Abstract
Statistics learners often bypass the critical step of understanding a problem before executing solutions. Worked-out examples that identify problem information (e.g., data type, number of groups, purpose of analysis) key to determining a solution (e.g., t test, chi-square, correlation) can address this concern. The authors examined the effectiveness of 3 kinds of worked examples on 96 college students’ problem understanding based on their problem categorizations and explanations: correct schema, contrasting schema (correct and incorrect features), and no schema (traditional solution-only examples). The contrasting- and correct-schema examples were more effective than were the no-schema examples—but neither schema was more effective than the other—in helping students construct accurate and complete problem schemata where data type and number of groups were prominently featured.