A journal for anger management: what the evidence says

A journal for anger management is a record. It can hold what set you off, how strong the anger was, what else was going on that day and what you expected to happen. Read back over a few weeks, a record like that can help you notice patterns. What the research behind this page didn’t turn up is a study showing that journaling reduces anger on its own.

I make Kettle, an anger journal app, so I have a reason to oversell journaling. I’m also not a clinician. So each finding here comes with its source and its limits.

Does journaling reduce anger?

Whether journaling reduces anger hasn’t been tested directly. The research for this page found no meta-analysis (a study that pools the results of many studies) and no randomised trial in a general adult sample that tests ordinary journaling as a way to reduce anger over weeks or months. A study like that may exist. This research didn’t find one.

What does exist is three neighbouring kinds of research:

• trials of “expressive writing”, where the outcome was rarely anger
• one-session laboratory experiments, mostly with undergraduates, on how people think or write about something that angered them
• anger programmes based on cognitive behavioural therapy (CBT), where a written log is one part of a package

Many of these papers could only be read as abstracts, so some findings come without effect sizes.

What expressive writing studies found

Expressive writing means writing about a stressful or traumatic experience over a small number of sessions. A daily journal kept for months is a different activity, so these results don’t carry straight across.

Review	What it found
Frattaroli, 2006: 146 randomised studies	A positive, significant but very small average effect: r = .075 (d = .151)
Meads and Nouwen, 2005: 61 trials	No clear improvement in objectively measured physical health or most other outcomes
Mogk and colleagues, 2006: 30 randomised trials	No significant effect on physical or psychological health
Reinhold and colleagues, 2018: 39 randomised trials	No significant long-term effect on depressive symptoms in physically healthy adults
Guo, 2023: 31 randomised studies	A small reduction in depression, anxiety and stress symptoms combined: Hedges’ g = -0.12, 95% CI [-0.21, -0.04]

The letters r, d and g are measures of effect size, meaning how much difference the writing made. All of these are small.

Anger barely featured in the studies pooled in 2006. Only 2 of Frattaroli’s 146 looked at anger-related outcomes. They gave r = .183, and two studies are too few to rest a claim on.

One later trial did measure anger, and it was large. In a 2015 trial of 1,292 US veterans who reported difficulty readjusting to civilian life, those assigned to write expressively online about that transition had greater reductions in anger six months later than those who wrote factually or didn’t write at all. The effects were small: d = 0.13 to 0.20 against factual writing and 0.22 to 0.35 against no writing, across the outcomes measured. That’s structured writing about one subject, in veterans. It isn’t a daily anger journal, but it’s the largest trial I’ve found that measured anger after writing.

Mood often dips, and some people did worse

Expressive writing typically has short-term costs. Frattaroli describes more negative mood and less positive mood shortly after writing.

Two individual trials also found people who did worse months later. In a trial of 116 young adults, expressive writing had no significant overall effect on anxiety, depressive symptoms or physical symptoms, and participants low in emotional expressiveness showed a significant increase in anxiety at three months (those high in it showed a reduction). In a trial of 90 recently separated adults, people judged to be actively searching for meaning in the separation reported significantly worse emotional outcomes after expressive writing than after control writing, up to nine months later.

Neither trial measured anger, and each finding is about a subgroup in one study. That’s a reason for caution. It doesn’t show that writing harms people.

What happens when people write about anger itself?

Research on writing about anger specifically is thin. The closest test is a 2012 experiment with 121 undergraduates. Each recalled a memory that made them angry, then spent 20 minutes on one of four tasks: analytical rumination (dwelling on the memory), reappraisal (looking for another way to see it), distraction, or writing about their thoughts with no instructions for handling the emotion. Rumination maintained anger. People in the other three conditions reported less anger after the writing task.

That was one session with students, with anger measured after the task and not weeks later, and the abstract gives no effect sizes. Ruminative writing kept anger where it was. The study doesn’t say it made anger worse.

Venting on paper versus reflecting

Venting on paper and reflecting on paper haven’t been compared directly in any study the research for this page found. The evidence here comes from neighbouring research.

What the venting research found

In a 2002 experiment with 600 undergraduates, people who hit a punching bag while thinking about the person who’d angered them felt angrier than people in a distraction group or a group that did nothing, and were more aggressive than the group that did nothing. The differences were small (d = 0.22 to 0.31).

