In the traditional publishing model, people do research, write a paper describing their results, and submit it to a journal. The paper undergoes peer review (see the Peer Review module for more information), and the reviewers and editor decide whether to accept the paper as is, request revisions, or reject the paper.
A problem many people have noted with this model is that reviewers and editors will be tempted to judge a paper not by how good the research itself was, but by how good the results are. If a study is nicely designed but the results don't support their preferred theory, they might find excuses to criticize the paper and block it from being published. Or if the effects of interest are not statistically significant, they might reject the paper. In the worst-case scenario, this may lead researchers themselves to conduct bad research practices (see the "Research Practices" module) in order to get better-looking results. For example, researchers might look for excuses to throw out inconvenient data in order to get results that look more convincing and thus will be more likely to be published.
One way people have addressed this problem recently is by creating a new publishing format: registered reports. In registered reports, unlike traditional publishing, a paper is not judged by how good the results are; it is only judged by how good the research plan is. To do a registered report, a researcher first submits a "Stage 1 manuscript", which describes the research question, motivation, and the plans for how the research question and hypotheses will be tested (in other words, it's a paper with only the Introduction and Methods section). It's more like a proposal than a full paper. The reviewers and editor evaluate this to see if the study plan is good; they may suggest ways to improve the plan. Eventually, if the reviewers and editor are convinced that the plan is good, they conditionally accept the paper: this means that they are promising to accept the final paper, no matter how the results turn out. Then the researcher goes and does the actual research, and the researcher does not need to be tempted to manipulate the results, because the researcher already knows the paper will be published even if the results don't look good.
Another benefit of registered reports is that a researcher can avoid spending a long time doing a crappy study that reviewers are going to criticize; the researcher can get the reviewers' criticisms first, before investing a lot of time in the study, and then make the necessary modifications before wasting a lot of time doing the study in a way that reviewers will not accept.
Registered reports are a powerful and exciting new way to improve the way research is done. In linguistics and psycholinguistics they are not very widely used yet (even I, a big supporter of registered reports in theory, have only ever written one once), but each year there are more and more journals starting to offer them.
Imagine you would like to submit a registered report to a journal in your research area. Is there any journal in your research area that offers registered reports? Which one?
(The Center for Open Science maintains an information page about registsered reports. In addition to having a lot of information and links about registered reports, there is also a tab there labelled "Participating journals", which includes a list of journals that offer a registered reports format.)
What do you think about registered reports? Can you foresee any other benefits to them, beyond the ones I described in the reading before this self-test? Can you foresee any problems with registered reports, or any reasons some of your research might not be appropriate for a registered reports format?
When you have finished these activities, continue to the next section of the module: "Predatory publishers".
by Stephen Politzer-Ahles. Last modified on 2021-05-05. CC-BY-4.0.