Gratitude List, How Long And How Often To Write

Search this phrase and you get handed a hundred things to be grateful for. Copy somebody else's hundred items into your notebook and you have produced a shopping list, which is roughly the opposite of the exercise.

A gratitude list is just the format, three to five things written down, and the only two variables that actually distinguish one from another are how many items you write and how often you write them. On both of those, the honest answer is thinner than you'd expect. The one direct comparison I could find reports that people benefited more from longer weekly entries than from shorter, more frequent ones. And the headline effect sizes for this whole family of exercises got substantially smaller once somebody corrected them for small sample bias, dropping to around r equals .10 for wellbeing with the depression effects becoming generally non-significant. Below is what survives, plus a table for picking a list length that isn't just three because everybody says three.

What The Ranking Pages Hand You

Two of the top three results for this phrase are lists of a hundred or more items. One is titled around 101 things to be grateful for, another offers a hundred, a third gives 65 examples. The intent behind them is obvious and slightly sad, which is that people arrive at this search stuck for what to write.

The trouble is that a supplied list solves the wrong problem. Being unable to think of something is not a failure of imagination, it's usually a sign that you're scanning categories rather than days. And a copied item carries no information, because the whole mechanism rests on noticing something that would otherwise leave no trace.

Schueller and Parks, in their review of the area, put the point about strength of exercise plainly, noting that although brief techniques can benefit people, the strategies have to push participants to make large changes rather than small ones (JMIR, 2012). Ticking items off somebody else's hundred is about as small a change as it's possible to make while still holding a pen.

So the useful question isn't what to put on a gratitude list. It's how long the list should be and how often you write it, which nothing in the top ten addresses at all.

Longer And Weekly Beats Shorter And Daily

This is the closest thing to a direct answer, and I want to be careful about how strong I make it sound.

In the same 2012 paper, reviewing the gratitude literature, Schueller and Parks report that a further investigation of gratitude journalling found participants benefited more when completing longer weekly entries as opposed to shorter, more frequent entries. That's their summary of prior work rather than a result of their own trial, so treat it as a signpost rather than a finding. I could not get the underlying paper open in this session, so I'm not going to quote numbers off it that I haven't seen.

Still, the direction matters, because it's the opposite of the default advice. Every app and every printed journal in this category pushes daily. Daily is a product cadence. It generates streaks, it generates notifications, and it generates the exact repetition problem that makes week three feel like paperwork.

The same review notes something else worth carrying. Their description of the three good things exercise is that benefits didn't show up until one month after the intervention finished, and then held. If that pattern is real, a person evaluating their gratitude list on day nine is evaluating nothing.

There's a longer treatment of the underlying trials in the gratitude journal write-up, which goes through the original weekly protocol and the meta-analysis specific to gratitude. This page is about the list itself.

What The Numbers Look Like After Correction

Here's the part that no listicle will ever tell you, and it's the single most useful thing on this page.

Carmela White, Bob Uttl and Mark Holder published a reanalysis in PLOS ONE in May 2019 (the full paper). They took the two most cited meta-analyses of positive psychology interventions, which is the family gratitude lists belong to, and redid the arithmetic accounting for small sample size bias.

Read the first column top to bottom. The wellbeing effect goes from .29 to about .10 as the analysis gets more careful, which is roughly a third of what the widely quoted figure suggests. The depression column is worse, ending at generally not statistically significant.

Their own summary is that many primary studies used small samples, that small sample size bias was pronounced in many analyses, and that once it was accounted for the wellbeing effects were small but significant while depression effects were variable and generally not significant. Their recommendation for the field is bluntly about sample sizes.

I want to be fair to the practice here rather than smug about it. Small but significant is not nothing, and this costs three minutes and no money, which is a completely different value proposition from a paid intervention. What it isn't is the thing the marketing describes. If you've been running a gratitude list and quietly wondering why you don't feel transformed, that arithmetic is your answer and it isn't a personal failing.

Over Half Of Them Stopped

Numbers on who actually keeps doing this are rare, so I'll give you the ones I found.

Schueller and Parks recruited 1,364 people who consented to an online trial of positive psychology exercises delivered in packages of two, four or six over a six week period. Attrition ran at 55.5% for the two-exercise group, 55.8% for four, and 52.7% for six. The assessment-only control group lost 42.5%, significantly fewer, which tells you that being given exercises to do made people more likely to disappear than being given nothing.

