An Analysis of the Effects of High- and Low-stakes Testing on Student Achievement
Frankie Eubanks Mathis
University of West Florida Libraries
Doctor of Education (EDD), University of West Florida
2013
Metrics
15 File views/ downloads
97 Record Views
Abstract
With increased emphasis on accountability, the use of low-stakes test data to make high-stakes decisions about program effectiveness is on the rise. In order to make valid inferences about what students know and can do, it is crucial to understand the consequences of low and high stakes in testing contexts. As a result, with a sample comprised of 49 eleventh grade students enrolled in Saraland High School, this study indicates that relationships between student performance on state-mandated testing programs in Alabama and grades earned in corresponding mathematics courses exist. Through regression analyses, it was determined that grades earned in mathematics classes are predictors of scores earned on the state-mandated tests. Additionally, in this study, the sample data revealed that student achievement did not increase as the individual consequences associated with the test increased. Rather, student achievement on the state-mandated tests was consistent with the overall mathematics achievement levels demonstrated in the classroom.