Mental processes underlying STEM problem solving in exceptional and typically developing children

What mental processes do children spontaneously engage when facing STEM problems? The question is under-researched not only in exceptional children with autism and high intelligence but also in their typically developing counterparts. It has been argued that the cognitive and emotional processes behind STEM problem solving are mutually conditioned and may influence learning outcome. Given this background, the present study aims to examine: (1) the cognitive and emotional processes behind STEM problem solving in typically developing, autistic, and high-intelligence students, (2) STEM performance differences among the three groups, (3) the role of these processes in explaining the group differences in performance, (4) students’ feelings about learning before and after the STEM experience. We recruited 88 male and 47 female Hong Kong 5th and 6th graders (Mage = 138.10±10.02 months), including 50 typically developing (TD), 35 autistic-normal-intelligence (ASD), and 50 high-intelligence students (HI, with a Raven’s Standard Progressive Matrices (RPM) score > 45). The students first attended a think-aloud workshop to learned to verbalize what they were thinking when solving physics problems. They proceeded to the STEM lesson on the next day starting with the Achievement Emotions Questionnaire (AEQ). They then worked on two physics experiments respectively on sound and electricity (5th grade), or friction and leverage (6th grade). Students verbalized their thoughts and emotional responses during the experiments and such verbalization was audio-recorded. Recordings were transcribed and coded into four types of processes: cognitive, metacognitive, emotional, and other (i.e., task-irrelevant thoughts), using the scheme devised by Montague and Applegate (1993). Finally, all the students took a multiple-choice physics quiz on what they just learned from the experiments and completed the Peabody Picture Vocabulary Test (PPVT) and the AEQ for a second time. The HI group used more cognitive strategies than the TD and ASD group; the TD and ASD group engaged in more irrelevant processes than the HI group. The ASD students ranked higher on the proportion of metacognitive processes than the HI students, and the TD students ranked higher on emotional processes than the HI students. The HI group obtained higher quiz (performance) scores than the TD group. Mediation analyses showed that relative to TD, HI’s more frequent use of cognitive strategies and less frequent engagement in irrelevant processes explained their higher quiz scores. The HI students had more positive feelings for STEM learning than the ASD students as indicated by the AEQ scores. The present study provides evidence for the differential involvements of types of mental processes in STEM learning among students with diverse backgrounds, and the potential roles of these processes in mediating between student background and performance. Cognitive and meta-cognitive thinking processes are likely to facilitate STEM problem solving while emotional and irrelevant processes may hinder it. We conclude that students with diverse backgrounds can take STEM lessons in one classroom because they share similar mental processes when engaged in STEM problems. They may learn from one another regarding what processes to focus on when approaching STEM problems. Copyright © 2023 SRCD.