Training Methods for the Child Directed Interaction (CDI) in Parent-Child Interaction Therapy (PCIT) and Parenting Skill Acquisition

Material Information

Training Methods for the Child Directed Interaction (CDI) in Parent-Child Interaction Therapy (PCIT) and Parenting Skill Acquisition
O'Brien, Kelly Ann
Place of Publication:
[Gainesville, Fla.]
University of Florida
Publication Date:
Physical Description:
1 online resource (58 p.)

Thesis/Dissertation Information

Doctorate ( Ph.D.)
Degree Grantor:
University of Florida
Degree Disciplines:
Clinical and Health Psychology
Committee Chair:
Eyberg, Sheila M.
Committee Members:
Pereira, Deidre B.
Boggs, Stephen R.
Smith, Stephen W.
Graduation Date:


Subjects / Keywords:
Child psychology ( jstor )
Coaching ( jstor )
Magnetic storage ( jstor )
Mothers ( jstor )
Parent training ( jstor )
Parenting skills ( jstor )
Parents ( jstor )
Teaching methods ( jstor )
Toys ( jstor )
Training methods ( jstor )
Clinical and Health Psychology -- Dissertations, Academic -- UF
coaching, training
Electronic Thesis or Dissertation
bibliography ( marcgt )
theses ( marcgt )
Psychology thesis, Ph.D.


Parent-Child Interaction Therapy (PCIT) uses several methods of training to promote parents' successful acquisition of skills involved in two basic parent-child interactions, child-directed interaction (CDI) and parent-directed interaction (PDI). Coaching is a core feature of PCIT that allows therapists to prompt and reinforce parents' use of their new skills during in-vivo parent-child practice. Coaching is assumed to be essential for accurate and efficient mastery of the PCIT skills that comprise the CDI and PDI, but there has been no empirical examination of the association between the coach training method and parenting skill acquisition. We were interested in the additive effects of both the performance feedback (via a skills frequency chart) and the coach training methods, over and above the effects of unaided parent practice following didactic training. A community-sample of 42 mothers of 3- to 6 year-olds with scores on the Eyberg Child Behavior Inventory (ECBI) Intensity Scale below the clinical cut-off ( < 132) for disruptive behavior participated in an experimental analog of CDI treatment by first viewing an 18-minute videotaped didactic presentation with explanations and brief therapist modeling of CDI parenting skills. Following didactic instruction, mothers were randomized to one of three CDI practice conditions with their child: (a) a 20-minute period of unaided practice (PRAC); (b) performance feedback (via frequency chart) on their skill acquisition from pre- to post-didactic training prior to a 20-minute period of unaided practice (FDBK; or(c) performance feedback on their skill acquisition prior to a 20-minute period of coached practice (COACH). A significant main effect for time was found for two of three 'Positive Following' skills (labeled praises and behavioral descriptions) and all 'Negative Leading' verbalizations (questions, commands, and critical statements) following didactic training alone. Following the practice period, a group by time interaction was found for the composite Positive Following skills, with the COACH condition demonstrating greater change than the PRAC or FDBK conditions, whereas no differences were found between the PRAC and FDBK conditions. Examination of the individual Positive Following skills showed a large effect size for change in behavioral descriptions in the COACH condition compared to the PRAC and FDBK conditions. More mothers in the COACH condition were labeled training responders, defined as having five of each Positive Following skill in the post-practice assessment. Training condition was not related to maternal change in Negative Leading behaviors. No between-group differences were found in training satisfaction ratings or the number of days of CDI practice during the two-week period following training. Higher yearly family income and maternal education were positively related to change in Negative Leading behaviors following didactic training. Future directions include examination of change in CDI Negative Leading behaviors with further coaching versus unaided practice, as well as examination of PCIT training methods for the acquisition of the PDI skills. Treatment component research examining the relations between PCIT training methods, skill acquisition, and family characteristics with clinical samples will help promote more efficient and effective treatment. ( en )
General Note:
In the series University of Florida Digital Collections.
General Note:
Includes vita.
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.
Thesis (Ph.D.)--University of Florida, 2008.
Adviser: Eyberg, Sheila M.
Electronic Access:

Record Information

Source Institution:
Rights Management:
Copyright Kelly Ann O'Brien. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Embargo Date:
LD1780 2008 ( lcc )


This item has the following downloads:

Full Text




2008 Kelly Ann OBrien 2


To my nephew, Ethen OBrien 3


ACKNOWLEDGMENTS I thank my graduate mentor, Dr. Sheila Eybe rg, for her guidance, support, and enthusiasm. I am appreciative of my doctora l committee, Drs. Stephen Boggs, Deidre Pereira, and Stephen Smith, for their contributions to this project a nd their support. Special thanks go to those who helped make this study possible, including Phillip Clemons, Courtney Ingalls, Melanie Fernandez, Kristen Marciel, Rhea Chase, Mary Brinkmeyer, Ashley Butler, Mary Keeley, and Monica Stevens. Finally, I thank my family and friends for their unconditional love and encouragement. 4


TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4 LIST OF TABLES................................................................................................................. ..........7 LIST OF FIGURES.........................................................................................................................8 ABSTRACT.....................................................................................................................................9 CHAPTER 1 INTRODUCTION................................................................................................................. .11 Methods for Parent Child Inter action Therapy (PCIT) Training............................................17 Study Objectives and Hypotheses..........................................................................................20 2 METHOD....................................................................................................................... ........23 Participants.............................................................................................................................23 Therapists........................................................................................................................23 Mother-Child Dyads........................................................................................................23 Procedures..................................................................................................................... ..........28 Recruitment.................................................................................................................... .28 Training....................................................................................................................... ....28 3 RESULTS...................................................................................................................... .........32 Observational Data Analysis..................................................................................................32 Skill Acquisition following Didactic Presentation.................................................................32 Parenting Skill Acquisition and Training Condition..............................................................33 Family Characteristics and Skill Acquisition.........................................................................35 Treatment Responders versus Nonresponders Analyses........................................................36 Parent Satisfaction............................................................................................................ ......38 Homework Practice.............................................................................................................. ..38 Post-Hoc Assessment of Child Behavior Change..................................................................38 Qualitative Findings........................................................................................................... .....39 4 DISCUSSION................................................................................................................... ......43 APPENDIX A SUMMARY OF CHILD DIRECTED INTERACTION (CDI) SKILLS...............................49 B CHILD DIRECTED INTERACTION (CDI) DIDACTIC PRESENTATION......................50 C SKILLS FREQUENCY CHART FOR PERFORMANCE FEEDBACK..............................54 5


LIST OF REFERENCES...............................................................................................................55 BIOGRAPHICAL SKETCH.........................................................................................................58 6


LIST OF TABLES Table page 2-1 Demographic characte ristics by training group.................................................................25 2-2 Mean Frequency of the Child Directed Interaction (CDI) beha viors pre-training and after didactic instruction....................................................................................................4 1 2-3 Mean frequency of the Child Directed Interaction (CDI) behaviors at each assessment for each training condition..............................................................................42 7


LIST OF FIGURES Figure page 2-1 Positive Following skills composite score.......................................................................34 2-2 Negative Leading behaviors composite score.................................................................35 8


Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy TRAINING METHODS FOR THE CHILD DIRECTED INTERACT ION (CDI) IN PARENT-CHILD INTERACTION THER APY (PCIT) AND PARENTING SKILL ACQUISITION By Kelly Ann OBrien May 2008 Chair: Sheila M. Eyberg Major: Psychology Parent-Child Interacti on Therapy (PCIT) uses several me thods of training to promote parents successful acquisition of skills involved in two basic parent-chi ld interactions, childdirected interaction (CDI) and pa rent-directed interaction (PDI). Coaching is a core feature of PCIT that allows therapists to prompt and reinforce parents use of their new skills during invivo parent-child practice. Coaching is assumed to be essential for accurate and efficient mastery of the PCIT skills that comprise the CDI and PD I, but there has been no empirical examination of the association between the coach training method and paren ting skill acquisition. We were interested in the additive effects of both the performance feedback (via a skills frequency chart) and the coach training methods, over and above the effects of unaided parent practice following didactic training. A community-sample of 42 mothers of 3to 6 year-olds with scores on the Eyberg Child Behavior Inventory (ECBI) Intensity Scale belo w the clinical cut-off (<132) for disruptive behavior participated in an e xperimental analog of CDI treatme nt by first viewing an 18-minute videotaped didactic presentation with explanations and brief ther apist modeling of CDI parenting skills. Following didactic instruction, mothers were randomized to one of three CDI practice 9


conditions with their child: (a) a 20-minute period of unaided pr actice (PRAC); (b) performance feedback (via frequency chart) on their skill acqui sition from preto post-didactic training prior to a 20-minute period of unaided practice (FDBK) ; or (c) performance f eedback on their skill acquisition prior to a 20-minute peri od of coached practice (COACH). A significant main effect for time was found for two of three Positive Following skills (labeled praises and behavioral descriptions ) and all Negative Leading verbalizations (questions, commands, and critical statements) following didactic training alone. Following the practice period, a group by time interaction was found for the composite Positive Following skills, with the COACH conditi on demonstrating great er change than th e PRAC or FDBK conditions, whereas no differences were f ound between the PRAC and FDBK conditions. Examination of the individual Positive Following sk ills showed a large effect size for change in behavioral descriptions in the COACH conditi on compared to the PRAC and FDBK conditions. More mothers in the COACH condition were labele d training responders, defined as having five of each Positive Following skill in the post-pr actice assessment. Training condition was not related to maternal change in Negative Leadi ng behaviors. No between-group differences were found in training satisfaction ratings or the numbe r of days of CDI pract ice during the two-week period following training. Higher yearly family in come and maternal education were positively related to change in Ne gative Leading behaviors fo llowing didactic training. Future directions include exam ination of change in CDI Ne gative Leading behaviors with further coaching versus unaided practice, as well as examination of PCIT training methods for the acquisition of the PDI skills. Treatment comp onent research examining the relations between PCIT training methods, skill acquisition, and fam ily characteristics with clinical samples will help promote more efficient and effective treatment. 10


