The Problem
Psychology’s replication crisis has generated considerable soul-searching and a wave of proposed remedies: preregistration, open data, larger samples, more robust statistics. Uher and colleagues (2025) argue that these responses, however well-intentioned, are addressing the wrong problem. The crisis, they contend, does not primarily stem from researcher misconduct or poor statistical practice. It stems from something more fundamental: the theoretical, conceptual, and measurement foundations of quantitative psychology are themselves deeply problematic. They call these the Questionable Research Fundamentals, or QRFs.
The Four Problem Areas
The paper is a collaborative effort by twelve researchers from across psychology, philosophy, measurement theory, and related disciplines, organised around four main areas.
The first concerns the scientific status of quantitative psychology itself. Contributors argue that psychology has embraced a form of scientism, pursuing the outward appearance of a hard science through quantification and statistical analysis, without adequately examining whether its fundamental concepts, measurement approaches, and philosophical assumptions are coherent or adequate for the phenomena being studied (Uher et al., 2025).
The second concerns psychometrics and what the paper carefully refers to as psychological “measurement” in inverted commas. The central claim is that psychological rating scales and questionnaires do not constitute genuine measurement in any sense comparable to physical measurement. Physical measurement requires that the thing being measured possesses the mathematical properties of quantity: equal intervals, a true zero, and additive structure. Psychological variables do not demonstrably possess these properties, and the numbers generated by rating scales cannot legitimately support the arithmetic operations routinely applied to them (Uher et al., 2025). Statistics and measurement, the paper insists, are different scientific activities serving different epistemic purposes, and conflating them has produced decades of quantitative results that appear more rigorous than they are.
The third area concerns the role of language in psychological inquiry. Much of psychology’s measurement relies on participants responding to verbal items, and the meaning of those items is not neutral. The semantic relationships among words in a questionnaire shape the statistical relationships among responses, meaning that many of the correlations psychology reports may reflect the structure of language rather than the structure of psychological reality (Uher et al., 2025). This connects directly to the earlier Arnulf and Smedslund findings reviewed in this series, where natural language analysis alone replicated patterns typically attributed to empirical data collection.
The fourth area concerns the generalisation of findings from samples to individuals. The paper addresses the ergodic fallacy directly: the mathematical assumptions required to justify inferring individual-level patterns from group-level statistics are not met by psychological phenomena, which are non-ergodic. When psychology claims that a finding about average group differences tells us something about how individuals function, it is making an inferential leap that its own mathematical foundations do not support (Uher et al., 2025).
The Conclusion
Uher and colleagues are blunt about what follows from this analysis. Simply minimising questionable research practices, as the reform movement has largely focused on, will not resolve psychology’s crises. The problems are located at a deeper level: in how constructs are defined, how measurement is conceptualised, how language is used, and how findings are generalised. Tackling these requires critical self-reflection and a fundamental rethinking of what doing science in psychology means (Uher et al., 2025).
Excerpt
Psychology’s replication crisis has been blamed on researcher misconduct and poor statistics. A sweeping 2025 paper by twelve researchers argues the real problem runs far deeper: the theoretical and measurement foundations of quantitative psychology are themselves fundamentally flawed.
Reference
Uher, J., Arnulf, J. K., Barrett, P. T., Heene, M., Heine, J.-H., Martin, J., Mazur, L. B., McGann, M., Mislevy, R. J., Speelmann, C., Toomela, A., & Weber, R. (2025). Psychology’s questionable research fundamentals (QRFs): Key problems in quantitative psychology and psychological measurement beyond questionable research practices (QRPs). Frontiers in Psychology, 16, 1553028. https://doi.org/10.3389/fpsyg.2025.1553028
