Objectives: We examined the validity and reliability of two coding schemes for parent-child relatedness in school-age children with autism, and conducted a preliminary study on the feasibility of these measures for assessing treatment outcome in future studies.
Methods: Participants were 40 children between the ages of six and 14, and their parents. Twenty of the children (16 boys, 4 girls) were diagnosed with autism, and twenty of the children (16 boys, 4 girls) formed a heterogeneous comparison group. The groups were matched for IQ (Autism M = 91.6; Comparison M = 88.9). Dyads participated in a thirty-minute semi-structured assessment of parent-child interaction involving opportunities for joint attention, experience sharing, co-creation and collaboration (RDA; Gutstein & Sheely, 2002). The RDA was subsequently coded by separate sets of blind judges for time spent in states of Joint Engagement (Adamson et al., 2009) and states of a) Interactive Regulation and b) Intersubjective Engagement. Inter-rater reliability was excellent (ICC range .75 - .82).
Comparisons were made to standardised measures of autism severity (Autism Diagnostic Observation Schedule, ADOS: Lord, Rutter, DiLavore & Risi, 2001 and the Social Responsiveness Scale, SRS: Constantino & Gruber, 2005), the Parent Child Relationship Inventory (PCRI: Gerard, 1994), and a more global coding of quality of parent-child interaction (Dyadic Coding Scales, DCS: Humber & Moss, 2005) which has been applied successfully to school-age children with autism (Beurkens, Hobson, & Hobson 2012).
Five of the children with autism receiving a relationship-based intervention were prospectively matched with five children with autism not receiving the intervention, and the measures were repeated one year later (with blind coding of baseline and outcome data), in order to provide preliminary information on the feasibility of this approach for assessing change over time during treatment.
Results: Dyads containing a child with autism spent more time in the following states:
-- Joint Engagement: ‘Supported Engagement’, suggesting caregivers were working harder to sustain joint focus.
-- Interactive Regulation: ‘Contingency without Elaboration’, an interaction-pattern characterised by rigidity and lack of flexibility/variation.
-- Intersubjective Engagement: ‘Coordination of Actions’, where partners engage by telling each other what to do, and attempt to control each other’s actions, rather than engage with each other’s intentions or feelings.
For the ten children with autism and their parents, taking part in the preliminary feasibility study, outcome coding is in progress and will be reported.
Conclusions: Coding schemes for parent-child interaction will allow better evaluation of intervention studies, and will offer clinicians tools to plan and evaluate relationship-based interventions.