Poster #7 - NIH Toolbox Executive Function Measures – Developmental Extension: Reliability and Validity with Disadvantaged Preschoolers

• In Event: 2-110 - Poster Session 07
  In Poster Session: PS 07 Section - Attention, Learning, Memory

Integrative Statement

Executive function (EF) skills are neurocognitive processes that support goal-directed action and are linked to academic success and socioemotional competence (Best, Miller, & Naglieri, 2011; Carlson & Moses, 2001). EF may be particularly important in supporting school readiness in children facing adversity and socioeconomic disadvantage (Masten et al., 2012; Obradović, 2010). However, one critical issue to studying EF in disadvantaged children is the lack of reliable and valid measurement. Although dozens of EF tasks exist (Carlson, 2005), many of them are not appropriate for younger or disadvantaged populations. To remedy this problem, developmental extension versions of the NIH Toolbox Dimensional Change Card Sort (DCCS) and Flanker Inhibitory Control and Attention task were created to make the tasks easier and alleviate floor effects. The current study examined the test-retest reliability and construct validity of the developmental extension versions of the DCCS and Flanker task in a sample of children experiencing homelessness.
Participants were 89 preschoolers living in an emergency homeless shelter with their families during the summers of 2015 and 2016. Children (57% female) were between the ages of 36 and 72 months (M = 55.86, SD = 9.72). The sample included 65% Black/African American, 17% Multiracial, 6% White/European American, and 3% American Indian or Alaska Native children. Data were collected in a designated research room at the shelter, and children participated in 2 study sessions about a week apart (M = 8.41 days, range = 3 – 23 days). During the first session, participants completed the developmental extension versions of the DCCS and Flanker on a touch-screen tablet, as well as Peg-tapping – an EF task in which children tap 1 time if the experimenter taps 2 times and 2 times if the experimenter taps 1 time (Diamond & Taylor, 1996). Children also completed the Woodcock Johnson Applied Problems subtest (Blackwell, 2001) as a measure of academic knowledge. During the second session, children were given the DCCS and Flanker again to assess test-retest reliability. Although this sample was highly mobile, 84% of children returned for their second session.
We found that the developmental extension was essential for using these tasks with children in the shelter: 69% of children required the DCCS extension and 46% needed the Flanker extension. Because we had some missing data due to non-compliance or children not returning for their second session, we multiply imputed 10 datasets, which were pooled and used in the subsequent analyses. Results indicated that both measures had good test-retest reliability (DCCS ICC = .72; Flanker ICC = .81). Construct validity was assessed by examining correlations between all variables (see Table 1). As expected, DCCS and Flanker were significantly correlated with child age, Peg-tapping, and Applied Problems. However, DCCS and Flanker were only significantly associated at Time 2, after controlling for child age, suggesting that children in this context may need some practice with the tablet-based measures to accurately assess their skills. Overall, the developmental extension versions of the DCCS and Flanker task appear to work well in an emergency shelter context.

Authors

• Rebecca Linnea Distefano, University of Minnesota Twin Cities
  Presenting Author

• Aria E Fiat, University of Minnesota Twin Cities
  Non-Presenting Author

• Jillian S. Merrick, University of Denver
  Non-Presenting Author

• Jerry Slotkin, University of Delaware
  Non-Presenting Author

• Philip David Zelazo, University of Minnesota Twin Cities
  Non-Presenting Author

• Stephanie M. Carlson, University of Minnesota Twin Cities
  Non-Presenting Author

• Ann S. Masten, University of Minnesota Twin Cities
  Non-Presenting Author