Computer-based guidance to support students' revision of their science explanations

As they encounter new ideas, students need to make integrated revisions to their science explanations, a key aspect of science learning. This involves filling gaps, resolving inconsistencies with evidence, and strengthening connections among ideas. Rather than making integrated revisions, even after automated, adaptive guidance, students typically add disconnected ideas or fix mechanical errors. The knowledge integration framework, supported by new technologies including natural language processing, guided the design of the Annotator, a tool that models the revision process for students' written explanations. This research investigates the added value of the Annotator compared to automated, adaptive guidance to support students to make integrated revisions to their science explanations and to strengthen knowledge integration. 798 6th and 7th-grade students from 4 schools participated in a study featuring pretests, posttests, embedded student explanations, student interviews and observations. Students using the Annotator who initially displayed unintegrated ideas were more likely to make integrated revisions to their explanations, than students receiving automated, adaptive guidance. These students also made greater knowledge integration revisions on the posttest one week later. Thus, modeling revision with the Annotator strengthened the ability of students who started with unintegrated ideas to explain scientific phenomena.