Interpreting the direct- and derived-Trail Making Test scores in Argentinian children: Regression-based norms, convergent validity, test-retest reliability, and practice effects

Abstract

Objective: We carried out two research studies (using a cross-sectional and a longitudinal design) to establish regression-based normative data, and to examine the convergent validity, the test-retest reliability, and the changes in performance with practice (1-month and 1-year interval) of the direct- and derived- (B-A, B + A, and B/A ratio) Trail Making Test (TMT) scores in Spanish-speaking children. Method: In S1 (n = 413 6- to 15-year-old<br />children and adolescents), regression-based norming procedure and partial correlations were employed. In S2 (n = 110 6- to 12-yearold children), intraclass correlation coefficient (ICC2,1), Pearson’s r-product-moment correlations, and paired t-tests were used. Results: S1 demonstrated that age was associated with better performance on both TMT-A and TMT-B. This improvement was more strongly related to direct scores than derived measures B-A and B/A. Additionally, TMT-B was found to be more related to other executive functions (EF) compared to the performance of TMT-A. Among the derived scores, only the B-A was related primarily to cognitive flexibility, while the B + A index was associated with most<br />EF, suggesting a general measure of cognitive functioning. In S2, fair to good test-retest reliability coefficients were found at Time 2 for TMT-A and TMT-B, as well as the B + A index, both in 6 to 8-year-olds (ICCs .61 to .74) and 9 to 12-year-olds (ICCs .53 to .65).<br />There was a significant increase in performance on TMT-A and TMT-B from the first assessment (Time 1) to the follow-up testing at Time 2 in older children. However, this significant improvement was not observed for TMT-B in the younger group. Test scores on<br />the TMT direct measures improved significantly across 1-year. The B-A and B/A ratio scores did not change across examinations. Conclusions: These findings have important implications for assessing EF and developing interventions that target cognitive flexibility in pediatric populations.