Assessing planning and set-shifting abilities in autism: are experimenter-administered and computerised versions of tasks equivalent?

Across studies, analysis of performance on classic measures of executive functioning (EF) among individuals with autism spectrum disorder (ASD) suggests that people with this disorder may be impaired only when tasks are experimenter-administered, but not when the same tasks are computer-administered. This would imply that the underlying cause of apparent executive dysfunction in ASD is a diminished ability to engage with another person/comprehend what another person expects, rather than a diminution of the control processes that typically underpin EF task performance. However, this suggestion is limited because, to our knowledge, no study has directly compared the equivalence of computer-administered and standard experimenter-administered versions of EF tasks that have been presented in counterbalanced order among a common sample of individuals with ASD. In the current study, 21 children with ASD and 22 age- and intelligence quotient (IQ)-matched comparison participants completed, in counterbalanced order, computerised and manual versions of both a planning task and a cognitive flexibility/set-shifting task. Contrary to expectation, results indicated that participants with ASD were equally impaired in terms of the key dependent variable on standard and computerised versions of both tasks. Practically, these results suggest that computer-administered and experimenter-administered versions of planning and set-shifting tasks are equivalent among individuals with ASD and can be used interchangeably in studies of EF among this population. Theoretically, these results challenge the notion that poor performance on EF tasks among school-aged children with ASD is only the result of a limited ability to engage with a human experimenter/comprehend socially presented rules.