20 years on: Confirmation of P. Anderson’s (2002) paediatric model of executive functioning in a healthy adult sample

Abstract

Executive Functioning (EF) is a construct that encompasses multiple interrelated higher order skills, however, conceptualising this nebulous construct remains challenging. This study aimed to confirm the validity of Anderson’s (2002) paediatric model of EF in a healthy adult sample using congeneric modelling. Measures of EF were selected based on utility with adult populations giving rise to minor methodological differences from the original paper. Separate congeneric models were constructed using each of Anderson’s constructs in order to isolate the sub-skills represented by each (Attentional Control-AC, Cognitive Flexibility-CF, Information Processing-IP, Goal Setting-GS), with a minimum of three tests per subskill. One hundred and thirty-three adults (42 males and 91 females) aged between 18 and 50 (M = 29.68, SD = 7.46) completed a cognitive test battery comprising 20 EF tests. AC revealed a good fitting model χ2(2) = 1.61, p = .447, RMSEA = 0.000, CFI = 1.000, after removing the non-significant indicator Map Search (p = .349), and BS-Bk as BS-Bk was required to covary with both BS-Fwd (M.I = 7.160, Par Change = .706), and TMT-A (M.I = 5.759, Par Change = −2.417). CF revealed a good fitting model χ2(8) = 2.90, p = .940, RMSEA = 0.000, CFI = 1.000 after covarying TSC-E and Stroop (M.I = 9.696, Par Change = .085). IP revealed a good fitting model χ2(4) = 1.15, p = .886, RMSEA = 0.000, CFI = 1.000 after covarying Animals total and FAS total (M.I. = 4.619, Par Change = 9.068). Lastly, GS indicated a good fitting model χ2(8) = 7.22, p = .513, RMSEA = 0.000, CFI = 1.000 after covarying TOH total time and PA (M.I = 4.25, Par Change = −77.868). Therefore, all four constructs were reliable and valid, and the utility of a parsimonious EF battery is suggested. Investigation of the inter-relationships between the constructs using regression techniques, de-emphasises the role of Attentional Control and argue instead for capacity bound skills.