260 likes | 945 Views
What Is PCIT?. Parent Child Interaction Therapy EVT Two phases (CDI and PDI)Treatment goals: Improve the quality of the parent-child interactionImprove parenting skillsDecrease negative child behaviors . PCIT Continued. Parent and child play with toys that encourage interactionAssessment based protocolECBIDPICSTAI.
E N D
1. The Dyadic Parent-Child Interaction Coding System II (DPICS II): Reliability and Validity with School Aged Children
Shelli Deskins, M.S.
Auburn University
West Virginia University Health Sciences Center
2. What Is PCIT? Parent Child Interaction Therapy
EVT
Two phases (CDI and PDI)
Treatment goals:
Improve the quality of the parent-child interaction
Improve parenting skills
Decrease negative child behaviors Empirically Based Treatment (EBT)Empirically Based Treatment (EBT)
3. PCIT Continued Parent and child play with toys that encourage interaction
Assessment based protocol
ECBI
DPICS
TAI
4. PCIT Continued Targeted problems and populations
Externalizing behavior problems
Abused and neglected children
FAS children
5. DPICS Dyadic Parent Child Interaction Coding System
Coding system for direct behavioral observations of parent-child interactions
Assessment tool for measuring pre/post and ongoing treatment changes
Allows for detailed analysis of behaviors
6. DPICS Categories Acknowledgement
Answer
Behavioral Description
Compliance
Contingent Labeled Praise
Criticism
Descriptive/Reflective Question
Destructive
Direct Command
Indirect Command
Information Description
Information Question Labeled Praise
Laugh
No Answer
No Opportunity for Answer
No Opportunity for Compliance
Noncompliance
Physical Negative
Physical Positive
Play Talk
Reflective Statements
Smart Talk
Unlabeled Praise
Yell
Whine
Warning
7. DPICS Manual
Workbook
Criterion tape
Transcripts
8. Physical Abuse-Children PA children demonstrate a host of problems:
Physical and verbal aggression
Depression
Peer rejection
Decreased school achievement
Compulsive compliance
9. Physical Abuse-Parents PA parents tend to:
Interact less physically and verbally with their children
Be more negative
Be less positive
Give more commands
Have a lower tolerance for child behavior
10. DPICS & DPICS II with Abused and Neglected Children Aragona and Eyberg (1981) found DPICS was useful with neglected children aged 3-7 yrs.
Timmer et al. (2002)
Chaffin et al. (2004) used DPICS II as an outcome measure for physically abusive families for children 4-12 years.
11. Current Study Provides normative data for 8 to 12-year-old children and their parents
Norms only available for 3 to 7-year-old children and their parents
Examines DPICS codes for physically abusive and nonabusive dyads Except for Chaffin et al. 2004, no large scale study compares PA and non-PA dyads using DPICS…
Chaffin does not compare PA to non-PA families.
Except for Chaffin et al. 2004, no large scale study compares PA and non-PA dyads using DPICS…
Chaffin does not compare PA to non-PA families.
12. Method Participants
n=38
19 PA dyads: 1 documented incident of PA and court ordered for treatment
19 Non PA dyads: Community samples from Auburn University and Northern Illinois
Children in matched group had no report of abuse by parent or child
Children 8-12 years
Groups were matched on demographics
“Parent gender” refers to the parent participating in the interaction (i.e. the abusive parent)
Procedure
Measures-KBIT, CAP, & CTS
Pretreatment observation-CDI, PDI, CU with the abusive parent
If children finished the CU early, told to remain in room.
All parents and children complied with the instructions.
Interactions coded using both written transcripts and videotapes
33% segments (one segment per family) randomly selected for reliability
13. Sample Characteristics Variable Abused (n=19) Nonabused (n=19) t
Gender of Child 7 female 7 female
12 male 12 male
Mean Child Age 10.21 9.94
Standard Deviation (1.79) (1.33) -.53
Gender of Parent 12 female 12 female
7 male 7 male
Mean Parent Age 35.24 38.60 1.09
Standard Deviation (10.32) (8.70)
14. Variable Abused (n=19) Nonabused (n=19)
Ethnicity
African American 6 5
Caucasian 13 14
Marital Status
Single 3 3
Married 8 10
Separated 2 1
Divorced 4 5
Other 2 0
Types of Education# Yrs. of Education #
<9 1 12 2 <12 3 13 2 High School Diploma 4 14 4 Vocational/Tec School 6 15 4
Some College 3 16 5
College 1 18 2
Unknown 1
15. Sample Characteristics Cont’d Variable Abused (n=19) Nonabused (n=19) t
KBIT IQ Composite (Mean)
Child Score 96.00 106.47 2.63*
SD 8.37 15.18
Parent Score 95.11 108.18 3.21*
SD 16.36 6.45
16. Analyses Reliability
Kappa
Insufficient data= <5 occurrences.
