How to Stop Being So Nice, and What the Research Says It Costs
The first validated measure of people-pleasing arrived in 2025, and what it tracks is not agreeableness. The fix everyone recommends has never been tested.
Before the article. If saying no to someone has a reliable consequence — an argument that lasts days, a silence you have to work off, money withheld, your phone checked, being told afterwards that you caused it — then what you have is not a problem with your personality, and a later section says so plainly. Find A Helpline, in the sources at the bottom, lists free and confidential services by country. In the United States the National Domestic Violence Hotline is 1-800-799-7233, or text START to 88788, 24 hours a day.
With that said. This is one of the few questions where the gap between how confidently the internet answers it and how little has been measured is enormous, and the gap is the reason this page exists.
The vocabulary the question is usually asked in — people-pleasing, the fawn response, codependency — did not come from research. The fawn response was named by a psychotherapist in a newsletter article in 2003, and his own piece cites a dictionary rather than a study. The questionnaire that tells people whether they have “people-pleasing syndrome” descends from a 2001 paperback, and the first proper psychometric validation of it was published in 2025.
What has been measured sits under other names, and it is better than the folk version in three specific ways. The trait that goes with distress is not warmth or agreeableness — it is a particular bundle of needing other people's reactions in order to know your own worth. The cost of giving something up for somebody turns out to sit in the motive rather than the quantity. And the famous doormat effect is not about being too forgiving: in every study, the thing that decided whether forgiving cost the forgiver anything was whether the other person made amends.
So what follows is twenty-two sections, covering what has been measured, what it does and does not license, who pays more for taking the standard advice, and the one piece of advice that is everywhere and has never been tested on anyone. Our editorial policy explains why every empirical claim here carries a linked source.
The thing you are asking about has four names, and only some of them were measured
Start with the awkward fact that organises everything below: the words this question is usually asked in are not research words.
Search for how to stop being so nice and the answers arrive in a shared vocabulary — people-pleasing, the fawn response, boundaries, codependency. That vocabulary comes from trade books and clinic blogs rather than from a literature, and it is worth knowing which parts of it anyone has actually put a number on before taking instructions in it.
Here is one way to see the gap, and it is a check anybody can repeat. On 28 October 2026, a search of PubMed for the exact phrase fawn response in titles and abstracts returned zero records. The word fawning in titles and abstracts returned 32, and all but one of them are about deer. Tonic immobility — the freeze state, the part of the same folk model that does come from the laboratory — returned 683. Sociotropy, the clinical construct closest to chronic approval-seeking, returned 136.
That is not an argument that the experience is imaginary. Far more people have the experience than have a name for it, and the usefulness of a word does not depend on a citation. It is an argument about where to take advice from, because the gap between how confidently this is explained and how little of it has been measured is as wide as anything this site has looked at.
So this page does something slightly different from the first page of results. It sets out what has been measured, under the names researchers actually used, and it says at each point how far the finding stretches. Some of it is more useful than the folk version. Some of it points the other way. And the single most popular piece of advice on the subject turns out never to have been tested on anyone. More of this kind of thing under Mindset, and our editorial policy explains why every claim here carries a linked source.
Where the fourth F came from, and what it rests on
The fawn response deserves its own section, because it is the explanatory frame most of this genre now runs on, and its origin is completely traceable.
It was named by Pete Walker, a psychotherapist in Berkeley, in an article titled Codependency, Trauma and the Fawn Response, published in The East Bay Therapist in January/February 2003. The sentence is not ambiguous: “I have named it the fawn response…the fourth ‘f’ in the fight/flight/freeze/fawn repertoire of instinctive responses to trauma.” The definition that follows is taken from a dictionary — Walker quotes Webster's for to act servilely; cringe and flatter — and the evidence that follows is clinical observation from his own practice, offered as exactly that.
Read the article for yourself and the thing you will not find is data. There is no study, no sample, no measure, no comparison group and no citation to an empirical paper anywhere in it. The references are to two trade books. That is not a criticism of Walker, who wrote a clinical essay for a therapists' newsletter and said so. It is a description of what is underneath the diagram that now gets reproduced, usually without his name on it, as though it were a finding about the nervous system.
The academic literature nearest to it is small and candid. A 2023 proposal in the European Journal of Psychotraumatology argued for replacing Stockholm syndrome with appeasement, explained through Polyvagal Theory, and its own background section notes that scarce empirical research supports the phenomenon it is replacing — and the theory it leans on has itself been challenged premise by premise in Biological Psychology. A 2026 review in Current Addiction Reports that builds a fight–flight–freeze–fawn framework for substance use ends by saying that fawning “remain[s] underdeveloped and understudied.” That is the field describing its own foundations, and the construct it proposed to replace turns out to be in no better shape: a systematic review of Stockholm syndrome found twelve papers, mostly case reports, and no validated diagnostic criteria, which is one of the findings in what trauma bonding actually rests on. Which leaves the question of what has been measured, and the answer is: quite a lot, under other names.
The measured construct is not kindness, and the distinction is the whole point
The closest thing to a worked-out science of being too accommodating is forty years old and almost never appears in the search results for this question.
In 1998 Vicki Helgeson and Heidi Fritz proposed a distinction that sounds like hair-splitting and is not. Communion is a positive, caring orientation towards other people. Unmitigated communion is a focus on others to the exclusion of the self. Their argument was that research had failed to separate the two, which is why measures of traditionally feminine traits kept failing to predict depression: the warm part is not the part that hurts.
The companion paper, in the Journal of Personality and Social Psychology the same year, tested it across four studies. The two constructs correlate, as you would expect. But unmitigated communion separates from communion along three specific lines: a negative view of the self, turning to other people for self-evaluative information, and psychological distress. And the authors went further than documenting the association. In their analysis, relying on others for self-esteem leads to overinvolvement with others and neglect of the self, and those two together account for the link between unmitigated communion and distress.
That gives the question a shape it does not have in the popular version. The target is not warmth, generosity, or being easy to get along with. It is a specific bundle: needing other people's reactions in order to know what you are worth, and then arranging your life around producing those reactions. You can be extremely kind and have none of it. You can have all of it and not be especially kind.
The honest limits: these are self-report measures in American samples, the framing is explicitly about gendered traits, and the analysis is cross-sectional, so the arrow between self-esteem, overinvolvement and distress is a model rather than a demonstration. Both papers are behind the APA paywall and what is reported here comes from the published abstracts, which is stated in the source list rather than hidden in it.
What it looked like when somebody followed the trait into a hospital
If the distinction were only a scoring artefact it would stay inside personality questionnaires. One study took it somewhere with harder outcomes.
