HKIER

Abstract

Creating new ideas that are correct (NICs) helps groups solve difficult problems, so modeling group processes that affect the likelihood of NICs might help students create NICs. 80 high school students worked in groups of 4 to solve an algebra problem and were videotaped. Group mean mathematics grade and percentage of NICs predicted solution score at the group level. Using a new statistical method for analyzing group processes in large data sets, I examined predictors of a NIC at the speaker turn level. Correct evaluations raised the likelihood of NICs over the next three turns. Furthermore, wrong, new ideas facilitated NICs by the next speaker. Justifications also predicted NICs by both the current speaker and the next speaker. Meanwhile, speakers who agreed or rudely disagreed were less likely to create a NIC in the same turn. Unlike unsuccessful groups, successful groups' justifications were more likely to accompany NICs, and their questions more often received satisfactory responses. Breakpoint analyses showed that NICs occurred in clusters for each group, averaging 3 time periods per group. Correct evaluation (lag 2) and agreement showed different size effects across time periods. Otherwise, all other predictors showed similar size effects across groups and across time periods.

Keywords: cooperative learning; social interactions; hierarchical linear modeling; time series analysis

摘要

N.A.