<![CDATA[Mathematical anxiety constitutes an affective barrier capable of undermining students' numeracy competence and overall mathematics achievement. This study was designed to examine the effect of utilizing the Edcafe platform on the level of mathematical anxiety among senior high school students within a statistics learning framework that integrates Problem-Based Learning with STEM (PBL-STEM). A quantitative method with a one-group pretest-posttest design was employed, involving 31 senior high school students as participants who completed the entire Edcafe PBL-STEM learning process. Mathematical anxiety was measured through a questionnaire encompassing cognitive, affective, and physiological dimensions, administered at two time points before and after the intervention. Data were analyzed descriptively and inferentially using the Wilcoxon Signed Rank Test. Findings revealed a significant decrease in the mean mathematical anxiety score from 56.65 at pretest to 51.10 at posttest (p < 0.05), accompanied by a distributional shift toward lower anxiety categories. These results indicate that the AI features embedded in Edcafe comprising automatic feedback, an adaptive learning chatbot, and contextual PBL-STEM scenarios are capable of constructing a more psychologically conducive learning ecosystem, thereby meaningfully reducing students' mathematical anxiety.]]>
Copyrights © 2026
