How do 3-year-olds do on the NIH Toolbox cognitive subtests?

• In Event: Poster Session 08
  In Poster Session: PS 08 Topic - Methods, History, Theory

Abstract

Background: The NIH Toolbox for the Assessment of Neurological and Behavioral Function is an initiative of the NIH Blueprint for Neuroscience Research (www.neuroscienceblueprint.nih.gov) to create common metrics of functioning allowing for comparison across studies. Since the NIH Toolbox was designed to be used in large longitudinal studies and clinical trials, continuity of assessments within the 3 – 85 years age range was emphasized in its development. The Early Childhood Cognitive Battery (ECCB) is a shortened version of the full cognitive battery for children ages 3 - 6 and consists of four subtests measuring language, episodic memory and executive function, specifically cognitive flexibility and inhibitory control. Executive functions are thought to be still developing within the 3 to 6 age range (Wiebe et al., 2011; Zelazo, Frye & Rapus, 1996), creating a challenge for developing a feasible, developmentally sensitive, psychometrically valid measure. The cognitive subtests were normed on a relatively small number of 3-year-olds (<100 English speakers for most subtests) and were developed for a computer; however, they were reformatted to an iPad, with limited exploration of feasibility of this format, including among 3-year-olds. No previous study has reported exclusively on the performance of 3-year-olds on the NIH Toolbox cognitive battery.
Objective: To examine feasibility and performance of 3-year-olds on three subtests from the NIH Toolbox ECCB.
Methods: NIH Toolbox ECCB subtests were attempted in 609 English-speaking 3-year-olds as part of the National Children’s Vanguard Study. Administration and device status coding was examined to determine participants who completed three subtests: the Flanker Inhibitory Control and Attention Test (Flanker), the Dimensional Change Card Sort Test (DCCS), and the Picture Vocabulary Test (PVT). First, the proportion of participants that failed the practice was calculated. Then the proportion of valid scores that were likely performing by chance, based on the 95th percentile of the binomial distribution with a chance probability of success at 0.5, was calculated for the DCCS and Flanker. Chance levels and accuracy scores could not be examined for the PVT because it is a computer adaptive test.
Results: Of those who attempted the Flanker, 19.38% did not pass the practice. 29.27% did not pass the practice on the DCCS. 15 participants (2.66%) did not complete the PVT practice. 66.91% of scores on the Flanker and 71.47% on the DCCS were below chance. The distribution of PVT computed scores seemed relatively normally distributed.
Conclusions: Based on the high proportion of 3-year-olds performing below chance on the Flanker and DCCS, the measures may not accurately measure differences in ability within children this young. The more variable distribution of scores on the PVT indicates that this subtest may be better able to differentiate ability. Ultimately future iterations of the NIH Toolbox should include more developmentally appropriate measures for young children.

Authors

• Lindsey Becker, Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD/NIH)
  Presenting Author

• Emma Condy, National Institute of Mental Health
  Non-Presenting Author

• Audrey Thurm, National Institute of Mental Health
  Non-Presenting Author