Does Archival Data Provide Useful Insights for Management Accounting Research?

Abstract

Empirical analysis using archival data has become one of the dominant methodological approaches in contemporary management accounting research in Japan. However, the specific characteristics and limitations remain underexplored. This paper examines how large sample sizes, which are a typical feature of archival data, influence hypothesis testing. Such hypothesis testing conducted in large samples can lead to results that are statistically significant but lack substantive meaning. To address this issue, we propose two directions for more constructive use of archival data. First, we emphasize the importance of testing hypotheses derived from rigorous theory. Second, we advocate for macro-level descriptive analyses that highlight clear differences or observable trends in the data.