CHAPTER 1 INTRODUCTION There are several evidence-based parent tr aining programs that aim to reduce child disruptive behaviors arising from early negative parent-child in teractions (Brestan & Eyberg, 1998). These treatment programs teach parents to use more effective parenting skills that improve the quality of parent-child interactions and decrease nega tive child behaviors. Studies of parent training programs have demonstrated that the changes in parenting behaviors are the causal mechanisms in producing changes in child outcomes (Forgatch & DeGarmo, 1999; Reid, Eddy, Fetrow, & Stoolmiller, 1999). Thus, instructional methods for training parents in new parenting skills are key to the efficiency and eff ectiveness of these treatments for children with disruptive behavior. Evidence-based parent training programs for young children with conduct problems have formats ranging from indi vidual parent training ( Parent Management Training Oregon Model (PMTO) ; Patterson, 2005), individual pa rent and child training ( Parent-Child Interaction Therapy (PCIT) ; Eyberg, 1999) and group parent training ( Incredible Years Parenting Program ; Webster-Stratton, 1990). In structional methods used within these training formats can include verbal instruction, written ha ndouts, discussion, videotape or live modeling, and role-play. The therapist(s) typically incorpor ate a combination of these inst ructional methods to facilitate parenting skills acquisition. A unique instructional component of PCIT is in-vivo coachin g, which allows parents to receive immediate therapist feedback as they practice new skills with their child during the treatment session (Eyberg, 1999). In dividual parent training with coaching has been found more effective in producing behavior change in parents and childre n than didactic-focused group parent training (Chaffin et al., 2004: Eyberg & Matarazzo, 1980). Advantages of the coach 11


training method include correction of parent mistakes the instant they happen, possibly hastened skill acquisition due to the immediacy and intens ity of coaching feedback, and tailored skill application to individual familie s (Herschell, Calzada, Eyberg & McNeil, 2002). Adaptations of standard PCIT, such as abbreviated and group form ats of PCIT, have retained the coach training component due to the impact of coaching on pare ntal skill development (Niec, Hemme, Yopp & Brestan, 2005; Nixon, Sweeney, Erickson & Touyz, 2003). Parent training research has traditionally focused on changes in child behavior, and sometimes parent behavior following treatment, but measures of parent be havior change or skill acquisition as a function of the training methods employed are infrequently included (Moreland, Schwebel, Beck, & Well, 1982). The earliest studie s of parent training programs instructional components assessed the relative efficacy of vari ous instructional methods, such as written materials, videotape modeling, lecture presen tation, role-playing, live modeling, and rehearsal for parenting skills acquisition (Flanagan, Ad ams & Forehand, 1979; ODell, Mahoney, Horton & Turner, 1979; ODell et al., 1982). These studies establishe d the efficacy of all of the instructional components over control conditi ons for time out and positive parenting skills acquisition. For time-out skills, Flanagan and co lleagues (1979) found videotape modeling (i.e., a videotape which consisted of short verbal descriptions and modeling) was more effective than written instruction and no different than the ro le-playing instructional condition (i.e., therapist briefly described time-out proced ure and then a group of parents we re divided into dyads to role play) for parent skill acquisition, while O Dell and colleagues (1979) found that videotape modeling plus a brief individual checkout (i.e., a training film was viewed and then parents received 7 minutes of therapist time to check their understanding of the material and provide a brief rehearsal of the techniques) was more effec tive than written instruction alone or individual 12


therapist modeling and rehearsal (i.e., therapist m odeled behaviors and role-played the part of the child so parents could practice and receive feedback during a 20-mi nute session). The brief check-out may have been more effective because parents were able to demonstrate initial skill gains from the videotape modeling instruction an d therapists could focus on areas of continued need in the brief time allotted. The investigators also noted that therapists reported that they spoke more quickly in the brief check-out cond ition than in the 20-mi nue individual modeling and rehearsal condition. For parenting reinforcem ent skills, no significant differences were found when the instructional methods of a written manual, videotape modeling or therapist live modeling and rehearsal were co mpared (ODell et al., 1982). The written manual, videotape modeling, and therapist m odeling and rehearsal instructional c onditions were all developed with parallel materials to standardize the information parents received. Parents received two treatment sessions, where instructional methods were first deliv ered in the clinic setting and then repeated in a second home booster training session. For the therapist live modeling and rehearsal condition, in the first clinic session therapists demonstrated reinforcement principles by modeling desired behaviors and then role-played the part of the child while the parent practiced, while during the home booster session the trainer observed the parent playing with his or her child and prompted and provided feedback. The i nvestigators speculated that differences for the training methods may have been reduced by the hi ghly parallel content of the training methods and the repeated exposure to training methods in each condition (ODell et al., 1982). The relative absence of some of the parenting skills targeted, su ch as asking questions and touch, in all of the treatment conditions was also noted. Thus, there was evidence for rather complex interaction effects between target ed parenting skills and outcome. 13


For successful treatment programs that include multiple instructional components, it is difficult to identify the instruc tional components responsible for the parents skill acquisition or the relative effectiveness of th e various procedures used. Two re cent studies have demonstrated the additive effects of additional therapist a ssistance following didactic instruction to systematically examine the association between different instructional methods and parenting skills acquisition (Foster & Robe rts, 2007; Lerman, Swiezy, Perkins-Parks, & Roane, 2000). Lerman and colleagues (2000) examined the relative effectiveness of low-cost training methods (i.e., written and verbal instruc tions) and therapist f eedback (i.e., verbal corrective feedback on observed parent-child interactions) for three famili es that were taught multiple parenting skills to decrease child problem behaviors. Based on the childs presenting problems, parents were taught targeted skills that included ignoring as a respon se to inappropriate behavi or, praise as a response to appropriate behavior, a nd communication prompts (e.g., as king what do you want?). A multiple baseline design was employed to assess the effectiveness of the training methods for acquisition of the targeted parenting skills for each family. Mothers were first presented with verbal and written instructions, and therapist feedback was introdu ced only if the parent did not reach a pre-set mastery criterion within 1 to 3 sessions. Results for individual families indicated that the efficacy of written and verbal instructions was variable; all parents achieved a high level of accuracy in the use of at le ast one parenting skill following written and verbal instruction alone, but across the three families, the speci fic skill that was accurately learned was not consistent. Each parent also required therapist fee dback to meet the training criterion for at least one of the skills. The inclusion of only three families and slight variations in targeted parenting skills across the families limits broad conclusions However, the preliminary findings suggested that the effectiveness of speci fic instructional methods (writt en and verbal discussion only 14


compared to additional therapist feedback) was re lated to the type of pa renting skill taught and varied for individual families. A second study also examined parental acquisition of specific parent training skills by assessing parent behavior change following vi deotape modeling alone and then with therapist assistance (Foster & Roberts, 2007 ). Foster and Roberts (2007) hypothesized that some parents would have difficulty implementing parenting st rategies with videotap e modeling alone, and would require therapist support and guidance to use new skills accurately. Maternal acquisition of responsive play skill s (descriptions, praise, and imita tion) and compliance training skills (effective commands and time-out sequence) was fi rst examined following videotaped modeling. After reading materials and vi ewing a 25-minute videotape th at included a rationale and graduated demonstrations of responsive play ski lls, four of ten mothers met pre-set behavioral mastery criteria. Mothers demonstrated significantly improved rates of positive attention and significantly reduced intrusive ve rbalizations during play afte r the videotape modeling alone; however, to meet mastery crit eria 60% required one standard parent training session, which included a variety of training procedures (e.g, discussion, live m odeling, guided practice with invivo coaching). For the compliance training sk ills of effective commands and a time-out sequence, only one mother demonstrated pre-set behavioral mastery cr iteria after reading materials and viewing a 45-minute videotape that included a rati onale and graduated demonstrations of compliance-eliciting skills. The remaining ni ne mothers required one to two parent training sessions with a dditional instructional methods of discussion, live modeling, and guided practice with in-vivo coach ing to meet mastery criteria. The investigators found that a variety of errors that could have potentially lim ited the long-term benefits of compliance training were observed following videotap ed modeling alone. Overall, thes e preliminary findings suggest 15


the relative effectiveness of th eir standard parent training inst ructional methods (discussion, live modeling, in-vivo coaching) compared to vide otaped modeling alone for parents skill acquisition. However, the combination of methods used in the second session does not explore the specific instructiona l technique(s) responsible for pare nts additional skill acquisition after videotape modeling. Small sample sizes have limited researchers ability to examine the relations between socio-demographic variables, in structional methods, and skill acquisition. One study found that parent variables, including educat ion level, socioeconomic status (SES), and reading level, were related to skill acquisition (ODell et al., 1982) A study that examined the interaction of mothers SES and instructional methods for sk ill acquisition found that mothers with lower SES that received modeling and role playing inst ruction were observed to use the behavior management skills they were taught with highe r frequency than mothers with lower SES who received reading and discussion training (Knapp & Deluty, 1989). In addition to behavioral skill acquisition outcome s, it is also important to examine parent satisfaction with instructional methods. The early studies of va rious instructional methods and skill acquisition found no differences between gro ups for parent attitudes toward training or satisfaction ratings (Flanagan et al., 1979; ODell et al., 1979 ; ODell et al., 1982). A recent study comparing 4-session PCIT groups in primar y care with a correspondi ng self administered program with PCIT bibliotherapy also showed no differences in a measure of parental satisfaction (Harwood & Eyberg, in prep). Howe ver, Nicholson and Sanders (1999) found that parents rated higher satisf action with a therapist-directed beha vioral family intervention than a matching self-directed intervention with written mate rials. Parental expectat ions for interventions and acceptability of training format s are important to consider, for self-directed interventions or 16


those with minimal therapist feed back require parental willingness to carry out behavior changes with little or no support from a th erapist (Elgar & McGrath, 2003). Methods for Parent Child Interaction Therapy (PCIT) Training In standard PCIT parents initially receive verbal instruction, therapist modeling and roleplay, and written handouts that review concepts. Af ter initial didactic inst ruction, parents receive weekly performance feedback on their skills obse rved during parent-child interaction via a skills frequency chart followed by guided practice with in-vivo coaching. Each of these instructional methods is described in more detail below: Didactic Presentation and Brief Modeling. A PCIT didactic presentation is given in the first treatment session for the first phase of tr eatment, the Child Directed Interaction (CDI). Therapists provide a definition and rationale for using Positive Following CDI skills (labeled praise, reflective statement, beha vioral description) and avoidi ng Negative Leading behaviors (information and descriptive/refl ective questions, direct or in direct commands, and critical statements). See Appendix A for a summary of PC IT CDI skills. Therapists also provide several examples of each Positive Follo wing skill to use and each Negativ e Leading behavior to avoid. Following the didactic presentation, therapists br iefly model the use of the CDI skills with the co-therapists playing the parent and child role. Parents are given written handouts that summarize all of the skill definitions, rationales, and examples to take home. Performance Feedback. Following the initial didactic session which includes the didactic presentation and modeling, the remaining sessio ns include performance feedback and coaching. For the feedback, therapists begin by observing and coding (i.e., tally freq uencies on a chart) the CDI Positive Following skills and Negative Lead ing behaviors during a 5-minute observation. After the behavioral coding pare nts are provided with feedback in which they are praised for their effort, skill frequencies in the 5-minute tim e period are described (i.e., you gave three 17


