Best practice self commitment
Many standards in psychological research concerning the study design, data analysis, reporting and publishing have changed in the light of the so-called replication crisis. The following overview attempts to take this development into account to foster transparency, robustness, and integrity of our own research. My research team and I took the form of a voluntary commitment. The aim is to commit ourselves publicly and then work accordingly for all upcoming projects, starting Oct-21.
- We use pre-registration or registered reports, wherever sensible (for a current list of supporting journals).
- We estimate sample size prior to data collection, with gPower or other appropriate software, for example, pwrSEM or simsem in a SEM context.
- Our work adheres to the Ethical Principles of Psychologists and Code of Conduct of the American Psychological Association.
- We double-check all syntax (four-eyes-principle).
- We use dynamic fit indices, wherever possible.
- We check for inconsistencies of inference statistics with statcheck and degrees of freedom in SEM models.
- We make (co)authors contribution for all publications transparent using Contributor Roles Taxonomy (CRediT).
- We use the Quantitative Design Reporting Standards (JARS-Quant, section 3.5, p. 78ff.) and Qualitative Design Reporting Standards (JARS-Qual, section 3.14, p. 95ff.) as outlined in the publication manual of the American Psychological Association (7th edition). For meta-analyses, we use the PRISMA guidelines.
- We provide open syntax and data. We upload everything that is necessary to reproduce the results to OSF. For sensitive data that cannot be anonymized, we consider generating synthetic data (e.g., synthpop).
- We save output of the analyses (e.g., html files via knitr, for an example).