher reports of caregiver satisfaction, greater child behavior improvements, and more permanent placements focus on developing caregiver parenting skills, motivation, and self eff icacy while facilitating decreases in family stress and increases in social support (Buehler et al., 2003; Denby et al., 1999; Timmer, Urquiza, & Zebell, 2006). Live behavioral training that increases empathy, parenting skills, motivation, self efficacy, a nd social support while decreasing stress should be considered in the development of training program with kinship foster caregivers. Parent Child Interaction Therapy Parent Child Interaction Therapy (PCIT) is an evidencebased treatment for children wit h disruptive behavior problems (Eyberg et al., 2008) and has been identified as one of the best practices for treatment of maltreated children (Chadwick Center on Children and Families, 2004). PCIT includes invivo behavior al parent training which has b een found efficacious in foster caregiver training (Van Camp et al., 2008) and in parent training interventions with biological parents (Kaminski, Valle, Filene, & Boyle,

PAGE 21

21 2008) Previous studies of PCIT have also indicated that it is effective in improving positive parenting skills (Eyberg et al., 2008), and discipline strategies (Bhuiyan, 2012), which are components identified as important in all parent training interventions (Kaminski et al., 2008), while decreasing parenting stress (Eisenstadt, Eyberg, McNeil, Newcomb & Funderburk, 1993; Eyberg & Robinson, 1982; Schuh man n Foote, Eyberg, Boggs, & Algina 199 5 ). PCIT holds promise as an efficacious training program for kinship foster caregivers because it contains many of the elements that have shown to b e effective in other foster caregiver training models. Two studies have examined PCIT with both kinship and nonkinship foster caregivers. Timmer et al., (2006) used a quasi experimental design to investigate PCIT with nonkin foster parents and nonabusi ve biological parents. This study found significant preto post treatment decreases in child behavior problems and caregiver distress for the nonkin foster parents. These results were not statistically different than changes detected with nonabusive biological parents. An earlier study also conducted by Timmer et al., (2004) compared kin and nonkin foster parents receiving PCIT and found that kinship caregivers rated their childs behavior as significantly better and their own stress as significantly worse than nonkin foster caregivers at the beginning of treatment. T he outcomes of parental distress and child behavior problems were not reported, but results did indicate that kinship caregivers were more likely to complete treatment than nonkinship care givers (Timmer, et al., 2004). PCIT should be further investigated as a training intervention with kinship caregivers because nonkin foster caregivers have demonstrated both caregiver and child benefits from the intervention

PAGE 22

22 and kinship caregivers appear more likely to complete this intervention, despite high levels of initial distress. PCIT takes an average of 12 to 16 weeks to complete for families with oppositional and defiant children (Eyberg et al., 2008), a duration that can be difficult for agenci es to fund (Chadwick Center on Children and Families, 2004). To address this limitation, McNeil, Herschell, Gurwitch, and Clemens Mowrer (2005) adapted PCIT for administration in a twoday group training program with nonkinship foster caregivers to det ermine if a shorter length intervention could be effective for managing child behavior problems. The first day consisted of didactic instruction and roleplay of CDIT treatment components. The second day involved didactic instruction and coaching of an adapted time out procedure with the foster parent child dyad. Twenty seven of 30 families available at the 1month follow up assessment reported significantly fewer child behavior problems on a child behavior rating scale than before training; however, only 8 of the 30 families participated in the 5month follow up, so it was unclear if treatment gains were maintained over this extended timeframe. A structured interview assessed self reported satisfaction with skills learned in the workshop one month later and results indicated that only 68% of caregivers were satisfied with the timeout skills taught in the workshop and satisfaction with the workshop decreased significantly at follow up. It is difficult to draw conclusions from the limited amount of data repor ted, but it appears that the brief workshop was not sufficient in maintaining foster caregiver satisfaction with behavior management strategies taught in session, even though an initial reduction in behavior problems was perceived by parents

PAGE 23

23 Results from studies of PCIT with foster caregivers have indicated that PCIT reduces both caregiver stress and child behavior problems (Timmer, et al. 2006) and that kinship foster caregivers are likely to complete this intervention than nonkinship caregivers (Timme r et al., 2004). However, an initial attempt to translate PCIT into a more cost effective short term foster caregiver training program did not maintain caregiver satisfaction at the onemonth follow up, and initial improvements in child behavior were not followed past onemonth to determine the stability of these findings. Results from the McN eil et al. (2005) study indicated that several foster caregivers were also dissatisfied with the timeout procedure taught in this workshop. It is clear that future training models using PCIT principles need to extend training beyond a twoday intervention and need to use the components of PCIT that will facilitate the greatest change in attachment and responsive parenting practices for the kinship caregiver population. The Child Directed Interaction Training CDIT is the first phase of PCIT, which focuses on providing caregivers with concrete positive attention skills and strategies to withhold attention when undesirable behaviors occur. This exclusively positive play interaction, which parents practice at home with their child each day, improves the qual ity of the caregiver child relationship (Eyberg et al., 2008). By providing caregivers with the skills to respond selectively to child behavior, the caregiver learns to be more sensitive and emotionally available to the child, and to respond with greater w armth to creat e a more secure caregiver child attachment (Floyd & Eyberg, 2003). Caregivers first practice following the childs lead in play, which allows them to practice each of the positive attention skills; Praise, Reflect, Imitate, Describe, and

PAGE 24

24 Enj oy ( Boggs & Eyberg, 2008). They also learn to ignore any negative behaviors to reduce the attention children have previously received for these behaviors. Caregivers are also taught to avoid asking question, giving commands, or stating criticisms of the ch ilds work or attributes. These leading and intrusive verbalizations are typically used by caregivers with the intention of engaging the child in more appropr iate prosocial behaviors, but instead they facilitate attention to the negative behaviors, increasing their frequency. Coaching of these relationship building skills serves to rapidly concretize these skills for the caregiver through reinforcement by the constant praise of the caregivers parenting skills by the therapist and the childs natural response to differential social attention. The development of evidencebased trainings, like CDIT, that builds positive parenting skills associated with improved child outcomes is the next step towards agency wide implementation of brief and effective kinship f oster caregiver training The Child Directed Interaction (CDIT) phase h as not been independently investigated as a training program for kinship foster caregivers, but holds promise for this population because it focuses on increasing the caregiver child at tachment relationship and provides a behavioral management technique (differential social attentions) that is powerful but nonconfrontational (Boggs & Eyberg, 2008; Zi sser & Eyberg, 2010). This combination has been shown to result in high foster caregiver satisfaction ( Grimm 2003; Puddy & Jackson, 2003). Increasing warmth in the kinship caregiver child relationship is particularly important because these caregivers often have parenting styles characterized by low emotional availability and harsh discipline (Harden et al., 2004; Lawler, 2008). Improvements in CDIT, when measured before the

PAGE 25

25 second phase of PCIT has begun, ha ve been shown to reduce parenting stress and improve child behavior problems (Eisenstadt et al., 1993) while increasing attachment ( Fussell, Nzi, & Eyberg, in progress). These findings provide further suggestion that CDIT alone may be sufficient for i mproving parenting skills among kinship caregivers. Current Aims and Study Hypotheses The purpose of this study was to evaluate the impact of CDIT on the parenting skills of kinship foster caregivers and to determine factors that moderate and predict parenting skill acquisition and predict treatment completion. A randomized, wait list control trial design was used to evaluate the effectiv eness of CDIT as a parenting intervention for kinship caregivers who express difficulties managing their child s behavior problems Caregiver parenting skills acquisition was measured using both observational and self report methods. Caregiver depression, parenting stress, and motivation were measur ed using self report instruments Study objectives and aims were as follows: Aim 1. To compare betweengroup differences between the CDIT immediate training (IT) and wait list control (WLC) conditions. Post train ing outcome measures from the IT were compared with the WLC, controlling for pre assessment scores in the two conditions. Hypothesis 1.1. Kinship caregivers in the IT will demonstrate significantly greater use of observed positive attention skills and si gnificantly fewer negative leading skills than kinship caregivers in the WLC. Hypothesis 1.2. Kinship caregivers in the IT will report less critical verbal force and nonphysical verbal force and will use more limit setting in their discipline practices th an the WLC.

PAGE 26

26 Hypothesis 1.3. Caregivers in the IT will report greater reductions in parenting stress than caregivers in the WLC. Hypothesis 1.4. Caregivers in the IT will report greater reductions in depressive symptoms than caregivers in the WLC. Aim 2 To quantify the relationship between caregiver variables at pretraining and skill acquisition. Hypothesis 2.1. Caregivers with greater endorsements of parenting stress and depression at pretraining will demonstrate less skill acquisition that caregivers with lower self reported parenting stress and depression at post training when group assignment is controlled. Aim 3 To identify factors that moderate the acquisition of positive attention skill Hypothesis 3.1. The change in caregiver motivation from pre to midtraining will moderate the relationship between preand post positive attention skills when controlling for group assignment. Aim 4 To predict factors that influence attrition. Hypothesis 4.1. High self report ratings of depressive symptom s and caregiver stress as well as low levels of caregiver motivation at pretraining will predict attrition during CDIT. Aim 5 To explore the maintenance of self report ed discipline techniques, stress, and depression 3 months following treatment completion. Hypothesis 5.1. Within the IT group, reductions in stress, depression, and critical verbal force will maintain training gains at the 3 month follow up. Improvements in limit

PAGE 27

27 setting and noncritical verbal force discipline strategies will maintain training gains at the 3 month follow up.

PAGE 28

28 CHAPTER 2 METHOD Participants Participants were 12 female kinship caregivers and their 2to 7 year old children. A priori power analyses were conducted using G*Power 3.1.2 statistical software to determine the number of caregiver child dyads needed to detect statistical betweengroup differences. This analysis was conducted using the Parenting Stress Inventory Short Form data from a previous CDIT outcome study of biological mother child dyads participating in treatment for disruptive child behaviors (Harwood & Eyberg 2004). Power was set at .80 with an alpha value of .05 to yield an 80% confidence interval for statistically significant results. Based on this analysis, it was determined that a total sample size of 14 would be needed to detect between group differences. The estimate of needed sample size is higher than the obtained sample of 12 participants and the large number of analyses performed with this sample increases the likelihood of false positive results (Christley, 2010). This is a common problem faced in psychological studi es and clinical trials (Maxwell, 2004). Underpowered studies can also increase the likelihood of detecting false negative results when conducting a smaller number of analyses (Maxwell, 2004). Because both false positive and false negative results could occur in the data, the findings of this study may be less meaningful than those detected in a study of adequate sample size. Participants were referred in the following ways: two were referred after being screened out of another parent training study in the Child Study Laboratory at the University of Florida, two were self referred after learning of the study at the Library Partnership, two were referred by a nurse practitioner at Childrens Medical Services,

PAGE 29

29 three were self referred after seeing fliers, news paper ads, or other recruitment materials announcing the study, one was referred by a pediatrician, one was referred from the Psychology Clinic at the University of Florida, and one was referred by the University of Florida Multidisciplinary Diagnostic and Training Program. Kinship c aregiver child dyad inclusion c riteria Children eligible for this study met the following criteria: (a) were between the ages of 2 and 7, (b) resided in a kinship caregiver home for the duration of the study, and (c) had a caregiver rating at or above 15 on the Eyberg Child Behavior Inventory (ECBI) Problem Scale, a score 1 standard deviation above the normative mean for parents of young children on this scale. Kinship caregivers eligible for this study (a) were able to communic ate by phone or email, (b) agreed to attend twice weekly 60 minute sessions for four weeks, (c) agreed to 5minute daily practice of skills, and (d) agreed to complete the four training assessments. Kinship c aregiver c hild dyad exclusion criteria Childre n were excluded from this study if they had a history of major sensory impairment (i.e., deaf or blind) or a preexisting or suspected diagnosis of a Pervasive Developmental Disorder, as indicated by an elevated score on the Child Behavior Checklist (CBCL) Pervasive Developmental Disorders subscale. No child scored within the clinically significant range on the CBCL Pervasive Developmental Disorders subscale. One child was excluded during the screening process for a previous diagnosis of Autism and was referred to another research study. The mean age of children at the time of the phone screen was 5.3 years, and the range was 2.0 to 7.4 years. The children were 58% female. The racial/ethnic distribution in this study was 58% Caucasian, 25% African American, 8% Hispanic, and 8% biracial

PAGE 30

30 which is representative of North Central Florida. On average, externalizing behavior problem scores on the CBCL were clinically significant M t score = 67.00, SD = 6.84 and all but 3 of the 12 children were clinically signific ant at the baseline assessment, t score range = 56 80. On average, c hild behavior problems on the ECBI intensity scale, M Raw Score = 155.15, SD = 23.47 were also clinically significant with all but 2 of the 12 children with scores above the clinical cut off at the baseline assessment Raw Score range = 122 202 On average, c hildren were below the clinically significant range on the CBCL Internalizing subscale, M T score = 58.92, SD = 10.00 and only 4 of the 12 children had clinically significant scores tscore range = 43 77. Caregivers were grandmothers (83%) and great grandmothers ( 17%), and half the familie s were single caregiver homes. Single caregiver homes are common among kinship caregi vers and occur at a rate of 52 to 77% depending on the geographic region of t he study (Berrick, Barth, & Needell 1994; Dubowitz, Feigelman, & Zuravin,1993; McLean & Thomas, 1996).The average age of the caregivers was 57.5 years with a range of 45.9 to 73.0 years. This population was older than other demographic research studies, which report average kinship caregiver age of 48 years old (Berrick et al., 1994, Dubowitz et al., 1993). In this study, c aregivers were 58% Caucasian, 25% African American, and 17% Hispanic. Caregiver r eported education levels indicated that 8% had less than a high school education, 8% held a high school diploma, 33% had completed some college, 30% completed college, and 16% held a graduate degree. Studies reporting education levels of kinship caregivers vary significantly by region, and have ranged from as few as 20% (Dubowitz et al., 1993) to as high 53% (Berr ick et al., 1994) of kinship caregivers receiving post high school education. However, the present

