Measuring children's sustained selective attention and working memory: validity of new minimally linguistic tasks.

Abstract

This study introduces visual tasks using nonlinguistic stimuli that measure sustained selective attention (SSA) and working memory (WM), two constructs foundational to learning and associated with developmental disorders in children. Using an argument-based approach to validation, we examine whether each task (a) measures distinct constructs, (b) shows internal consistency, (c) captures a range of performance, and (d) relates to development as indexed by age. Participants included 71 children, ages 4-10, of whom 12 had parental concern for language/learning. The SSA task presented spatial locations within a long and uninteresting task, following the continuous performance task paradigm. The WM task presented paired location sequences of increasing length, incorporating key elements of the n-back and complex span paradigms. Controlling for age, tasks were found to be minimally associated with each other (r = .26), suggesting related but distinct constructs. Internal consistency was high, with split-half reliability of .94 (SSA) and .92 (WM); the stability of these estimates was supported by bootstrapping simulations. Task performance was evenly distributed, with minimal floor or ceiling effects within this age range. Performance was positively related to age (SSA r = .49; WM r = .53). Exploratory correlations with a measure of parental concern were significant for SSA but not WM. The results show that these new tasks can be used to measure children's SSA and WM in a visual domain with minimal linguistic influence. These tasks capture developmental changes in the early school years. Further investigation can examine their utility for classifying children with developmental disorders.