| The United States needs to develop a strong workforce in the STEM fields. To build a pipeline of STEM employees, high schools need to increase student success in mathematics. Improving success in mathematics requires that educational assessment tools help all students regardless of their race, gender, grade level, or socioeconomic status. One school district on the eastern shore of Maryland has developed local assessments designed to help teachers to impact achievement in algebra. The district's benchmark assessments are given in January and March each year. It is important to determine if the locally-made benchmark assessments are good predictors of the end-of-year state assessment results. The purpose of this study was to determine the relationship between students' scores on benchmark assessments for algebra and the scores from the Algebra high school assessment (HSA). Results for 816 secondary students were used in this study. Multiple linear regressions were used to determine if the local scores significantly predicted the state assessment results while controlling for demographic variables previously reported to impact student achievement in mathematics. The study found statistically significant regression models for both the March (R2 = .46, F(12, 803) = 58.12, p < .01) and January ( R2 = .40, F(12, 803) = 44.34, p < .01) benchmark assessments. Additionally, effect sizes were large for both March and January benchmark assessments (Cohen's f2 = 0.86 and 0.66, respectively). For the demographic variables of race and grade level, significant differences on the state assessment were found. Recommendations include continuing to use the benchmark assessments for predicting student performance on the Algebra HSA and continuing to use the results to inform the district's data-driven decision-making model. |