A 2024 review of 154 studies, covering 10,189 participants, tested two kinds of activity: ones that lower arousal (calming ones such as deep breathing, mindfulness and meditation) and ones that raise it (such as hitting a bag, jogging and cycling). The calming activities decreased anger and aggression (g = -0.63, [-0.82, -0.43]). The arousal-raising ones were ineffective overall (g = -0.02, [-0.13, 0.09]). The authors conclude that their findings don’t support venting or going for a run as effective anger management.

Neither tested writing. The 2002 experiment was about hitting a bag, and the review’s abstract doesn’t mention writing or journaling (its full text couldn’t be opened for this page). One experiment did test venting in writing. In a 2021 study in China, people who had been provoked and then wrote down their dissatisfaction behaved more aggressively afterwards than people who did a simple recall task. Its abstract gives no sample size, and the outcome was aggressive behaviour, not how angry people felt.

Put that beside the 2012 experiment above, where ruminative writing was the one kind that didn’t bring anger down, and there’s a reason to steer a journal away from replaying the grievance. Neither study shows that replaying does harm over weeks of journaling.

What stepping back does

Self-distancing means looking at what happened the way an onlooker would, not reliving it from the inside. In research published in 2005, people recalled an experience that had made them angry at someone. The researchers expected that asking why from a distanced perspective would let people reflect without stirring up intense negative feeling again, and the findings were consistent with that. The abstract gives no sample sizes or effect sizes, and the distancing itself was done in thought, not in writing.

A 2022 meta-analysis found that a distant perspective dampened emotional experience by a moderate amount (g = 0.52, k = 230). Its abstract gives no separate figure for anger.

Neither is about keeping a journal. The research for this page found no study testing distanced writing prompts for anger beyond a single session.

Naming the feeling has mixed evidence

Putting a name to a feeling is called affect labelling. In a 2007 brain imaging study of 30 adults, labelling reduced the brain’s amygdala response to negative emotional images. But participants were matching a word to a photograph of someone else’s face, not naming their own anger or writing, and the outcome was brain activity.

Two experiments in 2021, with 80 and 60 participants, point the other way. People who named their emotions before trying to reappraise them reported feeling worse than people who reappraised without naming. Those studies used unpleasant images, not personal anger. So the common advice to name a feeling in order to calm it isn’t settled.

How a CBT anger programme uses a log

Self-monitoring, which means keeping track of your own anger, is a taught part of the CBT programme in SAMHSA’s anger management manual. SAMHSA is a US federal agency, and the programme is for groups of people with substance use and mental health problems. It introduces two tools. The anger meter is a scale from 1 (calm) to 10 (intense anger, with the feeling that you may lose control). The anger awareness record is a written log with six columns:

• situation
• anger cues
• anger meter rating
• behaviour
• consequences, positive or negative
• strategies used

Its prompts include “What sets me up to become angry?” and “What did I tell myself?”. Group leaders also help people separate events, which are facts, from interpretations, which are opinions, value judgments or perceptions of those events.

Does an anger log work on its own?

An anger log hasn’t been tested against a control group, as far as the research for this page could find. SAMHSA’s manual calls monitoring “helpful in increasing the awareness of anger”, which describes how the programme uses it and isn’t a test of it. The evidence is for whole programmes. The manual reports that studies of its treatment model, in clinical groups, have consistently found significant reductions in self-reported anger and violence. A 2018 review of meta-analyses concluded that anger treatments, most often CBT, have consistently shown at least moderate effectiveness in both non-clinical and psychiatric populations.

A systematic review of 42 CBT studies adds a caution: all but one reported a statistically significant result, and the exception was the only one that had assessed and corrected for social desirability bias, the pull to answer in a way that looks good.

Two studies come closer. Adults with problem anger checked in on their anger repeatedly for 10 days and did nothing else: 60 Australian veterans in one, and 98 trauma-exposed adults using a smartphone and a wearable in the other. Problem anger fell over the 10 days in both. Neither had a control group, so the drop can’t be credited to the monitoring.

So a log borrows one component from treatments that have evidence. It doesn’t inherit their results.

How to keep an anger journal

On the evidence above, the cautious way to keep an anger journal is as a log for noticing what happened. The thing to avoid is using it to relive the day. None of these steps has been tested as a way to reduce anger.

