Gratitude Journals: What Two Meta-Analyses Measured, and Against What
Evidence Review · The Reading Room, Vol. 31 · 9 min read · The actual effect sizes, why the comparison group changes the answer, and what the authors themselves recommended
Based on the published research of Jennifer S. Cheavens, PhD — Professor, Department of Psychology, The Ohio State University. Faculty profile
Do gratitude journals work? Two independent meta-analyses covering more than 5,000 participants agree the effect is real and small, and that it shrinks when compared against any other activity rather than against nothing.
The instruction is everywhere and it is always given with confidence. Three things a night. It rewires how you see your life.
Two meta-analyses have now pooled the trials. Both found something, both found it small, and both found the size of it depended almost entirely on what the journalling was being compared against. That second detail is absent from nearly every page recommending the practice, and it is the one that determines whether the recommendation is honest.
The short answer:
- Cregg and Cheavens pooled 27 studies covering 3,675 participants and found a small effect on symptoms of depression and anxiety: g = −0.29 at post-test.¹
- At follow-up that effect was g = −0.23, so it did not vanish, and it did not grow.¹
- Effects were larger against waitlist controls than against active ones, which the authors read as a possible placebo influence.¹
- Davis and colleagues, pooling 26 articles across 32 samples, found d = .31 for psychological well-being against a measurement-only control, falling to d = .17 when compared against another activity.²
- Both author teams concluded the promise is real and limited, and Cregg and Cheavens recommended interventions with stronger evidence for anyone specifically trying to reduce depression or anxiety.¹²
Scope and method of this review
This piece covers whether structured gratitude exercises, of the kind sold as a nightly journalling practice, produce measurable change in mood and well-being. It draws on the two largest meta-analyses of the question, published in 2016 and 2021.¹²
Sources were selected on one criterion: pooled analyses of randomised trials rather than individual studies, because single gratitude trials are small, numerous, and easy to select from in whichever direction a writer prefers.
What this review does not cover: gratitude as a disposition rather than an exercise, which is a separate literature; gratitude expressed to another person, such as letter-writing, which several trials treat differently from private listing; and whether journalling of any other kind has value, which these trials did not test.
The effect sizes, and what they are measured against
| Analysis | Compared against | Finding | Limitation |
|---|---|---|---|
| Cregg & Cheavens 2021, 27 studies, n = 3,675 | Mixed controls, post-test | g = −0.29 on depression and anxiety symptoms | Pools two distinct symptom domains into one estimate |
| Cregg & Cheavens 2021, follow-up | Mixed controls, later timepoint | g = −0.23, effect persists at reduced size | Follow-up intervals varied between trials |
| Cregg & Cheavens 2021, moderation | Waitlist vs active control | Larger effects against waitlist at both timepoints | Consistent with placebo influence, not proof of it |
| Davis et al. 2016, 26 articles, 32 samples | Measurement-only control | d = .31 for psychological well-being | A control group that does nothing sets a low bar |
| Davis et al. 2016 | Alternative-activity control | d = .17 for psychological well-being | Roughly half the effect once anything else is compared |
| Davis et al. 2016 | Alternative-activity control | d = .11 for anxiety and stress | Confidence interval crosses close to zero |
Sources: Cregg & Cheavens (2021), Journal of Happiness Studies 22:413-445; Davis et al. (2016), Journal of Counseling Psychology 63(1):20-31.¹²
The pattern across both analyses is the same. Against nothing, gratitude journalling produces a small benefit. Against almost any other structured activity, that benefit roughly halves.
For a woman being told to add a nightly practice to a week that is already full, the comparison group is not an academic detail. It is the entire question, because her real alternative is never nothing. It is the ten minutes spent on something else.
If the ten minutes would be better spent seeing what is actually on the list, that is a different exercise. The Quiet Audit is a private ten-minute pass through the week as it currently stands. Start here.
Key terms
Meta-analysis is a study of studies: results from many separate trials are pooled statistically to produce one estimate, which is more stable than any individual trial and inherits the weaknesses of all of them.
Effect size (reported here as g or d) expresses how large a difference is in standardised units rather than in points on a questionnaire. By long convention, around 0.2 is small, 0.5 is moderate and 0.8 is large. Every figure in this review sits at or below 0.31.
Waitlist control is a comparison group that does nothing and is offered the intervention later. It measures the intervention plus the effect of doing something at all.
Active or alternative-activity control is a comparison group given a different task of similar length and effort. It is the harder and more informative test, because it isolates what the gratitude content specifically contributed.
How strong is this evidence?
Moderate. Two independent meta-analyses, published five years apart, covering more than 5,000 participants between them, agree on both the direction and the approximate magnitude. That is genuine replication. What they agree on is that the effect is small.¹²
What the grade does not cover: the underlying trials were run overwhelmingly on college students rather than on working adults in midlife, and Davis and colleagues name that limitation themselves.² Dosage varied little across studies, so nobody has established whether more or longer would do more. And an effect measured on a questionnaire at four weeks is not the same claim as a change in how a life feels at fifty-two.
What the authors actually recommended
Both teams were more restrained than the practice’s reputation suggests, and both said so in print.