PAGE 31

31 sample of caregivers in this study appear s to be more educated than other studies The mean annual family income was $43,000, the range was $11,000 to 80,000, and the median family income was $37,000. Caregivers demonstrated clinically significant levels of total stress o n the PSI SF, M = 90, SD = 14.97, and reported a clinically signif icant level of difficulty with managing their childs behavior on the ECBI Problem Scale, M = 22.25, SD = 4.03. Regarding characteristics specific to children and caregivers in the foster system, three caregivers reported receiving previous parent training. Two caregivers reported having custody of another foster child. The mean length of placement of the index child in the caregivers home was 3.1 years with a range of 5 months to 7.4 years. Sixteen percent of children were adopted, 16% were in permanent guardianship, and 68% were in temporary custody. Families were randomly assigned to the Immediate Training (IT; n = 6) or Waitlist Control group (WLC; n = 6). There were no significant differences between the two groups on continuous demographic variables (Table 21 ); betweengroup differences in categorical variables were not statistically analyzed because the sample size was too small to meet Chi squared assumptions. They are also reported in Table 21. Of the 13 participant dyads, one family that was assigned to the IT condition was lost to contact after signing informed consent but before being informed of their group assignment. Of the remaining 12 families, 6 were randomized to the IT condition and 6 were randomized to the WLC condition. Among the IT families that began the training all 6 completed the 8week training and post treatment assessment, and 5 IT families participated in their 3month follow up assessment. Among the 6 WLC families, all

PAGE 32

32 completed the post wait assessmen t after 8 weeks on the waitlist Fi gure 21 displays participant screening, enrollment, and randomization. There was no attrition from either the treatment or the wait condition after the family informed of their group assignment. Only one family was lost to follow up analys es. Finding such a low rate of attrition from PCIT (or CDI) has not previously been reported in any PCIT outcome study. Previous PCIT studies at the University of Florida have reported attrition rates ranging from 27% to 47%; (Boggs et al., 2004; Fernandez and Eyberg 2009; Schuhmann et al., 1998; Werba et al., 2006). Studies of PCIT outcomes within community settings report a wider range of attrition, from 12% to 77% (Chaffin et al., 2009; Phillips, Morgan, Cawthorne, & Barnett et al., 2008; Timmer, Urquiza Zebell, & McGrath, 2005). Trainers All CDIT sessions were conducted by graduate student trainers who met PCIT International standards for certification as a PCIT therapist with the exception of holding a professional license to practice independently (Eyberg, Funderburk, McNeil, Niec, Urquiza, & Zebell 2009). To meet these requirements, individuals must receive formal instruction in PCIT in a graduatelevel academic course or a week long (40 hour) inte nsive training and must complete two PCIT cases under the supervision of a qualified trainer. PCIT therapists in this study were called trainers instead of therapists to reduce stigma associated with mental health services. Once trainers were assigned a case, they attended a brief meeting with the principal investigator to review CDIT protocol, discuss special considerations of this population, and learn the librarys procedures for facility use. Therapists met weekly with a principal investigator to mo nitor progress and address any concerns. Weekly meeting s were held between

PAGE 33

33 the principal investigator and the dissertation supervisor, Sheila Eyberg, Ph.D., to monitor dyad progress and to address issues of clinical relevance. Assessors Undergraduate research assistants taking course credit in the Department of Clinical and Clinical Health Psychology Child Study Laboratory at the University of Florida served as the primary assessors in this project. Assessors on this project were trained by the principal investigator in how to administer study measures and met with the PI for supervision and consultation immediately before and after each assessment to ensure proper execution of the assessment protocol. Trained graduate and undergraduate assessors completed the coding in both coding systems used in this study: the Dyadic Parent Interaction Coding System 3rd edition (DPICS III) and the Parent Daily Report/Daily Discipline Interview (PDR/DDI). All coders for both systems were reliable at a level of 80% or hig her on training materials and were uninformed as to the treatment condition of the participants when coding dyad observations and parent report discipline practices. A reliability coder coded one third of all DPICS III and two fifths of all PDR/DDI observ ations. Coding activities were supervised by the principal investigator. Measures A list of all measures and a chart of when each measure was collected can be found in Table 22 Eyberg Child Behavior Inventory The ECBI (ECBI; Eyberg & Pincus, 1999) is a 36 item caregiver report measure of disruptive behaviors in children aged 2 16. It measures disruptive behaviors in terms of their frequency (Intensity Scale) and the degree to which these behaviors are

PAGE 34

34 problematic for the parent (Problem Scale). The problem scale was used in this study as an inclusion criteri on measure, and the intensity scale was administered at every training session to monitor training progress, as indicated in the PCIT protocol. The ECBI consists of 36 items where the caregiver respon ds yes or no to the question on the Problem Scale, Is this a problem for you?, and on the Intensity Scale marks the frequency with which problematic behaviors occur. On the Intensity Scale, the caregiver indicates the frequency of the behaviors on a 1{never) to 7 (always) Likert type scale. A screening ECBI Problem score of 15 or higher was required for inclusion in the study. Within a community sample, the problem and intensity scales have 12week test retest reliabilities of .85 and .80, and 10month test retest reliabilities of .75 and .75, respectively (Funderburk, Eyberg, Rich, & Behar, 2003). Cronbachs Alpha for this study was .82 for the Intensity scale and .72 for the problem scale. Child Behavior Checklist There are two forms of the CBCL (CBCL 1.5 5 years; 618 years; Achenbach, 1991; Achenbach, Dumenci, & Rescorla, 2003) adm inistered to the parents of children in different age ranges. The CBCL is a comprehensive instrument designed to assess the frequency of a variety of specific behaviors in c hildren during the past 2 months for the 1.55 year version and during the past 6 months for the 618 year version. Caregivers rate their childs behavior or emotional presentation on a 100item (1.5 5 year version) or a 113item (6 18 year version) 3point Likert type scale. The Pervasive Developmental Disorders subscale contains 13 items that were used to screen for Autism. This scale has 1week test retest reliability of .83. Cronbachs alpha in our sample was .30.

PAGE 35

35 Dyadic Parent Child Interaction Coding System: Third Edition DPICS III ( DPICS III; Eyberg Nelson, Duke, & Boggs, 2005) is an observational coding system that measures specific parent verbalizations representing positive and negative parent child communication skills. The communication categories are recorded during standard play situations that require varying degrees of parental control. In this study, only the 5minute child led play situation was observed. In childled play, parents are asked to follow along with their child in any play activity that the child chooses. This situation was selected to measure parent CDIT skill acquisition. In the DPICS III, e very verbalization is coded in one of the following categories: B ehavior D escription, R eflection, L abeled P raise, U nlabeled P raise, D irect C ommand, I ndirect Command, Neutral Talk, Descriptive Question, Informational Q uestion, or N egative T alk. Two common composite categories relevant to the play situation, positive attention and negative leading, were used to analyze changes in parenting communication (Eyberg et al., 2005). The first category, percent P ositive A ttentio n, represents the percentage of caregiver verbalizations indicating the parent is attending to the positive behaviors of the child. This category is calculated by summ ing the frequency of B ehavior D escriptions, L abeled P raises, U nlabeled Praises, and R efle ctions and then dividing by the total number of observed parent verbal izations. The second category, N egativ e Leading, represents the percentage of criticisms, questions and ineffective commands given in the child led play. These leading and intrusive verbalizations are typically used by caregivers with the intention of engaging the child in more appropriate prosocial behaviors, but they instead facilitate attention to the negative behaviors, increasing their frequency and detracting from the nurturing r elationship. Negative leading is calculated by sum ming Information Q uestions,

PAGE 36

36 Descriptive Q uestions, Indirect C ommands, an d N egative T alk divided by the total verbalizations. Graduate students served as primary coders. All coders completed sections of t he coder training workbook (Fernandez, Chase, & Eyberg, 2005) that correspond to the DPICS III categories used in the current study. They were also required to meet 80% accuracy with two criterion tapes coded by expert DPICS III coders. All coders in this study were uninformed of the experimental condition of the families. Inter coder reliability was calculated using both percent agreement and Kappa indices of reliability. The overall Kappa reliability was .86 and ranged from .80 to 1.00 for each category coded. Total Percent Agreemen t was .88 and ranged from .86 to 1.00 for each category c oded in this study (see Table 23 ). Parent Daily Report/Daily Discipline Inventory The PDR (PDR; Chamberlain & Reid 1987; Webster Stratton & Spitzer, 1991) is a 20item questionnaire administered to parents by telephone for 5 consecutive days to obtain information on the daily frequency of child disruptive behaviors. The PDR has been shown to have test retest reliability of .62 to .82 (Chamberlain & Reid, 1987). Raw scores on this instrument were not used in this study; it was administered to permit administration of the PDR/DDI a companion measure of parent responses to the negative child behaviors reported on the PDR. An adapted version of the DDI created for a previous PCIT study (Bhuiyan, 2012) was used to assess change in discipline practices. Caregiver discipline procedures were categorized into nine broad codes: P hysical F orce, C ritical V erbal Force, N on C ritical V erbal Force, L imit S etting, T eaching, Positive V erbal Reinforcement, P ositive N onverbal Response, Lack of Response/No F ollow Through, and P arents S tate of

PAGE 37

37 M ind. This study investigated the three DDI categories that demonstrated change after PCIT in a previous study, Critical Verbal Force, Noncritical Verbal Force, and Limit Setti ng (Bhuiyan, 2012). Appendix B gives examples of coding definitions for these categories. Scores o n the original DDI demonstrated inter rater reliability of .94 for Critical Verbal Force, and .97 for Limit Setting. Test retest reliability was .59 for Criti cal Verbal Force, and .43 for Limit Setting. Internal consistency coefficients were as follows: .74 for Critical Verbal Force, and .59 for Limit Setting. No psychometric data w ere reported in the original study for the Noncritical Verbal Force Category. In the current study overall percent agreement was .82 and percent agreement by category was .76 for Critical Verbal force, .66 for Noncritical verbal force and .91 for Limit S etting. The K appa statistic was .83 for the Time 1 administration of this instrument Beck Depression Inventory II The BDI II (BDI II; Beck, Steer, & Brown, 1996) is a 21item self report questionnaire assessing the severity of somatic affective and cogni tive symptoms of depression in adults. Each item assesses a particular depressive symptom ( e.g. Appetite, Sadness ), typically providing a four choice response option ranging in severity of symptoms on a scale ranging from 0 to 3. Total scores are computed by summing all individual item scores and can range from 0 to 63 depending on the severity of depressive symptoms ( e.g., 0 13 minimal, 14 19 mild, 20 28 moderate, 29 63 severe ). The BDI II has a mean of 28.64 with a standard deviation of 11.75 in a clinically depressed outpatient sample (Steer, Ball, Ranieri, & Beck, 1999). Cronbachs alpha for this questionnaire in the standardization sample was .91, and test retest reliability was .93. Cronbachs alpha for this sample was .91.

PAGE 38

38 In the event that a caregiver endorsed item 9 on the BDI II, which refers to suicidal thoughts or wishes, the assessor would have immediately performed a standard suicide assessment. If further action were indicated ( i.e., the participant was in danger of harming herself), the assessor would have consult ed by telephone with the faculty supervisor, Dr. Sheila Eyberg, regarding the optimal course of action. All caregivers who endorse this item, regardless of the outcome of the suicide assessment, would have been provided with contact information for the Alachua County Crisis Center as well as the National Suicide Prevention Lifeline. However, no caregivers endorsed this item during any of the assessments. Parenting Stress Index Short Form The PSI SF (PSI SF; Abidin, 1995) is a 36item self report questionnaire measuring child, parent, and situational characteristics associated with parenting stress and dysfunctional parenting. Each item is rated on a 5point scale ranging from (1) Strongly Disagree to (5) Strongly Agree to i ndicate the degree to which the caregiver agrees with the statement. The PSI SF manual suggests three factors; however, a recent study suggests that a two factor scale is more appropriate for the PSI SF (Haskett, Ahern, Ward, & Allaire, 2006). These two f actors include a Personal Distress factor ( i.e. feel trapped by parenting responsibility, never get to do things I want to do) and a Childrearing Stress factor ( i.e. child is moody and easily upset; my efforts for child arent appreciated) Cronbachs alpha for these two factors were .78 (Personal Distress) and .91 (Childrearing Stress) (Haskett et al. 2006). C ronbachs alpha in this sample wa s .66 for the C hildrearing S tress factor and .83 for the T otal S cale.