Heidi Fritz interviewed 65 patients one week and four months after a first coronary event and measured all three traits: agency, communion and unmitigated communion. The predictions were stated in advance. Agency should predict improved health. Communion should be unrelated. Unmitigated communion should predict worse health over time.
All three held. Unmitigated communion went with worse depression, anxiety and mood, worse mental and physical functioning on the SF-36, and more cardiac symptoms at four months. Communion — ordinary caring about other people — was unrelated to any of it. And the paper identified what the association ran through: unmitigated communion was linked with poor health behaviour and with negative social interactions, and those partly explained the link to depression and to cardiac symptoms.
That last clause is the useful part and it is not mystical. A person who cannot stop attending to everyone else does the things that follow from it: skips their own appointments, does not rest, keeps supplying people who are difficult to supply, and ends up in more interactions that go badly. Nothing in the finding requires a nervous-system story.
The limits are large and belong in the same breath. Sixty-five people is a small sample. Four months is a short window. It is observational, so nothing here shows that the trait caused the outcome rather than travelling with something else, and a first coronary event is a particular population at a particular moment. Treat it as one careful study pointing the same way as the personality work, not as a reason to believe that being accommodating damages your heart. What it does establish is that the distinction between communion and unmitigated communion keeps surviving contact with outcomes that are not questionnaires.
Self-silencing, and the one pooled number this field can offer
The second measured construct is narrower and closer to what people mean when they say they are too nice to speak.
Dana Crowley Jack and Diana Dill built the Silencing the Self Scale out of a longitudinal study of clinically depressed women, published in 1992. It measures schemas about how to make and keep intimacy — judging yourself by an external standard, putting other people's needs first, suppressing anger and self-expression to avoid conflict, and experiencing a divided self in which the outward compliant person and the inward one come apart. They validated it on three samples at once: 63 college students, 140 residents of refuges for women experiencing domestic violence, and 270 mothers. It was internally consistent, stable on retest, and correlated with depression scores in all three.
Thirty years of studies followed, and in 2021 Sebastian Pintea and Andreea Gatea pooled them. Thirty-one published and eleven unpublished studies, 10,108 participants and 131 effect sizes. The overall association between self-silencing and depression was r = 0.391 — a moderate correlation, and about as solid a number as this subject has. Which components of the scale a study used, and the age, education and sexual orientation of the sample, explained part of the variation between studies.
Now the limit, and it is the limit on this whole page. Those are correlations. Almost none of the underlying studies could say whether silencing yourself makes you depressed, whether being depressed makes you silence yourself, or whether a third thing produces both — and the plain reading of the original clinical work is that the traffic runs in both directions. Including a longitudinal design was not what most of these studies were for.
So the honest statement is this: the pattern is real, it is well documented, it is associated with being less well, and nobody has shown that changing it changes the outcome. That last clause is the one the advice pages skip.
The questionnaire everybody uses came out of a 2001 paperback
Here is the part nobody on the first page of results mentions, and it is the most useful thing on this page for judging what you read elsewhere.
Until last year, the term people-pleasing had no validated instrument at all. The measures in circulation came from popular psychology. Harriet Braiker's 2001 book supplied the questionnaire most often reproduced online, organised into thoughts, behaviours and feelings. A 2021 self-help title added a twenty-item “approval rating test” with score bands running from no self-sabotaging behaviour down to a severe case requiring intervention. A 28-item Chinese questionnaire published in 2021 proposed six dimensions.
None of that was tested in the way a measure is supposed to be tested, and the people who finally tested it said so. In the 2025 validation study described in the next section, the authors write that Braiker's tool lacks empirical evidence for its cutoff criteria, that the reliability and validity of the twenty-item quiz remain unverified, and that the six-factor structure of the 2021 questionnaire fitted its own data badly — a comparative fit index of 0.79 and a Tucker–Lewis index of 0.76, both well under the 0.90 that counts as acceptable.
Read that back slowly, because it changes how to treat the genre. The quiz that tells you whether you have people-pleasing syndrome, and the score band that tells you it is severe, descend from a paperback. The bands were never calibrated against anything. A number that sorts you into a category is doing the thing our attachment style test says on its own page that a score cannot do.
It also explains the texture of the advice. When the measurement is borrowed from a trade book, so is the model underneath it, and so is the prescription that comes out the other end. Which is why so much of this subject consists of confident instructions with nothing behind them.
What the first proper validation found, including the result about agreeableness
In 2025, Xiaoxue Kuang and colleagues did the work: a psychometric validation of the Chinese People-Pleasing Questionnaire, published in PsyCh Journal, open access.
The design was three-stage. 602 first-year students in pre-testing, 2,498 in formal testing, and 2,340 again a month later with additional measures; after exclusions the analytic sample was 2,203 students. The original 24-item version fitted badly (CFI = 0.817, TLI = 0.797) and had to be cut, first to 18 items and then to 13, which fitted well (CFI = 0.966, TLI = 0.957, RMSEA = 0.066) across three factors — thought, behaviour and feeling — with the structure holding across sex, discipline, urban or rural origin, and time.
The items that had to be thrown out are worth reading, because they are the sentences the popular version is built from. I believe that nice people get the approval, affection, and friendship of others. I have to give of myself all the time in order to be worthy of love. It is extremely important to me to be liked by nearly everyone in my life. I would think that I am a bad person if I didn't give of myself all the time. Those four were dropped for loading on more than one factor; two more went for loading on nothing.
And then the correlations, in 2,203 people. Higher people-pleasing went with neuroticism (r = .42), loneliness (r = .41), social avoidance (r = .29), lower core self-evaluation (r = −.38) and lower mental well-being (r = −.25).
Its correlation with agreeableness was −.075. Essentially nothing.
That is the Helgeson and Fritz distinction appearing again, from a different instrument, in a different language, on the other side of the world. Whatever this trait is, it is not being a nice person. It sits with anxiety, loneliness and a low opinion of yourself — and being pleasant to be around is not part of it.
It is not the giving. It is what the giving is for.
If the trait is not warmth, the next question is which part of the behaviour carries the cost. There is a clean answer, and it is older than most of the advice built on it.
In 2005 Emily Impett, Shelly Gable and Letitia Anne Peplau published what they described as the first empirical investigation of motives for sacrifice in relationships — which tells you that until then nobody had separated them. In the first study they catalogued what everyday sacrifices actually consist of between dating partners and built a measure of why people make them. In the second, a two-week daily-experience study with both members of the couple reporting, they tested what the motives predicted.