1. If you’re still wound up, do something calming first. That’s what the 2024 review supports. The entry can wait.
2. Write what happened as plain facts, then what you made of it, and keep the two apart. That’s the event and interpretation split from SAMHSA’s manual.
3. Rate it, on the same scale every time so entries can be compared.
4. Note what else was going on that day and what you expected to happen. Those two are Kettle’s own questions, and no study tested them.
5. When you get to why, describe it the way someone watching would. That borrows from the distancing experiments and is untested as a journal prompt.
6. Stop when the questions are answered. If you catch yourself building the case against someone, cut it short.

If a blank page stalls you, journal prompts for anger has more questions to work from.

When to write, and for how long

No study the research for this page found says what time of day is best for an anger journal, how long an entry should be, or whether writing daily beats writing only when something happens. For expressive writing, the 2006 meta-analysis found that sessions of at least 15 minutes did better than shorter ones. A log is a different task: a few short answers, which take a few minutes.

Habit research offers a little. In one study, 96 volunteers chose an eating, drinking or activity behaviour to do daily in the same context, such as after breakfast, for 12 weeks. The time participants took to reach 95% of their plateau of automaticity (how automatic the behaviour felt) ranged from 18 to 254 days, and missing one opportunity didn’t materially affect the process. In a 2024 review of 20 studies, only four reported how long habits took to form. Two gave medians of 59 and 66 days (the second is the study above), and individual times ranged from 4 to 335 days. The review rated 11 of the 20 studies at high risk of bias.

None of those studies involved journaling. What they support is modest: tie the entry to something you already do each day, expect it to take weeks or months to feel automatic, and don’t read much into one missed day.

Reading your entries back

Reading entries back after a few weeks is how you look for repeats: the same situations, the same people or places, the same expectations, higher ratings on high-stress days. The anger triggers worksheet is a place to collect them. Your own entries can show that two things tend to turn up together. They can’t show that one causes the other.

A dip in mood right after writing about something upsetting is common in expressive writing studies. If entries regularly leave you angrier, treat that as a sign to change how you’re writing, or to stop.

Paper, worksheet or app?

Paper and apps haven’t been compared for anger journals in any study the research for this page found. The evidence on apps is about mental health apps in general. A 2024 meta-analysis of 176 randomised trials found small effects on symptoms of depression (g = 0.28) and generalised anxiety (g = 0.26) compared with control groups. A 2026 meta-analysis of 8 trials of mood monitoring concluded that it doesn’t increase or decrease mood symptoms in people with bipolar disorder, with no robust evidence of such effects in depression.

The research for this page found no trial of a stand-alone anger tracking app. In one small trial, 58 US veterans received anger management treatment with or without an app that included symptom monitoring. Anger fell in both groups, and the app group did no better.

People also stop using these apps quickly. An analysis of 93 Android mental health apps found median retention rates, meaning the share of people still using an app, of 3.9% at 15 days and 3.3% at 30 days. For tracker apps the 30-day median was 6.1%.

A notebook is fine. If you’d like a form, the daily anger log is a free printable page.

Why Kettle asks five questions

Kettle is a daily anger journal for iPhone. It’s coming soon and isn’t on the App Store yet. It asks five questions: what frustrated me, how frustrated I got from 0 to 10, what else was going on, what I expected that didn’t happen, and how stressed I am overall from 0 to 10. Everything stays on the phone, with no account.

The questions overlap with part of an anger log like SAMHSA’s: the situation, a rating (Kettle’s starts at 0 and SAMHSA’s at 1), and prompts close to what set you up and what you told yourself. The stress rating has no counterpart there, and Kettle’s five questions leave out behaviour, consequences and strategies. That’s my design choice. It isn’t evidence, and I’m not claiming Kettle will reduce your anger. How Kettle works has more.

When a journal is not enough

A journal isn’t treatment, and it can’t assess you. Neither can this page. Only a professional can do that.

The NHS says to see a GP if you’re struggling to cope with anger, if the things you’re doing to help yourself aren’t working, or if you’re acting in an abusive or violent way. It also says most anger management programmes include CBT as well as counselling.

If your anger is hurting you or the people around you, Get help lists places to start. If what you want is to understand where the anger comes from, read Why am I so angry?.