PAGE 39

39 Readiness, Efficacy, Attributions, Defensiveness, and Importance Scale The R eadiness, E fficacy, A ttributions, D efensiveness, and I mportance scale (READI; Brestan, E. V., Ondersma, S. J., Simpson, S. M., & Gurwitch, R., 1999) is a 43 item questionnaire assessing parents motivation for chang e via behavioral parent training and includes an 18item validity scale. Scoring is completed by summing individual scores, excluding the lie scale, to receive a Total Motivation score. Items are rated on a 5point Likert type scale ranging from (1) Strongly Disagree to (5) Strongly Agree. This instrument did not have published psychometric properties, but there was no measure of motivation for parent behavioral training with published psychometric properties. This measure was selected because it contains items relevant to the kinship foster caregiver populations. Cronbachs alpha in this sample is .63. Procedures After receiving approval from the University of Florida Health Science Center Institutional Review Board, a randomized wait list controlled study was conducted comparing the IT group with the WLC, allowing investigators to determine the specific effects of treatment while controlling for the effect of time. This study design meets Brestan & Eyberg (1998) criteria for a well conducted study. Well conducted groupdesign studies include prospective study design, clear inclusion/exclusion criteria, an appropri ate control condition, random assignment, reliable measures, clearly specified sample characteristics, clearly described statistical procedures, and use of defined treatment protocol with ways to assess integrity (Brestan & Eyberg, 1998). Figure 22 displa ys the participant flow from the preassessment to follow up.

PAGE 40

40 Initial Phone Screening Initial study screening was conducted during the first contact with interested participants. Basic identifying information was collected including name, contact informati on, and the presence or suspicion of Autism or any major sensory impairment. Study procedures were reviewed with the caregiver, including time requirements, wait list design procedures, and inclusion and exclusion criteria. If caregivers indicated interest in participating, a phone screener consisting of the problem scale of the ECBI was administered and kinship caregivers endorsing the presence of 15 or more problematic behaviors were scheduled for assessment. Eligible families were scheduled for a Time 1 assessment at the Library Partnership (LP). Thirty caregivers were excluded from the study because they were biological parents or non kin adoptive parents. Kinship caregivers screened for the study were excluded for the following reasons: 5 were uninteres ted after hearing about study protocol, 1 had a grandchild with a previous diagnosis of autism, 5 were noncustodial grandparents, 2 never completed the screening process, and 1 did not meet criteria on the ECBI Problem Scale. All excluded families were pr ovided referrals for treatment from other providers. Time 1 Assessment All assessment and training sessions took place at the LP. C aregivers brought paperwork to the Time 1 assessment confirming custody or guardianship as proof that they were qualified to provide consent on behalf of the child to participate in this research study. The Time1 assessment consisted of review of Institutional Review Board consent procedures and the completion of the CBCL to screen for the presence of a possible pervasive developmental disorder. No participant was excluded from the study during the Time 1 assessment. The assessment continued with a 5minute warm -

PAGE 41

41 up and then 5minute video recorded, standard observation of the childled play situation, later coded using the DPICS III system. Caregivers were also administered the following questionnaires: PSI SF, BDI II, READI, and a Demographic Questionnaire. After completing questionnaires, a second standard childled play observation was conducted following the same procedures as the first observational assessment. These administrations were separated by approximately 180 minutes. The standard DPICS III observation consists of two observations separated by one week of time. Both observations were conducted during a single visit in order to minimize caregiver burden of multiple assessment visits and increase the likelihood of assessment completion. The family was then informed of the PDR/DDI assessment procedures and a working phone number was confirmed. The PDR/DDI was administered by telephone for 5 evenings following the Time 1 assessment. Caregivers were asked to report discipline procedures used to manage a standard list of inappropriate child behaviors (e.g. talking back, not following directions, refusing to take no for an answer, using bad language, Chamberlain & Reid, 1987). PDR/DDI assessors wrote down all responses verbatim empathically encouraged the caregiver to elaborate. All responses were later coded using the DDI coding system. After completion of all assessment proc edures, participants were randomized to either the IT or WLC condition using a random number generator. Participants assigned to the IT condition began training within one week of the completion of their assessment, and those in the WLC condition were scheduled for their Time 3 assessment 5 weeks later.

PAGE 42

42 Training Procedures Caregiver training consisted of 8 twiceweekly sessions of CDIT. The first session was 90 minutes in length and consisted of an unstructured clinical interview to establish rapport and t o gather information about the presenting problems of the family, types of discipline used by the caregiver, family functioning, caregiver expectations, circumstances associated with placement changes, and other relevant clinical information. Following the clinical interview, caregivers were taught the CDIT skills using multimodal instruction techniques including explanation, modeling, and role play, following the standard PCIT protocol (Eyberg & Child Study Lab, 1999). Caregivers were taught the Do ski lls including Labeled Praise, Reflection, Imitation, Behavioral Description, and Enjoy as well as the Dont skills, Questions, Criticisms, and Commands. In addition, caregivers were instructed in the principles of active ignoring as a method to manage negative, attentionseeking behaviors. At the end of the session, the trainer and caregiver determined when and where in the home the caregiver would practice 5minutes of special time, or CDIT practice, everyday, as indicated in the PCIT protocol. Caregivers were asked to record homework practice on a homework sheet to assist with tracking homework completion rates. The next two sessions of CDIT were scheduled and/or confirmed following each session, and reminder phone calls were typically made the day of the appointment to encourage attendance. All subsequent training sessions, called CDIT Coach sessions, were 60 minutes in duration and began with completion of the ECBI Intensity Scale, 5 10 minute discussions of factors unrelated to the child s behavior according to protocol, 5 to 10minute homework review, 5 minutes of coding of the CDIT skills, and an average of 22

PAGE 43

43 minutes of live coaching per session. Coaching occurred inroom instead of through a bug in the ear device since no twoway mir ror was available. Previous PCIT research has been conducted using a bug in the ear device and no studies to date have explored the use of inroom coaching versus bug in the ear so it is uncertain what impact, if any, there is on skills acquisition using t his coaching method. During in room coaching, the trainer positions themselves behind the caregiver to conduct coaching and the child can hear instructions given as opposed to bug in the ear coaching where the trainer is behind a twoway mirror and the chi ld cannot hear what they are saying. Coaching focused on increasing Do and decreasing Dont skills that which were not yet at mastery criteria during the live coding. Caregivers were coached in active ignoring when necessary to facilitate changes in child behaviors and parent attending. The first three coaching sessions focused solely on coaching the CDIT skills, and the remaining four sessions also included discussion and provision of the PCIT optional handouts on creating effective labeled praises, m odeling, other discipline tools, and kids and stress/seeking help. Upon completion of the seven coaching sessions, families were provided with a small prize for successful completion, and additional referral options were discussed if necessary. In standard PCIT protocol, caregivers continue in CDI sessions until meeting master y criteria which consist of 10 L abeled P raises, 10 R eflections, 10 B ehavior D escriptions and fewer than 3 C ommands, Q uestions, and C riticisms combined during the 5minute coding period. Previous studies of PCIT have indicated that it usually takes an average of 6 sessions to meet mastery criteria (Harwood & Eyberg, 2004). For this study, caregiver participated in timelimited CDIT which consisted of 8 twiceweekly

PAGE 44

44 sessions, however, fam ilies were still coached to and encouraged to meet mastery criteria. By the final training session, 4 of the 6 caregivers met CDIT criteria for the Do skills and 3 of the 6 met CDIT criteria for the Dont skills. Coaching was performed inroom instead of using the standard bug in the ear device due to facility restrictions at the LP. On homework sheets given to caregivers at each session to record betweensession practice of CDIT skills, caregivers reported completing homework, on average, 67% of t he assigned days with a range of 33% 93%. The ECBI intensity scale was completed at each training session. Prior to training, all caregivers in the IT group reported child behavior problems in the clinically significant range (ECBI Intensity score > 132) Following training, 67% of IT caregivers rated their childs behavior within the normative range and below the criteria for treatment completion in standard PCIT (i.e., ECBI Intensity score < 114). The change in ECBI scores was both clinically and statis tically significant, t (5) = 7.47, p = .001, r = .96. Figure 23 shows the change in ECBI scores by session. Training Integrity Training integrity was evaluated by examining session outline checklists. Trainers self recorded their training activities on provided training integrity checklists from the PCIT manual (Eyberg & Child Study Lab, 1999). Undergraduate research assistants calculated the total self reported, completed training activities and calculated a percentage of accuracy after dividing the report ed activities by the total possible activities. Self reported integrity was 96%.

PAGE 45

45 Time 2 Assessment The Time 2 midwait or mid treatment assessment occurred three weeks following the Time 1 assessment (at the end of the 3rd CDIT Coach session for the IT gr oup) and caregivers completed the READI by mail, phone, or in person during this assessment. Time 3 Assessment The Time 3 assessment occurred two weeks after the Time 2 assessment after treatment completion for the IT and before treatment began for the WL C). All assessment procedures and data coding were conducted by an undergraduate research assistant or a graduate student that was unaffiliated with the training of the IT participants. Caregiver child dyads participated in two standard 5minute observations of the child led play situation, one at the beginning of the assessment and one at the end, which were identical to the procedures of the Time 1 play observation. Caregivers completed the PSI SF and BDI II during this assessment, and the PDR DDI was adm inistered by phone for 5 days following this assessment. Following this assessment, the WLC began their training sessions following the same training protocol. The IT group was reminded that they would be contacted in 3 months to complete the follow up assessment. Both groups were compensated $10 at this assessment. Time 4 Assessment For the IT group, the Time 4 assessment was a 3month follow up assessment. Caregivers were contacted via telephone for administration of the PDR/DDI for 5 days and completion of the PSI SF and BDI II either by mail or by phone. Families were compensated $15.00 for their participation in this follow up assessment. The WLC group did not serve as a time control for the 3 month follow up assessment due to the ethical implications of delaying intervention for this group for such a great length in time.

PAGE 46

46 The WLC group received $15 for the completion of a few short post treatment questionnaires that were not included in the study. This final assessment of the WLC group was completed in order to balance the assessment number and the compensation amount for the two conditions. Statistical Analyses Data analysis was conducted using the Statistical Package for the Social Sciences, Version 20.0. Initially, data were screened for outliers 3 standard deviations above the mean. No outliers were present on any measures. Univariate normality was then assessed by visually scanning descriptive statistics, boxplots, and histograms with normality curves and by ensuring that all measures had absolute values of kurtosis and skewness less than one, Z score statistics below 2.58, and nonsignificant Kolmogorov Smirnov and ShapiroWilk tests (Field, 2005). All variables met normality assumptions. Descriptive statistics, including means and standard deviati ons, for the BDI II READI Total Motivation Score, PSI SF Childrearing Stress factor, PDR/DDI Critical Verbal Force, Noncritical Verbal Force, and Limit Setting categories, and the DPICS III Positive Attention and Negative Leading composites are reported for each analysis. To investigate Aim1, group difference at Time 3 in parenting stress, depression, observed positive verbalization, and reported discipline strategies were examined using within subject analyses of covariance, ANCOVA, controlling for Time 1 scores. This method is recommended for randomized controlled trials because it corrects for the shared variance between preand post scores and is, therefore, a more statistically powerful method than a repeated measures ANOVA (Rausch, Maxwell, & Kell ey, 2003). Effect sizes for the ANCOVAs were calculated using the more conservative r statistic

PAGE 47

47 rather than eta squared; the formula for the r statistic, / + df, is computed using the t statistic and an effect size greater than .50 was considered a large effect (Field, 2005). To examine Aim 2, an ANCOVA was used as a multiple linear regression to determine if childrearing stress and depression were predictors of positive parenting skill acquisition. Using the ANCOVA combines the oneway ANOVA with a regression analysis, making it possible to investigate continuous predictor variables while also facilitating the entry of dichotomous variables (Field, 2005). This statistical procedure prov ided the opportunity to control for group design in the model to determine if the continuous predictor variables, caregiver stress and depression, explained a significant proportion of the variance above and beyond that which was explained by experimental condition alone. Aim 3 of this study was to examine the change in caregiver motivation from Time 1 to Time 2 as a potential moderator affecting the relationship between caregivers positive attention skills from Time 1 to the Time 3 assessment. Previous P CIT research had indicated that low motivation caregivers in the child welfare system benefit from a motivational component at the beginning of treatment, providing evidence that a low versus high split was warranted (Chaffin et al., 2009). Before this analysis could be conducted, it was necessary to define the cutpoint defining high versus low motivation on the READI, because an already established cutpoint had not been determined. When a statistical procedure requires a dichotomous variable for a construct measured continuously, arbitrary guidelines are often applied in selecting cutpoints to test the moderation analysis, such as using a median split or the pick a point method. (Hayes

PAGE 48

48 & Mathes, 2009). This moderation analysis used the JohnsonNeyman technique (Johnson & Fay, 1950) to address the concerns with arbitrarily selecting a cutpoint using other methods. This approach is recommended for use with a continuous, moderator variable because it calculates regions of significance in the continuous v ariable that can be used to divide data into the high motivation and low motivation groups (Hayes & Matthes, 2009). The JohnsonNeyman technique also corrects for large amount of shared variance, multicolinearity, by centering the variables used in th e model. This moderation analysis was conducted using the MODPROBE macro for SPSS as recommended by Hayes and Matthes (2009). Aim 4 of this study was to examine the predictors of attrition. This aim could not be achieved because there was only one famil y lost to contact prior to notifying them of their group assignment. The final aim of this study was to explore the maintenance of treatment gains at a 3 month follow up assessment for families in the IT condition. The first step was to conduct repeatedmeasures t tests to determine if changes in caregiver stress, depression, or reported discipline practices from Time 3 to Time 4 were significant. However, maintenance of gains between Time 3 and Time 4 would result in a nonsignificant difference, from which conclusions cannot be drawn. Therefore, the second step was to conduct equivalency testing to determine if the nonsignificantly different scores between Time 3 and Time 4 were maintained at a statistically equivalent level (Rogers, Howard, & Vessy, 1993). Equivalence within a margin of 20% is the recommended percentage for determining statistical equivalence in clinical trials (Rogers et al. 1993) and was selected for this study.