The split is between approach motives — doing it to create something good, to make the other person happy, because you want to — and avoidance motives: doing it to prevent conflict, to stop the other person being upset with you, to avoid feeling guilty.
Approach motives went with better personal well-being and better relationship quality. Avoidance motives went with worse on both. And sacrificing for avoidance motives was particularly bad for keeping the relationship going over time. Perceptions of the partner's motives mattered too: what you think they gave it up for is part of what you are responding to.
Sit with how that reframes the search. The question people type is about how much they give. The variable that did the work in the measurement is why, and the two are close to independent — the same act, the same evening, the same cancelled plan, with opposite associations depending on what it was for.
It also explains why the standard advice tends to miss. Give less is an instruction about quantity. The thing the data separate is a motive, and you can reduce the quantity without touching it: four fewer favours a week, every one of them still done to stop somebody being annoyed with you.
The same split in couples, measured three different ways
A daily diary of college students in dating relationships is a thin base for a claim this large. The follow-up is sturdier, and it is the strongest piece of evidence on this page.
Impett, Judith Gere, Aleksandr Kogan, Amie Gordon and Dacher Keltner recruited 80 dating couples and measured the same thing in three different ways. Part one: both partners came to the laboratory and discussed a sacrifice one of them had made. Part two: fourteen days of daily reports. Part three: a follow-up survey three months later.
In the laboratory, couples who discussed a sacrifice made for approach goals experienced greater relationship quality; couples discussing one made for avoidance goals experienced worse. Outside observers watching the recordings produced the same result, which matters, because it means the effect is not purely a mood colouring how people filled in a form.
In the daily data, on days when someone sacrificed for approach goals both partners' relationship quality rose. On days when someone sacrificed for avoidance goals, the giver's fell. Positive and negative emotions mediated the two effects respectively.
And at three months, approach goals predicted increases in relationship quality and avoidance goals predicted decreases, as reported by both partners. The authors' own one-line summary is the one to keep: sacrifice as such does not help or harm relationships; what shapes the bond is the goal the person was pursuing when they gave something up.
The wider literature refuses to make this simpler. A meta-analysis of 82 datasets and more than 32,000 people found the facets of sacrifice pointing in opposite directions — willingness to sacrifice going with better well-being for both partners, the costs of sacrificing going with worse — and that evidence is set out alongside the rest of the investment-model work in why people stay in relationships that are not making them happy. Eighty couples is still eighty couples, mostly young, mostly students. But three methods agreeing is a different kind of result from one diary.
The counterweight: for some people, holding it in was not a cost
This is where a page that only wanted to sell you an identity would stop. The same research group published the result that complicates it, and leaving it out would be dishonest.
There is already good evidence that suppressing your feelings during a sacrifice is expensive. One study looked at it from three angles and found costs for the listening partner as well as the suppressor, declines in well-being and relationship quality reported by both members of the couple over fourteen days, and lower satisfaction and more thoughts of breaking up three months later — with authenticity as the mediator. That work, and what it implies about naming a state without attaching a demand to it, is in the habits of emotionally mature people.
Then Bonnie Le and Emily Impett ran a fourteen-day daily-experience study that asked whether that is true of everybody. It is not. They measured interdependent self-construal — the degree to which a person's sense of self includes their close relationships — and it moderated the whole effect. People higher in interdependence who suppressed negative emotions during sacrifice experienced boosts in personal well-being and relationship quality. People lower in interdependence who did the same thing experienced the drops. Feelings of authenticity mediated both directions: for the first group, holding it in felt like being themselves.
That is a result with teeth, and it is worth saying what it does and does not license. It does not mean that silence is good for you if you are the self-sacrificing type; that would be reading it backwards, and the self-silencing literature points the other way. It means the cost of not voicing something is not a constant of human nature. It depends on whether the act reads, to the person performing it, as a violation of who they are.
So stop being so accommodating is not a universal prescription, and anyone issuing it as one is going past the evidence.
The doormat effect, in more detail than the name usually gets
The most quoted finding in this whole area is also the most mangled, so it is worth going through properly.
In 2010 Laura Luchies, Eli Finkel, James McNulty and Madoka Kumashiro published The Doormat Effect in the Journal of Personality and Social Psychology. The claim is not that forgiving is bad for you. It is conditional, and the condition is the other person's behaviour: forgiving bolsters self-respect and self-concept clarity when the person who hurt you has signalled that you will be safe and valued in a continued relationship, and erodes both when they have not.
Study 1 is the one worth knowing. Seventy-two newlywed couples in Ohio — 144 people, assessed within six months of the wedding and then every six to eight months, yielding seven measurements of self-respect across roughly the first five years of marriage. The measured conditioner was the spouse's agreeableness, as a standing tendency to behave in a way that signals the partner is safe.
The three-way interaction between forgiveness, partner agreeableness and time was significant: β = .09, t(297) = 2.65, p = .009. Splitting it, for people who forgave readily, the trajectory of their self-respect depended on who they were married to. With an agreeable spouse, self-respect drifted up, though not quite significantly (β = .12, p = .07). With a disagreeable one, it fell, significantly (β = −.13, t(68) = −2.03, p = .05).
And the control analysis is the detail that makes it a finding rather than a mood. When the authors ran the same model predicting self-esteem instead of self-respect, the interaction vanished (β = −.004, p = .92). Something specific was moving: not how good people felt about themselves in general, but whether they regarded themselves as a person who had stood up for their own principles.
The two experiments, and the half that did not come out
A longitudinal association is not causation, so the paper ran two experiments. Both are worth reporting in full, including the parts that did not work, because the headline has travelled a very long way from them.
Study 2 recruited 49 Northwestern undergraduates; five were excluded for guessing the deception, leaving 44. Each recalled a real, recent, unresolved incident in which somebody close had hurt them — friends 50%, romantic partners 21%, family 25%. They then sat a fake “forgiveness test,” a doctored reaction-time task, and were given random feedback telling them they had or had not forgiven. Separately, they were told their perpetrator's amends ranked in the 17th or the 83rd percentile of everyone who had taken part.
For self-respect, the interaction was not significant: F(1, 43) = 1.04, p = .31. For self-concept clarity it was: F(1, 43) = 6.50, p = .02. Among people told the other person had made only weak amends, those told they had forgiven reported markedly lower clarity about who they were (M = 5.40) than those told they had not (M = 7.45), t(19) = −2.25, p = .04. On the strong-amends side, the difference ran the predicted way and was not significant.
