Improving Reading Literacy through Diagnostic Assessment-Based Differentiated Instruction: An Analysis of Three AKM Cognitive Levels among Fifth-Grade Primary School Students

This study examined reading literacy gains following diagnostic assessment-based differentiated instruction and identified achievement patterns across three cognitive levels of Indonesia’s Minimum Competency Assessment (AKM). A pre-experimental one-group pretest–posttest design was employed with fifth-grade students at SDN 53 Sawerigading, Palopo. Twenty students completed the initial assessment, and paired analyses included 18 students with complete data. Instruments comprised two reading test forms and a diagnostic questionnaire. Analyses compared all 20 items with an 18-item scoring scope that excluded two items according to a decision recorded in the research documents. Data were analysed using descriptive statistics, paired-samples t-tests, Wilcoxon signed-rank tests for the 18-item scope, Cohen’s dz, and normalized gain. For 20 items, mean scores increased from 69.09 to 84.47, with a mean individual normalized gain of 0.498. For 18 items, scores increased from 74.13 to 82.86, with a normalized gain of 0.337. Both differences were statistically significant, and both mean normalized gains were classified as moderate. Cognitive-level analysis within the 18-item scope showed the largest improvement in evaluating and reflecting, from 50.00% to 72.69%, while two indicators involving tabular information remained unchanged. The direction of improvement was consistent, although its magnitude depended on the scoring scope. Disaggregated analysis helped identify instructional priorities. The absence of a comparison group prevents causal attribution