Reproducibility

Abstract

The importance of reproducibility has long been recognized in the natural sciences. However, in recent times there has been an acknowledgement that the highest standards of reliability and reproducibility are not always attained, even in the most highly respected journals. For example, Nature has published a number of articles highlighting such shortcomings and an announcement (Anon, 2013) promised that the Nature research journals would introduce editorial measures to address the issue. These measures included ensuring that key methodological details are reported and encouraging authors to be more transparent, for example by including their raw data. In management mathematics, similar concerns have been expressed about the lack of research that is fully reproducible, in the sense that independent researchers are able to obtain the same results as the original study (using the same data and the same methods). For example, Boylan et al. (2015) pointed to the paucity of fully reproducible papers on forecasting, thereby precluding replication of findings on different data. Kendall et al. (2015) suggested that minimum standards are required in optimization research, to ‘enhance the reproducibility of results, the comparison of results, and the efficiency of individual researchers and teams’.

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 — 28 citations exist upstream.

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