Study 3 used a controlled vignette with 247 undergraduates: a partner tells a mutual friend something private, then either does or does not apologise and try to make up for it, and you either do or do not forgive them. Here both measures came through. Imagining forgiving without amends produced lower anticipated self-respect (M = 3.34 against 4.50), t(125) = −3.44, p < .001, and lower self-concept clarity (3.69 against 4.65), t(125) = −3.36, p < .001. Making the betrayal more or less upsetting changed nothing.
Note the pattern across both. The erosion half kept reaching significance. The bolstering half never did.
The pooled result, and the study where the effect went missing
Study 4 is the one that should temper how this gets quoted, and it is the one nobody mentions.
Sixty-nine Northwestern first-years in dating relationships filled in a questionnaire every two weeks for six months — fourteen waves. Eleven of them reported no betrayal at all during the half-year and were dropped, leaving 58 people and an average of 3.4 incidents each. For every incident they rated how distressing it was, how far they had forgiven, and whether their partner had tried to make amends.
For highly distressing betrayals the interaction appeared, for self-respect (β = .12, p = .04) and for self-concept clarity (β = .20, p = .002). But look at which side it came from. Forgiving when the partner had made strong amends predicted bolstered clarity (β = .27, p = .01) and nearly bolstered self-respect (β = .19, p = .07). Forgiving when they had not made amends — the doormat half, the whole reason the paper is famous — was in the right direction and not significant on either measure (p = .59 and p = .22).
The authors were straightforward about this. They state that all fourteen simple effects across the studies ran in the predicted direction but that not all reached significance, and they ran an internal meta-analysis of Studies 2 to 4 to settle it. Pooled, the doormat effect held: β = −.21, z = −3.30 for self-respect and β = −.25, z = −3.91 for self-concept clarity. The bolstering effect also held, more modestly: β = .14 and β = .17.
So the honest summary is: a real effect, modest in size, measured almost entirely with single-item questions, in American samples, and visible in the pooled analysis more reliably than in any one study. The authors' own stated limits are that everything is explicit self-report, that participants were all in the United States, and that the findings may not extend to betrayals by strangers or by people you are no longer close to. A correction notice was published in the same journal two months later; it is paywalled, and what it amends could not be read.
The variable is amends, not forgiveness
Strip the doormat studies down and what is left is a reframe that does more work than any instruction about being less nice.
In every study, the thing that decided whether forgiving cost the forgiver anything was not a property of the forgiver. Not how forgiving they were by disposition, not how quickly they did it, not how upset they had been. It was whether the other person had acted in a way that signalled the relationship was safe — dispositionally, through agreeableness, in the married sample; event by event, through amends, in the rest.
That relocates the question. Am I too forgiving? has no answer, because the trait was not the active ingredient. Did anything change? does. Did they take responsibility for the specific thing, rather than for being an imperfect person in general? Has it happened again since? The authors' own practical note is that it may not be prudent to recommend forgiveness without considering whether the perpetrator has acted in a way signalling the victim will be safe and valued.
This fits an older result that the same group's work grew out of: in the first two years of marriage, people whose spouses rarely behaved badly had steadier satisfaction the more forgiving they were, while people whose spouses behaved badly often declined faster the more forgiving they were. That finding, and what the Gottman repair research does and does not support about the conversation afterwards, is in how couples actually repair after a fight.
And it gives the one question worth carrying out of this page. Not how do I stop forgiving so easily, which is an instruction to become somebody else. But: when I let this go last time, did the behaviour change? If the answer has been no for a while, the measured problem is not your character. It is a situation in which forgiving is doing nothing, and the research says that is the condition under which it costs you something.
Your model of how other people will react has been tested
Underneath every version of this question is a prediction: that if you decline, something bad will happen. That prediction is itself a measurable thing, and it has been measured — though in a neighbouring situation, and the difference matters.
Francis Flynn and Vanessa Lake ran six studies on people asking strangers for help: fill in a questionnaire, lend a phone, walk them to a gym. Before each attempt, askers predicted how many people would say yes. Across experimental and real field settings, they underestimated compliance by as much as 50%. Roughly twice as many people agreed as the asker had expected.
Studies 4 and 5 showed the bias was produced by the position rather than the person: put the same participant into the helper's seat and the estimate moved. And Study 6 identified the mechanism, which is the part that belongs on this page. Help seekers were less willing than potential helpers to appreciate the social costs of refusing a direct request — how awkward it is to say no to someone's face — and attended instead to the instrumental costs of helping: the time, the effort, the inconvenience of saying yes.
Two people, one request, two completely different cost sheets. The asker is adding up what a yes would cost. The person being asked is feeling what a no would cost.
Now turn it around, carefully, because this is an inference and is labelled as one. A person who finds it hard to decline is someone for whom the social cost of refusing is permanently vivid — the exact quantity the research shows other people routinely underweight when they imagine being in your position. Nobody has run the mirror experiment and measured whether people overestimate how badly a refusal lands. That study does not exist, and this page is not going to pretend it does.
What the evidence establishes is narrower and still useful: predictions about how others will respond to a direct request are systematically wrong, in a direction set by which chair you are sitting in.
Saying it plainly feels like failure, and predicts change anyway
If you do decide to say something, there is one study that measures what happens next, and its result is almost designed to be ignored.
Nickola Overall, Garth Fletcher, Jeffry Simpson and Chris Sibley videotaped 61 couples while each partner tried to produce a change in the other. They coded the strategies on two dimensions: positive or negative, and direct or indirect. Then they measured actual change in the targeted feature at three-month intervals for a year.
The immediate verdict and the eventual outcome came apart completely. Direct strategies — stating the problem, naming what you want — were rated as relatively unsuccessful right after the conversation, and predicted more change over the following twelve months, as reported by the partner on the receiving end. Positive-indirect strategies — softening, hinting, being encouraging around the edge of it — felt more successful at the time and predicted no change at all.
That is the specific trap of being accommodating, stated in a measurement. The indirect approach is the one that produces the pleasant conversation. It is also the one that produces nothing twelve months later, which is why the same request can be made gently for years without either person being dishonest about it. And the feedback signal available to you — how the conversation felt — is pointing the wrong way.
It is worth being precise about the scope. Sixty-one couples, in New Zealand, trying to change existing features of a partner; this is not about declining a request at work or with a parent. There are situations where directness is unsafe, and a later section says so. The same study's nulls and its cautions turn up again in the evidence on asking an avoidant partner for something and on what actually settles an anxious one, where generic warmth also kept failing to do the work people expect of it.
What it supports is small: if you say the thing and it feels like it did not land, that feeling is not evidence.
There is a treatment for this, it has a name, and the field mislaid it
Here is a fact the first page of search results will not give you. The problem described in all those posts has a corresponding clinical intervention, with decades of trial evidence, and it has largely fallen out of use.
