<![CDATA[Numeracy among pre-service mathematics teachers is frequently evaluated through test scores alone, making changes in problem-solving processes difficult to capture. This study examined changes in patterns of numeracy difficulties and problem-solving processes from pre-test to post-test following the implementation of DIFUNDES-PROBLEMING. A mixed-methods approach involving 38 mathematics education students at Universitas Jambi was employed. Data were collected using parallel contextual numeracy instruments on systems of linear equations and statistical reasoning, a holistic scoring rubric on a 0–100 scale, coding and categorization informed by grounded theory procedures, and Wilcoxon signed-ranks tests. Pre-test results indicated incomplete problem-solving processes, with modelling, verification, and reasoning often absent or only partially demonstrated. Post-test results showed a more structured sequence of modelling, calculation, verification, and conclusion drawing, with mean scores increasing from 45.57 to 76.76. Wilcoxon signed-ranks tests indicated a statistically significant improvement. Qualitative analysis of changes across the problem-solving indicators showed the greatest improvements in modelling, calculation, and verification, whereas contextual reasoning remained the primary challenge, particularly on items 1.d, 2.c, and 2.d. These findings indicate that DIFUNDES-PROBLEMING contributed to meaningful improvements in both numeracy achievement and the systematic quality of problem-solving processes.]]>
Copyrights © 2026
