Positive psychology interventions: a meta-analysis of randomized controlled studies
Restricting to randomised controlled studies only, positive psychology interventions produced small effects on subjective well-being (0.34), psychological well-being (0.20) and depression (0.23) - effects that remained small but statistically significant at three to six months' follow-up.
The stricter cousin of Sin & Lyubomirsky - randomised trials only, and smaller numbers as a result.
Bolier and colleagues included 39 studies totalling 6,139 participants, restricted to randomised controlled designs, measuring subjective well-being, psychological well-being and depression.
- Subjective well-being: standardised mean difference 0.34
- Psychological well-being: 0.20
- Depression: 0.23
At three to six months’ follow-up, effects were still small but still significant for both well-being measures - which is the more meaningful result, because most interventions of this kind decay to nothing.
Then the honest part. The authors report indications of publication bias, considerable variation in study quality, and high heterogeneity. Among the factors that moderated the effect on depression was low study quality - worse studies produced better results.
That last point is worth sitting with. It is the single best reason to distrust any confident claim about how much journaling will change your life, including confident claims made by people selling journals.
The authors found indications of publication bias and note that study quality varied considerably, with low-quality designs among the factors that made interventions look more effective. Heterogeneity was high. They call for additional high-quality peer-reviewed studies before the evidence base can be considered strong.
Articles citing this study
Summarized by Tiago Martins · Last updated: · How we write these