PAGE 49

49 Table 2 1 Demographic Characteristics of Immediate Training and Wait List Groups Note. IT = Immediate Training; WLC = Wait List Control. It was not possible to analyze differences between groups for the categorical variables because the sample size was too small to meet the Chi Square calculation assumptions. an = 5 IT WLC Characteristic M SD M SD t ( 11 ) p Child age (months) 69.00 12.33 58.67 26.55 .87 .42 Child sex (% female) 67% -50% ---Placement Length (months) 25.55 26.62 48.87 31.21 1.39 .19 Caregiver age (years) 57.40 9.00 57.74 9.20 .063 .95 Caregiver ethnicity (% minority) 50% -44% ---Single parent home (%) 67% -44% ---Caregiver education ----
PAGE 50

50 Table 22 Measures and Administration Time Points Note. CBCL = Child Behavior Checklist; ECBI = Eyberg Child Behavior Inventory; DPICS=III = Dyadic Parent Child Interaction Coding System 3rd edition; PSI SF = Parenting Stress Index Short Form; PDR DDI = Parent Daily Report/ Daily Discipline Inventory; READI = Readiness, Efficacy, Attributions, Defensiveness, & Importance Scale; BDI II = Beck Depression Inventory II. Instrument Time 1 Time 2 Time 3 Time 4 CBCL X X X ECBI X X X X DPICS III X X PSI SF X X X PDR/DDI X X X READI X X BDI II X X X

PAGE 51

51 Table 2 3 Inter Coder Agreement Statistics for the DPICS III Categories at the Time 1 Assessment DPICS III Category Percent Agreement Kappa CDI T Do Skills Behavior Description a --Unlabeled Praise 0 .90 0 .90 Labeled Praise 1 .00 1 .00 Reflective Statement 1 .00 1 .00 CDI T Dont Skills Indirect Command 1 .00 1 .00 Direct Command 0 .87 0 .86 Descriptive/Reflective Question 0 .89 0 .85 Information Question 0 .88 0 .88 Criticism 1 .00 1 .00 Total 0 .88 0 .86 Note: DPICS III = Dyadic Parent Child Interaction Coding System Third Edition; CDI T = Child Directed Interaction Training aNo Behavior Descriptions were coded during the selected Time 1 Observations

PAGE 52

52 Phone Screened for eligibility ( n = 63) Completed Time 1 Assessment and Randomized ( n = 13) Excluded ( n = 40): Did not meet criteria ( n = 30) Refused to participate ( n = 6) Did not attend Time 1 assessment ( n = 4) Completed Post Training Assessment ( n = 6) Completed Post Waitlist Assessment ( n = 6) Assigned to IT ( n = 7) Completed intervention ( n = 6) Contact lost before family was informed of training condition( n = 1) Assigned to WLC ( n = 6) Completed wait (n = 6) 3 month follow up Completed ( n = 5) Did not complete ( n = 1) Figure 2 1 Sampling and Flow of Participants t hroughout Child Directed Interaction Training IT = Immediate Training Group; WLC = Wait List Control. Participant flow chart from phone screening through the final assessment for Child Directed Interaction Training.

PAGE 53

53 TIME 1 Pre assessment : Assignment to WLC or IT WLC: 3 week wait period IT: 1 week wait period + 2 week CDIT TIME 3 WLC: 2 nd Pre assessment TIME 4 3 month Follow up Assessment TIME 2 WLC: Mid wait assessment WLC: 2 week wait period TIME 2 IT : Mid train assessment TIME 3 IT: Post assessment IT: 2 weeks CDIT Figure 2 2 Child Directed Interaction Training Assessment and Training Flow Chart for Participants IT = Immediate Training; WLC = Wait list Control. Participant flow chart for training and assessment time points.

PAGE 54

54 Figure 2 3 Mean F requency Rating of C hild Behavior Problem s on the ECBI t hroughout the c ourse of CDIT for the Immediate Training Group ECBI = Eyberg Child Behavior Inventory ; CDIT = Child Directed Interaction Training. The scores presented in this graph are the means from the 6 Immediate Training participants at each training session. 80 90 100 110 120 130 140 150 160 170 180 Session 1 Session 2 Session 3 Session 4 Session 5 Sesson 6 Session 7 Session 8 ECBI Intensity Score Session Number Clinical Cut off

PAGE 55

55 CHAPTER 3 RESULTS The main study analyses evaluated the effectiveness of CDIT training in reducing caregiver reported critical verbal force, reported noncritical verbal force, observed negative leading, caregiver stress, and caregiver depression while increasing the frequency of observed positive attention skills and limit setting when compared to wait list controls. All analyses were conducted using the study completers, ( n = 6 IT completers, n = 6 WLC completers), unless otherwise noted. Aim 1 : Analyses of Change Following CDIT Caregiver Self Report of Depression and Parenting Stress After training, significant between groups differences were detected on the PSI SF Childrearing Stress subscale, F (1, 9) = 20.94, p = .001, r = 74 and the BDI II T otal scale F (1 9 ) = 12.34, p = .007, r = 82. Large effect sizes between g roups indicate caregivers in IT endorsed significantly lower parenting stress and depression than those in WLC, see Table 31. Graphical representations of the betweengroup differences with confidence intervals for the PSI SF Childrearing Stress subscale and the BDI II Total scale are displayed in Figure 31 and Figure 32 respectively Observed Parent Child Interaction One caregiver in the WLC and two caregivers in the IT group completed their post wait or post training assessment by mail or over the phone because they were unable to come in to the Library P artnership for the assessment, resulting in a total sample size of 9 for observational data. Results revealed significant group differences on the DPICS III Positive Attention skills composite (i.e. Lab eled Praise, Behavior Description, Reflection, and Unlabeled Praise), F (1,6) = 162.91, p <.001, r = .98 Significant group difference

PAGE 56

56 were detected for the DPICS III Negative Leading composite (i.e. Questions, Criticisms, and Indirect Commands) with the I T group demonstrating fewer negative leading statements at the Time 3 assessment, F (1,6) = 83.35, p < .001, r = .94 than the WLC group. The statistically significant findings in combination with the large effect sizes between groups suggests differences found are likely replicable. These results are shown in Table 32 and a graphical representation of the betweengroup differences with confidence intervals are displayed in Figure 33 Caregiver Report of Daily Discipline Practices Group differences on the PDR/DDI were significant for Critical Verbal Force, F (1 8 ) = 5.60, p = .05, r = .59, with the IT group displaying a lower percentage of Critical Verbal Force than the WLC The PDR/DDI Limit Setting category was a trendlevel findi ng suggesting the IT group likely reported greater use of this category of discipline when compared to the WLC, F (1,8) = 4.86, p = .06, r = .57. No difference was found on the PDR/DDI Non critical V erbal F orce category, F (1 8 ) = 1.94, p = .20, r = .40. Me an scores, standard deviations, and effect sizes for the PDR/DDI categories are shown in the Table 33. Graphical representation of the betweengroup differences with confidence intervals are displayed in Figure 3 4 Aim 2: Predictors of Positive Atten tio n Skills Acquisition Predictor analyses were conducted using ANCOVA to conduct a multiple regression in order to facilitate the use of continuous variables while being able to control for group assignment. Results indicated that the overall predictor model was significant, F (3 5 ) = 88.85, p = .000, Adjusted R2 = .98. However, analyses of individual contributions of each variable to the model indicate that the PSI SF C hildrearing Stress subscale did not significantly predict skill acquisition, t (8) = 1.50, p = .19 nor

PAGE 57

57 explain a significant proportion of the variance in skill acquisition R2 = .31, F (1 5 ) = 2.26, p = .19, in the post training or post wait DPICS III Positive Atten tion Skills. T he addition of the BDI II Total Scale also did not signific antly predict skill acquisition, .003, t (8) = 1.75, p = .14 nor explain a significant proportion of the variance in skill acquisition R2 = .38, F (1 5 ) = 3.06, p = .14, in the post training or post wait Positive Atten tion Skills. However, experi mental condition did significantly predict skill acquisition, t (8) = 16.23, p < .001, and explained a significant proportion of the variance, 98%, in skill acquisition, R2 = .98, F (1 5 ) = 263.28, p < .001, in the post training or post wait Positive Atten tion Skills, F (1 5 ) = 263.28, p = .000, R2 = .98. Aim 3: Moderators of Positive Attention Skills Acquisition T o explore the impact of motivation on the acquisition of P ositive A ttention skills from the pre to post training/wait assessme nts, a moderation analysis was conducted following the JohnsonNeyman method of assessing moderation with a continuous variable when no a priori hypotheses suggest a level of significance for dichotomizing the variable. Results suggested that the change in motivation was a significant moderator, t (11) = 5.30, p = .003 and the interaction term between the change in motivation and the pretraining or prewait P ositive A ttention skills explained a significant increase in the variance of post training or post wait P ositive A ttention skills R2= .15, F (3, 5) = 8.47, p = .03. However, further exploration of these findings, controlling for group assignment, indicated that the change in motivation was not a significant moderator for pre to post training/wait P ositive Attention skills 1.057, t (11) = 3.84, p = .720.

PAGE 58

58 Aim 4: Attrition Analyses Analyses exploring predictors of attrition could not be conducted because there was no dropout from the sample. Only one participant who attended a Time 1 assessment did not complete IT C ontact was lost with this family before they could be informed of their experimental condition. Aim 5: Follow up Analyses Three month follow up data w ere collected only for the IT group due to the ethical considerations involved in having the WLC group wait an additional three months to receive their training N o observational data w ere collected at the follow up due to the concern that families would have difficulties completing follow up assessments at the libra ry. Two of the IT families did not complete follow up measures, although one of these two families participate in the daily discipline inventories leaving 5 participants in the follow up assessment of change in the daily discipline strategies and 4 participants to provide data on the measures of depression and parenting stress. Resul ts indicate that the BDIII Total scale, t (3) = 1.12, p = .34, the Childrearing Stress scale on the PSI SF, t (3) = 0.34, p = .77 the Critical Verbal Force scale, t (4) = .57, p = .60, and the Limit Setting scale, t (4) = 9.8, p = .60 did not change at the 3month follow up. However, the Noncritical Verbal Force scale did significantly decrease at the 3month follow up suggesting a further decline in the number of repor ted commands, t (4) = 2.76, p = .05. Equivalency testing at 20% equivalence revealed nonsignificant results for the PDR/DDI discipline categories and the BDI II Total scale, however, PSI SF Caregiver Stress remained equivalent. Corresponding means standar d dev iations and z scores for equivalence testing with their corresponding p values can be found in Table 34. A graphical representation of the maintenance of treatment scores with error bars can be

PAGE 59

59 found in Figure 35 for the BDI II, Figure 3 6 for the PSI SF Childrearing Stress, and Figure 37 for the DDI categories.

PAGE 60

6 0 Table 31. Mean Scores for Caregiver Self Report Measures at Time 1 and 3 Assessments IT WLC Measure a Assessment M SD M SD p r b Beck Depression Inventory Total Time 1 8.50 8.26 7.67 8.38 --Time 3 4.33 7.06 9.33 8.23 .007 .82 Parenting Stress Index Short Form Childre aring Stress Time 1 57.67 8.16 59.83 9.11 --Time 3 43.17 8.66 60.83 5.60 .001 .74 Note: IT = immediate treatment ( n = 6), WL C = wait list control ( n = 6). aScores for all measures are reported as raw scores. b r = effect size between IT and WLC groups at the Time 3 assessment

PAGE 61

61 Table 32 Time 1 and 3 Scores for Caregivers during 5 Minutes of Observed ChildLed Play IT WLC Measure Assessment Mean SD Mean SD p r a Positive Atten tion Time 1 11.07% 7.27% 4.14% 3.54% --Time 3 52.53% 6.07% 5.87% 3.91% <.001 .98 Negative Leading Time 1 42.01% 8.71% 37.1 0 % 7.63% --Time 3 8.78% 6.1% 45.96% 5.90% <.001 .94 Note: IT = I mmediate Training ( n = 4), WLC = W ait list C ontrol ( n = 5). ar = effect size of change between IT and WLC groups at the Time 3 assessment

PAGE 62

62 Table 33 Time 1 and 3 Percentage of Reported Discipline Practices IT WLC Measure Assessment Mean SD Mean SD p r a Critical Verbal Force Time 1 25.53% 15.46% 25.83% 8.61% --Time 3 10.94% 10.14% 21.67% 6.09% .048 .59 Non critical Verbal Force Time 1 17.52% 7.58% 14.50% 9.56% --Time 3 16.20% 9.78% 24.00% 9.78% .201 .40 Limit Setting Time 1 31.09% 12.58% 31.33% 8.21% --Time 3 50.88% 15.56% 30.33% 21.97% .059 .57 Note: IT = I mmediate Training ( n = 6), WLC = Wait list Control ( n = 6). ar = effect size of change between IT and WLC groups at the Time 3 assessment

PAGE 63

63 Table 34 Equivalency Testing Analyses between Time 3 and Time 4 Caregiver Report Variables Time 3 Time 4 Measure a M SD M SD p1 z1 b p2 z2 Beck Depression Inventory Total 6.5 0 8.02 8.5 0 5.75 .89 .14 .50 .67 Parenting Stress Index Childrearing Stress 48 .00 4.24 49 .00 6.38 .02 2.25 .005 2.77 Daily Discipline Inventory Critical Verbal Force 10.94% 10.13% 13.80% 12.73% .92 .09 .48 .70 Non critical Verbal Force 16.20% 9.78% 7.00% 8.83% .98 2.11 .85 1.02 Limit Setting 50.88% 15.58% 59.20% 25.15% .56 .14 .08 1.4 aScores for all measures are reported as raw scores. bz = resulting equivalency test z value.