labeled praises), and the therapist identifies ski lls that will be targeted in coaching based on the skills coding (i.e., today we will work on increasing reflective statements and decreasing questions). The parents skill frequencies collect ed during coding are recorded on a chart or graph to show parents a visual representation of their progress, and their current performance is compared to pre-set behavioral goals that indicate mastery. The mastery criteria for the PCIT CDI phase are 10 labeled praises, 10 behavioral descriptions, and 10 re flective statements, and fewer than 3 questions (both information and de scriptive/reflective questions), commands (both direct and indirect) or critical st atements in a 5-minute time period. In-Vivo Coaching. The skills data from the 5-minute obs ervation are used to guide in-vivo coaching of the parenting skills while parents practice with their child. Coaching is conducted in PCIT using a bug-in-the-ear system and one-way mirror. This method allows the parent to be alone in the room with the child when practicing the skills while receiving verbal communication from the therapist who is monitoring the interac tion from the other side of the one-way mirror. When these technologies are not available, in-room coaching is possible in which the therapist sits slightly behind the parent opposite the side of the child so the therapist can speak quietly to the parent and remain unobstrusive to the parent and child interaction (Herschell, Calzada, Eyberg & McNeil, 2002). For two parent families, each parent is coached approximately 20 minutes so that the majority of the session involves coaching. For one parent families, coaching is typically 30 minutes in each session. The coaching instructional method provides th e parent with immediate feedback on their implementation of the interaction skills as the parent plays with their child. Verbal communication from coach to parent is active and intensive. Therapists are taught to use coaching verbalizations that are mostly brief, qui ck, and positive statements that train the parent 18


to increase the Positive Following skills (i.e., labeled praises, reflective statements, and behavioral descriptions) and decrease Negative Leading behaviors of CDI. Gentle corrections (e.g., briefly pointing out parent mistake, such as by saying oops, question ) of parent mistakes are used sparingly, and are used particularly for raising the pa rents awareness of habitual negative communication behaviors. A therapist must be aware of the timing and delivery of feedback to the parent, who is involved in ongoing transactions with the child. A coaching session constitutes a moment-by-moment functional analysis in which behaviors of both the parent and the child are shaped toward more positive interactions and collaboration. Thus, therapists must stay attentiv e to the individual interaction style of each dyad in each session while working toward the treatment goals of skil l mastery and improved pare nt-child interaction. The behavioral principles that parents are taught to use with their child in CDI to change behavior (differential social attent ion) are similar to the principles that the therapist uses to train parents during coaching (Borrego & Urquiza, 1998). In CDI, therapists use differential social attention by giving most attention to correct parenting skills with the assumption that parents will perform behaviors that gain the greatest support a nd approval from the therap ist. Less attention is given to incorrect parent behavior s. The therapist verbally shapes parent behaviors by cueing and reinforcing closer and closer approximations to CDI goal beha viors. The therapists shaping behavior is similar to the parents shaping of th e childs behavior. The consistent application of social reinforcement for correct parent beha viors increases the likelihood of positive parent behavior change, which then le ads to changes in the interac tion between parent and child. Therapists also use client centered ther apy principles during coaching in CDI by expressing empathy, genuineness, and positive rega rd to the parent. The coach demonstrates these qualities by accurately desc ribing and being sensitive to the parents feelings during 19


coaching, treating the parent with respect (i.e., not acting condescending or superior), and not disapproving of the parent. The didactic presentation, performance fee dback, and in-vivo coach ing methods are also used to teach parents the disc ipline skills of the second phase of PCIT, the Parent Directed Interaction (PDI). However, in PDI coaching parent mistakes are always corrected immediately by the therapist. Study Objectives and Hypotheses The additive effects of the in-vivo coaching method over a nd above the effects of the initial didactic training has not been examined empirically for the acquisition of the specific positive parenting skills taught in the first phase of PCIT. Parents in treatment are able to attain pre-set behavioral mastery criteria with the combination of training methods used. However, this study examines the additive value of therapist performance feedback and in-vivo coaching for parental skill acquisition. This study extends the research by Lerman and colleagues (2000) and Foster and Roberts (2007) by examining the specif ic positive following ski lls taught in PCIT. In addition, the performance feedback and coaching methods are used individually following didactic instruction to examine the relative valu e of these individual training methods. The larger sample size will also allow for examination of th e relations between socio-demographic variables and acquisition of positive parenting skills. The majority of participants in the existing parent training skill acquisition literat ure were mothers so we included mothers only for this preliminary examination of PCIT instructiona l methods and skill acquisition. Furthermore, a non-clinical sample was recruited due to the brief nature of the training. Mothers first watched an 18-minute videotaped CDI didactic presentation with skill explanations and examples, as well as brief ther apist modeling. After this standardized didactic instruction, mothers were randomly assigned to one of three practice conditions: (a) 20-minutes 20


of unaided practice (PRAC), (b) performance fee dback (via frequency chart) on their skill acquisition from pre-to post-did actic training prior to a 20-mi nute period of unaided practice (FDBK), or (c) therapis t performance feedback, followed by 20-minutes of guided practice with in-vivo coaching (COACH). The first objective was to examine the effect s of the CDI didactic presentation alone on mothers positive parenting skills acquisition. We expected mothers to demonstrate a significant increase in Positive Following skills and a significant decrease in Negative Leading behaviors following didactic presentation alone. The second objective was to examine the additiv e effects of the performance feedback and the coach instructional methods, over and above skill acquisition related to didactic instruction alone. It was predicted that mothers in the COACH condition would show greater skill acquisition, with greater increases in total Positive Following skills and greater decreases in total Negative Leading behaviors after the 20-minute pr actice period than mothers in the PRAC and FDBK conditions. We also hypothesized that more mothers in the COACH condition would be considered training responders, defined as havi ng five or more of each Positive Following skill and less than 6 total Negative Le ading behaviors in a 5-minute a ssessment, than mothers in the PRAC and FDBK conditions. Analyses for differences between the PRAC and FDBK conditions were exploratory because performance fee dback alone has not been used in PCIT. The third objective of this study was to exam ine the relations between socio-demographic variables and maternal skill acquisition. We predic ted that higher maternal education and higher family income would be related to greater skill acquisition based on past research findings (ODell et al., 1982). 21


The fourth objective of this study was to compar e maternal satisfaction ratings for the three instructional methods as well as their ratings of helpfulness of the skills when taught by way of the different methods. We expected that all moth ers would report generally high satisfaction with the parenting skills taught and the instructional methods. However, we hypothesized that mothers in the FDBK and COACH conditions would rate higher satisfaction with training than mothers who received videotape didactic training followed by unaided practic e. Exploratory analyses also examined differences in mothers daily CDI pr actice for the two weeks following their training visit. 22


CHAPTER 2 METHOD Participants Therapists Therapists were 6 graduate students who were well-trained PCIT therapists. All therapists had taken a PCIT course, participated in weekly PCIT supervision, and had been a therapist for at least 3 PCIT cases. These traini ng experiences assured that therapis ts were able to code parent behavior accurately to provide feedback and to guide coaching of CDI skills. All therapists served in both the FDBK and COACH conditions. Therapists were five Caucasian females and one African American female, ages 23 to 28 years old. They had completed from one to five years of graduate school tr aining in a Clinical and H ealth Psychology program. Mother-Child Dyads Forty-two mother-child dyads were recruite d from the Gainesville community. Male and female children between the ages of 3 and 6, and mothers age 18 and older were eligible for the study. Mothers only were included in this pr eliminary study of training methods and skill acquisition. Because maternal depression was a si gnificant predictor of impaired mother-child functioning prior to CDI in a sa mple of children with elevated behavior problems (Harwood & Eyberg, 2005), the Center for Epidemiologic Studies Depression Scale (CES-D; Radloff, 1977) was completed so that mothers with elevated le vels of depressive symptoms could be excluded (no mothers had elevated CES-D scores). Childre n with clinically elevated behavior ratings according to maternal report (Eybe rg Child Behavior Inventory in tensity score > 132) were also excluded due to the focus on maternal acquisition of skills in a brief training model (six mothers reported clinical range ECBI scores for their chil d). Mothers of children with clinically elevated 23


ECBI scores were given information about services offered in the health center psychology clinic or community settings. Of the 42 maternal caregivers who participated in the study, 93% were biological and 7% were adoptive mothers. Their mean age was 34.55 years ( SD = 5.90) and their self-identified racial/ethnic background was 67% Caucasian, 19% African-American, 10% Asian American/Pacific Islander, and 7% Hispanic. They reported a wide range in education level [10th grade (2%); high school diploma or GED (12%); some college or associates degree (33%); bachelors degree (29%); masters degree (19%); and doctoral degr ee (5%)] and in yearly family income [<$20,000 (19%); $20,001 to $40,000 ( 31%); $40,001 to 60,000 (26%) and >$60,000 (24%)]. Their CES-D (depression) scores averaged 6.36 ( SD = 4.46). The children included 21 girls and 21 boys, with a mean age of 50.81 months ( SD = 12.68 months). The childrens racial/ethnic compos ition was 62% Caucasian, 17% African-American, 14% bi-racial, 5% Hispanic, and 2% Asian-American/Pacific-Islander. Mean maternal rating of child disruptive behavior on the ECBI Intensity Scale was 103.26 ( SD = 20.08), which is very close to the normative mean of 97. Socio-demogra phic data were compared for the three training groups using one-way ANOVAs or 2 analyses. See table 1 for demographic characteristics of participants by training condition. No signi ficant group differences were found, although differences in marital stat us approached significance, 2 (2) = 4.90, p = .086, with 93% of mothers in the coach training c ondition classified as married. Measures Family Demographic Questionnaire A questionnaire was given to collect sociodemographic information about the mother and ch ild including age, ethn icity, highest education level (of the mother), sex (of the child), and family yearly income. 24


