Pamela M. Jakwerth

Curriculum sensitive assessment: Content does make a difference

Abstract This chapter explores the use of rankings and total test scores in large-scale international comparative studies and some of the issues related to them. It examines their value as a basis for policy decisions and educational research. Country ranks on total test scores have been seen to be relatively robust to changes in test content, even changes directed at greater curriculum relevance. Unfortunately, other analyses suggest this robustness is likely a consequence of high levels of aggregation and broad, shallow domain sampling. Student achievement in mathematics and science is inherently multi dimensional. Highly aggregated scores of broadly sampled domains are inherently misleading and mask fundamental, educationally relevant diversities at more specific levels of the curriculum.