You can probably name the change you have not made. It has a date attached, and the date keeps moving. If someone asked why, you would likely reach for a word about strength — discipline, follow-through, the word people use when they are being kind about it, which is willpower.
Jérôme Blondé and Juan Manuel Falomir-Pichastor, at the University of Geneva, put a different question to 1,936 regular smokers in a 2026 paper in the European Journal of Social Psychology: who would you be afterwards?
A questionnaire that reads like something people actually say
The measure that carries this research is unglamorous. Five statements, rated from “No, not at all” to “Yes, absolutely.” The first reads: “If I were to quit smoking, I feel like I would no longer be the same.” Another: “If I were to quit smoking, I would feel disconnected from who I am.”
Read those slowly and they stop sounding like a survey. They sound like the thing people actually say, sideways, when they are explaining why this year is not the year.
In the first study, of 154 smokers in the United Kingdom, the average answer was 3.11 on a seven-point scale, which sits below the midpoint. The feeling is a variation among smokers rather than a condition all of them share, and the research is about what that variation goes with. In that first study it went with lower motivation to quit overall, before any mediator is taken out of it, and that is the only one of the four studies where the overall link reached significance.
Alongside it came six more questions: three about seeing yourself as a smoker, three about seeing yourself as an ex-smoker. Those were differenced and averaged into a single relative score running from −6 to +6, where a negative number means the smoker self still sits closer to the centre. The average in that first study was −2.26. The measure is not asking whether you have become someone new. It is asking which of two selves currently sits closer to the middle of you.
A year of asking the same people
The second study is the one with weight. It is pre-registered, and it followed regular smokers for a year, surveying them four times at four-month intervals. A thousand and sixteen people started. After exclusions the baseline sample was 943, and 684 were still answering at the final wave. At that last check, 117 of them, about 17 percent, reported they had quit and were not currently smoking.
Here is where the popular version of a finding like this would run ahead of the paper.
People who reported that feeling at the start were not measurably less likely to have quit by the next wave. The total effect of the feeling on the behaviour was non-significant at every transition the researchers tested. At one of them, the effect that survived once identity was held constant ran the other way: a positive direct effect of 0.09, which the authors flag as unexpected and do not return to. The plain correlation between the feeling and later quitting was not positive anywhere in that study; where it reached significance at all, it was slightly negative.
The route held instead. Smokers who reported that quitting would leave them no longer quite themselves scored lower on the relative ex-smoker measure, and that score predicted cessation four months later at every wave. The indirect path was statistically reliable at all three transitions, at −0.13, −0.04 and −0.02. The later two are smaller, but they are also estimated with prior quitting held constant, which the first is not, and the paper does not read them as a decline.
So the sentence that survives is a two-step one, and it resists compression. The feeling did not show up as a standalone predictor of who stopped. It went instead with not yet having anywhere to stand as the person who quit, and that is what tracked whether they quit. The authors are explicit that a missing total effect does not rule mediation out, and that their own data cannot establish which of the two steps came first.
The same word, elsewhere, has meant the opposite
The opposing evidence is in the paper’s own introduction rather than left for a critic. Feeling discontinuous from a past self, the authors note, has elsewhere been found to push people toward change: work on temporal landmarks, they write, found that dates which symbolically separate the present from the past can raise motivation to take up something new. Their argument is that these are two different discontinuities, one that separates you from a self you have already left, one that threatens the self you are still living in. That distinction is reasoned, and it is theirs; it is not something any of the four studies settled.
Two experiments, one reassuring paragraph
The last two studies push on the idea. Smokers read a short testimony from someone described as a former participant who had quit. In one version that person says the change was smaller than expected: “I really feel like I’ve stayed the same person!” Participants then wrote a few lines about what in them would stay put if they stopped.
It moved the dial on the feeling. In the third study, smokers who read the reassuring version rated their sense of becoming someone else at 2.00, against 3.35 for those who read nothing and went straight to the questions. They also reported carrying more of an ex-smoker self, and that in turn went with stronger stated motivation to quit.
The fourth study took 596 smokers, gave 288 of them the same reassurance, and came back a month later. Of the 527 who answered, 39 said they had quit and were still not smoking. The indirect effect ran the predicted way and was reliable. Neither the total nor the direct effect of the reassurance on quitting was: on its own, the testimony changed nothing measurable about behaviour.
One of the third study’s two control conditions misfired. The group asked to dwell on how much would change reported numerically less of the feeling than the group given no manipulation at all, though that contrast fell short of significance; the authors suggest the exercise made those smokers notice what would not change. The fourth study then dropped the active control altogether, which the authors say means that generic motivational boosts from reading a personal testimony, or from thinking hard about quitting, cannot be ruled out as the thing that moved. Cessation was self-reported throughout, by ticking the box that best described the situation, and the paper reports no biochemical confirmation. The sense of threat itself was never measured directly; it was inferred from the discontinuity ratings.
All of it also comes from one behaviour. The paper is about cigarettes, and its authors say the mechanism may not fully generalise, naming physical activity and dietary change among the places to test it. We are writers who read the paper, not clinicians, and this is a reading of four studies rather than anything resembling treatment or advice. If you are trying to stop smoking, a doctor or a national stop-smoking service can help you do it. If the harder part is who you would be without cigarettes, that is work a therapist does.
The finding itself is a mediation model across correlational and short-horizon experimental data, and the effect it describes shows up in aggregate and disappears inside any single life. Nobody in these studies was cured by a paragraph.
What it does is move the question to the far end of the change, which is worth something on the morning the date slides again. Reading it, the thing we keep landing on is not in the data at all: it may be hard to walk out of a life you know without some picture of who walks out, and most of us have never been told what of us survives the trip. The question the research leaves sitting on the table, then, is a plainer one than whether you want it enough. What would stay the same, if you stopped?