Table 2-1. Demographic charac teristics by training group PRAC FDBK COACH f 2 p Maternal caregiver --2.12 .346 Biological mother 100% 93% 86% Adoptive mother 0% 7% 14% Marital status 57% married 64% married 93% married --4.90 .089 Mother age (years) M = 33.36 SD = 5.97 M = 35.36 SD = 6.20 M = 34.93 SD = 5.78 0.40 --.671 Mother education (years) M = 15.57 SD = 2.10 M = 15.86 SD = 2.77 M = 14.43 SD = 1.79 1.57 --.221 Mother ethnicity --2.10 .350 Caucasian 64% 57% 79% African-American 7% 36% 14% Hispanic 14% 7% 0% Asian-American 14% 0% 7% Bi-racial 0% 0% 0% Child age (months) Note. n = 14 in each training condition. PRAC = did actic instruction plus unaided practice. FDBK = didactic instruction pl us therapist performance feedb ack before unaided practice. COACH = didactic instruction plus therapist pe rformance feedback before coached practice. M = 46.93 SD = 12.72 M = 51.0 SD = 14.64 M = 54.5 SD = 10.00 1.27 --.293 Child gender 57% male 36% male 57% male --1.71 .424 Child ethnicity --.968 .616 Caucasian 64% 50% 71% African-American 0% 36% 14% Hispanic 14% 0% 0% Asian-American 0% 0% 7% Bi-racial 21% 14% 7% Yearly family income --2.63 .854 $0-$20K 29% 14% 14% $20K-$40K 28% 29% 35% $40K-$60K 14% 36% 29% above $60K 28% 21% 21% Mother CES-D M = 5.36 SD = 4.45 M = 7.93 SD = 4.75 M = 5.79 SD = 4.04 1.36 --.269 Pre ECBI Intensity M = 97.21 SD = 18.66 M = 112 SD = 19.76 M = 100.57 SD = 20.08 2.21 --.123 Eyberg Child Behavior Inventory (ECBI; Eyberg & Pincus, 1999). The ECBI is a 36item parent report measure of disruptive behavior in children between 2 and 16 years of age. It measures disruptive behaviors in terms of thei r frequency (Intensity Scale) and the degree to 25


which these behaviors are problematic for the pare nt (Problem Scale). Only the Intensity Scale was used in this study. On this 7-point scale, th e total intensity score can range from 36 to 252, with a normative mean of 96.6 and a standard deviation of 35.2. Within a community sample, 12-week test-retest reliability of .80, and 10-m onth test-retest reliability of .75 have been reported for the Intensity Scale (Funderburk, Eybe rg, Rich, & Behar, 2003). The ECBI was used as a screening measure in this study, as well as a two-week telephone follow-up measure of child behavior change followi ng the brief training. Center for Epidemiologic Studies Depression Scale (CES-D; Radloff, 1977) The CESD is a 20-item self-report questi onnaire assessing level of depr essive symptoms. The measure was developed by the Center for Epidemiologic St udies and is intended for research with the general population. Sixteen items express negative feelings or behaviors (e.g., I was bothered by things that usually dont bothe r me) and four items express positive feelings or behaviors (e.g., I enjoyed life). Each item is rated on a 4-point scale accord ing to how often the feeling or behavior has occurred in the past week (1 = rarely or none of the time, 2= some or a little of the time, 3 = occasionally or a moderate amount of ti me, and 4 = most or all of the time). A total score is calculated (rang e of 0 to 60), with higher scores i ndicating higher levels of distress. A score of greater than or equal to 16 indicates clinically elevated levels of distress, but does not necessarily mean that the participant has a cl inical diagnosis of de pression. High internal consistency has been reported for the measur e (Radloff, 1977), as well as good test-retest reliability (e.g., Hann, Winter, & Jacobsen,1999). The CES-D was used as the depression screening measure for the mate rnal caregiver in this study. Dyadic Parent-Child Interaction Coding System (Third Edition) (DPICS; Eyberg, Nelson, Duke, & Boggs, 2004). The Dyadic Parent -Child Interaction Coding System is a 26


behavioral coding system that measures the qual ity of parent-child so cial interaction during standard 5-minute situations that vary in the degree of parental control required. This study used the child led play situation in which parents are asked to follow along with their child in whatever game their child chooses. The coded pare nt behaviors of interest in this study were those that express reciprocity, nu rturance, and parental control. A strength of the DPICS is the inclusion of parent behaviors targeted for ch ange in PCIT, which permits parents skill acquisition to be monitored. Three graduate students and one post-baccalaureate student served as primary coders of the DPICS categories that comprised the Positiv e Following and Negative Leading behavior measures. All coders complete d a coder-training workbook (Fernandez, Chase, & Eyberg, 2005) and reached 90% accuracy with a criterion tape coded by expert DPICS coders. All coders were uninformed of assessment time (pre -, post-didactic, and post-practic e), and all coders with the exception of the first author were uninformed of training condition and study hypotheses. Onethird of taped segments (42 segments) were coded independently by two observers for the purpose of calculating kappa inter-rater reliabilit y. According to Fleisss (1981) convention for interpretation of kappa statisti cs, all individual categor ies were coded with good to excellent inter-observer reliability. Following is a listing of the specific behaviors with kappa coefficients included in parentheses: Information Question (.94), Descriptive Refl ective Question (.88), Direct Command (.88), Labeled Pr aise (.87), Behavioral Descri ption (.84), Indirect Command (.74), Reflective Statement (.70), and Negative Talk (.65). Satisfaction Questionnaire (SQ) The SQ was created for this study to measure maternal satisfaction with the instructi onal methods and maternal ratings of the helpfulness of the parenting skills. The SQ consists of 4 to 7 items with a 5-point response choice for each item (1 27


not at all to 5 extremely). All mothers received the first f our questions asking about the helpfulness of the skills (how helpful were the skills for improving communication with your child and how helpful were the skills for engaging your child), th eir likeliness to use the skills regularly in the future, and the helpfulness of the didactic instruction. Questionnaires given to mothers in the two therapist-assisted conditions had an additional question about the helpfulness of the performance feedback chart, and forms for mothers in the coaching condition had a question about the helpfulness of this instruction. All mothers we re also asked to rate their overall satisfaction with the training they rece ived. Space was also given for mothers to write additional comments about their experience. Procedures Recruitment The study protocol and informed consent were approved by the unive rsity Institutional Review Board. Mother-child dyads were recruited through flyers that invited mothers of 3to 6year-olds to participate in a study to learn new ways to comm unicate with and engage their young child during play. During the initial phone call information was provided about study eligibility and procedures. If mothers remained interested a visit was scheduled at their convenience. Fifteen of the 63 mothers who cal led for information did not schedule. Those mothers that provided reasons cited time demand or the difficulty of bringing their child with them to the visit as barriers to scheduling. Training At the beginning of the study visit, the in formed consent was reviewed, and screening measures were completed (ECBI, CES-D, fam ily demographic form). Mothers who met study inclusion criteria were then obs erved with their child in the st andard 5-minute DPICS Child Led Play (CLP) situation, in which they were instruct ed to follow along with their child in whatever 28


game their child chose to play (Time 1: pretraining). This first obs ervation provided the mothers baseline measures of Positive Follo wing and Negative Leading behaviors. The playroom was set up with three standardized toys (farm playset, car and garage playset or train playset, and colored foam building blocks). Following this initial observation, all mothers viewed an 18-minute videotaped CDI didactic pr esentation that reviewed the parenting skills taught in the CDI portion of PCIT (see appendix B for outline). Two PCIT therapists (including the first author) provided the information on the video to standardize the didactic training across all participants. Mothers were at this point randomized to one of two training conditions: One third of mothers were randomized to the PRAC condition, and twothirds were randomized to a therapist-aided condition. Mothers were randomized to the FBCK or COACH conditions after the performance feedback was given so that therapists would be uninformed of practice condition (unaided or with coaching) during feedback. Mothers then returned to the playroom for their second parent-child interaction assessment of CLP, after which they were as ked to practice the skills they learned from the video as best they could, aiming for the pre-set behavioral mast ery criteria. They were told that after a twominute warm-up, the therapist would observe and s core their skills for a 5 minute period to see how they were doing (Time 2: after didactic training). After this second observation was completed, mothers who were randomly assigned to the therapist-assisted group were gi ven performance feedback on how close their skills were to mastery criteria. Feedback was aided by a char t showing the frequenc y counts of the three Positive Following skills and the Negative Lead ing behaviors (see appendix C for performance feedback chart). Therapists were instructed to give mothers br ief general support and praise for their effort while reviewing their skill frequencies. If the mothers asked a question about a skill, 29


therapists were instructed to provide a simple definition of the skill first (e.g., A behavior description is a statement that de scribes what your child is doing). If parents inquired further, therapists provided one example (e.g., An example of a behavior description would be youre feeding the horse some hay). Twenty-one of the 24 feedback segments we re taped (recording was stopped, in error, immediately after the coding period for three fam ilies, so the feedback could not be integrity checked for these families). Data for the 21 mo thers indicated that they each received approximately two praises or supportiv e statements during the feedback ( M = 1.95, SD = 1.11; e.g., youre doing a great job prac ticing or it is hard to no t ask questions but youre doing a good job). One-third of mothers asked the therap ist one question and one mother (5%) asked the therapist two questions. Of the mothers who asked questions, 57% asked for an example of a Negative Leading behavior they engaged in (e.g., When did I give commands?) and 43% asked questions to clarify their unders tanding of a skill (e.g., A labe led praise is more specific right?). One mother asked what she should do if her child kept playing with the same toy because she felt it was too repetitive. Therapis ts provided one skill definition for two mothers and two skill definitions for one mother following parent questions during feedback. No skills were modeled during feedback. Twenty-four percen t of mothers commented to the therapist that they did not realize how often they asked questions or directed their childs behavior during play. Following performance feedback, mothers were randomized to either the FDBCK or COACH condition. Following feedback, mothers randomized to th e FDBK condition were asked to practice the CDI skills with thei r child for 20 minutes, after which th eir skills would be assessed again and their progress reviewed. Mothers randomized to the COACH condition were asked to 30


practice the CDI skills with thei r child for 20 minutes while being coached by the therapist via the bug-in-the-ear device. Coaches used a range of st rategies to assist the mothers in their use of the skills, including prompting for use of a pa rticular skill (Praise him for sharing), reinforcement for accurate application of skills (Wonderful behavior description), gentle corrections for mistakes (Oops, question), a nd observations regarding the childs response to the mothers behavior (He talks more when you reflect). All mothers were coded again for five minut es following the 20 minut es of practice (Time 3: post-practice assessment). At the end of the visit, mothers completed a satisfaction questionnaire. Mothers were asked to do a five -minute daily CDI practic e with their child and were given sheets to record thei r practice at home to he lp increase the accuracy of their report in the follow-up phone call. Handouts that reviewed the CDI strategies and appropriate CDI toys were given at the end of the visit. Approxima tely 19 days following their training visit ( M = 19.32, SD = 7.35), mothers completed the telephone follow-up, which involved administration of ECBI Intensity Scale and questions asking the number of days they had practiced. Mothers received $25 compensation for the training visi t and $5 for completion of the telephone followup. 31


