A historical perspective of testing and assessment including the impact of summative and formative assessment on student achievement.

摘要

In 2007, Georgia developed a comprehensive framework to define what students need to know. One component of this framework emphasizes the use of both formative and summative assessments as part of an integral and specific component of the teachers' performance evaluation. Georgia administers the Criterion-Referenced Competency Test (CRCT) to every elementary student in Grades 1 though 8. Before 2008, the state tested eighth-grade students on a quality core curriculum. In 2008, the state began testing students on a Georgia Performance Standard curriculum. A direct comparison of the curriculum change should have contained items to test both curriculums. However, this was not done. Therefore, the current study was designed to examine if differences in students achievement occurred because of the curriculum change.;Archival CRCT and ITBS data from 21 middle schools were analyzed to determine if assessment changes affected student achievement. Results of two doubly multivariate, repeated measures ANCOVAs found no statistically significant differences between the two curriculums. However, the lack of significance could be attributed to the small sample size. The increase in scores at the end of the three-year period measuring the quality core curriculum and at the end of the three-year period measuring the Georgia Performance Standard curriculum provided partial support to the hypothesis of a difference in achievement between eighth-grade students who were taught and then tested under different curriculums.;Recommendations for practice include the provision that educators be engaged in professional development in regards to the use of data. Most principals and district leaders do not have the skills to navigate high-stakes testing results. More importantly, though, are the university systems that should augment an instructional strategy class and add a data leadership class to the current list of courses needed to earn a leadership degree. Another recommendation is to those who develop criterion-referenced tests. Changing the score scales on the criterion-referenced competency tests when the curriculum changes make it very difficult to study data to determine progress over time. In the statistical world, this creates a confounding variable that may be hard to control.