Developing a middle school survey of student engagement: Implications for closing mathematics achievement gaps

The purpose of this study was to develop a tool to examine the relationships among student engagement, mathematics achievement, and mathematics self-efficacy of middle school students attending high and low-performing schools with significant minority and low income populations. Three objectives were targeted: (1) to construct and initiate the validation process of a multi-scale survey for assessing student engagement in mathematics for middle school students; (2) to obtain a reliability greater than 0.75 for the survey; and (3) to make a preliminary examination of differences in engagement scores derived from the survey of various groups of students and of relationships between student engagement benchmarks and academic achievement. The result of two pilot administrations (N=30 seventh graders; N=157 eighth graders), review and revisions by experts, and a series of factor analyses was a 36-item survey instrument with a reliability of 0.900 as measured by Cronbach's alpha. In addition, a set of demographic items was included in the Middle School Survey of Student Engagement in Mathematics (MSSSE-M).;Evidence of construct-related and content-related validity of the MSSSE-M was gathered through a step-by-step instrument development process. The multi-scale survey is designed to measure student engagement perceptions related to the educational experiences of middle school students studying mathematics. The survey enables researchers to generate numerical indices derived from students' educational experiences that can be related to outcomes in mathematics achievement.;Preliminary hypothesis testing of differences in mean student engagement scores by school type (low-performing vs high-performing) or student achievement level (state criterion-referenced tests) revealed significant differences on some student engagement benchmarks. High achieving students scored significantly higher than low achieving students on the academic challenge and enriching educational experience student engagement benchmarks (alpha=0.05). Students attending South Carolina schools that evidenced high achievement scores while serving typically low-performing students also scored significantly higher than students attending schools identified to receive state assistance for low achievement on these same two indices, but lower with regard to student-faculty interaction in mathematics. No significant correlations were found between the six student engagement benchmarks and academic achievement as measured by state standardized tests in mathematics. A larger field test is recommended.