Examining Convergent and Divergent Validity of the Repetitive Behavior Scale for Early Childhood (RBS-EC)

Background: Recent evidence suggests that aspects of restricted and repetitive behavior (RRB) manifest as early as 12 months of age in high-risk infant siblings subsequently diagnosed with autism, as measured with both direct observation and parent-report (Elison et al., 2014; Wolff et al., 2014). Improved measurement of RRBs could augment early identification.

Objectives: Following the initial psychometric validation of the Repetitive Behavior Scale – Early Childhood (RBS-EC; Wolff, Boyd, & Elison, 2016), we sought to explore the longitudinal continuity and convergent/discriminant validity between the RBS-EC and previously established behavioral measures. We predicted that RBS-EC scores from an initial time point would be positively correlated with the same participants’ RBS-EC scores an average of 7.8 month later (range 0.45-13.8 months, SD=2.28 months). We predicted that a composite score of all behaviors addressed on the RBS-EC would be associated with dysregulating behaviors (including aspects of sensory sensitivity) as measured by the Infant and Toddler Social Emotional Assessment (ITSEA; Carter et. al., 2003). For discriminant validity, we tested the association of the composite RBS-EC score with the internalizing sub-scale of the ITSEA.

Methods: A community sample of parents of 53 participants (27 male, 26 female) completed the RBS-EC online at a mean age of 22.81 months (range 15.4-26.9 months, SD=2.79 months) as part of a larger study. Participants then participated in a follow-up behavioral visit and additional questionnaires again at a mean age of 30.8 months (range 24.6-38.8 months, SD=3.22 months). The Mullen Scales of Early Learning was administered to participants, and the Vineland Adaptive Behavior Scales-II was completed with parents. Key metrics from these assessments were used as control variables in the analytic models.

Results: RBS-EC sub-scale scores of repetitive motor, restricted behavior, composite behavior frequency, and interference ratings at the first time point were significantly correlated with scores on the same sub-scales at time of behavioral assessment, (all r’s > 0.5, and p’s < 0.001, see Figure 1). This indicates that most RBS-EC sub-scales show continuity across an average of 8 months.

 After observing a zero order correlation between RBS-EC composite mean frequency at time point 1 and ITSEA Dysregulation at time point 2 (see Figure 2), a hierarchical regression analysis revealed a significant association between these variables when controlling for sex, VABS ABC, Mullen fine motor, and age at time point 2, F(5, 47)=4.177, p=.003, ΔR² = 0.149. This demonstrates that the composite of all items on the RBS-EC has moderate convergent validity with established measures of dysregulation behaviors, and other items capturing ASD traits.

Finally, a hierarchical regression indicated that ITSEA Internalizing scores were not significantly associated with RBS-EC composite scores, (F(5, 47)=1.287, p=.286, R² change=.007), controlling for the same variables as above. The lack of association between these scores and RBS-EC’s composite score provides preliminary evidence of divergent validity for the RBS-EC.

Conclusions: This study provides preliminary evidence for predictive, convergent, and discriminant validity for the RBS-EC. More work is needed to extend these findings to samples with higher risk individuals and those with current ASD diagnoses.