B-74Two-Factor Structure of the Comprehensive Trail Making Test in Adults

Objective: Previous evaluation of the Comprehensive Trail Making Test (CTMT) using a confirmatory factor analysis (CFA) revealed that, for children, a two-factor model provided a better fit than a one-factor model. The purpose of this study was to explore whether these findings suggested a developmental factor structure or if the two-factor structure would also be supported for adults. Method: A CFA was conducted using MPlus to examine the factor structure for adults in the standardization sample. Participants: The sample for this study included 695 adults, aged 18–74 (M =40.48, SD = 15.65). The participants were predominantly women (67.77%) and whites (88.35%). Selection: Data were collected across 16 states with all regions of the United States represented using stratified random sampling. Measures: All five trails tasks from the Comprehensive Trail Making Test (Reynolds, 2002) were considered. Results: The findings of the CFA indicated an excellent fit for the two-factor model: χ²(4) = 1.976, p = .741; CFI = 1.000; RMSEA = .000; SRMR = .005. A one-factor structure was not supported: χ²(5) = 63.636, p = .000; CFI = .965; RMSEA = .129; SRMR = .034. Results were the same when younger (under 50) and older (50+) were considered separately. Conclusions: These data suggest that a single score, the CTMT Composite Index, may not be a valid representation of executive functioning. Instead, the results support a two-factor structure, which represents attention/sequencing and set switching/inhibition, to identify an individual's deficits in functioning.