PAGE 64

64 Figure 31. Caregiver Depression at the Time 3 Assessment Error bars represent 95% confidence intervals. For significance, p < .05. Those receiving CDIT reported less depression than those in the wait list, F(1, 9) = 20.94, p= .001, r = .74, SE = 1.15. CDI T = Child Directed Interaction Training. BDI II = Beck Depression Inventory II. IT = Immediate Training group. WLC = Wait list Control group 0 2 4 6 8 10 12 14 Time 3 Assessment IT WLC Clinical Cut Off

PAGE 65

65 Figure 32 Childrearing Stress at the Time 3 Assessment Error bars represent 95% confidence intervals. For significance, p < .05. Those receiving CDIT reported le ss childrearing stress than those in the wait list, F(1, 9) = 12.34, p = .007, r = .82, SE = 2.55. CDIT = Child Directed Interaction Training. PSI SF = Parenting Stress Index Short Form. IT = Immediate Training group. WLC = Wait list Control group. 0 10 20 30 40 50 60 70 Time 3 Assessment IT WLC

PAGE 66

66 Figure 33 Percentage of observed caregiver verbalizations with their child by category at the Time 3 Assessment. Error bars represent 95% confidence intervals. For all categories, p < .05 was need for significance. Those receiving CDIT reported a higher percentage of Positive Attention to child behavior F(1 6 ) = 162.91, p <. 001, r = .98, SE = 2.70% and a lower percentage of Negative Leading in the childled play F(1 6 ) = 83.35, p < 001, r = .94, SE = 3.1% than the wait list controls. CDIT = Child Directed Interaction Training. DPICS III = Dyadic Parent Child Interaction Coding System 3rd Edition. IT = Immediate Training group. WLC = Wait list Control group. 0.00% 10.00% 20.00% 30.00% 40.00% 50.00% 60.00% Positive Attention Negative Leading IT WLC

PAGE 67

67 Figure 34 Percentage of discipline used by category at the Time 3 Assessment Error bars represent 95% confidence intervals. For all categories, p < .05 was need for significance. Those receiving CDIT reported a lower percentage of Critical Verbal Force, F (1 8 ) = 5.60, p = .05, r = .59, SE = .037 and a higher percentage of Limit Setting F (1 8 ) = 4.86, p = .06, r = .57, SE = 7.1%, than the wait list controls. The reported percentage of use of Non Critical Verbal Force was nonsignificant between those receiving CDIT and the wait list CDIT = Child Directed Interaction Training. DDI = Da ily Discipline Inventory. IT = Immediate Training group. WLC = Wait list Control group. 0.00% 10.00% 20.00% 30.00% 40.00% 50.00% 60.00% Critical Verbal Force Limit Setting Non Critical Verbal Force IT WLC

PAGE 68

68 Figure 35 Maintenance of Caregiver Depression from Post Training to the 3month Follow up assessment Error bars represent 95% confidence intervals. For significance, p < .05. Depression scores were nonsignificant between the two time points t (3) = 1.12, p = .34, SE = 1.78. BDI II = Beck Depression Inventory II. 0 2 4 6 8 10 12 14 Assessment Time Point Post Training 3month Follow Up

PAGE 69

69 Figure 36 Maintenance of Childr earing Stress from Post Training to the 3month Follow up Assessment. Error bars represent 95% confidence intervals. For significance, p < .05. Childrearing Stress scores were nonsignificant between the two time points, t (3) = .034, p = .77, SE = 3.19. PSI SF = Parenting Stress Index Short Form. 0 10 20 30 40 50 60 Assessment Time Point Post Assessment 3month Follow up

PAGE 70

70 Figure 37 Maintenance of Parent Reported Discipline Practices from Post Training to the 3month Follow up Assessment. Error bars represent 95% confidence intervals. For all categories, p < .05 was need for significance. DDI = Daily Discipline Inventory. Scores were nonsignificant between the two time points for Limit Setting, t (4) = 9.8, p = .60, SE = .12 and Critical Verbal Force, t (4) = .57, p = .60, SE = .02 There was a significant decrease in the caregiver report of Noncritical Verbal Force, t (4) = 2.76, p = .05, SE = .03 0.00% 10.00% 20.00% 30.00% 40.00% 50.00% 60.00% 70.00% 80.00% Post Follow Up

PAGE 71

71 CHAPTER 4 DISCUSSION Findings in this study provide promising evidence that timelimited CDIT may be an effective parent training model for kinship caregivers in decreasing parenting stress and depressive symptoms while facilitating positive changes in parenting skills and discipline practices. Following CDIT, families demonstrated less use of critical verbal force in discipline and showed fewer negative leading behaviors during play while increasing their positive attention to child appropriate behavior and demonstrating a significant trend towards more limit setting strategies in discipline. Findings should be interpreted with caution given the small sam ple size. Appropriate interpretations of results were facilitated by considering effect sizes (Maxwell, 2004) and all statistically significant findings were supported by large effect sizes, emphasizing the importance of replicating study with a larger sam ple to evaluate the stability of the findings and draw firmer conclusions regarding this brief intervention. The primary objective of this study was to explore group differences following CDIT using a randomized control trial with a wait list control desi gn. As expected, caregivers who received training reported fewer depressive symptoms than wait list controls and this improvement were maintained at the 3month follow up assessment. Although statistical equivalence was not attained, it is noteworthy that the one grandmother in the moderately depressed range on the BDI II was in the mild range following CDIT and remained in the mild range at the 3month follow up. Although most grandparents in this study endorsed minimal depressive symptoms before the inter vention, research on the grandparent kinship caregiver population suggests these caregivers typically report higher levels of depression than other caregivers

PAGE 72

72 (Cuddeback, 2004). Evidence that CDIT can aid in reducing depressive symptomatology at post trai ning and at 3month follow up suggests this intervention may be beneficial for depressed kinship caregivers Caregivers receiving CDIT reported significantly lower childrearing stress than wait list controls and maintained these improvements at a statistically equivalent level 3months following training completion. In addition to obtaining statistical equivalenc e, all four caregivers initially endorsing clinically significant parenting stress before training reported a reduction in parenting stress into the normative range following CDIT. These findings suggest that time limited CDIT can facilitate detectable imp rovements in childrearing stress, stress related to parenting difficult child behaviors and dysfunction in the parent child relationship, similar to those improvements reported in performancebased CDIT (Eisenstadt, et. al 1993, Harwood & Eyberg, 2004) and full PCIT (Timmer et.al, 2006). It is impressive that the reduction in parenting stress remained equivalent at follow up given the high amount of stressors that kinship caregivers report (Harden et al., 2004). Following timelimited CDIT, kinship caregiv ers increased their responsiveness to child appropriate behaviors while decreasing their intrusive behaviors in observed child led interactions, changes similar to those detected following full protocol PCIT (Eyberg, Nelson, Boggs, & Stevens, 2005). Kinshi p caregiver parenting styles are often characterized by low levels of emotional availability (Lawler, 2008) and training programs that facilitate an improvement in the warmth of the caregiver child relationship are needed. In this study, CDIT improved kins hip foster caregiver emotional availability and responsiveness by increasing their positive attention to child appropriate behaviors. This was achieved in CDIT by teaching caregivers to describe the childs

PAGE 73

73 appropriate behavior, reflect the childs appropr iate talk, and praise the childs specific prosocial and acceptable behaviors during the play. CDIT also improved emotional availability by decreasing intrusiveness in the play by reducing caregivers negative leading. This was achieved by teaching caregiv ers to avoid intrusive verbalizations including criticism, indirect commands, and questions during the play. Intrusive verbalization are used by caregivers in an attempt to involve a child in prosocial and appropriate interactions, however these comments c an be aversive for the child and can increase the likelihood of negative interactions between the caregiver and child. Increasing positive attention and decreasing negative leading in the caregiver child relationship facilitates the development of stronger attachment relationships which can reduce the negative psychological impacts of prior child abuse and neglect (Dozier et al., 2006; Schofield & Beek, 2005). Results showing change in the kinship caregivers daily report of discipline practices following child misbehaviors indicated a change in the pattern of discipline used by these caregivers after CDIT, even though discipline methods were not directly addressed in treatment. Specifically, the caregivers reported less use of critical verbal force and a trend level finding towards greater reliance on limit setting and these gains were maintained at the 3month follow up assessment. Discipline practices categorized as noncritical verbal force were nonsignificant at post training, but significantly decreased at the 3month follow up. Changes in discipline practices following CDIT were only partially supported by previous research on full protocol PCIT that revealed significant decreases in both critical verbal force and noncritical verbal force as well as significant increases in limit setting (Bhuiyan, 2012). The differences in the main

PAGE 74

74 behavior management strategies taught in the CDI versus the PDI are the most likely explanations for the discrepancies in these findings. The primary means of behavior management taught in CDIT is differential social attention, which combines consistent ignoring of negative attention seeking behaviors and positive reinforcement of acceptable and desirable behaviors (Boggs & Eyberg, 2008). Kinship caregiver report of disc ipline practices suggest that their ability to ignore and decrease negative attention to negative behaviors in the play generalized to their discipline practices in the home, thus reducing their use of critical verbal force (e.g. arguing, threatening, and criticism) to manage difficult child behaviors. Since differential attention is taught in both CDIT and full protocol PCIT, it is a logical that both types of parent training interventions would experience significant improvements in critical verbal force. However, caregivers use of limit setting in this study following CDIT was a significant trend and Bhuiyan (2012) found statistically significant improvements following fullprotocol PCIT. Limit setting includes ignoring, a skill taught in CDIT, as wel l as discipline strategies like time out and withdrawing privileges, which are not taught in CDIT. The PDI does teach discipline skills including how to deliver effective commands and follow through with a timeout procedure. The discipline methods taught in PDI may explain why there was a discrepancy in the findings in limit setting. The most surprising discipline finding was the significant reduction in Non critical Verbal Force (e.g. commands, repeated commands) at the 3month follow up, but not at the post training assessment. CDIT does not specifically address the use of commands to manage child behaviors whereas PDI teaches how to give effective commands and

PAGE 75

75 ensure compliance to these commands using a TimeOut procedure. However, it appears that over time kinship caregivers give fewer commands to manage misbehaviors, like we see in full protocol PCIT. It is possible that the consistent use of differential social attention, generally considered by parents to be the more acceptable form of behavior manag ement than timeout (Boggs & Eyberg, 2008), reduces the need for multiple commands to ensure compliance overtime. No observational data was collected at the 3month follow up to confirm the parent reported change, so findings may be due to parent report bi as as well. The second objective of this study was to explore factors that predict acquisition of the positive attention skills taught during CDIT. The greatest variance was explained by the training condition, as would be expected, and parent psychologic al variables were not predictors of caregiver positive attention skills in this study. The impact of kinship caregiver psychological variables on parent training skill acquisition and completion is understudied, but a previous study of PCIT with kinship caregivers has indicated that high levels of parenting stress in this population may actually increase parent training completion (Timmer et al., 2004).. Depression in kinship caregivers has not previously been explored in a parent training study, but a prev ious study exploring factors leading to positive change in dysfunctional mother child interactions, which was a composite category consisting of parenting stress, child behavior problems, and dysfunctional parenting practices, indicated that parent depress ion was unrelated to improvements following performancebased CDI alone (Harwood & Eyberg, 2004). Maternal social support was the only significant predictor of reductions in mother child dysfunction (Harwood & Eyberg, 2004).

