Does journaling actually work?

Journaling works well for learning and poorly for mood, and most people are using it for mood. Reflecting on experience beats accumulating more experience by around 23%, and structured debriefs improve performance at d = .67. Then you reach the wellbeing research and the numbers collapse: across 146 randomized studies, the pooled effect of written reflection on health and wellbeing is r = .075.

Both of those are real. They are measuring the same activity and finding almost opposite things, and the gap between them is the most useful fact about journaling that nobody puts on the cover of a notebook.

Name the mechanism

The variable is distance.

Undifferentiated self-focused attention, the ordinary act of turning your attention inward and leaving it there, is reliably associated with negative affect across 226 effect sizes. That is the failure mode. You open the notebook at 11pm, you write down what happened, and what you have actually done is run the evening again at the same camera angle you filmed it from. You were inside it the first time. Now you are inside it twice.

Reflection is the version with a step back in it. The distinction that holds up in the research is why rather than what: analyzing the causes of an event rather than re-describing its contents. Same event, same person, same ten minutes, and the two produce measurably different outcomes. Processing beats venting, and venting is the thing most journaling quietly becomes.

Then the depth

Consider an ordinary bad Tuesday. You did not speak in a meeting where you had something to say. You noticed it in the car afterwards, the specific hot feeling of having watched your own mouth stay shut.

The replay version writes: I didn’t speak up again. I hate that I do this. That is accurate, and it is the same sentence you wrote in March, and writing it has moved nothing. You have added one more entry to a file whose only finding is that you are a person who does not speak up.

The reflective version asks why that particular room. Not why you are like this, which is a question about your character and has no answer, but why that room, on that day. Because the person who spoke first was senior to you. Because you had not slept. Because you had already decided, before the meeting started, that your idea was probably obvious. The answer is smaller and more specific than the story, and unlike the story it names something you could change on Thursday.

That is the whole difference. One produces a record. The other produces a mechanism.

The naming itself does some of the work: putting an emotion into words changes how the brain responds to it, an effect robust enough to have its own name in the literature, affect labeling. And monitoring your own progress toward something is among the strongest behaviour-change levers measured anywhere, at d+ = 0.40, which is a large finding for how boring it sounds.

Where this stops

The honest caveats are worth more than the claims.

The wellbeing effect is small. It is reliably not zero across 146 randomized studies, and it is r = .075, which will not carry a difficult season on its own. You will still see d = 0.47 quoted from a 1998 paper. That figure is now generally regarded as an overestimate, and the field has revised toward the smaller number.

The physical-health evidence is actually stronger than the psychological, which is not the direction anyone expects. And nearly all of it studies writing, so any claim about speaking your reflection instead is an inference about the mechanism rather than a finding about the format.

None of this makes journaling a poor use of three minutes. It makes it a poor use of three minutes if what you want is to feel better tonight, and a good one if what you want is to understand the pattern you keep enacting.

What to do with this

Stop writing what happened. Ask why that room, why that day, why the specific version of you that showed up.

You can do this in a notebook. The reason most people do not is that the question has to come from somewhere, and at 11pm nobody generates a good question about themselves. Oghere asks it, listens, and tells you what it noticed across the last several nights. Five minutes, and the audio is never kept.