Implementation and initial validation of a computer-based system for the assessment of reading competencies

• Main Author: Sinatra, Gale Marie
• Language: ENG
• Published: ScholarWorks@UMass Amherst 1989
• Subjects: Educational psychology|Psychology|Experiments|Psychological tests|Educational evaluation
• Online Access: https://scholarworks.umass.edu/dissertations/AAI9011800

The purpose of the present study was to investigate the viability of a computer-based assessment system, called the System for the Assessment of Reading Competencies, to aid in the description and diagnosis of reading difficulties. The practical need for such a system, the theoretical bases that serve as its foundation, and the evidence for the system's validity are presented. Currently, there is little relation among diagnosis, assessment, remediation, and instruction in reading. The lack of integration of these areas and the need for assessment instruments that provide sufficient information for diagnosis and remediation led to this system's development. The present study involved the implementation and initial validation of the system. The system is a computer-based, componential, reading assessment instrument that is based on an information processing model of reading. One-hundred-and-twelve second, third, fourth, and fifth grade students were administered Sentence Verification Technique (SVT) Tests of listening and reading comprehension in addition to several computer tasks. These tasks included a response time measure, letter matching, word naming, pseudoword naming, category matching, syntactic analysis, and sentence comprehension. Both response times and response accuracies were collected on each task. The results were analyzed to examine the reliability and validity of the assessment system. The results showed that the assessment battery was reliable. Further, the results showed that the system successfully discriminated between students in different grade levels and between students of differing abilities within the same grade. The results showed a developmental trend such that the largest differences between ability groups were on the lower level tasks for second graders and on the higher level tasks for third and fourth graders. It was concluded that the evidence supports the reliability and validity of the system as a measure of reading ability, and further that the system has potential as a diagnostic instrument. It was also concluded that measures of response time may discriminate between ability groups even when measures of response accuracy do not. The development of a remedial component to the assessment system is discussed as a direction for future research.