CHAPTER 3 RESULTS Observational Data Analysis Mothers skills were examined during 5-minute taped segments before the didactic training (T1), after the didactic traini ng (T2), and following the 20-minute practice period (T3). The twoweek follow-up did not include be havioral observations; mothers were called and completed the ECBI and reported the number of days they practiced. Prior to quantitative analyses, the observationa l data were assessed for normality for each group at T1, T2, and T3. When indicated, log tr ansformation was used to bring skewness and kurtosis within acceptable limits. When log tr ansformations were unsuccessful in achieving normal distributions, nonparametric tests were used to analyze group differences. Table 2 shows the mean frequency for all Positive Following ski lls and Negative Leading behaviors at T1 and T2 for the entire sample. Table 3 shows the me an frequency of each Positive Following skill, Negative Leading behavior, and Positive Followi ng and Negative Leading composite scores for mothers in each training condition (PRAC, FD BK, COACH) at each assessment (T1, T2, T3). There were no pre-training group di fferences for the CDI skills. Skill Acquisition followin g Didactic Presentation Following didactic instruction alone, mothers de monstrated a significant increase in total Positive Following skills, t (41) = 5.84, p < .001, and a significant decrease in total Negative Leading behaviors, t (41) = 6.40, p < .001 (see Table 2). All individual behavioral categories significantly changed with the ex ception of reflective statements. Examination of individuals skill acquisition showed that 21% of mothers had si x or more reflective statements at pre-training and all but one of these mothers decreased from pre-training to the pos t-didactic assessment. Conversely, 20% of mothers increa sed their use of reflective stat ements by six or more following 32


didactic training alone. There was wide variability in mothers change for most targeted CDI behaviors. Mothers infrequently gave critical statements at the preor post-didactic assessments. Parenting Skill Acquisition and Training Condition The additive effects of performance feedb ack and in-vivo coaching were examined over and above the effects of unaided practice following didactic instru ction. A 2 (Time 2, Time 3) X 3 (PRAC, FDBK, COACH) repeated meas ures ANOVA was conducted to examine group differences in skill acquisition following the pr actice period. Boxs M and Levenes Test for Equality of Error Variances were not significant. The main effect for time was significant, F (1,40) = 30.66, p < .001, 2 = .44, power = 1.00, and there was a significant Time x Group interaction, F (2,39) = 11.00, p < .001, 2 = .36, power = .99. The change in Positive Following composite scores at each assessment for each tr aining condition is presented in Figure 1. Followup univariate analyses were conducted to exam ine group differences. Mothers in the COACH condition showed greater change in tota l Positive Following skills than the PRAC, t (26) = 3.17, p = .004, and FDBK conditions, t (26) = 4.12, p < .001. No differences were found between the PRAC and FDBK conditions, t (26) = 1.46, p = .158. Follow-up analyses were conducted to examine the interaction between the individua l Positive Following skills and instructional methods. Log transformation did not bring data normality within acceptable limits so the nonparametric Kruskal-Wallis test was used. Change in behavior descriptions was significantly affected by training group, H (2) = 14.8, p = .001. Mothers in the COACH condition improved more than mothers in the PRAC, U = 23.5, p < .001, and FDBK conditions, U = 29, p < .001. There was no significant difference between the PRAC and FDBK conditions, U = 89.5, p = .701, 2 = .10. Change in reflective statements was also significantly aff ected by training group, H (2) = 5.63, p < .05, with mothers in the COACH cond ition demonstrating greater improvement 33


than the FDBK condition U = 54.5, p = .04. No significant difference was found between the COACH and PRAC groups, U = 85, p = .57. The difference betw een the PRAC and FDBK groups approached significance, U = 60, p = .07. Change in labeled pr aise was not significantly affected by training condition, H (2) = 2.04, p = .36. 0 5 10 15 20 25 Pre After didacticAfter practice AssessmentFrequency PRAC FDBK COACH Figure 2-1. Positive Following skills composite score Log transformation corrected the significant Le venes test of homogeneity of variance found for the raw data Negative Leading compos ite scores. A 2 (T2, T3) X 3 (PRAC, FDBK, and COACH) repeated measur es ANOVA was conducted to examine group differences in changes in Negative Leading behaviors. Th e main effect of time was significant, F (1,39) = 10.27, p = .003, 2 = .21, power = .88. There was no Time x Group interaction, F (2,39) = 2.23, p = .12, 2 = .10, power = .43. Overall, mothers signif icantly decreased their Negative Leading behaviors from the post-didactic assessment to the post-practice assessment but no differences between training conditions were found (see Figure 2). 34


0 5 10 15 20 25 30 35 40 45 Pre After didacticAfter practice AssessmentFrequency PRAC FDBK COACH Figure 2-2. Negative Leading behaviors composite score Family Characteristics and Skill Acquisition No socio-demographic factors were significantly related to mothers increase in their composite Positive Following skills from the preto post-didactic assessment. However, yearly family income and maternal education were si gnificantly related to mothers decrease in composite Negative Leading behaviors. Specifically, mothers with high family incomes (>60,000) demonstrated greater change than mothers with low family incomes (<20,000), t (16) = 2.14, p = .048, and mothers with at least some college education demonstrated greater change than mothers with high school education or less, t (40) = 2.60, p = .013. In regard to individual skills, mothers change in critic al statements from the preto post-didactic assessment was significantly related to maternal rated child disruptive behavior scores on the ECBI, r = .424, p = .005. Mothers of children with higher ECBI scor es demonstrated greater change in critical statements than mothers with lower ECBI scores. Higher ECBI scores were also related to higher frequency of critical statements at the pre-as sessment, which indicated that these mothers had more opportunity to decrease this behavior given that the mean frequency of critical statements 35


was very low for the group. Mothers change in la beled praises from the preto post-didactic assessment was not related to ECBI score or mate rnal depressive symptoms, however change in labeled praises during the practi ce period was significantly rela ted to maternal depressive symptoms as rated on the CES-D, r (41) = -.524, p < .001. Higher maternal CES-D scores were related to less improvement in labeled prai ses from the post-didact ic to post-practice assessments. Treatment Responders versus Nonresponders Analyses In addition to examination of skill frequencies, maternal skill acquisition can be evaluated as meeting or not meeting behavior al mastery criteria. Mastery criter ia in clinical treatment is 10 of each Positive Following skill and less than three Negative Leading behaviors in a 5-minute time period. Due to the brief nature of the analog training situation, we examined the number of mothers who achieved 5 of each Positive Following sk ill or fewer than 6 total Negative Leading behaviors during the 5-minute post-didactic posttraining assessments. These criteria reflect a substantial change in the quality of the interaction from pre-training. Following didactic instruction alone, 31% of mothers had five or more labeled praises, 31% of mothers had five or more reflective stat ements, and 19% of mothers had five or more behavioral descriptions. Five percent of mothers were treatme nt responders who reached these criteria on all three of the Positive Following sk ills in a 5-minute time period, reflecting their ability to use the range of Positive Following sk ills taught. Ten percent of mothers had six or fewer Negative Leading behaviors in the post-didactic assessment. After the practice period, 60% of mothers had five or more labeled praises, 55% of mothers had five or more reflective statements, and 33% of mothers had five or more behavior descriptions in a 5-minute time period at the po st-assessment. Nineteen percent of mothers were treatment responders who reached these criteria on all three of the Positive Following skills in a 36


5-minute time period. This treatment response clas sification was significantly related to training condition, 2 (2) = 7.72, p = .021. The COACH condition had significantly more responders than the PRAC and FDBK conditions, 2 (1) = 4.76, p = .029. Forty-three percen t of mothers in the COACH condition were Positive Following treatment responders compared to seven percent of mothers in each of the FDBK and PRAC conditi ons. Twenty-four percent of mothers had six or fewer Negative Leading behaviors in the 5-minute time period. Training group was not significantly related to treatme nt response classification for Negative Leading behaviors, 2 (2) = 1.05, p = .592. Another way to examine treatment response is to evaluate individual cases that declined over the practice period despite overall trends of improvement. We examined frequencies of skills immediately following didactic instructio n compared to post-training frequencies for declines during the practice period. Twelve percen t of mothers gave fewer labeled praises at post-training than immediately following didactic instruction (1 in PRAC, 4 in FDBK; M decrease = 5.2, SD = 3.70), 21 percent of mothers declined in frequency of behavior descriptions (3 in PRAC and 6 in FDBK; M decrease = 2.89, SD = 1.45), and 29 percent of mothers declined in frequency of reflective statements (2 in PRAC, 7 in FDBK, and 3 in COACH; M decrease = 2.58, SD = 1.16). For the Negative Leading behaviors, 21 percent of mothers increased number of descriptive/reflective questi ons asked at post-training comp ared to immediately following didactic instruction (4 in PRAC 2 in FDBK, and 3 in COACH; M increase = 3.78, SD = .88), 21 percent of mothers increased number of inform ation questions asked (3 in each training condition; M increase = 1.44, SD = .24), 33 percent of mothers increased number of direct commands given (6 in PRAC, 6 in FDBK, and 2 in COACH; M increase = 3, SD = .66), 38 percent of mothers increased number of indirect commands given (4 in PRAC, 9 in FDBK, and 3 37


in COACH; M increase = 3.06, SD = .42), and 10 percent of mother s increased number of critical statements made (1 mother in FDBK condition had an increase of 9, and 3 mothers in COACH condition each had an increase of 1). Parent Satisfaction For the total sample of 42 mothers, mean ratings for all items on the Satisfaction Questionnaire (SQ) were approximately 4 or abov e on a 5-point scale. Ta ble 4 shows the mean SQ item ratings by training group. An average tota l score for the SQ wa s calculated and there were no significant differences in sa tisfaction between training conditions, F (2, 39) = .27, p = .75. The didactic instruction was ra ted as very to extremely helpful in the PRAC (M = 4.14, SD = .86) and FDBK (M = 4.5, SD = .85) conditions, compared to ratings of somewhat to very helpful ( M = 3.79, SD = .89) for the COACH condition. However, no group differences in helpfulness of didactic instruction were found, F (2,39) = 2.36, p > .05. Performance feedback was rated as very to extremely help ful for mothers in the FDBK condition ( M = 4.5, SD = .65) and mothers in the COACH condition ( M = 4.71, SD = .47) with no group differences found, t (26) = -1.00, p = .34. All mothers in the COAC H condition rated coaching as extremely helpful ( M = 5, SD = 0). Homework Practice Mothers reported practicing an average of 7.77 days ( SD = 4.37) out of a possible 14 days at the two week follow-up. There were no significant differences in the number of days practiced across training conditions, F (2, 36) = .166, p > .05. Post-Hoc Assessment of Child Behavior Change Child disruptive behavior scores on the ECBI Intensity Scale at pre-training ( M = 103.61, SD = 18.22) and two-week follow-up (M = 96.32, SD = 20.64) assessments were examined for 38


changes in child behavior after training. A significant decrease in disruptive behavior scores was found, t (35) = 2.65, p < .05, 2 = .17. A negligible effect si ze was found however this was a prevention sample with many children well belo w the ECBI normative (115) and clinical cutoffs (132). Twenty-five percent of children that had ECBI scores above the normative mean at pre-training were rated below the normative mean at post-training. Qualitative Findings Mothers were given space to make writte n comments at the end of the satisfaction questionnaire. Themes that appeared in short wr itten comments are discussed here. Thirty-six percent of mothers commented on their increase d awareness of their behavior during play following the training. They described not knowing be fore how much they directed or influenced free play, their tendency to make all activities t eaching, and their lack of awareness of how many questions they ask (I didnt realize how many questions I asked un til I tried not to!). Another major theme of comments were descriptions of childrens positive response to the skills, with twenty percent of mother s describing observations such as increased speech, more positive behavior, happier mood, increased self-esteem or engagement during play. Twelve percent of mothers commented about specific skills, particularly about the be nefits of labeled praises. One mother also thought it was helpfu l to learn the difference between direct and indirect commands. In regards to the training methods, four mothers ( 29%) noted that they woul d have liked to have written materials with them to reference duri ng the CDI practice. Two of the mothers in the feedback condition (14%) reported that they li ked having feedback on their performance. Four mothers in the coaching condition (29%) noted the helpfulness of this training method (e.g. would have been lost without coaching and coaching was the most helpful). 39


