Who affirmations help, and who they make feel worse
The same sentence lands differently depending on where you are starting from. What the Wood, Perunovic and Lee study actually found, how far it can fairly be pushed, and how to tell in about ten seconds which side of the line a given line puts you on.
Most writing about affirmations treats them as a thing that either works or does not. The more interesting finding in this area is that the same sentence can go two different directions depending on who is reading it, and that the direction is roughly predictable.
That is worth knowing before you spend a month on a practice, because the group most likely to be sold this technique is the group one well-known study found it went badly for.
The finding, stated plainly
Wood, Perunovic and Lee published "Positive self-statements: power for some, peril for others" in Psychological Science in 2009. Participants were asked to repeat, or to hold in mind, the sentence "I am a lovable person." Participants who already scored high on self-esteem measures came out marginally better in mood than controls. Participants who scored low came out worse — worse mood, and worse self-regard, than people who had done nothing at all.
A follow-up condition in the same line of work is the part people forget. When participants with low self-esteem were allowed to consider ways in which the statement was both true and untrue — that is, allowed to argue with it — the damage did not appear. It was not the topic that hurt. It was being required to assert something without being allowed to weigh it.
How far this can fairly be pushed
Not as far as either camp wants. Some honest limits:
- The samples were undergraduates, modest in size, and the outcome was self-reported mood measured shortly afterwards. This is a laboratory result about a short-term state, not a study of what happens to somebody over a year.
- It predates psychology's replication reckoning. It has been cited constantly and rarely replicated at scale, which is a description of the field rather than an accusation about these authors.
- "Self-esteem" here is a questionnaire score, not a diagnosis, and the split into high and low is a statistical convenience applied to a continuum.
- It shows a short-term mood effect. It does not show that affirmations are harmful in any lasting sense, and nobody should read it that way.
What it does establish, and what is worth carrying, is a direction: a statement far outside what you currently accept is not neutral. It provokes the rebuttal. And the person with the most detailed rebuttal available is the person the sentence was aimed at.
Why the mechanism makes sense on its own terms
Two other literatures point the same way, which is more reassuring than any single study.
Latitude of acceptance
Persuasion research going back to Sherif and Hovland describes a band around your current position within which a message can move you, and a region outside it where messages get rejected and can push you further away. Nothing about this band cares whether the message came from a stranger or from you. "I am completely confident" on a flat day is outside it.
Self-verification
Swann and colleagues have documented that people seek information consistent with their existing self-view, including an unflattering one, and rate it as more accurate. A stable self-concept is how you predict what happens next. Praise that contradicts it does not read as a gift; it reads as a claim that may be about to be tested.
Put together: the further a sentence is from what you believe, the more work your mind does to defend the existing picture, and the more of that work is composed of evidence against you.
The ten-second test
Read the line. Then notice which of these happened.
- Nothing much. Mild agreement, no argument. This is a sentence you can use.
- A specific counter-example arrived unbidden, with detail. That is the out-of-range case, and repeating it harder is the exact move the study suggests goes badly. Rewrite it smaller.
- Mild embarrassment, but no actual objection. This is unfamiliarity rather than disbelief. It usually passes if you stay with it for a few seconds, and it fades faster if you read silently instead of aloud.
Most people can sort a line into one of those three within a breath. The skill is not generating conviction; it is being willing to discard the sentence in case two instead of treating the discomfort as a personal failing to be overcome.
What to use instead when you are in the second case
The pattern across the sturdier findings is that the sentence gets smaller, more specific, or further away from a claim about your character.
- Evidence rather than trait: "I have got through worse weeks than this" instead of "I am strong."
- Permission rather than assertion: "I am allowed to stop before this is finished." There is nothing here for a counter-argument to attack.
- A small capability with a deadline you can check: "I can write one paragraph before lunch."
- A value rather than a quality: "what matters to me here is doing this carefully." This is the shape the actual self-affirmation research studies, and it makes no claim about you at all.
- A change of person: "you have handled worse than this." Kross and colleagues have reported across several studies that second-person and name-based self-talk produces less distress under pressure than the first person.
None of this is a treatment, and it should not be presented as one. When something is genuinely wrong, the right move is care, not a better-worded phrase.
Who does well with the standard version
It would be dishonest to only list the caveats. People who read a confident line and feel it click are usually in one of these positions:
- The claim is already broadly true for them and the sentence is a reminder rather than an installation.
- They are reading it as intention rather than description — hearing "I am calm" as "calm is what I am going for today" — which quietly converts it into a shape the mind does not argue with.
- They are in a decent stretch. Latitude of acceptance widens on a good day and narrows on a bad one, which is why the same card can feel true in June and absurd in November.
That last point is the practical one. Your range moves. A practice that assumes a fixed list of sentences will be right every day has assumed something false about you.
What this argues for in a daily-card habit
Three things, and they are all about being allowed to say no.
First, choose by where you actually are rather than where you would like to be. Confidence lines on a flat day are the ones that get argued with; a calm or gratitude line on the same day usually passes without resistance. Lumo asks for a mood — calm, healing, gratitude, strength, sleep and others — for this reason rather than as decoration, and the colour of the card follows the mood you picked.
Second, be able to write your own. A sentence in your own vocabulary about your own week is inside your latitude by construction, which is a structural advantage no general-purpose collection can match. In Lumo those live under Mine, in the same place as the ones you saved, and you can attach a private note to any card about what it brought up.
Third, do not let the streak become the reason. If you are opening something you are not reading, the counter has replaced the point. Lumo draws the days you showed up as a heatmap, but breaking the run locks nothing, deletes nothing and is never mentioned back to you. Everything stays on the phone: no account, no sign-in, no network requests.
Frequently asked questions
Can affirmations be bad for you?
One well-known study found a short-term cost for one group. Wood, Perunovic and Lee (2009) reported that participants with low self-esteem who repeated "I am a lovable person" ended up in a worse mood than people who did nothing. It was a laboratory study of a short-term state with modest samples, so it should not be read as evidence of lasting harm — but it is a real result that runs against the usual advice.
Why do affirmations make me feel worse?
The proposed mechanism is that a statement far outside what you currently believe does not get absorbed, it gets rebutted. Asserting it prompts you to retrieve the counter-evidence in detail, which leaves you having just reviewed a list of reasons the sentence is false. Using a line you can actually accept today avoids the mechanism entirely.
Do affirmations work better if you already have high self-esteem?
That is what the 2009 study found, though the improvement in that group was small. A likely reason is simply range: if the sentence is already roughly true for you, there is nothing to argue with, so it reads as a reminder rather than a claim.
What should I use instead if positive statements backfire?
Shapes that are harder to argue with: pointing at something that actually happened, granting yourself permission, claiming a small capability you can verify within the hour, or naming a value rather than a quality. Addressing yourself as "you" or by name also has better support than the first person.
Does this mean affirmation apps are useless?
It means the useful ones let you say no to a line and write your own. A collection of sentences written for a general audience is raw material rather than a prescription. If you notice a card is nearly right, the gap tells you what your own version should say.
Try it on an ordinary morning
Lumo shows one card at a time on a full screen — a first-person affirmation, or a quote with its author named — in a colour that follows the mood you chose. Save the lines worth keeping, write your own, and let a widget carry today's onto your home screen. No account, no network requests, no analytics, no ads. Free, with no in-app purchases in this version.