Abstract
Statement of Problem
Medical research indicates that adolescents have a late-to-bed, late-to-rise sleep cycle that is in direct conflict with the average start time for American high schools. Adolescent sleep patterns can have an important effect on academic performance. This thesis examines whether later high school start times result in improved standardized test scores.
Sources of Data and Methods of Analysis
I compiled data by contacting individual school districts and high schools previously identified as having a later start time. I also contacted state departments of education to obtain test score data for the later starting districts. Additional demographic data were obtained from the State Education Data Center. To analyze the data I used a combination of linear regression and review of a time series graph.
Conclusions Reached
Later starting high schools were grouped into three categories. Linear regression analysis results generally indicated a positive effect on test scores for later starting high schools, although the coefficient for the start time variable was not statistically significant for the first group. For the second group, later start time was associated with a 9.8 percent increase in test scores and this result was statistically significant. My time-series analysis found a noticeable increase in test scores after implementing a later start time for Group 3. These results suggest that later start times result in an improvement in high school test scores and provide strong support for further study of this issue.