Specification Curve Analysis
Definition
An analytic approach that consists of identifying, calculating, visualising and interpreting results (through inferential statistics) for all reasonable specifications for a particular research question (see Simonsohn et al. 2015). Specification curve analysis helps make transparent the influence of presumably arbitrary decisions during the scientific progress (e.g., experimental design, construct operationalization, statistical models or several of these) made by a researcher by comprehensively reporting all non-redundant, sensible tests of the research question. Voracek et al. (2019) suggest that SCA differs from multiverse analysis with regards to the graphical displays (a specification curve plot rather than a histogram and tile plot) and the use of inferential statistics to interpret findings.
References:
- Simonsohn, U., Simmons, J. P., & Nelson, L. D. (2015). Specification curve: Descriptive and inferential statistics on all reasonable specifications. http://sticerd.lse.ac.uk/seminarpapers/psyc16022016.pdf
- Simonsohn, U., Simmons, J. P., & Nelson, L. D. (2020). Specification curve analysis. Nature Human Behaviour, 4(11), 1208–1214. https://doi.org/10.1038/s41562-020-0912-z
- Voracek, M., Kossmeier, M., & Tran, U. S. (2019). Which Data to Meta-Analyze, and How? Zeitschrift Für Psychologie. https://doi.org/10.1027/2151-2604/a000357
Originally drafted by: Bradley Baker
Reviewed by: Tina B. Lonsdorf, Sam Parsons, Charlotte R. Pennington
Suggest an improvement
Spotted an error, a broken link, or think this entry is missing something? Let us know — this goes directly to the glossary team.