PAGE 76

76 Therefore, perceived social s upport may be more appropriate predictor of skill acquisition to explore. Change in caregiver motivation did not moderate the improvement in positive attention skills when controlling for group assignment. Motivation was originally selected as a variable impacting the strength of change in positive parenting skills because previous studies exploring motivation in the child welfare population found that a motivational enhancement component before beginning PCIT improved treatment completion (Chaffin et al., 2009). It was assumed that caregivers who experienced an increase in motivation from preto mid treatment would also demonstrate greater gains in parenting skills in a time limited intervention. However, this was not the case. Recent research further ex ploring the impact of a motivational component on PCIT treatment helps to clarify this finding. Chaffin et al. (2011) found that enhancing motivation with PCIT in the child welfare population improved treatment retention not because it facilitated improvements in the observed acquisition of parenting skills during session, but rather because it facilitated the generalization of treatment skills outside of the play in the home environment (Chaffin et al., 2011). Therefore, motivation in this study may bett er predict the generalization of skills outside of session than those observed during the assessment observations in the training environment. The final objective of this study was to explore attrition. However, no family dropped out after group assignment and only one family was lost to follow up. Every family that started treatment or their wait period completed the study. The lack of attrition is particularly intriguing because behavioral parent training with PCIT suggests an attrition rate around 36% ( Fernandez & Eyberg, 2009), and attrition in family therapy

PAGE 77

77 ranges from 30 to 65% (Wierzbicki & Pekarik, 1993). The high level of retention in this study may be in part a reflection of the careful, clinical considerations taken into account when working wit h the kinship caregivers in this study who endorsed clinically significant levels of parenting stress and largely came from single caregiver homes. Clinical considerations for the kinship caregivers in this study began during conceptualization of the study design. Sessions were held twice weekly for four weeks in part because the evidencebased interventions in foster care involve intensive service delivery with multiple therapeutic contacts per week (Fisher & Chamberlain, 2000; Fisher et al., 1999). Howev er, intensive service delivery is costly and a significant burden on kinship caregivers who must commit to attending these appointments while simultaneously experiencing greater stress (Timmer et al., 2004) and fewer financial resources than other caregivers (Harden et al., 2004). To decrease this burden, training was time limited to only 4 weeks of treatment. In addition, reminder phone calls were made to the caregivers before each session to facilitate problem solving of perceived barriers to treatment at tendance when applicable. In addition, CDIT was selected as the parenting intervention because of its focus on increasing the caregiver child bond nurturance and warmth within the interaction while providing behavior management strategies. These are elements of interventions that receive the highest satisfaction ratings from foster parents after program completion (Grimm 2003; Puddy & Jackson 2003). However, no measure of training satisfaction was administered during the study and future studies may benefit from administering a measure of satisfaction with various procedural components of intervention delivery in order to further understand the factors contributing to kinship caregiver satisfaction and completion of CDIT.

PAGE 78

78 Several steps were also taken to red uce potential stigma of receiving psychological services such as the selection of the training location and referring to therapists as trainers. All CDIT assessment and training visits occurred at a community resource center and library that is associated with the local child welfare system, the Library Partnership (LP). In addition to reducing the potential stigma of attending mental health clinics for parent training, the library setting increased kinship caregiver connectedness to a service agency and pr ovided peer and professional social support as well, all factors shown to improve foster caregiver satisfaction with training (Buehler et al., 2003; Denby, et al., 1999; Fisher et al., 1999, Nzi & Eyberg, in press). At the LP, families could connect with social workers who could assist in providing community and financial resources and access to a clothing closet. The library setting supported a community environment and several of the families were observed to read together and socialize with others at the library before and after training sessions. A scale of perceived social support would be useful in future studies to assist in understanding the role of treatment setting/location to perceived social/community support and reduced attrition. Limitations A major limitation of this study is the small sample size. Insufficient power may increase the likelihood that significant differences ware detected in error, further indicating the need for replication to determine the stability of these findings. Difficulties with maintaining consistent referral sources during the project largely impacted the sample size. Strategies to improve and sustain community recruitment and referrals are essential to address in future research.

PAGE 79

79 Most of the referrals for this study came from organizations and individuals affiliated with the University of Florida who had a preexisting relationship with the Child Study Lab, previous research studies using PCIT, and the effectiveness of the research interventions. Kinship caregivers are less likely to seek out mental health services on their own (Leslie et al., 2000) and may not have presented to a university facility for these services. Although studies on increasing kinship caregiver participation in research are very limited, research with low income and minority groups has suggested that these populations are more likely to engage in research if relationships are established with formal and informal community leaders (Dennis & Neese, 2000). The referrals sought from community leaders and from local church and community organizations where the concentration of kinship caregivers is likely higher, were largely unsuccessful. It is possible that community recruitment efforts were less fruitful because a strong relationship with these referral sources had not been established. It is also possible more frequent personal contact and relationship building during the recruitment process would have improved recruitment. In addition to improving relationships with community leaders, maintai ning a stronger relationship with staff and personnel at the local DCF agency, the Partnership for Strong Families (PSF), may have increased study referrals as well. The study was originally designed after consulting administrators at the PSF and the initi al recruitment plan was for referrals to come solely from that organization. However, by the end of this project, several administrators, social workers, and staff members who were present at the initiation of the project had left the organization. Social worker and agency attrition rates are major problems faced by child welfare agencies (Mor Barak, Nissly, & Levin,

PAGE 80

80 2001). Variables contributing to high attrition rates are reported by social workers to include burnout, job dissatisfaction, availability of employment alternatives, low organizational and professional commitment, stress, and lack of social support (Mor Barak et al., 2001). Although no specific research to our knowledge has been published on effective ways to increase referrals to this study, r eferrals from social workers might have been improved by a more collaborative relationship between our research team and the PSF social workers and their supervisors. Attending weekly staffing meetings held at the PSF building might facilitate referrals by establishing a consistent presence at the agency that serves as both a reminder of the research study opportunity for families and a resource for new social workers and case managers that may have otherwise been unaware of the research study and potential benefits. Social workers could immediately have their questions answered regarding the study as they arise and could conveniently be provided with feedback regarding referred families participation, which was a protocol approved by the IRB to help PSF sta ff with managing their cases. Improving involvement and relationships with local DCF agencies should be considered in future studies. A second limitation of this study was the selection of a motivation instrument without established psychometric propert ies. This unstandardized instrument restricted analyses of motivation of caregivers attending trainings, limiting the integrity of our moderation analysis. No measure was identified at the time of conceptualization of this project with strong psychometric properties and further research should be conducted on valid instruments for measuring parent motivation for parent training.

PAGE 81

81 Caregiver education and racial/ethnic composition of this study was different that other studies with kinship caregiver populatio ns. This sample was more educated and had a larger majority of Caucasians than demographic studies of kinship caregivers (Berrick et al., 1994, Dubowitz et al., 1993). These discrepancies may indicate that the kinship caregivers in this sample are not representative of all kinship caregivers and studies would need to be conducted with more diverse populations in order to further explore the impact of these demographic factors on the outcome of the study. Future Directions Additional study of kinship caregiver psychological factors is important in order to further understand the impact of CDIT on the mental health of these. The impact of parent training on kinship caregivers mental health has been understudied despite revie ws that indicate frequent anxiety and depression in this population (Grinstead, Leder, Jensen, & Bond, 2003). In this study, several of the grandmothers appeared easily overwhelmed, reported feeling uncertain about their parenting ability, and would call f requently worried about financial, time, and energy barriers to successfully attending treatment. The high levels of stress experienced by kinship caregivers (Harden et al., 2004) could contribute to the observed anxious presentation of these caregivers. F uture studies with CDIT and kinship caregivers should seek to explore not only the impact of anxiety on the caregiver child relationship but also the effect of CDIT on caregiver anxiety, particularly their anxiety related to parenting. Caregivers did experience a reduction in parenting stress and this reduction was likely facilitated by the acquisition of parenting skills taught in CDIT and resulting child behavior improvements. Understanding the mechanism of change and the factors impacting the degree of change a family may experience are important aspects to

PAGE 82

82 further explore in studies of CDIT with kinship caregivers. Future studies should explore mediators and moderators (i.e. caregiver education, child behavior problems ) of the changes in parenting stress and other psychological variables in order to further explore the mechanisms by which CDIT changes parent psychological variables. To further understand the impact of CDIT on caregiver discipline strategies it will be important to conduct observations of kinship caregivers in more high demand situations where they lead the play. High demand situations like the parent led play or cleanup would assist in better understanding the generalization of differential social attention techniques and changes in child behaviors, like compliance, as well. In order to further explore the parent training in different discipline procedures that the kinship caregiver population might need, future studies should compare CDIT with full protocol PCIT to fully understand t he differences in discipline practices of the two approaches and to assess kinship caregiver satisfaction with different types of behavior management techniques. Results from this study hold promise that significant changes in childrearing stress, caregiver depression, and positive parenting skills can happen within a brief period (4 weeks) when using timelimited CDIT with kinship caregivers. This is the first time CDIT has been studied with kinship foster caregivers. This cost effective and short term par ent training model may help to stabilize kinship caregiver placements through stress reduction and more positive parenting practices, while reducing more coercive and harsh practices. This is also the first time that CDIT has been studied in a community se tting, providing preliminary evidence that CDIT can be delivered with fidelity and yield positive results in the community. The training was adapted for this population in the

PAGE 83

83 following ways: a preset number of sessions were conducted; coaching was providing in room, and optional PCIT handouts on modeling, positive opposites, kids and stress, and other discipline tools were provided to each family. However, all other elements of were followed per standard PCIT protocol. The success of this study likely depended on the location of service delivery (i.e. a community resource center) and clinically sensitive accommodations (i.e. phone reminders). These elements should be considered when tailoring this intervention for foster parents. The long term outcomes of CDIT training are an important step in the future studies of this intervention, particularly for reducing multiple placements. Comparisons of timelimited to performancebased CDIT is another important step in determining which approach is more effectiv e with kinship foster caregivers, an underserved population of caregivers in need of appropriate parent training, to break generational cycles of abuse and neglect and facilitate the growth of appropriate reciprocal interactions while improving child behav iors.

PAGE 84

84 APPENDIX A DAILY DISCIPLINE INT ERVIEW CODING DEFINI TIONS 1. Critical Non Physical Force: Verbal criticism towards child or intimidating child a. Yell, shout, scold, scream b. Argue, fight c. Threaten with physical punishment or physical har m, throw away, destroy toy d. Reject child e. Humiliate child f. Express disapproval or criticize child g. Force child to do an activity h. Threaten to ignore, discipline, or punish in a nonphysical manner i. Negative command (Dont do that; Stop) j. Confront k. Make apologize or tell truth l. Guilt by induction m. Lecture; talk to them n. Critical non verbal forcestomp feet or glare 2. Non critical Verbal Force: Commands a. Command b. Repeat command c. Stick to ground, be firm 3. Limit Setting: a. Warning of Time Out or withdrawal of privileges ; c ounting b. Time Out, send to room, chair, or couch (actual or attempted) or added minutes c. Nose to corner d. Withdraw privileges or toys e. Logical consequences (warn or enforce) f. Remove child from situation or alter situation g. Separate children or separate child fro m parents h. Stop activity i. Distract, redirect j. Supervise, monitor k. Ignore (or attempt) l. Request compliance

PAGE 85

85 APPENDIX B READI Each statement describes how a person might feel when thinking about their parenting. Please indicate the extent to which you tend to agree or disagree with each statement. In each case, make your choice in terms of how you feel right now, not what you have felt in the past or would like to feel. Strongly Disagree = 1 Disagree = 2 Undecided = 3 Ag ree = 4 Strongly Agree =5 1. I'm ready to start working on my parenting 1 2 3 4 5 2. Somebody needs to fix my child's behavior 1 2 3 4 5 3. I sometimes act without thinking 1 2 3 4 5 4. I will work on my c hilds behavior problems later 1 2 3 4 5 5. The only way my child will ever behave is if I threaten to spank him or her. 1 2 3 4 5 6. If I decide to change my parenting, I can 1 2 3 4 5 7. I am always a good person 1 2 3 4 5 8. Changing how I parent will not help my childs behavior 1 2 3 4 5 9. Bad things could happen if my child's behavior doesn't get better 1 2 3 4 5 10. I know that I can successfully complete treatment 1 2 3 4 5 11. I never worry about my health 1 2 3 4 5 12. Most of the problem is that my child is strong willed 1 2 3 4 5 13. Being here is pretty much a waste of time for me; I already know how to parent my child... 1 2 3 4 5 14. I'm ready to change my parenting. 1 2 3 4 5 15. Spanking is the best way to teach kids right from wrong 1 2 3 4 5 16. I sometimes think of myself before others 1 2 3 4 5 17. It's worth it to spend money to help my child with his/her behavior... 1 2 3 4 5 18. I'm certain that I can learn new parenting skills 1 2 3 4 5 19. There is nothing wrong with how I'm raising my child.. 1 2 3 4 5 20. I sometimes think of mys elf first.. 1 2 3 4 5 21. It is very important that my child's behavior problems be fixed. 1 2 3 4 5 22. I don't believe in spanking 1 2 3 4 5 23. Sometimes I have bad thoughts 1 2 3 4 5 24. The way I'm parenting now is not working... 1 2 3 4 5 25. I need to learn to be more consistent 1 2 3 4 5 26. I always do what is right 1 2 3 4 5 27. I dont need to learn new parenting techniqu es; I already know what works 1 2 3 4 5 28. I'm eager to learn any skills the therapist can teach me 1 2 3 4 5 29. I sometimes fail to keep all of my promises 1 2 3 4 5 30. I am sure that I can get to my appointments on time 1 2 3 4 5 31. I have a lot to do with the fact that there is a problem with my child's behavior. 1 2 3 4 5 32. I never get mad at others 1 2 3 4 5 33. I want to change the way I discipline my child 1 2 3 4 5 34. I 'm sure I can use new parenting techniques at home 1 2 3 4 5 35. People sometimes take advantage of me 1 2 3 4 5 36. I dont really need to change my parenting 1 2 3 4 5 37. Spare the rod and spoil the child 1 2 3 4 5 38. I never listen to gossip.. 1 2 3 4 5 39. I'm sure I can learn to manage my childs behavior 1 2 3 4 5 40. I sometimes say bad words.. 1 2 3 4 5

