Loading Motifs…
An evening prompt asks for three things that went well today, and why. It is one of the most famous exercises in positive psychology — and three controlled replications found no depression benefit beyond placebo.
0 of 3Controlled replications that found a depression benefit beyond placebo or control (Mongrain 2012, Woodworth 2017, Gold 2023). The 2005 original reported six-month effects among the 71% who finished.
Built from published experiments and company reports across several products. Nobody has checked the screens on a dated day, so read each number with the caveat printed beside it.
An evening prompt asks for three things that went well today, and why. It is one of the most famous exercises in positive psychology — and three controlled replications found no depression benefit beyond placebo.
People adapt quickly to good things and drift back to their usual level of happiness, so it was an open question whether any exercise could make people lastingly happier (Seligman, Steen, Park and Peterson).
This play is not tied to a full playbook. Use it for this decision alone.
The original study found lasting gains in happiness and fewer depressive symptoms.
STRONGWhat it does not show Only the 411 of 577 participants (71%) who completed all five follow-ups were analysed. Convenience sample tilted toward people who wanted to be happier; mostly educated, white and financially comfortable; mildly depressed on average. The happiness scale was new; figure effect sizes are labelled with an unclear statistic and described as differences from baseline. In Figure 1, the three good things group's pre-test happiness bar is below the placebo group's. The continuation analysis is reported as ANOVAs with adherence and continuation as factors, not a formal mediation test. Reviewed as the authors' final manuscript hosted by Greater Good, not the typeset article.
Positive Psychology Progress: Empirical Validation of Interventions Seligman et al. · 2005
A replication with a better control found no depression benefit beyond it.
SUPPORTINGWhat it does not show Read as an abstract; sample size, randomization details and separate results for three good things were not available. A null against control, not evidence of harm.
Do positive psychology exercises work? A replication of Seligman et al. (2005) Mongrain and Anselmo-Matthews · 2012
Use the play for this decision alone, or combine it with an existing product strategy after checking for conflicts.
Stating a mental-health benefit as the app's effect is a health claim several replications do not support, and may draw app-store or regulatory review.
Every group improves right after the week, placebo included, so a before-and-after of your own users cannot show the prompt did anything.
Entries describe people and events from the user's day; sharing them, even anonymously, exposes that content.
Plus everything above, as plain text for your coding agent, with a short instruction on top: check the fit, say what the evidence doesn’t support, and adapt it rather than copy it. Paste it and you get an answer, no prompt to write.
The 2005 original, three controlled replications and one uncontrolled healthcare-worker programme · evidence reviewed 25 September 2026.
Explore company-inspired strategies that connect your product, experience, and growth.
Browse the playbooks ↗An identical replication found effects indistinguishable from placebo.
STRONGWhat it does not show The trial report was read as an abstract; design details come from the authors' data paper. Only 72 of 295 participants completed the measures on all occasions. 85% female; separate figures for three good things were not read.
Web-Based Positive Psychology Interventions: A Reexamination of Effectiveness Woodworth et al. · 2017
A text-message version briefly lifted mood, with no depression effect.
STRONGWhat it does not show The authors report that about 458 participants per group would have been needed to detect a 1-point PHQ-9 change; the trial was underpowered for depression. Single department during COVID-19 surges; participants not blinded; three evenings a week rather than nightly. Gift-card drawing for completing surveys and texts.
"\"Three Good Things\" Digital Intervention Among Health Care Workers: A Randomized Controlled Trial" Gold et al. · 2023
Healthcare workers improved after a 15-day email programme, with no control group.
SUPPORTINGWhat it does not show Uncontrolled prospective pre-post design; the placebo-group improvements in E1, F1 and F2 show why that matters. Recorded as peer_reviewed_research because the ladder has no rank for uncontrolled intervention studies. Results are limited to those who completed assessments. The larger subgroup effects are also consistent with regression toward the mean. Bundles the prompt with emotion tagging and optional sharing.
Forty-five good things: a prospective pilot study of the Three Good Things well-being intervention in the USA for healthcare worker emotional exhaustion, depression, work-life balance and happiness Sexton and Adair · 2019