Split View: 감사 일기는 정말 효과가 있을까 — 축복 세기 실험과 정직한 효과 크기

Does the Gratitude Journal Actually Work — The Counting Blessings Experiment and Honest Effect Sizes

Introduction — Tracing the Source of the Most Common Prescription

"Write down three things you are grateful for every day." It is hard to avoid this prescription in self-help content. And because it is so common, doubt comes naturally. Is there actually any evidence behind that familiar line?

There is. The source is "Counting blessings versus burdens," published in JPSP in 2003 by Robert Emmons and Michael McCullough. The reason this paper closes out Psychology, Straight from the Papers is that it is the best possible case for practicing the method of this series — reading famous studies without the hype. The conclusion lands neither on "it works" nor on "it does not work," but on it works, conditionally, and at a humble size.

The Three Experiments in the Original Paper

Experiment 1 — weekly journals, 10 weeks. Some 190 college students were randomly assigned to three conditions. Once a week, the gratitude condition wrote down "5 things from the past week you were grateful for," the hassles condition "5 things that annoyed you," and the neutral condition "events that happened that week." After 10 weeks, the gratitude condition rated their lives more positively than the hassles condition, felt more optimistic about the coming week, and — most interestingly — logged about 1.5 more hours of exercise per week. They also reported fewer physical symptoms.

Experiment 2 — daily journals, 2 weeks. Raising the frequency from weekly to daily made the difference in positive affect sharper. This time, though, the exercise effect did not reappear, and when the comparison condition was switched to "writing ways you are better off than others (downward comparison)," the gap with the gratitude condition shrank on some measures. What you compare against begins to steer the result.

Experiment 3 — adults with neuromuscular disease, 3 weeks. Adults living with chronic pain were assigned daily gratitude records. The gratitude condition reported more positive affect and life satisfaction than controls, felt more connected to other people, and improved even in sleep quality and sleep duration. The differences were confirmed in observer reports rated by spouses. It is this third experiment that gives the paper a weight beyond a small-pleasures study for healthy college students.

The 20 Years Since — The Effect Is Real, but Smaller Than Advertised

Stop reading here and it looks like a cure-all. Following the rules of the series, we balance it against the follow-up literature.

The sober numbers from meta-analyses. Syntheses such as the 2016 meta-analysis by Davis and colleagues put the average effect of gratitude interventions at small to below-medium. The estimate gets humbler the more the comparison is not "doing nothing" but an active control (keeping an ordinary journal, for example). The 2010 review by Wood and colleagues likewise flagged design limits in the early studies (a patchwork of measures, short follow-ups). A gratitude journal is not a treatment for depression, and it does not fight in the same weight class as an intervention like exercise.

What remains even so. The direction of the effect is consistently positive across dozens of studies, the cost is 3 minutes a day, and there are no known side effects. A small effect × a near-zero cost is the same arithmetic we saw in the growth mindset installment. Adjust the expectation from "this will change my life" to "this shifts the baseline of my mood by a few degrees," and the gratitude journal becomes a scientifically honest habit.

The mechanism is plausible, too. Human attention defaults to threats and problems (negativity bias), and we adapt to good things so fast that they stop registering (hedonic adaptation). A gratitude record is attention training that pushes against both defaults: deliberately hunting for the good (attention reallocation) and lingering over it as you write (savoring). In the language of the wandering mind study, it is practice in bringing attention back to a good present — even for just 3 minutes a day.

How to Keep It from Going Stale — Conditions of Use from the Research

The greatest enemy of the gratitude journal is turning it into a formality. The moment you copy-paste "family, health, coffee," the effect is gone. The conditions of use distilled from follow-up studies look like this.

• Specificity is everything. Not "grateful for a colleague" but "grateful to O, who spent 30 minutes on my code review today." The resolution of one item matters more than the number of items.
• 1-3 times a week beats every day. In an experiment by the Lyubomirsky team, the once-a-week group reportedly gained more happiness than the three-times-a-week group. The interpretation: the moment it becomes a duty, the savoring disappears. Sparse and rich beats frequent and thin.
• Items about people carry the largest effects. The effect grows when the object of gratitude is a person rather than a thing — and when, now and then, the record turns into expression (a short message of thanks). In the gratitude visit experiment by Seligman, expression is what produced the largest short-term effect as well.
• Mix in mental subtraction. Imagining "what if this had never happened" — mental subtraction — is a variant technique that resets hedonic adaptation. The same thing becomes vivid again once you route it through a world in which you lost it.

Closing the Series — Five Habits for Reading Papers

With this installment, the 10-part "Psychology, Straight from the Papers" series comes to a close. Five lines for the reading habits that ran through all ten.

1. The more famous the conclusion, the sooner you should turn to the method section of the original paper. The marshmallow, the 10,000 hours — the myth always came apart in the details of the method.
2. Effect size over p-value, and cost-effectiveness over effect size. A small effect is useful when it is cheap, and a large effect that will not replicate might as well not exist.
3. The number of papers is not the amount of evidence. As the publication bias simulation showed, one preregistered large-scale replication outweighs hundreds of cherry-picked papers.
4. The moment you ask under what conditions, a fight becomes a discovery. The ending of the income and happiness debate is the model.
5. And practice always arrives as a system. The bridge between knowing and doing was never willpower but design — in every installment of this series.

Guide to Reading the Originals

• Original paper: Emmons, R. A., & McCullough, M. E. (2003). Counting blessings versus burdens: An experimental investigation of gratitude and subjective well-being in daily life. Journal of Personality and Social Psychology, 84(2), 377-389.
• The sober synthesis: Davis, D. E., et al. (2016). Thankful for the little things: A meta-analysis of gratitude interventions. Journal of Counseling Psychology, 63(1), 20-31.
• Review: Wood, A. M., Froh, J. J., & Geraghty, A. W. A. (2010). Gratitude and well-being: A review and theoretical integration. Clinical Psychology Review, 30(7), 890-905.

Reading tip: in the 2003 paper, the results table of Experiment 3 (the patient sample) is the core. Reading it against the effect patterns in the healthy student samples (Experiments 1 and 2) is good practice in watching how the sample changes the conclusion. And if you want to actually start a gratitude journal — tonight, one specific item is enough.