Poster #50 - The Stability of Joint Engagement States in Infant Siblings of Children with ASD: Implications for Measurement Practices

• In Event: 1-115 - Poster Session 03
  In Poster Session: PS 03 Section - Developmental Disabilities

Integrative Statement

Background: Previous research indicates that caregiver-child engagement around objects (i.e., ‘joint engagement’) offers an important context for social communication and language development, in typically developing (TD) children and children with autism spectrum disorder (ASD) (Adamson, Bakeman, Deckner, & Romski, 2009). For researchers who study development or who study interventions that promote joint engagement, obtaining stable estimates of this phenomena is of interest. Considering the stability of constructs derived via observational measurement systems is important, because unstructured or semi-structured measurement contexts can introduce unwanted variance due to session (i.e., day-to-day variation in measurement not due to the construct of interest) and coder (i.e., variation in human judgment). These unwanted variations are associated with increased Type II error. However, few studies give guidance on the numbers of sessions and coders needed to obtain sufficiently stable estimates of these constructs. In this study, we explore the stability of two joint engagement states - lower- and higher- order supported joint engagement (LSJE and HSJE, respectively; Bottema-Beutel et al., 2014) in infants at low- and high- risk for ASD. In HSJE, the adult and child reciprocally play together through observable processes such as turn-taking, imitating, or collaborating to achieve a commonly held goal. In LSJE, the adult influences the child’s play, but there is no such reciprocity.
Method: We carried out a generalizability study (a method derived from G theory) to partition error variance between two facets of our measurement system - session and coder. To do this, two 15 min parent-child free play sessions were video recorded for 20 caregiver-child dyads. These sessions were then coded for engagement states by two coders. ANOVAs were used to generate sums of squares for each facet, and for interactions between facets. A decision study was then conducted to determine the number of sessions and coders required to obtain g coefficients of 0.80, which is considered sufficiently stable. This process was conducted separately for 10 infants at high risk for ASD (i.e., infant siblings of children with ASD) and for 10 infants at relatively lower risk for ASD (i.e., infant siblings of TD children) for LSJE and HSJE. Groups did not significantly differ on chronological age, mental age, language age, or sex.
Results: Analyses indicated that in high risk infants, 4 sessions and 1 coder was required for HSJE; 1 session and 1 coder was required for LSJE. In low risk infants, 2 sessions and 1 coder was required for HSJE; 7 sessions and 2 coders were required for LSJE. See Figures 1 and 2 for g coefficients associated with various numbers of coders and sessions according to risk group.
Conclusion: This study will aid researchers who are interested in deriving stable estimates of joint engagement states in planning for future studies by offering guidance on the number of sessions and coders to include in their measurement systems.

Authors

• Kristen Bottema-Beutel, Boston College
  Presenting Author

• So Yoon Kim, Boston College
  Non-Presenting Author

• Shannon Crowley, Boston College
  Non-Presenting Author

• Ashley Augustine, Vanderbilt University
  Non-Presenting Author

• Jacob I Feldman, Vanderbilt University
  Non-Presenting Author

• Bahar Keceli-Kaysili, Vanderbilt University
  Non-Presenting Author

• Pooja Santapuram, Vanderbilt University
  Non-Presenting Author

• Sarah Bowman, Vanderbilt University
  Non-Presenting Author

• Carissa Cascio, Vanderbilt University
  Non-Presenting Author

• Tiffany G Woynaroski, Vanderbilt University
  Non-Presenting Author

• Alexandra Golden, Vanderbilt University
  Non-Presenting Author

• Neill Broderick, Vanderbilt University
  Non-Presenting Author