In 2018 Brittany Speed, Brandon Goldstein and Marvin Goldfried published a review in Clinical Psychology: Science and Practice titled, flatly, Assertiveness training: a forgotten evidence-based treatment. Their case is that a substantial body of research links assertiveness to a range of clinical problems and populations, that assertiveness training works as a transdiagnostic intervention, and that research on it as a stand-alone treatment has diminished drastically since its peak.
The shape of the literature bears that out. On 28 October 2026, PubMed held 251 records with assertiveness training in the title or abstract, 21 of them mentioning randomisation, and nine systematic reviews or meta-analyses. But look at what those reviews are about: nurses speaking up about medical errors, communication training for healthcare students, interventions for outwardly-directed aggression in people with intellectual disabilities. The technique survives inside occupational and clinical niches.
Nobody has run a trial of assertiveness training on the population asking this question — ordinary adults who feel they accommodate too much in their relationships. The thing the search is for has a treatment, and the treatment has never been tested on the thing the search is for.
That is still more actionable than anything else on this page, for one reason: it is a name. It is far easier to ask a therapist about a structured intervention with a forty-year literature than to arrive with I think I'm a people-pleaser, a phrase whose first validated questionnaire is from last year. And it points at skills — making a request, refusing one, disagreeing out loud — rather than at an identity to adopt.
Nothing here is a diagnosis or a treatment recommendation, and this site does not make either. It is a pointer to the vocabulary that gets you a better conversation.
The bill for being less nice is itemised, and it is not split evenly
Every page on this subject tells you to be more assertive. Almost none of them mention that the cost of doing so has been measured, and that it is not the same for everyone.
Hannah Riley Bowles, Linda Babcock and Lei Lai ran four experiments on what happens when someone initiates a negotiation rather than accepting what they are offered. In the first two, participants read accounts of candidates who did or did not ask for more. Evaluators penalised female candidates more than male candidates for asking. In the third, with videotaped candidates, male evaluators penalised women more than men; female evaluators penalised everybody for asking. And the mediator the authors identified is the word this whole article is about: perceptions of niceness and demandingness explained the resistance. In the fourth experiment, facing a male evaluator, women were less inclined to negotiate than men, and nervousness explained the difference. Facing a female evaluator, there was no gender difference at all.
Emily Amanatullah and Michael Morris found the same machinery from the inside. Women hedged their assertiveness when bargaining for themselves and obtained worse outcomes — and did not when bargaining on someone else's behalf, where they did not expect to be judged for it. Anticipated backlash statistically mediated the effect. Laurie Rudman's experiments showed the same trade in impression management: self-promotion raised competence ratings for women and cost them social attraction and hireability.
So the instruction has a price tag, the price is partly set by whoever is in the room, and in these studies it fell unevenly. That is not a reason to keep quiet. It is a reason to stop treating the difficulty as a personal failing, and to notice that the nervousness in experiment four was tracking something real.
The limits: these are laboratory and vignette studies in workplace settings, not relationships, and they are about negotiation rather than everyday accommodation. The related argument about who is asked to do the invisible adjusting, and at whose expense, is set out in what happens when friendliness gets read as flirting and is not repeated here.
The dating version of this question is about a label, not a behaviour
A share of the people asking how to stop being so nice are not asking about work or family. They are asking why it has not worked out romantically, and the folklore has an answer ready: nice guys finish last.
That has been studied, and the studies cannot carry what is asked of them. Edward Herold and Robin Milhausen surveyed 165 university women in 1999 and asked directly about the nice guy stereotype. More than half agreed that nice men have fewer sexual partners. More than half also said they would prefer a nice man over a bad boy as a date. Women who put less emphasis on sex, had fewer partners, and were less approving of men with many partners were more likely to choose the nice one. The authors' summary was that nice men are likely to have fewer sexual partners but are more wanted for committed relationships.
Two things follow. The first is that the folklore is not supported in the form it circulates — the same respondents who endorsed the stereotype preferred the person it describes. The second is more important and applies to the whole nice-guy literature: these studies manipulate a described character in a vignette, a label attached to a stranger on a page. They do not observe anyone accommodating anybody. So they cannot answer the question this article is about, which is what happens to a real person who cannot say no.
There is also a premise here worth refusing outright. The complaint is usually that kindness was spent and not repaid, and that frame — decency as a deposit, interest as the return — is the thing that makes the whole genre unpleasant. Nobody owes anybody attraction. The evidence on why that pattern actually persists, and on the single move that resolves it, is in why people keep ending up in the friend zone, and the reason intensity organised around uncertainty keeps getting mistaken for depth is in how love and anxious attachment come apart. More of this under Dating.
When this stops being a personality question
There is a version of this situation that no amount of work on yourself will touch, and the vocabulary of people-pleasing is very good at hiding it.
Everything above assumes a world in which saying no is uncomfortable. Some people are asking this question from inside a different one, where saying no has a consequence: an argument that lasts days, a silence that has to be worked off, money withheld, something broken, or being told afterwards that you caused it. If declining reliably costs you something, you are not failing to be assertive. You are responding accurately to your situation, and the research on traits has nothing to say about it.
The markers are specific rather than atmospheric. Needing to account for where you were and who you spoke to. Your phone being checked. Being discouraged from seeing friends or family. Decisions about money, work or time being made for you. Being told that your reaction to something is the real problem. The National Domestic Violence Hotline publishes this list and notes that even one or two of these is a warning sign.
If any of this is familiar, that is the thing to act on, before anything else on this page. Find A Helpline, in the sources below, lists free and confidential services by country. In the United States the National Domestic Violence Hotline is 1-800-799-7233, or text START to 88788, around the clock. If what you have is a difficult relationship rather than that, our free conflict style test takes about two minutes and says plainly what it cannot tell you.
The distinction between a feeling and a pattern of control is not a formality — in the meta-analytic literature jealousy turns up alongside controlling behaviour rather than instead of it, and the psychology of jealousy goes through what has been measured. And it runs the other way too: being told you are too nice to everyone can be a way of narrowing who you are allowed to be warm to.
What the evidence actually supports doing
Everything above produces a short list, and it is shorter than the genre promises. Each item is tied to something measured, and the inferences are flagged as inferences.
Audit the motive, not the quantity. The variable that separated good outcomes from bad in two studies and a three-month follow-up was approach versus avoidance: doing it to make something good happen, or doing it to stop something bad. That question can be asked of a single favour, tonight, and it does not require you to give less of anything.
Ask whether anything changed, not whether you forgive too easily. The doormat studies found the cost sitting in the other person's amends, not in your disposition. Repeated forgiveness where nothing changes is the measured condition, and it is a fact about the situation.
