Getting Stuck: Children with High Functioning Autism Spectrum Disorders Demonstrate Impaired Cognitive Flexibility on the Flexible Item Selection Task (FIST)

Background: Individuals with an autism spectrum disorder (ASD) show a lack of cognitive flexibility across different settings (Geurts, Corbett & Solomon, 2009).  Inflexible thinking can have a significant impact on social functioning and development of adaptive skills.  There are two types of tasks designed to assess flexibility:  (1) Inductive (i.e., tasks which provide feedback to participants and require them to demonstrate flexibility in problem-solving by inferring a new rule (e.g., Wisconsin Card Sorting Test [WCST]); and 2) Explicit (i.e., tasks which do not require reasoning, but provide explicit cues to guide the participant, e.g., Preparing to Overcome Prepotency [POP]).  A third type of task that is relatively absent in ASD research is deductive flexibility tasks (i.e., provide a general rule and children demonstrate flexibility by executing new behaviors within the rule).

Objectives: To measure flexibility in pre-adolescent children with ASD relative to typically developing (TD) children matched on verbal mental age with a deductive flexibility task.

Methods: Forty-four children participated in this study. The ASD group (n=22) and the TD group (n=22) were matched on verbal mental age (ASD M(SD)=7.92(2.14; TD M(SD)=7.16(1.18) t(42)=1.59, p>.05) and gender (ASD M/F= 18/4; TD M/F= 16/6; Ⅹ²(N=44)=0.47, p>.05), but not chronological age (ASD M(SD)=8.48(1.52; TD M(SD)=6.26(0.82) t(42)=6.02, p<.05).  Children completed the Flexible Item Selection Task (FIST; Jacques & Zelazo, 2001). In the FIST, children are presented three simple pictures (e.g., small blue shoe, small yellow teacup, small yellow shoe) and asked to group two items that go together.  Children are given two opportunities to select pairs (pair A and pair B), and pair B is  “two things that go together but in a different way” from pair A.  In order to focus deliberately on switching we examined the percentage correct of selecting Pair B after correct selection of Pair A (hereafter, referred to as shift score).

Results: The ASD group had a lower shift score than the TD group (ASD M(SD)=76.08%(29.33); TD M(SD)=89.48%(9.77) even when controlling for differences in chronological age (Fs>10, ps<0.001).  

Conclusions: Pre-adolescent, school-aged children demonstrated impaired performance on the FIST, a deductive measure of cognitive flexibility, relative to verbal-age matched controls.  There are few executive control tasks for this age range that do not require inductive reasoning or provide explicit instructions regarding the new rule (e.g., now sort by shape). This form of flexibility may have greater ecological validity because many everyday tasks require a self-selected rule (e.g., putting toys in a toy chest based on toy theme) and then adaptation (e.g., putting toys in a toy chest so that the lid actually closes!), and provides a useful metric for evaluating flexibility in younger children with ASD.