“The Role of Executive Function in Mathematics and Science Learning Difficulties of Students with Disabilities”

PI: Paul Morgan (Penn State University)
CO-PIs: Yoonkyung Oh, George Farkas, Marianne M. Hillemeier 

Funder: National Science Foundation (NSF)

Duration: 2016-2019


We will investigate whether executive functions (EF) deficits are related to lower mathematics and science achievement, with a particular focus on those with or at risk for disabilities. We will also use growth mixture modeling (GMM) to identify early development trajectories of EF as well as mathematics and science achievement. For students with and without disabilities, and with particular conditions, we will identify characteristics of homes, classrooms, schools, and neighborhoods that most strongly relate to their EF as well as mathematics and science achievement. Our analyses will make extensive use of newly available data from a large, diverse, and nationally representative cohort of U.S. schoolchildren followed from kindergarten entry until the end of fifth grade. Assessments of three specific types of EF (i.e., working memory, cognitive flexibility, and inhibitory control) were individually and repeatedly administered, as were individually administered and untimed assessments of reading, mathematics, and science achievement. Characteristics of the students’ homes, classrooms, schools, and neighborhoods were surveyed. The cohort’s very large sample size, six-year timeframe, and rich data collection will allow us to make use of a range of quasi-experimental methods to investigate the project’s objectives. These methods include growth mixture modeling to identify latent classes of students displaying persistently lower EF as well as lower mathematics or science achievement over time. We will identify factors that increase the risk of these class memberships. We will also use fixed effects and covariate adjustment for previously identified confounds (e.g., autoregressive achievement, behavioral inhibition and self-regulation, family socioeconomic status) in multivariate regression models to identify potential targets of STEM interventions that might be delivered during the elementary grades.

Our project will provide important new information about how best to help students with disabilities who are at risk of experiencing persistently low levels of mathematics and science achievement as they age, thereby limiting their STEM-related educational and career opportunities. Our project will help answer whether underlying deficits in EF may be resulting in these students’ learning difficulties, and so might be targeted by early STEM interventions. We will further inform these efforts by providing new knowledge as to which type of EF deficit (i.e., working memory, cognitive flexibility, or inhibitory control) is more strongly related to learning difficulties in mathematics and science. Our project will inform the timing of these STEM intervention efforts by clarifying how early in schooling persistent learning difficulties in mathematics and science emerge, and whether and to what extent these difficulties begin to negatively impact students’ feelings and views towards these subjects. Thus, we will address whether STEM interventions should be designed to address attitudinal as well as academic barriers by the elementary grades. Collectively, the project’s findings will strongly inform screening, monitoring, and intervention efforts for students with or at risk for disabilities when these efforts may be most effective.