Expect directness to feel like it failed. In the one study that tracked it for a year, direct requests felt unsuccessful and predicted change; the pleasant indirect version felt successful and predicted none. The in-the-moment signal is unreliable, so do not use it to decide whether to try again.
Distrust your forecast of the reaction. People asking for things underestimated compliance by up to half, because they were pricing the cost of a yes while the other person was feeling the cost of a no. Your forecast is produced by which seat you are in.
Do not assume silence is costing you the same as it costs other people. For people whose sense of self includes their relationships, holding something in during a sacrifice went with higher well-being. This part is genuinely individual.
And if the pattern is constant and old, there is a name to take to a clinician — assertiveness training — rather than a label to take on. The single most repeated instruction on the subject, be less warm, appears nowhere on this list, for the reason in the next section.
What nobody has measured
The limits belong on the page in the same size type as the findings, because this is a subject where the confident version is always wrong.
Nobody has tested the fix. There is no study in which people were assigned to accommodate less and researchers measured what happened to their relationships or their well-being. Every instruction on the first page of results, the restrained list above included, is reasoning from adjacent findings. Assertiveness training is the closest thing to an exception and has never been trialled on this population.
Almost all of it is cross-sectional. Self-silencing and depression correlate at r = .39 across 10,108 people, and that number cannot tell you which way the traffic runs. The same is true of the people-pleasing questionnaire's correlations, and of the unmitigated communion work.
The samples are narrow. First-year students at one Chinese university. Northwestern undergraduates, three times. Seventy-two newlywed couples in Ohio. Sixty-five cardiac patients. Eighty dating couples. The backlash experiments are workplace vignettes with student evaluators, not relationships.
The doormat effect rests on single-item measures, and its headline half was not significant in the study with real betrayals in real relationships — it is visible in the pooled analysis.
The fawn response has no instrument. No validated measure, no prevalence estimate, no study separating it from the constructs that do have measures. The review that most recently built a framework around it says so itself.
And the measurement is one year old. The first properly validated people-pleasing questionnaire was published in 2025, in one cultural context whose authors note that the same behaviour is read there as a social virtue. What that means for anybody else has not been established.
What survives is still more than the genre offers: the trait that tracks distress is not kindness and barely correlates with agreeableness; the cost of giving something up sits in the motive; the cost of forgiving sits in the other person's amends; your prediction of how a refusal lands is made in the wrong chair; and the one sentence that works will feel like it did not. More under Mindset, and the claims this site has had to retract are listed on the corrections page.
Pulling it together
If one thing survives from all of this, it is that the question is aimed at the wrong variable. Being kind, warm, generous or easy to get along with is not the thing that tracks distress in any of this research — in the 2025 validation sample of 2,203 people, people-pleasing and agreeableness correlated at −.075, which is nothing. What tracks distress is needing other people's reactions in order to know what you are worth, and then organising a life around producing them.
The second thing is that the usable levers are smaller and more specific than the genre's. Not give less, but notice what a particular act was for — approach or avoidance, to make something good happen or to stop something bad. Not forgive less, but ask whether anything changed after the last time. Not be more confident, but expect the direct sentence to feel like it failed, because in the one study that tracked it for a year, that is exactly how it felt and it was the version that worked.
And the third is the limit. Nobody has run the experiment. No study has assigned people to be less accommodating and measured what happened next. Anyone telling you otherwise, in a confident voice, with a numbered list, is reasoning from adjacent findings — including this page, which says so each time it does it.
Frequently asked questions
How do I stop being too nice?
The research does not support the instruction in that form, because the thing that goes with distress is not niceness. Across four studies Helgeson and Fritz separated communion — a caring orientation towards others — from unmitigated communion, a focus on others to the exclusion of the self, and only the second went with psychological distress. The 2025 validation of the Chinese People-Pleasing Questionnaire in 2,203 students found the same shape from a different instrument: people-pleasing correlated .42 with neuroticism, .41 with loneliness and −.38 with core self-evaluation, and −.075 with agreeableness. What the evidence does support is narrower. Impett, Gable and Peplau separated approach motives for sacrifice from avoidance motives and found the two associated with opposite outcomes, so the useful audit is what a given act was for rather than how much you give. And if the pattern is old and constant, assertiveness training is a real intervention with a real literature, which is a better thing to take to a clinician than a label.
Is people-pleasing a trauma response?
Nobody has shown that, and the claim is much newer and thinner than it sounds. The fawn response was named by the psychotherapist Pete Walker in an article in The East Bay Therapist in January 2003; he writes “I have named it the fawn response,” defines it from Webster's dictionary, and offers clinical observation rather than data. A PubMed title-and-abstract search for the exact phrase “fawn response” returned zero records when this page was written, and a 2026 review that builds a fight–flight–freeze–fawn framework states that fawning remains underdeveloped and understudied. The nearest academic proposal, replacing Stockholm syndrome with “appeasement,” rests on Polyvagal Theory, which has itself been challenged premise by premise in Biological Psychology. None of that means the experience is not real. It means there is no measure, no prevalence figure and no study separating it from constructs that do have measures, so a page explaining your behaviour through it is not reporting evidence.
Is being too nice a sign of low self-esteem?
The two travel together in every sample anyone has measured, but the direction has not been established. Fritz and Helgeson's model has unmitigated communion individuals relying on others for self-esteem, which produces overinvolvement with others and neglect of the self, and those in turn account for the link to distress — but it is a cross-sectional model rather than a demonstration of cause. In the 2025 people-pleasing validation, core self-evaluation correlated −.38 with the total score. And the self-silencing literature, pooled across 31 published and 11 unpublished studies with 10,108 participants, puts the correlation with depression at r = .391, again without being able to say which produces which. The honest answer is that they are strongly associated, that the clinical accounts describe traffic in both directions, and that nobody has run the study that would settle it.
How do I say no without feeling guilty?
The research does not offer a way to remove the feeling, but it does say something about the prediction underneath it. Flynn and Lake ran six studies on people making direct requests and found they underestimated compliance by as much as 50%. Study 6 located the reason: people in the asking position were less willing to appreciate the social cost of refusing, and attended instead to the instrumental cost of helping. Two people, one request, two different cost sheets — the asker pricing a yes, the asked feeling a no. Someone who cannot decline is someone for whom the cost of refusing is permanently vivid, which is the quantity other people routinely discount. Note the limit honestly: nobody has run the mirror experiment measuring whether people overestimate how badly a refusal lands, so that last step is an inference and is labelled as one here.
Is forgiving too easily bad for you?
