<![CDATA[Purposes: This study aimed to examine seventh-grade students’ response patterns in integer learning using Newman-informed multiple-choice distractors and to investigate whether these patterns varied across items, diagnostic indicators, difficulty levels, and gender. Method: The study employed a quantitative descriptive design supplemented by chi-square tests of association. The dataset included 1,094 seventh-grade students who answered 27 valid four-option integer diagnostic items. Each distractor was preassigned during instrument development to one of five Newman Error Analysis categories: reading, comprehension, transformation, process skill, and encoding. Correct and blank responses were summarized separately. Findings: From 29,538 possible item responses, 29,501 were nonblank, comprising 17,822 correct and 11,679 incorrect responses. The dominant wrong-response categories were process skill errors (29.79%), transformation errors (28.94%), and comprehension errors (25.44%). Contextual financial-literacy and LCM items produced the highest wrong-response percentages. Association tests showed moderate associations between Newman error type and both item (Cramér’s V = .49) and diagnostic indicator (V = .30), a small-to-moderate association with difficulty level (V = .22), and trivial gender-related associations (V = .04 for error type and V = .02 for response status).\ Significance: The findings suggest that NEA-informed distractors may enhance the diagnostic value of multiple-choice mathematics assessment by linking incorrect options to hypothesized sources of difficulty. The results support prioritizing item- and indicator-specific diagnostic feedback over interpretations based primarily on gender. In practice, remediation should prioritize guidance on selecting appropriate operations for transformation errors and guided procedural practice for process-skill errors.]]>
Copyrights © 2026
