Poor methodological detail precludes experimental repeatability and hampers synthesis in ecology

• Published in: Ecology and evolution 2015-10, Vol.5 (19), p.4451-4454
• Main Authors: Haddaway, Neal R., Verhoeven, Jos T.A.
• Format: Article
• Language: English
• Subjects: Archiving; Carbon; Collaboration; Ecological monitoring; Ecological research; Evidence synthesis; experimental design; Land use planning; Literature reviews; Meta-analysis; Original Research; reliability; Reproducibility; research legacy; Statistical analysis; Studies; susceptibility to bias; Systematic review; transparency
• ISSN: 2045-7758; 2045-7758
• DOI: 10.1002/ece3.1722

Despite the scientific method's central tenets of reproducibility (the ability to obtain similar results when repeated) and repeatability (the ability to replicate an experiment based on methods described), published ecological research continues to fail to provide sufficient methodological detail to allow either repeatability of verification. Recent systematic reviews highlight the problem, with one example demonstrating that an average of 13% of studies per year (±8.0 [SD]) failed to report sample sizes. The problem affects the ability to verify the accuracy of any analysis, to repeat methods used, and to assimilate the study findings into powerful and useful meta‐analyses. The problem is common in a variety of ecological topics examined to date, and despite previous calls for improved reporting and metadata archiving, which could indirectly alleviate the problem, there is no indication of an improvement in reporting standards over time. Here, we call on authors, editors, and peer reviewers to consider repeatability as a top priority when evaluating research manuscripts, bearing in mind that legacy and integration into the evidence base can drastically improve the impact of individual research reports. Despite the widely accepted need to ensure that scientific research is both reproducible and repeatable, currently published ecological research still fails to provide sufficient methodological detail. Recent systematic reviews highlight the extent of the problem, which affects the ability to verify the accuracy and assess the reliability of any analysis, to repeat methods used, and to assimilate the study findings into powerful and useful meta‐analyses. We call on authors, editors and peer‐reviewers to consider repeatability as a top priority when evaluating research manuscripts, bearing in mind that legacy and integration into the evidence base can drastically improve the impact of individual research reports.