Ronald J Prinz1, Carol W Metzler2, Matthew R Sanders3, Julie C Rusby2, Chao Cai1. 1. Department of Psychology, University of South Carolina, Columbia, SC, USA. 2. Oregon Research Institute, Eugene, OR, USA. 3. School of Psychology, The University of Queensland, St Lucia, Qld, Australia.
Abstract
BACKGROUND: This study evaluated whether an evidence-based parenting intervention, when delivered online, could effectively address disruptive behavior problems in young children and yield outcomes comparable to in-person delivery of the same intervention. METHODS: Families (n = 334) of children (3-7 years; 63% White, 22% African American, 15% other races; 63% male) with disruptive behavior problems were randomized to online-delivered intervention (ODI) or staff-delivered intervention (SDI), resulting in baseline and demographic equivalence. Primary outcome measures for child disruptive behavior (independent observation, parent report) and secondary outcome measures of parenting and family impact were assessed at baseline, postintervention, and follow-up. Conducted using intent-to-treat (ITT) as well as per-protocol (PP) methods, noninferiority analyses, which drew on an HLM framework with repeat measures across three timepoints and on REML to provide unbiased estimates of model parameters, tested whether the outcome-difference CI did not exceed the a priori noninferiority margin. RESULTS: For ITT and PP analyses, the ODI was found to be noninferior to the SDI on the primary outcome: independently observed child disruptive behavior and parent-reported child behavior problems. The pattern for secondary outcomes was more varied: (a) noninferiority for observed positive and aversive parenting; (b) noninferiority for observed quality of parent-child relationship at post but not follow-up assessment; (c) noninferiority for parent-reported inappropriate/inconsistent discipline for PP but not ITT analyses; and (d) noninferiority not confirmed for parenting daily hassles and adverse family quality of life, despite large effect sizes for the ODI (Cohen's d .75-1.07). Finally, ODI noninferiority was found for teacher-reported child disruptive behavior. CONCLUSIONS: The tested online-delivered parenting intervention demonstrated clear noninferiority with the corresponding staff-delivered parenting intervention on the primary outcome, child disruptive behavior problems, and reflected substantial though nonuniform noninferiority and meaningful effect sizes for secondary outcomes related to parenting and family. Future research will guide optimization of online interventions.
BACKGROUND: This study evaluated whether an evidence-based parenting intervention, when delivered online, could effectively address disruptive behavior problems in young children and yield outcomes comparable to in-person delivery of the same intervention. METHODS: Families (n = 334) of children (3-7 years; 63% White, 22% African American, 15% other races; 63% male) with disruptive behavior problems were randomized to online-delivered intervention (ODI) or staff-delivered intervention (SDI), resulting in baseline and demographic equivalence. Primary outcome measures for child disruptive behavior (independent observation, parent report) and secondary outcome measures of parenting and family impact were assessed at baseline, postintervention, and follow-up. Conducted using intent-to-treat (ITT) as well as per-protocol (PP) methods, noninferiority analyses, which drew on an HLM framework with repeat measures across three timepoints and on REML to provide unbiased estimates of model parameters, tested whether the outcome-difference CI did not exceed the a priori noninferiority margin. RESULTS: For ITT and PP analyses, the ODI was found to be noninferior to the SDI on the primary outcome: independently observed child disruptive behavior and parent-reported child behavior problems. The pattern for secondary outcomes was more varied: (a) noninferiority for observed positive and aversive parenting; (b) noninferiority for observed quality of parent-child relationship at post but not follow-up assessment; (c) noninferiority for parent-reported inappropriate/inconsistent discipline for PP but not ITT analyses; and (d) noninferiority not confirmed for parenting daily hassles and adverse family quality of life, despite large effect sizes for the ODI (Cohen's d .75-1.07). Finally, ODI noninferiority was found for teacher-reported child disruptive behavior. CONCLUSIONS: The tested online-delivered parenting intervention demonstrated clear noninferiority with the corresponding staff-delivered parenting intervention on the primary outcome, child disruptive behavior problems, and reflected substantial though nonuniform noninferiority and meaningful effect sizes for secondary outcomes related to parenting and family. Future research will guide optimization of online interventions.