The Importance of Sample Weights and Plausible Values in Large-Scale Assessments

Abstract

International large-scale assessments such as PISA (The Programme for International Student Assessment), PIAAC (The Programme for the International Assessment of Adult Competencies) and TIMSS (Trends in International Mathematics Science Study), play a key role in determining educational policies besides their primary objectives of measuring, evaluating and monitoring the educational process. Therefore, it is critical to analyze the data gathered from the large scale assessments using scientifically accurate statistical methods as the results have the potential to influence millions of stakeholders through major policy changes. Analysis of these data that consists of hundreds of different genuine variables requires expertise and using specific methods. This study illustrates issues to be considered while analyzing PISA, PIAAC and TIMSS data by presenting relevant syntax and exemplifying the possible incorrect results that might be encountered. In Turkey, there are very limited courses that focus on large scale data analysis. Workshops are also very limited to reach major groups. The aim of this study is to raise awareness related to sample weights and plausible values. Comparative findings of the study showed that without using sample weights and plausible values there is a high probability to get incorrect results. In this study, t-test and multiple regression analyses conducted by IDB Analyzer and multilevel regression analysis by Mplus were exemplified.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 26 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.