Push button replication: Is impact evaluation evidence for international development verifiable?

Le résumé fourni par la source

OBJECTIVE: Empirical research that cannot be reproduced using the original dataset and software code (replication files) creates a credibility challenge, as it means those published findings are not verifiable. This study reports the results of a research audit exercise, known as the push button replication project, that tested a sample of studies that use similar empirical methods but span a variety of academic fields. METHODS: We developed and piloted a detailed protocol for conducting push button replication and determining the level of comparability of these replication findings to original findings. We drew a sample of articles from the ten journals that published the most impact evaluations from low- and middle-income countries from 2010 through 2012. This set includes health, economics, and development journals. We then selected all articles in these journals published in 2014 that meet the same inclusion criteria and implemented the protocol on the sample. RESULTS: Of the 109 articles in our sample, only 27 are push button replicable, meaning the provided code run on the provided dataset produces comparable findings for the key results in the published article. The authors of 59 of the articles refused to provide replication files. Thirty of these 59 articles were published in journals that had replication file requirements in 2014, meaning these articles are non-compliant with their journal requirements. For the remaining 23 of the 109 articles, we confirmed that three had proprietary data, we received incomplete replication files for 15, and we found minor differences in the replication results for five. CONCLUSION: The findings presented here reveal that many economics, development, and public health researchers are a long way from adopting the norm of open research. Journals do not appear to be playing a strong role in ensuring the availability of replication files.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.