Not by itself — what the research found is conditional on the other person. In Luchies, Finkel, McNulty and Kumashiro's doormat studies, forgiving eroded self-respect and self-concept clarity only when the person who hurt you had not signalled you would be safe and valued. Among 72 newlywed couples followed for roughly five years, people who forgave readily gained self-respect if their spouse was agreeable and lost it if the spouse was not; the same model predicting general self-esteem found nothing, so something specific was moving. In two experiments, imagining forgiving without amends produced significantly lower self-respect and self-concept clarity. The active ingredient was never the forgiver's disposition. So the useful question is not whether you forgive too easily, it is whether anything changed after the last time.
What is the difference between being kind and being a people-pleaser?
The measurement answers this unusually cleanly. Helgeson and Fritz built their whole theory on separating communion — caring about other people — from unmitigated communion, which is a focus on others to the exclusion of the self and shows up alongside a negative view of the self and reliance on others for self-evaluation. Only the second one is associated with distress. In Fritz's study of 65 patients after a first coronary event, unmitigated communion predicted worse depression, functioning and cardiac symptoms four months later, while communion was unrelated to all of it. And in the 2025 people-pleasing validation, the correlation with agreeableness was −.075. The practical reading: kindness that leaves you intact and kindness used to find out whether you are acceptable are different things, and only one of them appears in the findings.
Does giving things up for a partner damage a relationship?
It depends entirely on why, and this is the best-supported result on the subject. Impett, Gable and Peplau measured approach motives for sacrifice — doing it to create something good — separately from avoidance motives, doing it to prevent conflict or guilt, in a two-week daily study of couples. Approach motives went with better well-being and relationship quality; avoidance motives went with worse, and were particularly bad for keeping the relationship going. A later study of 80 couples replicated it three ways: a laboratory discussion scored by outside observers, 14 days of daily reports, and a three-month follow-up in which approach goals predicted increases and avoidance goals decreases, as reported by both partners. The authors' own summary is that sacrifice as such neither helps nor harms; the goal being pursued is what shapes the bond.
Will being more assertive make people like me less?
In the measured settings it carried a cost, and the cost was not distributed evenly. Bowles, Babcock and Lai ran four experiments on initiating a negotiation and found evaluators penalised female candidates more than male candidates for asking, with perceptions of niceness and demandingness explaining the resistance; with male evaluators women were less inclined to negotiate and nervousness explained the gap, while with female evaluators there was no gender difference at all. Amanatullah and Morris found women hedged when bargaining for themselves and did not when bargaining for someone else, mediated by anticipated backlash. Rudman found self-promotion raised women's competence ratings while costing social attraction and hireability. Those are workplace vignettes and laboratory negotiations rather than relationships, so they do not transfer directly. What they establish is that the difficulty is not purely in your head.
Do nice people really finish last in dating?
The studies usually cited cannot answer it, and they do not say what the folklore says. Herold and Milhausen surveyed 165 university women: more than half agreed that nice men have fewer sexual partners, and more than half also said they would prefer a nice man over a bad boy as a date, with the authors concluding that nice men are more wanted for committed relationships. More importantly, that literature manipulates a described character in a vignette — a label attached to a stranger on a page — rather than observing anyone actually accommodating anyone, so it cannot speak to what happens to a real person who cannot say no. The premise is also worth refusing: kindness is not a deposit and attraction is not the interest on it. Why the pattern persists, and the one move that resolves it, is covered separately.
Is there any treatment for this?
There is one with a name and a long evidence base, and it has largely dropped out of view. Speed, Goldstein and Goldfried's 2018 review in Clinical Psychology: Science and Practice is titled “Assertiveness training: a forgotten evidence-based treatment,” and argues that a substantial body of research supports it as a transdiagnostic intervention while research on it as a stand-alone treatment has diminished drastically. The gap is real: PubMed holds 251 records with assertiveness training in the title or abstract, but the systematic reviews are in specific niches — nurses speaking up about medical errors, communication training for healthcare students, aggression in intellectual disability. No trial has been run on ordinary adults who feel they accommodate too much in their relationships. This page is not a diagnosis or a treatment recommendation. It is a pointer to the vocabulary that gets a better conversation with a clinician.
Sources
What this article draws on. For the broader bibliography see our sources page.
- Walker, P. (2003) — Codependency, Trauma and the Fawn Response. The East Bay Therapist, Jan/Feb 2003 — the article in which the term is coined (“I have named it the fawn response”); it cites a dictionary definition and clinical observation, and no empirical study
- PubMed (searched 28 October 2026) — Title/abstract search for the exact phrase "fawn response" — zero records; "fawning" returns 32, of which all but one are veterinary papers about deer
- Algarin, A. B., Lee, J. & Zhan, X. (2026) — Fight, Flight, Fawn, Freeze: Rethinking Substance Use Through a Stress Response Lens. Current Addiction Reports — open access; states that fawning and strategic assimilation remain underdeveloped and understudied
- Bailey, R., Dugard, J., Smith, S. F. & Porges, S. W. (2023) — Appeasement: replacing Stockholm syndrome as a definition of a survival strategy. European Journal of Psychotraumatology, 14(1), 2161038 — open access; a proposal built on Polyvagal Theory, noting that scarce empirical research supports the phenomenon it replaces
- Grossman, P. (2023) — Fundamental challenges and likely refutations of the five basic premises of the polyvagal theory. Biological Psychology, 180, 108589 — the challenge to the theory the appeasement proposal rests on
- Helgeson, V. S. & Fritz, H. L. (1998) — A theory of unmitigated communion. Personality and Social Psychology Review, 2(3), 173-183 — the distinction between communion and a focus on others to the exclusion of the self. Paywalled; reported here from the published abstract
- Fritz, H. L. & Helgeson, V. S. (1998) — Distinctions of unmitigated communion from communion: Self-neglect and overinvolvement with others. Journal of Personality and Social Psychology, 75(1), 121-140 — four studies; reliance on others for self-esteem leads to overinvolvement and self-neglect, which account for the link to distress. Paywalled; reported from the published abstract
