Understanding students' statistical thinking: An error analysis approach

This exploratory study examined errors that students commit solving multiple-choice questions about descriptive statistics and basic concepts in research methods. The sample consisted of 81 undergraduate students in an introductory statistics course. The results indicated that the most frequently detected errors were confusing concepts, misinterpreting descriptive information, applying inappropriate procedures and applying partial information. Analysis reveal potential sources of students' errors include assimilation of statistical concepts into inappropriate schemata, failure to use knowledge sources, and lack of ability to relate and combine knowledge from different sources.