ABSTRACT
There is significant overlap between reading and writing, but no known standardized measure assesses these jointly. The goal of the present study is to evaluate the properties of a novel measure, the Assessment of Writing, Self-Monitoring, and Reading (AWSM Reader), that simultaneously evaluates both reading comprehension and writing. In doing so, we evaluate reliability (Cronbach's alpha) and various aspects of construct-related validity, including separate criterion measures of reading and writing, and the AWSM Reader's relations with other variables, including language and executive function (EF), both of which are implicated for both reading and writing. Participants were 377 3rd, 4th, and 5th graders with or at-risk for reading and writing difficulties. Reliability was low for the AWSM Reader reading comprehension (α = .58), but good for writing (α = .75-.80). Criterion-related validity indices revealed moderate correlations with other standardized and commonly used reading and writing measures, r = .47 to .50 (all ps < .001). Additionally, validity data supported the relations of both language and EF to AWSM Reader reading and writing, with EF showing unique prediction in both reading and writing domains. Results provide initial support for the measure per se but stress the difficulty in constructing combined reading and writing measures; directions are given for future work. Results also add to data on the contributions of language and EF to both reading and writing.
ABSTRACT
PURPOSE: Acute lymphoblastic leukemia (ALL) is the most common pediatric cancer diagnosis. Cognitive late effects develop in 20%-40% of ALL survivors, but the course of declines is unclear. The aim of this paper is to characterize cognitive functioning, and its association with patient-reported outcomes, early in treatment. PATIENTS AND METHODS: A total of 483 children with high-risk ALL, aged 6-12 years at diagnosis, consented to the neurocognitive study embedded in a prospective therapeutic trial, Children's Oncology Group (COG) AALL1131. A computerized neurocognitive battery (Cogstate) was administered 3 months post diagnosis assessing reaction time, visual attention, working memory, visual learning, and executive functioning. Parent-reported executive functioning and patient-reported physical symptoms were also collected. RESULTS: Data from 390 participants (mean age at diagnosis = 9.2 years, 55.4% male) were obtained. Relatively few patients reported pain (16.0%) or nausea (22.6%), but a majority (68.5%) reported feeling at least some fatigue at testing. Mean Cogstate Z-scores were within normal limits across tasks; however, rates of impairment (Z-scores ≤ -1.5) for reaction time, working memory, visual learning, and visual attention were all higher than expected compared to the standardization sample. Patients reporting fatigue were significantly more likely to have impaired reaction time and visual attention compared to those reporting no fatigue. CONCLUSION: Findings support feasibility of computerized cognitive assessments and suggest higher-than-expected rates of impaired cognitive performance early during treatment for pediatric ALL, notably within 3 months of diagnosis, suggesting intervention efforts may be indicated. These results also highlight acute factors that may impact reliability of "baseline" assessments conducted soon after diagnosis.