Written comments of the nine mothers (21%) that rated their overall satisfaction as a little or somewhat were examined. Two mother s, both in the feedback plus practice groups, rated their overall satisfaction as a little. Both mothers written feedback included their desire to have more trainer guidance (it would have been nice to have someone in the room with us to help model the correct behavior and would rath er have a hands on contact demonstration) all mothers were informed of randomization to one of the three tr aining groups during the informed consent process. One of these mothers also indicated that she felt the play was too repetitive and that she enjoys asking her son learning questions. Seve n mothers 4 in PRAC, 2 in FDBK, and 1 in COACH rated their overall satisfaction as somewhat. These mothers written comments included found the new communication akward, found it hard because child always wants to do pretend play, and would have liked to learn additional strategies. The one mother in the coaching group that rated her overall satisfaction as somewhat commented that she felt the CDI skills d id not add anything new to our play. At the two-week follow-up, twenty-nine percen t of mothers spontaneously reported that they found themselves generalizing their use of the skills (e.g., The skills are helpful at other times like when we are walking somewhere and allow me to really listen to him, I find myself using the skills throughout the day, and I find myself using the advice and skills at other times) and two mothers noted that their child asked for special playtime. Three mothers reported that they had shared the information w ith family members. Two mothers described what they perceived to be substantial improvements in their relationship with their child and their childs cooperation during the follow-up. No moth ers asked for additional guidance with their use of the strategies at home during the follow-up phone calls. 40


Table 2-2. Mean Frequency of the Child Di rected Interaction (CDI) behaviors pretraining and after di dactic instruction Pre-Training After Di dactic Instruction M SD M SD t p Positive Following Skills Labeled Praise 0.16 0.48 3.28 3.14 6.49 <.001 Behavior Description 0.23 0.43 1.93 2.32 4.59 <.001 Reflective Statement 2.91 3.58 3.84 4.09 1.36 .182 PF Composite 3.43 3.68 9.33 6.29 5.84 <.001 Negative Leading Behaviors D/R Question 15.79 9.43 10.28 5.46 3.66 .001 Information Question 7.86 5.83 1.86 2.70 6.81 <.001 Direct Command 6.28 6.28 2.56 3.40 3.61 .001 Indirect Command 3.53 3.05 2.21 2.67 2.60 .013 Critical Statement 0.81 1.12 0.30 0.80 2.42 .020 NL Composite 34.52 15.89 17.45 10.66 6.40 <.001 Note. n = 42. PF Composite = positive following composite (labeled praise + behavior description + reflective statement). NL Composite = negative leading composite (D /R question + information question + direct command + indirect command + critical statement). D/R question = descriptive/reflective question. 41


Table 2-3. Mean frequency of the Child Directed Interaction (CDI) behaviors at each assessment for each training condition Pre-Training After Didactic Instruction After Practice PRAC FDBK COACH PRAC FD BK COACH PRAC FDBK COACH Positive Follo wing Labeled Praise M=0.36 SD=.745 M=0 SD=0 M=0.13 SD=0.35 M=2.35 SD=1.82 M=4.36 SD=4.16 M=3.36 SD=2.87 M=4.79 SD=3.47 M=4.57 SD=2.17 M=6.00 SD=4.26 Behavioral Description M=0.21 SD=0.43 M=0.14 SD=0.36 M=0.33 SD=0.49 M=1.71 SD=2.97 M=2.71 SD=1.98 M=1.43 SD=1.91 M=2.43 SD=2.17 M=3.86 SD=4.87 M=8.60 SD=6.13 Reflective Statement M=4.57 SD=3.20 M=2.57 SD=4.94 M=1.67 SD=1.35 M=2.79 SD=3.24 M=4.93 SD=4.87 M=3.92 SD=4.14 M=4.21 SD=4.46 M=4.50 SD=5.35 M=6.73 SD=4.35 Positive Following Composite M=5.14 SD=3.35 M=2.71 SD=4.95 M=2.29 SD=1.20 M=7.29 SD=4.01 M=12.00 SD=7.85 M=8.71 SD=5.88 M=11.43 SD=7.23 M=12.93 SD=8.95 M=21.71 SD=9.38 Negative Lea ding Descriptive/Reflective Question M=19.57 SD=12.61 M=14.14 SD=8.71 M=13.80 SD=5.25 M=10.07 SD=5.20 M=10.21 SD=5.71 M=10.79 SD=5.96 M=8.86 SD=3.68 M=6.50 SD=6.54 M=6.80 SD=4.65 Information Question M=10.86 SD=7.89 M=5.93 SD=3.20 M=6.87 SD=4.55 M=1.43 SD=1.95 M=2.07 SD=2.84 M=2.07 SD=3.36 M=1.21 SD=1.31 M=1.14 SD=1.29 M=0.93 SD=1.28 Direct Command M=4.57 SD=4.22 M=8.93 SD=8.07 M=5.40 SD=5.51 M=1.86 SD=2.31 M=3.14 SD=4.82 M=2.86 SD=2.71 M=2.07 SD=2.37 M=3.21 SD=3.49 M=1.47 SD=1.60 Indirect Command M=3.29 SD=3.24 M=3.00 SD=2.77 M=4.27 SD=3.17 M=2.36 SD=2.84 M=1.57 SD=2.14 M=2.71 SD=3.10 M=2.50 SD=2.59 M=2.79 SD=2.67 M=1.87 SD=1.96 Critical Statement M=0.64 SD=.75 M=1.36 SD=1.50 M=0.47 SD=0.83 M=.214 SD=.802 M=.357 SD=.929 M=.357 SD=.745 M=0.14 SD=0.53 M=0.71 SD=2.40 M=0.33 SD=0.62 Negative Leading Composite M=38.92 SD=19.67 M=33.57 SD=15.38 M=31.07 SD=11.78 M=16.21 SD=8.68 M=17.36 SD=12.24 M=18.79 SD=11.38 M=14.78 SD=7.50 M=14.36 SD=12.63 M=11.50 SD=7.37 42 Note. n = 14 for each training condition. Positive Following Composite = labeled praise + behavior description + reflective statement. Negative Leading Composite = descriptive/reflective question + informa tion question + direct command + indirect command + critical state ment.


CHAPTER 4 DISCUSSION The additive training value of the coach in structional method was supported for maternal acquisition of the Positive Following CDI sk ills, over and above initial gains following a videotaped CDI didactic presentation. Specifical ly, there was an interaction between training method and skill acquisition for mothers who were coached during their 20-minute practice after the didactic presentation. Mothers in the COACH condition demonstrated greater skill acquisition with higher frequency of total Positive Following skills and coached mothers were more likely to use five of each Positive Following skill (labeled praises, reflective statements, and behavioral descriptions) at the training post-assessment than mothers who engaged in unaided practice or received therap ist performance feedback before unaided practice. In regard to specific skills, coaching had the greatest impact on mothers change in be havioral descriptions following the practice period. Mothers in th e COACH and PRAC conditions had greater acquisition of reflective statements during th e practice period than mothers in the FDBK condition. There were no differences in the acquisition of labeled praises by training condition following this brief training. Theref ore, findings also suggest inte raction effects between specific skills and outcome, regard less of training condition. The additive value of performance feedback before unaided practi ce over unaided practice alone was not supported. Therapists provided enc ouragement and behavioral feedback with the performance feedback chart, aimed at increasi ng the mothers awareness of the discrepancy between their current behavior and the behavioral goals of the child-directed interaction (i.e., pre-set mastery criteria). Therapist feedback on their performance-goal di screpancy prior to the mothers practice did not impact subsequent sk ill acquisition over the practice period. Of note, only one-third of the mothers asked any question about their performance when they were shown 43


a visual representation of their current skill strengths and deficits during the therapist performance feedback. Our hypothesis that mothers in the COACH condition would decrea se Negative Leading behaviors more than mothers in the PRAC and FDBK conditions was not supported in this study. At the end of the 20-minute practice period, moth ers in all training condi tions had significantly decreased their use of these le ading behaviors. However, only 24% of mothers were able to decrease these leading behaviors in the CDI to fewer than six in the 5-minute assessment period, which was the definition for a treatment responder in this brief analog study. In regard to specific Negative Leading behaviors, mothers in each training condition rarely asked information questions at the training post-assessment and critic al statements were infrequent at pre-training and remained low. The use of direct and indi rect commands in each training condition at the training post-assessment was also quite low. Descriptive/reflective question was the highest frequency Negative Leading behavior demonstrat ed by mothers post-training. The majority of CDI coaching is focused on reinforcing parental approximations of the Positive Following skills and Negative Leading behaviors in the CDI are in itially ignored with the expectation that the parent behaviors gaining most therapist attention will increase and replace parent behaviors receiving less attention. Therap ists gradually introduce gentle corrections for Negative Leading behaviors that remain in later CDI coaching sessions, to raise pa rental awareness of habitual leading behaviors used in interactions with thei r child. The additive value of coaching parents to decrease their leading behaviors may not have been supported in this study due to the brief nature of the training. Further research is required to determin e if parents are able to meet mastery criteria for avoiding the negative leadin g behaviors with continued self instructional methods and unaided practice, or if therapist coaching is necessary. 44


