Insights into survey error of large scale educational achievement surveys

While educational achievement surveys revolutionised research on education crossnationally, the surveys have been repeatedly subject of heated debate since first results were published. This paper reviews existing research examining the design and methodology of educational achievement surveys. Results are reported by allocating them to the specific survey error component of achievement estimates they address. Different error components from the design, collection, processing and analysis of survey data constitute the total survey error, which is an error difficult to quantify but important for assessing the overall accuracy of the surveys' achievement estimates. The review shows that there are many reasons to assume that the total survey error associated with countries’ educational achievement estimates is likely to be inflated by other errors besides the standard error reported by survey organisers. Given the policy relevance of the surveys’ estimates, policy makers and the research community would greatly benefit from survey organisers providing more transparency on the different potential errors of educational achievement estimates. Without this information the debate about the fitness of educational achievement data for policy making is unlikely to dissolve.