Open AccessPreregistration of exploratory research: Learning from the golden age of discovery
Author(s) -
Ulrich Dirnagl
Publication year - 2020
Publication title -
plos biology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 4.127
H-Index - 271
eISSN - 1545-7885
pISSN - 1544-9173
DOI - 10.1371/journal.pbio.3000690
Subject(s) - soundness , data science , biology , flexibility (engineering) , trustworthiness , open science , exploratory research , statistical power , scientific discovery , computer science , psychology , internet privacy , statistics , cognitive science , social science , mathematics , sociology , programming language
Preregistration of study protocols and, in particular, Registered Reports are novel publishing formats that are currently gaining substantial traction. Besides rating the research question and soundness of methodology over outstanding significance of the results, they can help with antagonizing inadequate statistical power, selective reporting of results, undisclosed analytic flexibility, as well as publication bias. Preregistration works well when a clear hypothesis, primary outcome, and mode of analysis can be formulated. But is it also applicable and useful in discovery research, which develops theories and hypotheses, measurement techniques, and generates evidence that justifies further research? I will argue that only slight modifications are needed to harness the potential of preregistration and make exploratory research more trustworthy and useful.