Abstract:

Game-based learning and formative assessment are important elements of learning for today's students. This study brings the two together using the RPG MAKER software (Gotcha Gotcha Games Inc.). With this software, we created a role-playing game that incorporates multiple formative assessments. However, what kind of feedback is given in formative assessment must be carefully considered. Therefore, this study refers to the article "Focus on formative feedback" by Shute (2008), from which five feedback types are selected. 144 fifth-grade elementary school students (47% female) were recruited from four classes in northern Taiwan. This is a one-group pretest-posttest design where students play the same game. Because each feedback type has a matching paper-based test item, we can derive a student's progress scores on a certain feedback type. In addition to feedback types, the study also considered gender, and therefore, a two-way mixed design ANOVA was performed. A two-way ANOVA was performed to analyze the effect of feedback types and gender on progress scores. A two-way ANOVA revealed that there was not a statistically significant interaction between the effects of feedback types and gender (F(3.65, 517.55) = 0.75, p = .55). Simple main effects analysis showed that feedback types did have a statistically significant effect on progress scores (p < .001). The verification feedback and topic contingent feedback had worse performance. Presumably, verification feedback simply tells subjects whether their responses are correct or not, so they have no way of knowing why they answered correctly or incorrectly, which has a limited impact on learning. The reason why the topic contingent feedback is less effective may be because the feedback is detailed feedback, re-teaching the concept of the question, and the subjects do not have the patience to read all the content. Try again feedback had better performance, perhaps through this mechanism, students can have a deeper impression of the question. In addition, simple main effects analysis showed that gender did have a statistically significant effect on progress scores (p < .001). The reason may be due to students' attitude towards learning. Male subjects tended to complete learning tasks quickly, while female subjects tended to complete learning tasks steadily.

Keywords:

Feedback type, Formative assessment, Game-based learning.