This analog study standardized the amount of pa rent practice in order to compare directly the effects of guided versus unguided practice fo llowing gains made from didactic presentation. One observation of note was our finding that some mothers skills declined from the assessment immediately after didactic traini ng to the final assessment after practice. Of the 26% of mothers that decreased their use of one of the Positive Following skills, 90% of these mothers were in the PRAC and FDBK conditions. A major advantage of coaching is that therapists are able to explain the effects of the CDI skills on the chil ds behavior as it happens, which may increase the mothers buy-in or motivation to work on using the range of skills they were taught. The intensive coach training method may be more eff ective because of the therapists ability to encourage and support the parents perseverance in concentrated skills practice throughout the 20-minute practice period and final assessment in th e context of the range of child behaviors that may present. Coaching assists parents with inco rporating and sustaining new parenting behaviors in on-going interactions with their child. Maternal satisfaction with the parenting skills they learned or the instructional components was not related to training group. Ge nerally parents reported that they felt the skills were helpful in their interaction with their child and that they were satisfied with the instructional components, which suggests a ceiling effect. Th is replicates previous research finding no relations between parents overa ll impression of training, how well they felt they understood the content, or how likely they were to use the techniques and four different instructional methods for positive reinforcement skills (ODell et al., 1982), as well as no relations found between parental attitudes and training methods for time out skills (Flanagan et al., 1979; ODell et al., 1979). The brief nature of the trainings in the sk ill acquisition literature and the types of skills taught may have influenced mothers generally high satisfaction ratings. 45


The amount of home practice of the skills repo rted during the two-w eeks post-training was also not related to training condition in this study. Variable s unrelated to a treatments instructional components, such as the type of sk ills taught or parent a nd child characteristics, may relate to parents continued practice. Daily pr actice is an important c linical goal and further research is needed to determine the most successful ways to promote parent practice. A number of group and abbreviated adaptati ons of PCIT have retained the core instructional components of PCIT but have altere d the dose of coaching, and some abbreviated PCIT programs have used video di dactic instruction and therapist telephone contacts to decrease therapist contact time (e.g., Ni ec et al., 2005; Nixon et al., 2003). This study provides preliminary evidence for the additive effects of the coach training method for faster acquisition of the Positive Following skills over practice alon e, with or without brie f therapist performance feedback. Further treatment comp onent research must continue to examine systematically potential adaptations that may improve the efficiency or effectiveness of PCIT (Eyberg, 2005). For example, in the current study mothers demonstr ated significant improve ments in most of the targeted parenting skills follo wing the 18-minute CDI didactic vi deo alone. When delivered by the therapist, the length of the CDI Teach sessio n is typically 60 to 90 minutes. For some parents, watching a didactic video prior to their first session with the therapist may improve treatment efficiency and allow the therapist to focus trea tment time based on parent improvement from the self-guided didactic inst ruction. Indeed, the favorable outcome s for the abbreviated PCIT format that used didactic videos to supplement faceto-face treatment time (Nixon et al., 2003) and the earlier finding that a brief 7-minute therapist check-out following a didactic video was more efficacious than a 30-minute therapist didactic and role-play session for acquisition of time-out skills (ODell et al,. 1979) woul d support further study of the role of didactic videotapes and 46


therapist coaching to promote parenting skill acq uisition and improve treatment efficiency. Adaptations that deviate from the use of the coac h training method to assist parents with meeting PCIT behavioral mastery criteria should exam ine questions such as the minimal amount of coaching necessary to support skill acquisition and maintenan ce for individual families. It must be noted that parents exhibited great variability in skill acquisition regardless of training group. Morowska and colleagues (2005) ha ve described efficient modes of treatment delivery in which parents are assigned to self-dir ected or intensive treatments based on a number of individual difference characteristics, such as level of child behavior problems or parental mental health. Others have propos ed that clinicians begin with the least expensive method of training, with supplemental instruction provide d only for those skills that fail to meet performance criteria establishe d by the therapist and parent (L erman et al., 2000). These methods allow for the clinicians time and efforts to be focused more precisely where they are most needed. The current study suggests that it may be beneficial, at least for preventive intervention, for families to begin with didactic instruction pl us a brief, intensive coach training procedure to promote fast and accurate skill acquisition. Future research could help determine which families would require repeated coaching and which famili es might continue to improve on their own or with minimal therapist guidance. The current findings are restricted by the use of only one post-assessment observation immediately following training. Multiple observations, including in the home environment, are recommended for future research to better unde rstand mothers generalization and maintenance of skill gains based on instructiona l methods. Findings are also restricted to the specific way that the training conditions were defined and implemented in this study. Changes in the 47


implementation of the didactic instruction, pe rformance feedback, or coaching methods may impact the relationship between these traini ng methods and parent skill acquisition. Although our sample size was small, it was substantially larger than most previous research in this area and allowed for further examination of socio-demographic variables in relation to speed of skill acquisition. Higher family income and maternal education were positively related to decreases in Negative Lead ing behaviors following didactic training alone. Although we excluded mothers with elevated matern al depressive symptoms, change in labeled praises over the practice period was negatively related to mate rnal depressive symptoms reported. Further study of the impact of parent and child characteristics on parenting skill acquisition is warranted to learn more about the efficacy of instructiona l formats for acquisition of specific parenting skills with different popula tions presenting for beha vioral parent training. The analog study methods allowed for the examin ation of the interaction between the well defined training methods used in PCIT and the sp ecific parenting skills of the CDI. Preliminary evidence suggests the additive value of the co ach training method over therapist performance feedback prior to unaided practice or unaided pr actice alone for greater acquisition of the range of CDI Positive Following skills (defined as usin g five or more of each positive following skill), and for greater increases of beha vioral descriptions in particular. These gains were over and above gains made with didactic training alone. Further research of the multiple variables (and the interaction of these variables) important in sk ill acquisition parent a nd child characteristics (including differences for mothers and fathers) specific parenting skills, and the range of training methods will help to id entify those factors that will improve treatment efficiency and effectiveness for individual families. 48


APPENDIX A SUMMARY OF CHILD DIRECTED INTERACTION (CDI) SKILLS Positive Following Skills Labeled praise : A labeled praise to the child provid es a positive evaluation of a specific behavior, activity, or product of the child. Reflective statement: A reflective statement is a declar ative phrase or statement that has the same meaning as a preceding child verb alization. The reflection may paraphrase or elaborate on the childs verb alization but may not change the meaning of the childs statement or interpre t unstated ideas. Behavioral description : Behavioral descri ptions are non-evalua tive, declarative sentences or phrases in which the subject of th e sentence is the child and a verb describes the childs ongoing or immediately completed observable verbal or nonverbal behavior. Negative Leading Behaviors Question : Questions are verbal inquiries that are distinguishable from declarative statements by having a rising inflection at the end and/or by having th e sentence structure of a question. Questions request an answer but do not suggest that a behavior is to be performed by the child. There are two types of questions: Descriptive/Reflective Questions are usually closed-ended questions that call for no more than brief acknowledgement in response. Information Questions are open-ended questions that ask for specific information. Command : Commands are statements in which th e parent directs the vocal or motor behavior of the child. There are two types of commands: Direct Commands are statements that indicate a specific action to be performed by the child. Indirect Commands are often in question form and sugge st that compliance is optional. Criticism : Critical statements are verbal expres sion of disapproval of the child or the child's attributes, activities, products, or choice s. This also includes sassy, sarcastic, rude, or impudent speech. 49


APPENDIX B CHILD DIRECTED INTERACTION (CDI) DIDACTIC PRESENTATION Give overview of CDI strategies Teaches you the kinds of skills that play therapists use with children to build a good relationship with them and help them feel safe and calm. Improves your child's self-esteem. Improves your childs social skills, like sharing, which children need to get along with other children and have friends. Results in a secure, warm relati onship between you and your child Explain that the rules of CDI apply only to th e short play therapy sess ions (special time) that you will ask parents to have at home each day with their child. Indicate that while some of the ru les you will be describing are good general parenting skills, CDI is a special therapeu tic playtime and that the rules are not intended for use throughout the day when pare nts have to carry on regular routines or direct the child's activity. Next explain behaviors that should be avoided during this special time: Avoid Commands Commands are statements that try to dire ct the play by suggesting what the child should do. There are two types of commands: o Direct: Sit down. Please hand me the car. o Indirect: Would you like to sit down? Lets put the cars away. Commands take over the lead of the play. If the child doesnt obey, th e play can stop being fun. Avoid Questions A question asks the child to give an answer. There are two types of questions: o Questions that ask for information who, what, where, when, how. o Insincere questions voice tone goes up at end of sentence. Questions are often hidden comman ds Would you like to clean up? Questions take over the l ead of the conversation. Questions sometimes suggest disapproval. Questions often suggest you arent really listening to your child. 50


Avoid Criticisms Criticisms are negative or cont radictory statements about yo ur child or his actions o Youre not nice, or That doesnt go that way. Criticism points out mistakes rather than correction: o Thats wrong is a criticism. o It goes like this allows you to correct without criticizing. Criticisms lower your childs self-esteem. Criticism creates an unpleasant interaction. Explain what parents are to do during special play: Praise your child's appropriate behavior. Giving your child a compliment about his behavior. There are two types of praise: o Labeled praise is specific about what you like. For example, You choose such pretty colors! You're being so careful with your car! I like it when you draw neatly! o Unlabeled praise is nonspecific praise. For example, Good! That's great! Nice job. Labeled praise is better because it lets your child know exactly what you like. Increases the behavior that is praised. Increases child's self-esteem. Adds more warmth to relationship. Makes both parent and child feel good! Reflect appropriate talk. Repeating/paraphrasing what your child sa ys. Yes, thats a blue crayon. Allows child to lead the conversation. Shows child you're really listening. Shows you accept/understand what he is saying. Improves and increases child's speech. May feel awkward for you, but children love it! Imitate appropriate play Doing the same thing the child is doing, su ch as drawing a tree if your child is drawing a tree. o Remember to keep your attention/comme nts focused on what your child is doing. Lets child lead. Approves of child's activity. Shows child you're involved. Teaches child how to play with others (for example, taking turns). Describe appropriate behavior State exactly what your child is doing: You're drawing a sun. 51


