Using OSF to increase the accessibility, credibility, transparency, and reproducibility of our research project: From preregistrations to preprints

Open Research objectives/practices

The objectives of documenting, registering, archiving, and sharing our research project on OSF (Open Science Framework) were threefold. Firstly, by documenting our projects’ sampling and data collection procedures, descriptions of our datasets, and codebooks, we aimed to be transparent about our research processes and methodologies. Secondly, by preregistering the designs of our individual studies (i.e., each study’s research questions, hypotheses, methods, and analysis plan), we intended to make our studies more reproducible. Thirdly, by sharing our data, preprints, and publications, we aspired to make our research materials and the outputs of our research freely accessible. By applying these various principles and behaviours, we strived to be as transparent and open as possible across all phases of the research cycle, so that others can appropriately evaluate our work.

Introduction

We created a public project on OSF for our Caribbean Research Program on “Father absence and consequences for reproductive behavior and psychosocial development among Caribbean, Caribbean-Dutch and native Dutch youth”. We started with uploading a project-preregistration, in which we summarized the full project and included our main questions, hypotheses, and analyses. Subsequently, we preregistered each individual study, and shared its preprint on PsyArXiv upon submission. Of note, we did so after consulting Sherpa Romeo to be sure our intended journal allowed preprints. We also added information on our sampling and data collection procedures, descriptions of our datasets, and codebooks for both our quantitative and qualitative data on OSF. In the near future, we also plan to write a data paper and to make our data available to other researchers.

Motivation

In the long tradition of quantitative research into father absence, this family form has been negatively associated with multiple domains of offspring development. Unfortunately, most of this research has not been preregistered. By preregistering our project and individual studies, we were able to be transparent about our planned analyses before we conducted them, which benefited clarity among collaborators and prevented questionable research practices (e.g., not reporting all variables measured, HARKing: hypothesizing after the results are known, p-hacking). In doing so, we treated our participants’ data in an ethical way. That is, we did not exhaustively search for statistical significance, but tested and reported the analyses we planned, which were based on our theoretical and conceptual models. In addition, by sharing our preprints independently upon submission, we aimed to benefit from a wider range of feedback than afforded through peer review, and to share our findings and interest as soon as possible.

Lessons learned

By preregistering our studies, I learned to design and evaluate our planned research at an early stage of the research process. Though creating a (first) preregistration took some time and effort, it saved time once all collaborators evaluated our research plans and we could conduct the research as preregistered. I am very grateful to my supervisor dr. Tina Kretschmer, who teached me about the importance of high quality study designs and the publication of null findings (e.g., through preprints). Specifically regarding our research project, father absence not being related to offspring outcomes, is also a very valuable result and worth publishing. I hope that the so-called “disconnection between what is good for scientists and what is good for science” will disappear in the near future.

URLs, references and further information

https://osf.io/txzvc/