Impact of intelligent vs non-intelligent mCSCL on mathematics performance: A comparative study of grade 5 students

This study investigated the effects of using the intelligent and non-intelligent versions of Ibigkas! Math-a mobile-based, computer-supported collaborative learning platform for Grade 5 mathematics on the mathematics performance of Grade 5 students. Out of 155 Grade 5 students from four universities, only 119 participated in the five consecutive-day experiment. Ethical approval was granted before data collection. The study employed a randomized pretest-posttest control group design, in which participants were randomly assigned to either the control (non-intelligent version) or experimental groups (intelligent version). They played the game for three days, and performance on a pre- and post-test was utilized as a measure of mathematics performance. In-game interaction logs were used to assess patterns of behavior. Hierarchical linear models estimated the effects of group and individual behaviors on learning gains. It was discovered that both software versions can be utilized for classroom settings. Students tend to complement each other's behaviors in the game, with no significant difference observed in their in-game behaviors. Students in the intelligent version demonstrated a more methodical and analytical approach, selecting correct answers in a relatively shorter time, thus making the intelligent version more preferable than the non-intelligent version. It is worth noting that the combinations of in-game interactions were shown to affect group learning gain negatively. Therefore, as tasks become more complex, the mathematics performance of the group may also be negatively affected. This study referred to this phenomenon as collaboration fatigue. Considering these findings, the study calls for debriefing sessions before proceeding to the next round of game sessions. Based on the positive outcomes of the tests, Ibigkas! Math is recommended as a math mCSCL tool for Grade 5 students.