Does a Gratitude Journal Work? What the Research Actually Shows

Search for "gratitude journal" and you will find promises that a few lines a night will change your life. The research is more interesting than that, and more modest. Writing down what went well does seem to move well-being a little in the right direction. It is not a cure for anything, and the size of the effect depends heavily on what you compare it with.

Here is what the best known studies found, and what their numbers mean once you translate them out of statistics.

The study that started it: counting blessings versus burdens

In 2003, psychologists Robert Emmons and Michael McCullough published three randomized experiments in the Journal of Personality and Social Psychology. In the first, 192 undergraduates (after dropping incomplete data) were randomly assigned to write, once a week for nine weeks, either up to five things they were grateful for, up to five hassles, or five events that had affected them.

The gratitude instruction was simple: "There are many things in our lives, both large and small, that we might be grateful about." Students listed things like "waking up this morning" and "the generosity of friends."

At the end, the gratitude group rated their life as a whole at 5.05, against 4.67 for the hassles group and 4.66 for the events group. They reported fewer physical symptoms (3.03 versus 3.54 and 3.75) and more exercise.

Our computation: the life rating was a seven-point scale running from minus 3 (terrible) to plus 3 (delighted). The gap between the gratitude and hassles groups is 5.05 minus 4.67, or 0.38 points. On a seven-point scale that is a bit more than a third of one step. Real, statistically, and small.

The authors were careful too. In their own words, the gratitude groups did better "across several, though not all, of the outcome measures." In the first study, positive and negative mood overall did not differ between groups. The most robust finding across the three studies was a lift in positive mood.

What happens when you pool dozens of studies

Single studies bounce around. In 2016, Don Davis and colleagues pooled the gratitude intervention literature in a meta-analysis in the Journal of Counseling Psychology. Their key move was to separate studies by what the comparison group did, because "write anything" and "do nothing" are very different baselines.

The pooled effects on psychological well-being, measured as Cohen's d:

• Gratitude versus a measurement-only control: d = 0.31.

• Gratitude versus an alternative activity: d = 0.17.

• Gratitude versus an activity-matched comparison: d = 0.14.

• Gratitude versus a psychologically active comparison: d = minus 0.03, which is no difference.

• On anxiety, against an alternative activity: d = 0.11, not statistically significant.

Turning d into odds you can feel

A d value is hard to picture, so we converted each one into what statisticians call the probability of superiority. The question it answers: if you picked one random person from the gratitude group and one from the comparison group, how often would the gratitude person score higher?

Method: divide d by the square root of 2, then look up that value on the standard normal curve. If there were no effect at all, the answer would be 50 percent, a coin flip.

1. d = 0.31 versus doing nothing: 0.31 divided by 1.414 is 0.219, which corresponds to about 58.7 percent.

2. d = 0.17 versus another activity: 0.17 divided by 1.414 is 0.120, about 54.8 percent.

3. d = 0.14 versus a matched activity: about 53.9 percent.

4. d = minus 0.03 versus an active comparison: about 49.2 percent, a coin flip.

So the honest summary is this: a gratitude practice beats doing nothing by a modest margin, beats a neutral writing task by a slim one, and does about as well as other deliberate positive exercises. It is a small, cheap nudge, not a transformation.

The meta-analysis that checked depression and anxiety

A later meta-analysis by David Cregg and Jennifer Cheavens at Ohio State, published in the Journal of Happiness Studies, looked specifically at symptoms of depression and anxiety. According to the Ohio State summary, it covered 27 studies with 3,675 participants, which works out to roughly 136 people per study. The verdict on symptoms was "a small difference," and co-author Jennifer Cheavens said directly that "It would not be something you would recommend as a treatment."

That matters for how anyone should talk about a gratitude journal. It is a wellness habit for noticing good moments. It is not therapy, and if you are struggling with depression or anxiety, evidence-based treatment from a professional is the right path. If you are in crisis, call or text 988.

So does it work?

For what it is, yes, modestly. The pattern across studies is consistent enough to take seriously: people who regularly write down good things tend to report slightly higher well-being and positive mood than people who do not. Emmons and McCullough also found that daily listing produced a stronger gratitude effect than weekly listing in their later studies.

What the studies cannot do is make the practice happen. The Emmons and McCullough experiments all gave participants a structure: a form handed in at class, a packet mailed back each week, a set schedule. Outside a study, the hardest part is simply remembering to write something tonight.

That is the gap Gratitude Textapp is built for. It is an AI journal that runs entirely by text message. Each evening it asks for one to three things you are grateful for, small things included. You reply in your own words, it saves the entry, and it answers with one short line that picks up what you wrote. It is an AI, not a person or a therapist, and it can get things wrong.

A real exchange looks like this:

App (8pm): Gratitude time from your AI journal: What's something good, big or small, from today?

You: my kids laughing at dinner, a long walk

App: Kids laughing at dinner and a long walk, those small joys add up.

Try it by text

Text Hi to (872) 999-4985. From then on, each evening at 8pm, or a time you choose, it texts you a prompt asking what went well. Reply in a few words. Later you can ask it to show you a good moment from a past entry, ask for your streak and the themes that come up most, or move the time. If you already wrote that day, it skips the evening prompt.

It works on any phone that can text, with no app to install and no account to create. Your first 20 messages are free, and after that one $9.99/month subscription covers every Textapp. Entries are stored by Textapp to run your journal and are never shared or sold. Message and data rates may apply, and you can text STOP at any time. Details are at textapp.co/apps/gratitude.