PAGE 86

86 41. I think that I can l earn to be a better parent... 1 2 3 4 5 42. I don't think my parenting is a problem... 1 2 3 4 5 43. I never do anything that is bad for my health 1 2 3 4 5 44. It's time to change the way my child and I get a long 1 2 3 4 5 45. If things don't change, my child's future could be hurt 1 2 3 4 5 46. I am always happy with what I have... 1 2 3 4 5 47. Things with my child's behavior have to change very soon 1 2 3 4 5 48. Other people think I need to change how I am as a parent, but I dont think so. 1 2 3 4 5 49. The only way my child will ever behave is if I spank him or her. 1 2 3 4 5 50. It's very important that my child and/or I get help 1 2 3 4 5 51. I would rather try to live with my faults as a parent than try to change them 1 2 3 4 5 52. I never raise my voice in anger 1 2 3 4 5 53. I sometimes act silly.. 1 2 3 4 5 54. I am willing to do whate ver it takes to be sure that we get help. 1 2 3 4 5 55. I am certain that I can do positive things to help my child behave 1 2 3 4 5 56. I sometimes lose my temper 1 2 3 4 5 57. I always tell the truth.. 1 2 3 4 5 58. I have problems that are more important than my childs behavior right now. 1 2 3 4 5 59. It might be hard, but I'm ready to parent differently. 1 2 3 4 5 60. Id like to learn what will work to change my childs behavior 1 2 3 4 5 61. Spanking may work in the short run, but I worry that in the long run it teaches the wrong thing.. 1 2 3 4 5

PAGE 87

87 LIST OF REFERENCES Abidin, R. (1995). Parenting Stress Manual (3rd ed.). Odessa, FL: Psychological Assessment Resources, Inc. Achenbach, T.M. (1991). Manual for the Child Behavior Checklist 4 18 and 1991 profile. Burlington, VT: University of Vermont, Department of Psychiatry. Achenbach, T., Dumenci, L., Rescorla, L. ( 2003). DSM oriented and empirically based approaches to constructing scales from the same item pools. Journal of Clinical Child and Adolescent Psychology, 32(3) 328340. doi:10.1207/S15374424JCCP3203_02 American Academy of Pediatrics. (2000). Development issues for young children in foster care. Pediatrics, 106, 11451150. doi:10.1542/peds.106.5.1145 Barth, R., Courtney, M., Berrick, J., & Albert, V. (2004). From child abuse to permanency planning child welfare services and pathways, Aldine de Gruyter, New York, NY Beck, Steer, & Brown, (1996). Manual for Beck Depression Inventory II. San Antonio, TX: Psychological Corporation. Berrick, J.D. (1998). When children cannot remain home: Family care and kinship care. Protecting Children From Abuse and Neglect, 8, 72 87. doi:10.2307/1602629 Berrick, J.D., Barth, R.P., Needell, B. (1994). A comparison of kinship foster homes and foster family homes: Implications for kinship foster care as family preservation. Children and Youth Services Review, 16, 33 63. doi: 10.1016%2F01907409%2894%2990015 9 Bhuiyan, N. (2012) Maternal disciplinary behaviors and disruptive behaviors in Parent Child Interaction Therapy. (Unpublished masters thesis). University of Florida, Gainesville, Florida. Boggs, S.R., & Eyberg, S.M. (2008). Positive attention. In W. ODonohue, & J.D. Fisher (Eds.). Cognitive behavior therapy: Applying empirically supported techniques in your pract ice (2nd ed). New York: Wiley. Boggs, S.R., Eyberg, S.M., Edwards, D., Rayfield, A., Jacobs, J., Bagner, D., & Hood, K. (2004). Outcomes of parent child interaction therapy: A comparison of dropouts and treatment completers one to three years after treatm ent. Child & Family Behavior Therapy 26 1 22. doi:10.1300/J019v26n04_01

PAGE 88

88 Brestan, E. V. & Eyberg, S. M. (1998). Effective psychosocial treatments of conduct disordered children and adolescents: 29 years, 82 studies, and 5,272 kids. Journal of Clinical Child Psychology 27, 180 189. doi:10.1207/s15374424jccp2702_5 Brestan, E. V., Ondersma, S. J., Simpson, S. M., & Gurwitch, R. (1999, April). Application of Stage of Change Theory to Parenting Behavior: Validating the Parent Readiness to Change Scale. P oster presented at the Florida Conference on Child Health Psychology, Gainesville, Florida. Brown, J. & Calder, P. (2000). Concept mapping the needs of foster parents. Child Welfare, 79, 729 746. PMid:11104156 Buehler, C., Cox, M.E., & Cuddeback, G. (20 03). Foster parents perceptions of factors that promote or inhibit successful fostering. Qualitative Social Work, 2, 61 83. doi:10.1177/1473325003002001281 Carr, A. (2004). Interventions for post traumatic stress disorder in children and adolescents. Pediatric Rehabilitation, 7, 231244. doi:10.1080/13638490410001727464 Chadwick Center on Children and Families. (2004). Closing the quality chasm in child abuse treatment: Identifying and disseminating best practices. San Diego, CA: Author. Chaffin, M. & Friedrich, B. (2004). Evidencebased treatments in child abuse and neglect. Children and Youth Services Review, 26, 10971113. doi:10.1016/j.childyouth.2004.08.008 Chaffin, M., Funderburk, B., Bard, D., Valle, L.A., & Gurwitch, R. (2011). A combined motivation and parent child interaction therapy package reduces child welfare recidivism in a randomized dismantling field trial. Journal of Consulting and Clinical Psychology, 79, 84 95. doi: 10.1037/a0021227 Chaffin, N., Valle, L.A., Funderburk B., Gurwitch, R., Silvosky, J., Bard, D., McCoy, C., Kees, M. (2009). A motivational intervention can improve retention in PCIT for low motivated child welfare clients. Child Maltreatment, 14, 356 368, doi: 10.1177/1077559509332263. Chamberlain, P. (2 003). The Oregon multidimensional treatment foster care model: features outcomes, and progress in dissemination. Cognitive and Behavioral Practice, 10, 303 312. doi:10.1016/S10777229(03)800482

PAGE 89

89 Chamberlain, P., Price, J., Leve, L., Heidemarie, L., Lands verk, J., & Reid, J. (2008). Prevention of behavior problems of children in foster care: outcomes and mediation effects. Prevention Science, 9, 1727. doi:10.1007/s1112100700807 Chamberlain, P.& Reid, J.B. (1987). Parent observation and report of child symptoms. Behavioral Assessment, 9, 97 109. Christley, R.M. (2010). Power and error: Increased risk of false positive results in underpowered studies. The Open Epidemiology Journal, 3, 16 19. doi:10.2174/1874297101003010016 Clausen, J. M., Landsverk, J., Ganger, W., Chadwick, D., & Litrownik, A. (1998). Mental health problems of children in foster care. Journal of Child and Family Studies, 7 283 296. doi: 10.1023/A:1022989411119 Cuddeback, G.S. (2004) Kinship family foster care: a methodological and substantive synthesis of research. Children and Youth Services Review, 26 623639. doi:10.1016/j.childyouth.2004.01.014 Denby, R., Rindfleisch, N., & Bean, G. (1999). Predictors of foster parents satisfaction and intent to continue to foster. Journal of Child Abuse and Neglect, 23, 287303. doi:10.1016/S01452134(98)001264 Dennis, B.P. & Neese, J. B. (2000). Recruitment and retention of African American elders into community based research: Lessons learned. Archives of Psychiatric Nursing, 14, 3 11. doi: 10.1016/S08839417(00)800035 Dozier, M., Peloso, E., Lindhiem, O., Gordon, M.K., Manni, M., Sepulveda, S., & Ackerman, J. (2006). Developing evidencebased interventions for foster children: An example of a randomized clinical trial with infants and toddlers. Journal of Social Issues, 62(4), 767 785. doi:10.1111/j.15404560.2006.00486.x Dubowitz, H. Feigelman, S., & Zuravin, S. (1993). A profile of kinship care. Child Welfare, 72, 153 169. Eisenstadt, T., Eyberg, S., McNeil, C., Newcomb, K., and Funderburk, B. (1993). Parent Child Interaction Therapy with behavior problem children: Relative effectiveness of two stages and overall treatment outcome. Journal of Clinical Child Psychology, 22(1) 4251. doi:10.1207/s15374424jccp2201_4 Eyberg, S.M., Boggs, S., & Algina, J. (1995). Parentchild interaction therapy: A psychosocial model for the treatment of young children with conduct problem behavior and their families. Psychopharmacology Bulletin, 31 83 91.

PAGE 90

90 Eyberg, S.M. & Child Study Lab (1999). Parent Child Interaction Therapy Manual. Unpublished manuscript, Department of clinical and Health Psychology, University of Florida, Gainesville, Florida. Eyberg, S.M., Funderburk, B.W., HembreeKigin T.L., McNeil, C.B., Querido, J.G., & Hood, K. (2001). Parent child interaction therapy with behavior problem children: One and two year maintenance of treatment effects in the family. Child & Family Behavior Therapy, 23, 1 20. doi:10.1300/J019v23n04_01 Eyberg, S.M., Funderburk, B., McNeil, C., Niec, L., Urquiza, A., & Zebell, N. (2009). Training Guidelines for Parent Child Interaction Therapy Retrieved from Parent Child Interaction Therapy International, Inc. website: http://www.pcit.orgwpcontent/uploads/ 2011/01/PCIT_Training_Guidelines_2009.pdf Eyberg, S.M., Nelson, M. M., & Boggs, S.R. (2008). Evidencebased treatments for child and adolescent disruptive behavior disorders. Journal of Clinical Child and Adolescent Psychology, 37, 213 235. doi:10.1080/15374410701820117 Eyberg, S., Nelson, M. Boggs, S. & Stevens, M. (2005). [Dyadic Parent Child Interaction Coding System Third edition parent and child verbalizations]. Unpublished raw data. Eyberg, S. M., Nelson, M. M., Duke, M., & Boggs, S. R. (200 5). Manual for the Dyadic Parent Child Interaction Coding System (3rd ed.). Eyberg, S. M., & Pincus, D. B. (1999). Eyberg Child Behavior Inventory and Sutter Eyberg Student Behavior Inventory: Professional Manual Odessa, FL: Psychological Assessment Resources. Eyberg, S.M., & Robinson, E. (1982). Parent child interaction training: Effects on family functioning. Journal of Clinical Child Psychology, 11, 130 137. doi:10.1080/15374418209533076 Fernandez, M., Chase, R., & Eyberg, S. M. (2005). The workbook: Coder training manual for the dyadic parent child interaction coding system. 3rd ed. Fernandez, M.A., & Eyberg, S.M. (2009). Predicting treatment and follow up attrition in Parent Child Interaction Therapy. Journal of Abnormal Child Psychology, 37, 431 441. doi:10.1007/s10802008 92811 Field, A. (2005). Discovering Statistics Using SPSS SAGE Publications Inc., Thousand Oaks, CA. Fisher, P.A., Burraston, B., & Pears, K. (2005). The Early Intervention Foster Care Program: Permanent placement outcomes from a randomized trial. Child Maltreatment, 10(1), 61 71. doi:10.1177/1077559504271561

PAGE 91

91 Fisher, P. & Chamberlain, P. (2000). Multidimensional treatment foster care: A program for intensive parenting, family support, and skill building. Journal of Emotional & Behavioral Disorders, 8(3) 155 165. doi:10.1177/106342660000800303 Fisher, P. Ellis, H., & Chamberlain, P. (1999). Early intervention foster care: A model of preventing risk in young children who have been maltreated. Childrens Services: Soc ial Policy, Research, and Practice, 2(3) 159 182. doi:10.1207/s15326918cs0203_3 Fisher, P. & Kim, H. (2007). Intervention effects on foster preschoolers attachment related behaviors from a randomized trial. Society of Prevention Research, 8, 161170. doi:10.1007/s11121007 00665 Fisher, P.A., Kim, H.K., Pears, K.C. (2009). Effects of multidimensional treatment foster care for Preschoolers (MTFC P) on reducing permanent placement failures among children with placement instability. Children and Youth Ser vices Review, 31, 541 546, doi:10.1016/j.childyouth.2008.10.012 Fisher, P. A., & Stoolmiller, M. (2008). Intervention effects on foster parent stress: Associations with child cortisol levels. Development and Psychopathology, 20, 10031021. doi:10.1017/S 0954579408000473 Floyd & Eyberg, (2003). Testing the attachment theory of parent child interaction therapy, Poster session presented at the annual meeting of the American Psychological Association, Toronto. Funderburk, B.W., Eyberg, S.M., Rich, B., Behar L. (2003). Further psychometric evaluation of the Eyberg and Behar rating scales for parents and teachers of preschoolers. Early Education & Development 14, 67 81. doi:10.1207/s15566935eed1401_5 Fussell, R.E., Nzi, A.M., & Eyberg, S.M. (2012). Emotional abuse and attachment: A parent child interaction therapy study. Manuscript in preparation, Department of Clinical and Health Psychology, University of Florida, Gainesville, Florida. Grimm, B. (2003). Child & Family services reviews: Part II in a series. Foster parent training: What the CFS reviews do and dont tell us. Youth Law News: Journal of the National Center for Youth Law, (2). Grinstead, L.N., Leder, S., Jensen, S., & Bond, L. (2003). Review of research on the health of caregiving grandparents Journal of Advanced Nursing, 44, 318 326. doi: 10.1046/j.13652648.2003.02807.x Harden, B.J., Clyman, R.B., Kriebel, D.K., & Lyons, M.E. (2004). Kith and kin care: Parental attitudes and resources of foster and relative caregivers. Children and Youth Services Review, 26, 657 671. doi:10.1016/j.childyouth.2004.02.001

