Aptitude-treatment interactions revisited: effect of metacognitive intervention on subtypes of written expression in...

We examined the effectiveness of a metacognitive intervention for written language performance, based on the Hayes model of written expression, for 73 fourth-grade (n = 38) and fifth-grade (n = 35) students. The intervention consisted of twenty 45-min writing lessons designed to improve their awareness of writing as a problem-solving process. Each of the lessons addressed some aspect of planning, translating, and reflecting on written products; their self-regulation of these processes; and actual writing practice. All instruction was conducted in intact classrooms. Prior to the intervention, all students received a battery of neurocognitive tests measuring executive functions, attention, and language. In addition, preintervention writing samples were obtained and analyzed holistically and for errors in syntax, semantics, and spelling. Following the intervention, the writing tasks were readministered and cluster analysis of the neurocognitive data was conducted. Cluster analytic procedures yielded 7 reliable clusters: 4 normal variants, 1 Problem Solving weakness, 1 Problem Solving Language weaknesses, and 1 Problem Solving strength. The response to the single treatment by these various subtypes revealed positive but modest findings. Significant group differences were noted for improvement in syntax errors and spelling, with only spelling showing differential improvement for the Problem Solving Language subtype. In addition, there was a marginally significant group effect for holistic ratings. These findings provide initial evidence that Writing Aptitude (subtype) x Single Treatment interactions exist in writing, but further research is needed with other classification schemes and interventions.