Validity
Discriminate (Criterion ?) Validity
Differences between groups in respective situations and across situations (using t-test)
Differences with summary variables across situations (using t tests)
Classification with significant codes (logistic regression)
Convergent Validity
Correlations with the CAP & CTS scores (using Pearson)
17. Results Reliability
Parent-child interactions can be reliably coded using the DPICS-II for this age group
Kappa reliability estimates across codes and summary variables were above .50
Hypothesis Supported
Validity
PA children will demonstrate greater compliance and fewer inappropriate behaviors than non-PA children
Non-PA children were significantly more compliant during PDI
In addition, when looking across situations, Non-PA children demonstrated significantly higher rates of compliance than PA children.
No significant differences between PA and Non-PA children on Inappropriate Behavior summary variable.
Hypothesis Not Supported
18. Results-cont’d PA parents will display more commands, inappropriate behavior, and negative talk than non-PA parents.
PA parents displayed a significantly higher Direct Command ratio than Non-PA parents.
Hypothesis supported
No differences in the groups on Negative Talk or Total Commands issued.
Hypothesis not supported
PA children will demonstrate fewer Prosocial behaviors than Non-PA children
Non-PA children demonstrated significantly more Prosocial behaviors than PA children
Hypothesis supported
19. Results-cont’d
PA Parents will show fewer Prosocial behaviors than Non-PA parents
No differences in the groups on parental Prosocial behavior
Hypothesis not supported
DPICS II codes and summary variables will correctly classify dyads into their groups using logistic regression
Parents-Laugh and Unlabeled Praise correctly classified observations into groups 82% of the time
Child -Acknowledgement, Laugh, and Compliance correctly classified observations 71% of the time.
Parent Direct Command Ratio and Child Prosocial correctly classified dyads into their groups 72% and 68% of the time, respectively.
Hypothesis supported
20. Results-cont’d 6. DPICS II categories and summary variables for parents will demonstrate convergent validity by correlating with CAP total Abuse score
No differences in groups on the CAP total Abuse score
Groups differed on the Rigidity scale, but did not correlate with summary variables
Hypothesis not supported
7. DPICS II categories and summary variables for parents will demonstrate convergent validity by correlating with CTS scores
On the CTS, groups only significantly differed on the Physical Maltreatment Frequency and Occurrence Scales
Correlations with summary variables and these two scales not significant
Hypothesis not supported
21. Discussion Solid evidence for reliability and criterion validity of this coding system for use with school aged PA and Non-PA dyads.
16 child categories and 17 parent categories had “excellent” or “good” reliability; another 5 child and 4 parent categories had “fair” reliability.
Many of the categories that have demonstrated difficulties for coders and poor reliabilities in past, continued to do so in this study.
Developmental considerations
Demand characteristics
Recommendations
22. Discussion-cont’d In general, results point to the tone of the interaction/tone that exists in Non-PA families as the factor that is important.
Ex. Non-PA children and parents show more positive behaviors, ex. AK and UP during CDI; LA and CO during PDI; LA during CU
Appears that interactions b/w abusive dyads not overly negative, but that Non-PA interactions are just more positive
PA children do not show more negative behaviors, but show fewer prosocial behaviors.
PA parents issue more DC and children are less compliant—may help explain precursors to an abusive incident
Need to take development into consideration
23. Discussion-Cont’d Parent DC ratio and child Prosocial variables are most useful in discriminating PA and Non-PA dyads
Poor convergent validity demonstrated in this study
Possible explanations
24. Limitations Sample size
Manner in which PA was ruled out in the community sample
Type I error
Few differences on the measures selected
Transcripts
Intraclass correlations???
These data are nominal; however, DPICS studies have traditionally used ICC as a way to asses reliability.
25. Future Directions Study investigating the use of transcripts
Use of other behavioral measures (ECBI, BASC, PSI)
Limiting and combining categories
Sequential analysis study
Treatment outcome study using the 8-12 year old age range