Like a sports announcer, a running commentary. Lets your child lead. Lets your child know you're interested and paying attention to him. Lets your child know you approve of what she is doing. Models speech and teaches vocabulary and concepts. Holds your childs attention to the task, and teaches your child how to hold her own attention to a task Be Enthusiastic! Let your voice show excitement about your childs appropriate behavior. For example, You are being SO nice to share with me! Lets your child know that you enjoy the time you are spending together. Increases the warmth of your play. Review PRIDE acronym Describe Ignoring inappropriate behavior when child misbehaves. Children usually like CDI a lot and show good behavior, but we will practice using ignoring if your child misbehaves. Serves to decrease ONLY attention-seeking behavi ors (yelling, sassing, whining, crying for no good reason); not for hitting, stealing, etc. Any attention, positive or negative, can increase attention-seeking behaviors. Avoid any verbal or nonverbal reaction to inappropriate behavior (e.g., looking at the child, smiling, frowning). Once you begin ignoring, you must continue until the behavior stops (explain consequences of stopping i gnoring too soon and increasing the negative behavior). Continue ignoring until your child is doing something appropriate. Praise your child immediatel y for appropriate behavior. Helps your child notice the difference be tween your responses to good and bad behavior. Ignored behavior gets worse before it gets better, so only ignore a behavior if you can continue to ignore it when it gets worse. Consistent ignoring eventually decreases many behaviors. Describe how to combine ignoring with the PRIDE skills. While ignoring the negative behaviors, l ook for any positive behavior occurring at the same time, and comment on it. If a negative behavior stops look at your child with a friendly look, and comment on what your child is doing that is the opposite of negative behavior. Any time you see behavior that is opposite to the negative behaviors you have to ignore, give your child BIG labeled praises for the positive opposite. Explain that if the behavior cant be ignored, the parent must stop the play These behaviors include: o Aggressive behaviors (e .g., hitting and biting). 52


o Destructive behaviors (e .g., drawing on the wall). Teaches your child that good behavior is required during special time. Shows your child that you ar e learning to set limits. Tell your child, Special time is stopping because you hit me. Maybe next time you will be able to play nicely during special time. Try to initiate CDI again late r in the day, if possible. Model CDI for parents Model what CDI looks like using all the PRIDE skills and avoiding negative leading behaviors [3 minutes] Describe the kinds of toys that are good a nd not good to use for CDI at home and why Toys that dont have rules ar e best, so that parents can let the child lead the play without worrying about the child breaking rules. Construction toys, such as Legos, blocks, tinker toys. Play sets, such as farms, houses, and towns. Creative toys, such as crayons and paper are good. Many objects around the house can make excellent creative toys, like pots and pans. Avoid board games. Structured rules prevent free play. Avoid pretend-talk toys su ch as puppets, toy telephones. You want to communicate directly with your child. Avoid toys that encourage rough play (balls), aggressive play (Super-hero figures), or messy play (finger paints). These increase the chances of behavior problems, and you want this special time to be very positive. Explain how to set up the CDI play session at home. Ask parents to think about the best place a nd time each day to have this special time with their child. Minimize distractions (siblings, telephone, TV, etc.). Place two or three appropriate toys in CDI area before starting the session. Let the child choose from your selection once CDI begins. Ask each parent when is the best time for them. Explain the importance of practi cing CDI every day for 5 minutes. Long enough to provide therapeu tic effect for the child. Long enough for parents to be able to learn the skills. Short enough not to be too time consuming. Short enough that parents will not become fr ustrated with the c oncentration required at first to learn the skills. 53


APPENDIX C SKILLS FREQUENCY CHART FOR PERFORMANCE FEEDBACK After Didactic Instruction After Practice Labeled Praises Reflections Behavior Descriptions Descriptive/ Reflective Questions Information Questions Direct Commands Indirect Commands Criticisms 54


LIST OF REFERENCES Borrego, Jr., J., & Urquiza, A. J. (1998). Importance of therapist use of social reinforcement with parents as a model fo r parent-child relationships: An example with Parent-Child Interaction Therapy. Child and Behavior Therapy, 20, 27-54. Brestan, E. V., & Eyberg, S. M. (1998) Effective psychosocial treatments of conduct-disordered children and a dolescents: 29 years, 82 studies, and 5,272 kids. Journal of Clinical Child Psychology, 27, 180-189. Chaffin, M., Silovsky, J. F., Funderburk, B., Vall e, L. A., Brestan, E. V., Balachova, T., et al. (2004). Parent-child interaction therapy with phys ically abusive parents: Efficacy for reducing future abuse reports. Journal of Consulting and Clinical Psychology, 72, 500-510. Eisenstadt, T. H., Eyberg, S. M., McNeil, C. B., Newcomb, K., & Funderburk, B. (1993). Parent child interaction therapy w ith behavior problem children: Relative effectiveness of two stages and overall treatment outcomes. Journal of Clinical Child Psychology, 22, 42-51. Eyberg, S. M. (1992). Assessing therapy outcome with pres chool children: progress and problems. Journal of Clinical Child Psychology, 221 306-311. Eyberg, S. M. (1993). Consumer satisfaction measures for assessing parent training programs. In L. VandeCreek, S. Knapp, & T.L. Jackson (Eds .), Innovations in Clinical Practice: A Sourcebook (Vol. 12). Sarasota. Professional Resource Press. Eyberg, S. M. (1999). Parent-child interaction therapy manual. Retrieved from the World Wide Web: Eyberg, S.M., Bessmer, J., Newcomb, K., Edwards, D., & Robinson, E. (1994). Manual for the Dyadic Parent-Chi ld Interaction Coding System-II Social Documents (Ms. No. 2897). Eyberg, S. M. & Boggs, S. R. (1997). Parent-child interaction therapy for oppositional preschoolers. In Handbook of parent training: parents as cotherapists for childrens beha vior problems. New York: Wiley. Eyberg, S. M., Duke, M., McDiarmid, M ., Boggs, S.R., & Robinson, E.A. (2004). Dyadic parent-child intera ction coding system manual (3rd rev.). Retrieved from the World Wide Web: Eyberg, S. M (2005). Tailoring and adapting Pa rent-Child Interac tion Therapy to new populations. Education and Treatment of Children, 28 197-201. 55


Fernandez, M., Chase, R., & Eyberg, S.M. (2005). The workbook: Coder training manual for the dyadic parent-child interaction coding system (3rd ed.). Retrieved from the World Wide Web: Flanagan, S., Adams, H. E., & Forehand, R. (1979). A comparison of four instructional techniques for te aching parents to use time-out. Behavior Therapy 10, 94-102. Foote, R. C., Schuhmann, E. M., Jones, M. L., & Eyberg, S. M. (1998). ParentChild Interaction Therapy: A Guide for Clinicians. Clinical Child Psychology and Psychiatry, 3, 1395-1045. Forehand, R. L. & McMahon, R.J. (1981). Helping the noncompliant child. New York: Guilford. Forgatch, M. S. & DeGarmo, D. S. (1999). Parenting through cha nge: An effective prevention program for single mothers. Journal of Consulting and Clinical Psychology, 67, 711-724. Funderburk, B. W., Eyberg, S. M., Ric h, B. A., & Behar, L. (2003). Further psychometric evaluation of the Eyberg a nd Behar rating scales for parents and teachers of preschoolers. Early Education & Development, 14, 67-81. Hann, D., Winter, K., & Jacobsen, P. (1999). Measurement of depressive symptoms in cancer patients: evaluation of the Center for Epidemiological Studies Depression Scale (CES-D). Journal of Psychosomatic Res earch 46, 437-43. Harwood, M. & Eyberg, S. M. (in prep). Ea rly identification a nd intervention for disruptive behavior in primary care: A randomized controlled trial. Herschell, A. D., Calzada, E. J., Eyberg, S. M. & McNeil, C. B. (2002). Clinical Issues in Parent-Child Interaction Therapy, Cognitive and Behavioral Practice, 9, 16-27. Knapp, P. A. & Deluty, R. H. (1989). Relativ e effectiveness of two behavioral parent training programs. Journal of Child Clinical Psychology, 18, 314-322. Lerman, Swiezy, Perkinds-Parks, & Roan e, H.S. (2000). Skill acquisition in parents of children with developmental di sabilities: interacti on between skill type and instructional format. Research in Developmental Disabilities, 21 183-196. Morawska, A., & Sanders, M. R. (2006). Self-Administered Behavioral Family Intervention for Parents of Toddlers: Part I. Efficacy. Journal of Consulting and Clinical Psychology, 74, 10-19. 56


Moreland, J. R., Schwebel, A. I., Beck, S., & Well, R. (1982). Pare nts as therapists: A review of the behavior therapy pare nt training litera ture 1975 to 1981. Behavior Modification, 6, 250-276. Niec, L., Hemme, J., Yopp, J., & Brestan, E. ( 2005). Parent-Child Interaction Therapy: The rewards and challenges of a group format. Cognitive & Behavioral Practice, 12, 113-125. Nixon, R.D.V., Sweeny, L., Erickson, D.B ., & Touyz, S.W. (2003). Parent-Child Interaction Therapy: A comparison of st andard and abbreviated treatments for oppositional defiant preschoolers. Journal of Consulting an d Clinical Psychology, 2, 251-260. ODell, S. L., Mahoney, N. D., Horton, W. G., & Turner, P. E. (1979). Media-assisted parent training: Alternative models. Behavior Therapy, 10, 103 -110. ODell, S. L., OQuin, J. A., Alford, B. A., OBriant, A. L., Bradlyn, A. S., & Giebenhain, J. E. (1982). Predicting the acquisition of parenting skills via four training methods. Behavior Therapy, 13, 194-208. Patterson, G.R. & Guillion, M.E. (1968). Living With Children: New methods for parents and children. Research Press: Champaign, IL. Radloff, L. S. (1977). The CES-D scale: a self-report depression scale for research in the general population. Applied Psychological Measurement 1, 385401. Reid, J. B., Eddy, J. M., Fetrow, R. A. & Stoolmiller, M. (1999). Description and immediate impacts of a preventive intervention for conduct problems. American Journal of Community Psychology, 27 483-517. Webster-Stratton, C. H. (1990). Enhancing th e effectiveness of self-administered videotape parent training for families with conduct-problem children. Journal of Abnormal Child Psychology, 18, 479-492. Webster-Stratton, C. H. & Reid, M. J. (2003) The Incredible Years Parents, Teachers and Child Training Series: A multifaceted treatment approach for young children with conduct problems. In A. E. Kazdin & J. R. Weisz (Eds.), Evidence-based psychotherapies for children and adolescents (pp. 224-240). New York: Guilford. 57


BIOGRAPHICAL SKETCH Kelly OBrien was raised in Maryland. Afte r graduating from Catoctin High School she attended St. Marys College of Maryland in St. Marys City, MD. She graduated in May 2002 with a psychology major. Her interests in early intervention for child behavior problems and treatment outcome research led he r to seek a position in the Child Study Lab at the University of Florida under the mentorship of Dr. Sheila M. Eyberg. 58