Studies // Cited work
Written emotional expression: Effect sizes, outcome types, and moderating variables
Smyth, J. M. (1998). Written emotional expression: Effect sizes, outcome types, and moderating variables.
Level A Meta-analysis doi:10.1037/0022-006X.66.1.174
What this study supports
Supports the claims that writing was followed by improved reported physical health, psychological well-being, physiological and general functioning but not health behaviours, that it increased immediate distress unrelated to health outcomes, and that several moderators were identified. Fixed-effects synthesis, later criticised as limiting generalisability.
- Who was studied
- Studies of a written emotional expression task in healthy participants; the number of studies is not stated in the abstract.
- Design
- Meta-analysis
- Level of evidence
- Syntheses of controlled studies: meta-analyses and systematic reviews of experiments and trials.
Claims that rest on it
Each claim states what the site asserts, at what strength, and with what caveat.
Across 146 randomised studies, experimental disclosure had a positive and significant but small average effect (r = .075), with several moderators; an earlier synthesis found benefits across reported physical health, well-being and functioning but not health behaviours.
- Scope
- Randomised studies of writing or talking about personal, meaningful topics, largely in non-clinical samples.
- Caveat
- The average conceals heterogeneity, moderators are not itemised in the abstract, and a small average does not predict benefit for an individual.
Last reviewed · Review due · expressive-writing-small-average-effect
Expressive writing tends to increase immediate distress, and in a research synthesis that increase was unrelated to later health outcomes.
- Scope
- Healthy participants in written emotional expression studies.
- Caveat
- One early synthesis with fixed-effects methods; the dissociation was measured across studies, not within individuals.
Last reviewed · Review due · writing-distress-unrelated-to-health-outcomes
This site reads the dissociation between immediate distress and later health outcomes as meaning that relief or discomfort at the end of a writing session is weak evidence of benefit; the inference is the site's interpretation and was not tested.
- Scope
- How readers should weigh how a writing session feels at its end.
- Caveat
- An interpretation reasoned from a study-level dissociation; no study tested end-of-session feeling as a predictor of benefit.
Last reviewed · Review due · session-end-feeling-weak-evidence-of-benefit