The result that surprised me is the shape. Depressive symptoms fell in every condition, but the reduction was larger in the groups given two or four exercises than in the group given six or the control. More was not better past a point.

If that transfers to list length, and I'm speculating rather than reporting, it argues against the hundred-item version fairly directly.

The Study Where The Arrow Pointed Backwards

I went looking for a trial that used a list specifically, daily, with a stated maximum number of items, and found one that also happens to contain the most uncomfortable result in this whole area.

Izabela Krejtz, John Nezlek, Anna Michnicka, Pawel Holas and Marzena Rusanowska ran a two week daily diary study with 58 adults in Poland, published open access in the Journal of Happiness Studies (the paper). Twenty-nine people were assigned to the gratitude condition and twenty-nine to a control. Both groups filled in daily measures of events, stress and wellbeing. The gratitude group, before completing each diary, wrote what they were grateful for that day, and the instruction let them describe up to 6 things. Fifty-seven of them produced 781 days of usable data.

The main finding is the one in the title. The link between daily stress and daily wellbeing was weaker in the gratitude group. On worry, the slope from stress ran at .10 for the gratitude group against .26 for the control. On the Beck triad measure the slopes were minus .06 against minus .16. So the list didn't reliably make people happier on average, it made bad days hurt less, which is a subtler and honestly more believable claim.

Then the lagged analysis, which nobody quotes. They modelled each day as a function of the previous day, in both directions. Their sentence is that they "found no support for a gratitude to well-being causal link in our data" while finding some support for a wellbeing to gratitude link. Relaxed people with fewer negative thoughts became more grateful the next day, rather than the reverse.

That's one small study and the authors treat it carefully, so I'd not overread it. But it's a real result from a real design, and it suggests part of what a gratitude list measures is how your week is going rather than how your week is going to go.

Six items, incidentally. Not three.

Picking A Length That Isn't Just Three

Three items is the default because one famous exercise used three. It is not a law. Here's how I'd actually choose, and the reasoning column is what makes it usable rather than arbitrary.

The last row is the one I'd defend hardest. A list is enumeration. It has no mechanism for working anything out, and using it on a stuck situation produces a page of items you already knew about, which then reads back as evidence that you should be fine. If that's the job, long form writing is the format with the protocols attached.

The enjoyment row comes straight out of the 2012 trial, where enjoyment of an exercise correlated with still using it at six week follow-up across five of the six exercises, with correlations from .12 to .39. Modest correlations. But it's the only lever in the whole paper that's under your control, and it argues for doing the version you don't dread rather than the version with the best paper behind it.

When A List Is The Wrong Instrument

Directly, and in the body rather than in a footer.

Writing things down is not care and it isn't a substitute for help from a qualified professional. The reanalysis above found the depression effects for this whole family of exercises to be variable and generally not significant once corrected, so a gratitude list is a poor candidate for the job of lifting a depressive episode, and anybody presenting it that way is going past the evidence. If you're in immediate crisis, 988 takes calls, texts and chats around the clock in the US, and it's explicitly for emotional distress rather than only for people who are actively suicidal.

There's a quieter failure specific to this format too. A list of good things, written during a week that is genuinely bad, can read as an instruction to stop being upset about something you're entitled to be upset about. That's a known problem rather than a character flaw, and it's covered properly in the gratitude journal piece along with why the founding researchers saw it coming. If what's happened is a loss, writing through grief is a different practice with a different shape and I'd go there instead.

What I'd Write Tonight

Five items, Sunday, one sentence of because on each, and no attempt at nice phrasing.

The habit I'd actually try to build isn't the writing, it's the noticing during the week, because the list is only a container. My own record here is one-directional. I've shipped seven paid toolkits, watched a site sit at average position 47 in search for months, and I've got a channel with 7 subscribers, so I'm well practised at cataloguing what isn't working and out of practice at the other direction entirely. That asymmetry is the whole argument for having a form, and it's also why I'd keep the form small.

If the list goes stale, change the constraint rather than the length. Rotate what you're allowed to write about, since a category you can't repeat forces you to actually look at the week. There's a version of that rotation borrowed from what Tim Ferriss does with the same five lines, and it's the most practical thing anybody in this genre has published.

For how the wording of a prompt changes what comes out, the prompt guide is the longer piece, and gratitude journalling for beginners has a day by day plan for the first week if you'd rather be told exactly what to do.