Study type: Critical and conceptual review.
In plain language
A treatment can help people on average while helping some much more than others, doing little for others, or working differently in different contexts. This review asks whether clinical psychology has relied too heavily on group averages when its practical task is to help a particular person, couple, or family. Hayes and colleagues identify the ergodic error: assuming that a relationship found across people also describes how the same processes operate within a particular person over time.
The authors propose idionomic analysis as a way forward. Researchers first model the dynamics of each person’s own experiences and behavior. They then look for shared patterns and broader principles, retaining generalizations that improve understanding of individual cases. The aim is to make personalization an empirical process: measure what matters to this person, test candidate relationships, and revise the treatment plan as evidence accumulates.
This is a critical and conceptual review, rather than a systematic review or a new treatment trial. It connects a historical argument about psychology’s reliance on population-ranking methods with evidence from intensive longitudinal studies and a practical overview of methods for studying individual change.
Key points
- Averages can conceal opposing relationships. A small or zero pooled association can combine meaningful positive relationships for some people with negative relationships for others. The distribution of individual effects matters alongside the pooled estimate.
- General evidence remains useful. The authors describe group findings as starting hypotheses or Bayesian priors, and recognize their value for appropriately framed population-level decisions. Their argument concerns when such findings justify conclusions about particular people.
- Repeated measurement supports personalization. Experience sampling, daily diaries, and other repeated observations can connect personally relevant processes with outcomes at the timescale on which they occur.
- Methods answer different questions. The review compares i-ARIMAX, iBoruta/tsBoruta, GIMME and S-GIMME, growing self-organizing maps, multilevel vector autoregression, random-effects meta-analysis, and client-generated networks such as PECAN and PLAN.
- Assess the spread of effects. The review emphasizes the between-person standard deviation and prediction interval when assessing how well a pooled effect represents individual effects. It cautions that I² alone does not describe their absolute spread.
What the paper does and does not establish
- The review argues for a research and clinical framework; it does not establish that every idionomic method improves treatment outcomes or that every average effect fails to generalize.
- Methods differ in their assumptions, data requirements, interpretability, and ability to handle trends and temporal dependence. A model should fit both the question and the available observations.
- Within-person associations are candidates for understanding and intervention. Establishing that changing a process causes improvement requires additional evidence, including appropriate experimental designs.
How to cite
APA 7
Hayes, S. C., Ciarrochi, J., Sahdra, B. K., Hofmann, S. G., Ong, C. W., & Hernández, C. (2026). One size fits none: A call for an idionomic revolution. Clinical Psychology Review, 129, Article 102800. https://doi.org/10.1016/j.cpr.2026.102800
BibTeX
@article{hayes2026one,
title = {One size fits none: A call for an idionomic revolution},
author = {Hayes, Steven C. and Ciarrochi, Joseph and Sahdra, Baljinder K. and Hofmann, Stefan G. and Ong, Clarissa W. and Hernández, Cristóbal},
journal = {Clinical Psychology Review},
year = {2026},
volume = {129},
pages = {102800},
doi = {10.1016/j.cpr.2026.102800}
}
Related work
- Evolving an idionomic approach to processes of change
- Time-series machine learning: the tsBoruta tutorial
- Heterogeneity in psychological flexibility and well-being
- All publications by Joseph Ciarrochi
The hosted PDF contains the published article followed by two pages of reference corrections. The correction notice states that these changes do not alter the article’s claims. Summary prepared from the full paper on 29 September 2026. Please cite the published article linked above.