A comparative and interpretive study of the effects of traditional multiple-choice assessment with generative alternative assessment on state anxiety and mathematics achievement of sixth grade students

This researcher described and compared the effects of traditional multiple-choice assessment and generative alternative assessment on sixth grade students' mathematics achievement and state anxiety level. She looked at how verbal ability, quantitative ability, self-esteem, self-concept, mathematics anxiety, trait anxiety, school failure tolerance and gender affect achievement and state anxiety on the two types of assessment instruments. Error patterns in procedural questions were also examined and compared for the traditional multiple-choice assessment and generative alternative assessment.;The subjects used in the study were predominantly White students in an affluent suburban middle school in Nassau County, New York. The sample consisted of 274 sixth grade students. Participants completed the following affective inventories: the State-Trait Anxiety Inventory for Children by Spielberger (1973); the Self-Esteem Inventory: School Form by Coopersmith (1989); the Mathematics Anxiety Scale for Children by Chiu & Henry (1990); the Self Concept Scale; and the School Failure Tolerance Scale by Clifford (1991). Students also took the Cognitive Abilities Test by Thorndike and Hagen (1993) as a part of the regular school assessment program. Due to attrition of two of the mathematics teachers' classes, the sample that completed the assessments was comprised of 135 students.;The curricular area chosen for this study was measurement (area and perimeter of squares, rectangles, parallelograms, and triangles, as well as area and circumference of a circle). Assessment was implemented at the completion of the unit's instruction. The assessment consisted of a traditional multiple-choice assessment and a student generative alternative assessment. The two assessment measures consisted of six high-level problem-solving questions. The assessments were administered to participants with a two-day time frame between the two test administrations. Students' state anxiety level was measured after completing each assessment format. Analysis of the data was conducted to answer the research questions and hypotheses. The findings supported the view that students exhibited higher levels of achievement and higher levels of state anxiety under the generative alternative assessment than the traditional multiple-choice assessment. Alternative assessment was found to be a better format over multiple-choice assessment for high-order problem solving. Affective factors were also examined and recommendations for future research were made.