Compared to the wealth of debate surrounding replicability and transparency, relatively little attention has been paid to the issue of generalisability – the extent to which research findings hold across different samples, cultures, and other parameters. Existing research suggests that researchers in psychology are prone to generalisation bias, relying on narrow samples (e.g. drawn predominantly from US or European undergraduate samples) to draw broad conclusions about the mind and behaviour. While recent attempts to address generalisability concerns have been made – such as journals requiring explicit statements acknowledging constraints on generality – addressing this bias at root, and developing truly generalisable methods and results, requires a deeper understanding of how researchers perceive generalisability in the first place.

 

In the current study, Schmidt et al. (2023) tackle the issue of cross-cultural generalisability using four large-scale international studies that are being conducted as part of the Psychological Science Accelerator (PSA) – a globally distributed network of researchers in psychology that coordinates crowdsourced research projects across six continents. Specifically, participants (who will be PSA research members) will estimate the probability that an expected focal effect will be observed both overall and within regional subsamples of the PSA studies. They will also predict the size of these focal effects overall and by region.

 

Using this methodology, the authors plan to ask two main questions: first whether researchers can accurately predict the generalisability of psychological phenomena in upcoming studies, and second whether certain researcher characteristics (including various measures of expertise, experience, and demographics) are associated with the accuracy of generalisability predictions. Based on previous evidence that scientists can successfully predict the outcomes of research studies, the authors expect to observe a positive association between predicted and actual outcomes and effect sizes. In secondary analyses, the authors will also test if researchers can predict when variables that capture relevant cultural differences will moderate the focal effects – if so, this would suggest that at least some researchers have a deeper understanding as to why the effects generalise (or not) across cultural contexts.

 

The Stage 1 manuscript was evaluated over two rounds of in-depth review. Based on detailed responses to the reviewers' comments, the recommender judged that the manuscript met the Stage 1 criteria and therefore awarded in-principle acceptance (IPA).

 

URL to the preregistered Stage 1 protocol: https://osf.io/vwqsa (under temporary private embargo)
 
Level of bias control achieved: Level 6. No part of the data or evidence that will be used to answer the research question yet exists and no part will be generated until after IPA.
 
List of eligible PCI RR-friendly journals:
 

 

 

1. Schmidt, K., Silverstein, P. & Chartier, C. R. (2023). Registered Report: Researcher Predictions of Effect Generalizability Across Global Samples. In principle acceptance of Version 3 by Peer Community in Registered Reports. https://osf.io/vwqsa