PAGE 92

92 Harwood, M.D., & Eyberg, S.M. (2004). Child Directed Interaction: Prediction of change in impaired mother child functioning. Journal of Abnormal Child Psychology,34, 335347 doi:10.1007/s10802006 9025z Haskett, M.E., Ahern, L.S., Ward, C.S., & Allaire, J.C. (2006). Factor structure and validity of the Parenting Stress Index Short Form, Journal of Clinical and Child Psychology, 35 302312. doi:10.1207/s15374424jccp3502_14 Hayes, A .F. & Matthes, J. (2009). Computational procedures for probing interactions in OLS and logistic regression: SPSS and SAS implementations. Behavioral Research Methods, 3 924 936. doi:10.3758/BRM.41.3.924 Hegar, R.L., (1999). The cultural roots of kinship care. In: Hegar, R.L. and Scannapieco, M., Editors, 1999. Kinship foster care: policy, practice, and research, Ox ford University Press, New York doi:10.1093%2Facprof%3Aoso%2F9780195109405.003.0005 Howe, D. & Fearnley, S. (2003). Disorders of attachment in adopted and fostered children: Recognition and treatment. Clinical Child Psychology and Psychiatry, 8, 369386. doi:10.1177/1359104503008003007 Hurlburt M Leslie L Barth R & Landsverk J. (in press). National variation in efforts to train child welfare workers and foster caregivers on children's mental health and development. Journal of Social Service Research. James, S., Landsverk, J., Slymen, D.J., & Leslie, L.K. (2004). Predictors of o utpatient mental health service use: The role of foster care placement changes. Mental Health Services Research, 6, 127 141. doi : 10.1023/B:MHSR.0000036487.39001.51 Johnson, P.O., & Fay, L.C. (1950). The Johnson Neyman technique, its theory and applicat ion. Psychometrika 15, 349 367. doi:10.1007/BF02288864 Kaminski, J.W., Valle, L.A., Filene, J.H., & Boyle, C.L. (2008). A metaanalytic review of components associated with parent training program effectiveness. Journal of Abnormal Child Psychology, 36, 567 589. doi: 10.1007/s10802007 92019 Landsverk, J., Litrownik, A. J., Newton, R., Ganger, W., & Remmer, J. (1996). Psychological impact of child maltreatment. Washington, DC: National Center on Child Abuse and Neglect. doi:10.1016/01907409(96)00014X Lansford, J. E., Dodge, K. A., Pettit, G. S., Bates, J. E., Crozier, J., & Kaplow, J., (2002). A 12 Year prospective study of the long term effects of early child physical maltreatment on psychological, behavioral, and academic problems in adolescence. Ar chives of Pediatric and Adolescent Medicine, 156, 824 830.

PAGE 93

93 Lawler, M.J. (2008). Maltreated children's emotional availability with kin and nonkin foster mothers: A sociobiological perspective. Children and Youth Services Review, 30, 11311143. doi:10.1016/j.childyouth.2008.02.007 Lee, J. & Holland, T. (1991). Evaluating the effectiveness of foster parent training. Research on Social Work Practice, 1, 162174. doi:10.1177/104973159100100204 LeProhn, N. (1994). The role of the kinship foster parent: a comp arison of the role conceptions of relative and nonrelative foster parents. Children and Youth Services Review, 16, 65 84. doi: 10.1016/01907409(94)900167 Leslie, L., Landsverk, J., Ezzet Lofstrom, R., Tschann, J., Slymen, D., & Garland, A. (2000). Chil dren in foster care: Factors influencing outpatient mental health service use. Child Abuse and Neglect, 24, 465 476. doi:10.1016/01907409(94)900167 Linares, L.O., Montalto, D., Rosbruch, N., & Li, M. (2006). Discipline practices among biological and foster parents. Child Maltreatment, 11, 157167. doi:10.1177/1077559505285747 Marcus, R. F., (1991). The attachments of children in foster care. Genetic, Social & General Psychology Monographs 11 7 365394. Maxwell, S.E. (2004). The persistence of underpowered studies in psychological research: Causes, consequences, and remedies. Psychological Methods, 9, 147 163. doi:10.1037/1082 989X.9.2.147 McLean, B., & Thomas, R. (1996). Informal and formal kinship caregiver populations: A study in contrasts. Child Welfare: Journal of Policy, Practice, and Program, 7, 489 505. McNeil, C., Herschell, A., Gurwitch, R., Clemens Mowrer, L. (2005). Training foster parents in Parent Child Interaction Therapy. Education and Treatment of Children, 28(2), 182 196. Mor Barak, M.E., Nissly, J.A., & Levin, A. (2001). Antecedent to retention and turnover among child welfare, social work, and other human service employees: What can we learn from past research? A review and metanalysis. Social Service Review, 75, 625661. doi : 0.1086/323166 Nzi, A.M. & Eyberg S.M. (2012). Tailoring Parent Child Interaction Therapy for oppositional defiant disorder in a case of child maltreatment. In ODonohue and Lilienfeld (eds.), Case studies in clinical psychological science ( in press). New York, NY: Oxford University Press.

PAGE 94

94 Patterson, G. R., Reid J. B. & Dishion T. J. (1992). Antisocial boys Eugene, OR: Castalia Press Pears, K., & Fisher, P.A. (2005). Developmental, cognitive, and neuropsychological functioning in preschool aged foster children: Associations with prior maltreatment and placement history. Journal of Developmental and Behavioral Pediatrics, 26, 112 122. doi:10.1097/00004703 20050400000006 Phillips, J., Morgan, S., Cawthorne, K., & Barnett, B. (2008). Pilot evaluation of parent child interaction therapy delivered in an Australian community early childhood clinic setting. Australian and New Zealand Journal of Psychiat ry, 42, 712719. doi:10.1080/00048670802206320 Pine, D.S. & Cohen, J.A. (2002). Trauma in children and adolescents: Risk and treatment of psychiatric sequelae. Biological Psychiatry, 51, 519531. doi:10.1016/S00063223(01)01352X Price, J.M., Chamberlain P., Ladsverk, J., Reid, J.B., Leve, L.D., & Laurent, H. (2008). Effects of a foster parent training intervention on placement changes if children in foster care. Child Maltreatment, 13, 64 75 doi:10.1177/1077559507310612 Puddy, R. & Jackson, Y. (2003). The development of parenting skills in foster parent training. Children and Youth Services Review, 25(12) 987 1013. doi:10.1016/S01907409(03)001063 Rausch, J.R., Maxwell, S.E., & Kelly, K. (2003). Analytic methods for questions pertaining to a randomized pretest, posttest, follow up design. Journal of Clinical Child & Adolescent Psychology, 32, 467 486. doi: 10.1207/S15374424JCCP3203_15 Rogers, J.L., Howard, K.I., & Vessey, J.T. (1993). Using signi ficance tests to evaluate equivalence between two experimental groups, Psychological Bulletin, 113 553565. doi:10.1037//00332909.113.3.553 Rubin, D. M., Alessandrini, E. A., Feudtner, C., Mandell, D.S., Localio, A. R., & Hadley, T. (2004). Placement st ability and mental health costs for children in foster care. Pediatrics, 113 13361341. doi:10.1542/peds.113.5.1336 Schofield, G. & Beek, M. (2005). Providing a secure base: Parenting children in long term foster family care. Attachment & Human Development 7 (1), 3 26. doi:10.1080/14616730500049019 Schuhmann, E.M., Foote, R., Eyberg, S.M., Boggs, S., & Algina, J. (1998). Parent child interaction therapy: Interim report of a randomized trial with short term maintenance. Journal of Clinical Child Psychology, 27, 34 45. doi:10.1207/s15374424jccp2701_4

PAGE 95

95 Smith, D.K., Stormshack, E., Chamberlain, P., & Whaley, R.B. (2001). Placement disruption in treatment foster care. Journal of Emotional and Behavioral Disorders, 9, 200 205, doi: 10.1177/106342660100900306. Steer, R.A., Ball, R., Ranieri, W.F., Beck, A.T. (1999). Dimensions of the Beck Depression Inventory II in clinically depressed outpatients. Journal of Clinical Psychology, 5, : 117 128. doi:10.1002/(SICI)10974679(199901)55:1<117::AID JCLP12> 3.0.CO;2 A Timmer, S. G., Sedlar, G., & Urquiza, A. J. (2004). Challenging children in kin versus nonkin foster care: Perceived costs and benefits to caregivers. Child Maltreatment 9 251 262. doi:10.1177/1077559504266998 Timmer, S.G., Urquiza, A.J., & Zebell, N.M. (2006). Challenging foster caregiving maltreated child relationships: The effectiveness of parent child interaction therapy. Children and Youth Services Review, 28, 1 19. doi:10.1016/j.childyouth.2005.01.006 Timmer, S.G., Urquiza, A.J., Zebel l, N.M., McGrath, J.M. (2005). Parentchild interaction therapy: Application to maltreating parent child dyads. Child Abuse & Neglect, 29, 825842. doi:10.1016/j.chiabu.2005.01.003 U.S. Department of Health and Human Services (USDHH S ). (201 0 ). Child Maltre atment 2010. Retrieved May 28 2012, from http://www.acf.hhs.gov/programs/cb/pubs/cm10/index.htm U.S. Department of Health and Human Services (USDHH S ). (2011). The AFCARS Report: Prelimi nary FY 2010 Estimates as of June 2011 (18). Retrieved March 05, 2012, from www.acf.hhs.gov/programs/cb/stats_research/afcars/tar/report18.htm Van Camp, C., Vollmer, T., Goh, H L., Whitehouse, J., Montgomery, J., & Borrero, J. (2008). Behavioral parent training in child welfare: Evaluations in skills acquisitions. Research on Social Work Practice, 18, 377 391. doi:10.1177/1049731507314008 Webste r Stratton & Spitzer, (1991). Development, reliability, and validity of the Daily Telephone Discipline Interview. Behavioral Assessment, 13, 221 239. Werba, B., Eyberg, S.M., Boggs, S.R., & Algina, J. (2006). Predicting Outcome in Parent Child Interaction Therapy: Success and Attrition. Behavior Modification, 30, 618646. doi:10.1177/0145445504272977 Wierzbicki, M. & Pekarik, G. (1993). A metaanalysis of psychotherapy dropout. Professional Psychology: Research and Practice, 24, 90 195. doi: 10.1037/07357028.24.2.190

PAGE 96

96 Zinn, A. (2012). Kinship foster family type and placement discharge outcomes. Children and Youth Sciences Review, 34, 602 612. doi:10.1016/j.childyouth.2011.11.015 Zisser, A., & Eyberg, S.M. (2010). Treating oppositional behavior in children using parent child interaction therapy. In A.E. Kazdin & J.R. Weisz (Eds.) Evidencebased psychotherapies for children and adolescents ( 2nd ed ., pp. 179 193). New York: Guilford.

PAGE 97

97 BIOGRAPHICAL SKETCH Amanda May Nzi (formerly Seib) was born in Aurora, Colorado. She was raised in Aurora where she graduated from Grandview High School in 2004 with high honors. She earned her Bachelor o f S cience in psychology in May 2007 from Milligan College in Milligan College, Tennessee, where she graduated magna cum laude. During her time at Milligan, Amanda received the Senior Psychology Major of the Year award for her work in the psychology field and her research in empathy and forgiveness. In August of 2007, she enrolled in a dual Mas ter of Science and Doctor of Philosophy program in the Department of Clinical and Health Psychology at the University of Florida in Gainesville, Florida. During her time at the University of Florida Amanda served as a Graduate Research Assistant in the Chi ld Study Lab under the mentorship of Sheila M. Eyberg, Ph.D. She was awarded the University of Florida Florence Shafer Memorial Award for Excellence in Psychotherapeutic Counseling and the College of Public Health and Health Professions PHHP Graduate Research Grant. Amanda received her Master of Science degree from the Department of Clinical and Health Psychology at the University of Florida in May 2009. Amanda went on to complete her predoctoral internship at University of Oklahoma Health Science Center She received her Ph.D. in psychology from the University of Florida in the summer of 2013.