Tuning in to Teens (TINT) Replication: A Study of a Universal Intervention for Families Adapted for an at-Risk Population of Adoptive and Guardianship Families

Background/Purpose: Tuning in to Teens (TINT) is a six-week parent education program designed to proactively prepare parents to support their teens in managing the complex developmental task of adolescence by developing the youth’s emotional intelligence through a technique known as emotion coaching. Previous experimental research with a generalist population, parenting children between the ages of 10-13, demonstrated a reduction in emotion dismissing behavior by parents and related reductions in youth internalizing and externalizing behavior. The current study sought to replicate these findings with a post-adoption/guardianship population identified as at-risk for a return to state care. Adaptations were made to the intervention to incorporate material specific to an adoption/guardianship population, resulting in one additional session added to the six-session format. Groups were facilitated by teams of two: (1) an experienced clinician and (2) agency adoption/guardianship staff, with regular coaching of the facilitators by the purveyor of the program.

Methods: A post randomized consent trial was used, with 1,232 families in the target population (784 intervention, 448 comparison). The program successfully made contact with 452 (58%) of intervention cases, and 94 (12%) participated. Adoptive/ guardianship caregivers were included in the study if permanence occurred for their child after age 5 or the child had been in congregate care, and the child was currently between the ages of 10-13. 22 groups (154 sessions) were run, organized in 7 cohorts. A short questionnaire (10 questions) was distributed to all intervention and control cases after the fourth cohort, after it was determined that just 12% of eligible families were participating in the intervention, to assess differences that might emerge between groups in the study. The survey assessed such things as warmth in the relationship, struggling to manage the child’s behavior, confidence in meeting the child’s needs, and thoughts about ending the adoption/guardianship. Forty-three percent (n=244) responded to the questionnaire.

Results: Bivariate analysis revealed randomization created balanced intervention and control groups. However, important differences were found between those who self-selected out of the intervention sample (i.e., eligible intervention cases who did not participate) and those participated. Intervention participants were more likely to struggle to manage their child’s behavior (t=-2.52, df=98, p≤.05) and were less confident that they could meet the child’s needs (t=2.37, df=100, p≤.05) than parents/guardians in the control group. Differences were also found between intervention participants and intervention group non-participants. Intervention participants noted more struggles managing their child’s behavior (t=-2.31, df=165, p≤.05), more struggles to appropriately respond (t=-2.02, df=162, p≤.05), and were less confident that they could meet the child’s needs (t=2.76, df=168, p<.01). Based on these results, a propensity score model (PSM) was used to approximate a counterfactual within the comparison group, which was then used to create a data set that was balanced on covariates.

Conclusions/Implications: Randomization provided balanced samples, however, parents/guardians that were motivated to participate in the intervention differed noticeably from the control group overall and differed from the non-participants within the intervention sample, suggesting that randomization might not provide an appropriate counterfactual for intervention participants.