- Helgeson, V. S. & Fritz, H. L. (1999) — Unmitigated agency and unmitigated communion: Distinctions from agency and communion. Journal of Research in Personality, 33(2), 131-158
- Fritz, H. L. (2000) — Gender-linked personality traits predict mental health and functional status following a first coronary event. Health Psychology, 19(5), 420-428 — 65 patients interviewed one week and four months after a first coronary event; unmitigated communion predicted worse outcomes and communion was unrelated
- Jack, D. C. & Dill, D. (1992) — The Silencing the Self Scale: Schemas of intimacy associated with depression in women. Psychology of Women Quarterly, 16(1), 97-106 — validated across three samples: 63 students, 140 refuge residents and 270 mothers
- Pintea, S. & Gatea, A. (2021) — The relationship between self-silencing and depression: A meta-analysis. Journal of Social and Clinical Psychology, 40(4), 333-358 — 31 published and 11 unpublished studies, 10,108 participants, 131 effect sizes; r = 0.391
- Kuang, X., Li, H., Luo, W., Zhu, J. & Ren, F. (2025) — The mental health implications of people-pleasing: Psychometric properties and latent profiles of the Chinese People-Pleasing Questionnaire. PsyCh Journal, 14(4), 500-512 — open access; 2,203 students, the 24-item version cut to 13, and the correlation with agreeableness of −.075
- Impett, E. A., Gable, S. L. & Peplau, L. A. (2005) — Giving up and giving in: The costs and benefits of daily sacrifice in intimate relationships. Journal of Personality and Social Psychology, 89(3), 327-344 — the first study to separate approach from avoidance motives for sacrifice. Paywalled; reported from the published abstract
- Impett, E. A., Gere, J., Kogan, A., Gordon, A. M. & Keltner, D. (2014) — How sacrifice impacts the giver and the recipient: Insights from approach-avoidance motivational theory. Journal of Personality, 82(5), 390-401 — 80 dating couples across a laboratory discussion, a 14-day diary and a three-month follow-up. Paywalled; reported from the published abstract
- Righetti, F., Sakaluk, J. K., Faure, R. & Impett, E. A. (2020) — The link between sacrifice and relational and personal well-being: A meta-analysis. Psychological Bulletin, 146(10), 900-921 — 82 datasets, 9,547 effect sizes, N = 32,053; willingness to sacrifice and the costs of sacrificing point opposite ways
- Le, B. M. & Impett, E. A. (2013) — When holding back helps: Suppressing negative emotions during sacrifice feels authentic for people high in interdependence. Psychological Science, 24(9), 1809-1815 — a 14-day daily-experience study; the association reverses by interdependent self-construal, mediated by authenticity
- Impett, E. A., Kogan, A., English, T., John, O., Oveis, C., Gordon, A. M. & Keltner, D. (2012) — Suppression sours sacrifice: Emotional and relational costs of suppressing emotions in romantic relationships. Personality and Social Psychology Bulletin, 38(6), 707-720 — the costs of suppression during sacrifice, for both partners
- Luchies, L. B., Finkel, E. J., McNulty, J. K. & Kumashiro, M. (2010) — The doormat effect: When forgiving erodes self-respect and self-concept clarity. Journal of Personality and Social Psychology, 98(5), 734-749 — four studies; the full text, read here, is posted by the second author
- Luchies, L. B., Finkel, E. J., McNulty, J. K. & Kumashiro, M. (2010) — Correction to Luchies, Finkel, McNulty, and Kumashiro (2010). Journal of Personality and Social Psychology, 99(1), 119 — a correction notice published two months after the article; it is paywalled and what it amends could not be read
- Flynn, F. J. & Lake, V. K. B. (2008) — If you need help, just ask: Underestimating compliance with direct requests for help. Journal of Personality and Social Psychology, 95(1), 128-143 — six studies; people underestimated compliance by as much as 50%, and Study 6 traced it to discounting the social cost of refusing
- Overall, N. C., Fletcher, G. J. O., Simpson, J. A. & Sibley, C. G. (2009) — Regulating partners in intimate relationships: The costs and benefits of different communication strategies. Journal of Personality and Social Psychology, 96(3), 620-639 — 61 couples videotaped, with change rated every three months for a year; direct strategies felt unsuccessful and predicted change
- Speed, B. C., Goldstein, B. L. & Goldfried, M. R. (2018) — Assertiveness training: A forgotten evidence-based treatment. Clinical Psychology: Science and Practice, 25(1) — the review arguing that research on assertiveness training as a stand-alone treatment has diminished drastically
- Bowles, H. R., Babcock, L. & Lai, L. (2007) — Social incentives for gender differences in the propensity to initiate negotiations: Sometimes it does hurt to ask. Organizational Behavior and Human Decision Processes, 103(1), 84-103 — four experiments; perceptions of niceness and demandingness explained resistance to female negotiators
- Amanatullah, E. T. & Morris, M. W. (2010) — Negotiating gender roles: Gender differences in assertive negotiating are mediated by women's fear of backlash and attenuated when negotiating on behalf of others. Journal of Personality and Social Psychology, 98(2), 256-267
- Rudman, L. A. (1998) — Self-promotion as a risk factor for women: The costs and benefits of counterstereotypical impression management. Journal of Personality and Social Psychology, 74(3), 629-645 — three experiments; self-promotion raised competence ratings and cost social attraction and hireability
- Herold, E. S. & Milhausen, R. R. (1999) — Dating preferences of university women: An analysis of the nice guy stereotype. Journal of Sex & Marital Therapy, 25(4), 333-343 — N = 165; more than half endorsed the stereotype and more than half preferred the nice guy as a date
- National Domestic Violence Hotline (United States) — Warning signs of abuse — 24/7 confidential support: 1-800-799-7233, or text START to 88788
- Find A Helpline — Free, confidential support lines by country, including domestic abuse services
Keep reading
- MindsetWhat Actually Helps an Anxiously Attached Partner, and What Only Helps TodayReassurance lowered attachment anxiety on the day and not six months later, in the one study that measured both. Something else predicted the decline.
- MindsetWhat Actually Helps an Avoidant Partner, and Why the Usual Advice Lands in the Worst ZoneAcross four studies and 298 couples, low and moderate practical support made avoidant partners measurably worse. Only high levels helped.
- MindsetTrauma Bonding: What the Term Actually Rests On, and Where the Seven Stages Came FromThe seven stages appear in no study. Four measures of trauma bonding do exist, and no page on the first results page mentions any of them.
- SignsWhen Friendliness Gets Read as Flirting, and What Can Actually Be ChangedMost people have had it happen, in both directions. In the one experiment that held the behaviour constant, what moved the reading was not the behaviour.
- DatingCushioning in Dating: What Has Actually Been Measured About Keeping Someone in ReserveThe word has no studies behind it. The behaviour has six, under another name — and the number of people someone keeps in reserve did not predict commitment.
- DatingBreadcrumbing in Dating: Where the 30% Figure Comes From, and What Was Actually MeasuredIt comes from 626 Spaniards surveyed in 2019, 86% of whom had never heard the word. In the same table, more said they had done it than had it done to them.