Error-flagging support for testing and its effect on adaptation

Abstract

The effect of providing error-flagging support during tests was studied in spring 2009 with two tutors on while and for loops. A partial crossover design was used for the study and mixed-factor ANOVA was used to analyze the students' score per problem, number of revised problems, effect of revision on score and time spent per problem, and finally, the effect of errorflagging on adaptation and learning. Students scored better on tests with rather than without error-flagging support. This can be attributed to the fact that students revised their answers on more problems when error-flagging feedback was provided. But, they did not necessarily revise more often per problem with error-flagging feedback. Students scored the same with revisions as without revisions, whether or not error-flagging support was provided. They spent less time per problem when error-flagging support was provided, whether or not they revised their answers. One explanation for this is that error-flagging may speed up the problem-solving process. Finally, if error-flagging feedback is provided during pre-test, students solve significantly fewer problems during the subsequent practice session which uses the outcome of the pre-test as the basis for adaptation. Therefore, providing error-flagging feedback during the pre-test improves adaptation. But, it does not result in greater learning - learning was not significantly different with versus without error-flagging feedback. © Springer-Verlag Berlin Heidelberg 2010.