Davis and colleagues wrote that the first fruits of gratitude interventions are in, and that they show positive but limited promise, and recommended tempering enthusiasm until longer and more powerful interventions demonstrate stronger evidence.²
Cregg and Cheavens went further, because their outcome was clinical rather than general well-being. Having found the small effect described above, they recommended that individuals seeking to reduce symptoms of depression and anxiety engage in interventions with stronger evidence of efficacy for those symptoms.¹
That is a specific and useful instruction. It does not say the practice is worthless. It says it should not be the intervention someone relies on when the target is a clinical symptom, and it identifies the failure mode: a woman using three lines a night as her whole answer to something that warranted a different one. The distinction between thinking that moves and thinking that circles is a more useful place to start for most of what shows up at eleven at night.
When this belongs with a doctor
A nightly writing exercise is not a treatment for depression or an anxiety disorder, and the authors of the larger meta-analysis said as much.¹ Low mood present most of the day for two weeks or more, loss of interest in things ordinarily enjoyed, a change in sleep or appetite that has settled rather than passed, or difficulty functioning at work or at home are reasons to book an appointment rather than reasons to write more lines. Effective treatments exist and have considerably larger effect sizes than anything in the table above.
The one move
The one move
If you keep the practice, change one thing: write what you did rather than what you are grateful for. One line, specific, from today. It costs the same ten minutes, it is not vulnerable to the comparison-group problem, and it produces a record you can read back in March.
If the ten minutes would be better spent seeing the week whole, the Quiet Audit is a private pass through what it is actually costing. Start here.
The source for this piece
Jennifer S. Cheavens, PhD
Professor in the Department of Psychology at The Ohio State University, and co-author of the largest meta-analysis of gratitude interventions on depression and anxiety symptoms.
What we read: Cregg, D. R., & Cheavens, J. S., Journal of Happiness Studies (2021).¹
Where to follow her work: Faculty profile
Who else has measured this
Two meta-analyses converge on a small effect. It is worth knowing who established gratitude as a subject of serious study in the first place.
Robert A. Emmons, PhD (Wikipedia · faculty profile · Greater Good profile), Professor of Psychology at the University of California, Davis, is the field’s most cited researcher on gratitude and co-authored the original trials the later meta-analyses pooled.² His subsequent work has been consistent with the more modest, comparison-dependent effect this review reports.
The founding researcher and the two later meta-analyses land in the same place: real, small, and dependent on what it is compared against.
Against almost any other structured activity, that benefit roughly halves.
Questions this review is asked
Do gratitude journals work?
They produce a small, replicated benefit. Two meta-analyses put it between g = −0.29 and d = .31 against inactive controls, dropping to around d = .17 against another activity.¹² That is a real effect at the low end of the scale, not the change most articles promise.
Why does the comparison group matter so much?
Because a control group that does nothing measures the practice plus the benefit of doing anything at all. A control group given a different task of similar length isolates what gratitude specifically added, and by that measure the effect roughly halves.²
Is a small effect the same as no effect?
No. Both analyses found the effect reliably and it persisted at follow-up in the larger one.¹ Small means the average person changes a little. For a free exercise with no side effects, small can still be worth having, provided nobody is relying on it for something larger.
Would doing it longer or more often help?
Unknown. Davis and colleagues specifically noted that dosage varied little across the trials, so the literature cannot answer whether more would do more.² Anyone claiming it would is going beyond the data.
Does this apply to women in their fifties?
That is the largest gap in the evidence. The trials ran mostly on college students, which the authors flag.² Extending a student sample to a working woman managing a household and a parent is an extrapolation, and it should be described as one.
Is writing to another person different from listing privately?
Possibly, and the trials treat the two formats differently. This review pooled the literature as the meta-analyses did and did not separate them, so it cannot make a claim either way.
On the researcher. Jennifer Cheavens is a clinical psychologist, and her own meta-analysis recommends stronger-evidence interventions for anyone specifically targeting depression or anxiety symptoms; nothing here should be read as a substitute for that guidance.
Sourcing, disclosure and disclaimers
This is not medical advice. Blue Leaf Journal publishes general information for a general readership. Nothing here is a diagnosis, a treatment recommendation, or a substitute for care from a clinician who knows your history.
We are not clinicians. Blue Leaf Journal is an independent publication. We read published research and translate it. The findings belong to the researchers and institutions named and linked above; the plain-English rendering is ours, and so is any error in it.
No affiliation and no endorsement. Blue Leaf Journal is not affiliated with Jennifer S. Cheavens, Robert A. Emmons, The Ohio State University or the University of California, Davis. None of them has reviewed, approved or endorsed this article, and none of them is responsible for it. They are cited because their published work is the evidence for what it says.
No commercial relationship. Nobody named above paid for or was paid for this coverage. There are no affiliate links, no sponsored placements and no gifted products in this article.
How this was checked. Every figure above is drawn from the primary papers, each linked in the references, and can be verified there. Sources were checked on 6 September 2026. If you find something we have got wrong, write to miriamalderton@blueleafjournal.com and we will correct it and say that we did.
References
1. Cregg, D. R., & Cheavens, J. S. (2021). Gratitude Interventions: Effective Self-help? A Meta-analysis of the Impact on Symptoms of Depression and Anxiety. Journal of Happiness Studies, 22, 413-445. Read the meta-analysis
2. Davis, D. E., Choe, E., Meyers, J., et al. (2016). Thankful for the Little Things: A Meta-Analysis of Gratitude Interventions. Journal of Counseling Psychology, 63(1), 20-31. Read the meta-analysis
Miriam Alderton is Research Editor at Blue Leaf Journal. She reads the methods section first and the abstract last, and every figure in this piece is linked to its source above.
Written by Miriam Alderton, Research Editor, for Blue Leaf Journal. Updated: 29 August 2026.
