Walter Mischel's study into impulse control began in the 1960s. He created what he called his "Surprise Room" at Stanford University's creche (amazing how sinister that sounds now). There, he tested pre-school kids on their self-control. It was simple: they could have one marshmallow immediately, or wait, alone in a room, for a given number of minutes, ring a bell and the researcher would give them two.

The results were astonishing to the team: good impulse control turned out to be a predictor of success in disparate areas, from academic results to health in later life. We have become used to the truth of this: deferred gratification enables the fulfilment of long-term goals; it is so commonplace an observation that you might find it in a fortune cookie or a posh Christmas cracker. So Mischel's description of the research team's surprise struck me as a little overdone, until I remembered that he was the one behind the results that we now hold as truisms. After that, his irksome habits in this book seemed less important. It is undeniable that he is not sceptical enough, but considering the impact he has had on psychological and criminological terrain, the humility in his tone is much more notable than his writing's longueurs.

To call Mischel the trailblazer would be to airbrush out the work of Freud, who intuited pleasure deferral as a component of self, a marker of increased maturity, in 1911. His argument was that the infant created a mental, "hallucinatory" image of the object of desire; the child's libidinal or sexual energy was directed at this hallucinatory image. Such a visual representation allowed "time-binding"; it enabled the infant to delay and temporarily inhibit the impulse for immediate gratification.

Mischel's is a much more practical approach than Freud's – his concern is with the lived experience of impulse control, the moment of failure, the moment of success, what can be learned and what unlearned. His first distinction is between "hot" responses to stimuli – dominated by the limbic system and irresistible – and "cool" responses, led by the pre-frontal cortex and eminently resistible. Think, if you've ever been a smoker, of that moment when you crumble: it is not when you conjure up the image of a cigarette, its size, shape and colour (these are supplied by your pre-frontal cortex, to help you distinguish between a cigarette and a straw); rather, it is when you imagine the way it hits the back of your throat, engulfs your breath with delicious smoke (that was my amygdala talking). Children who can be persuaded to think of marshmallows as puffs of cloud are more likely to overcome their temptation than children encouraged to think of the squidge and the sweetness. But, as per Freud's observations, children who were able to imagine the marshmallow with a frame around it, as an image rather than the thing itself, found it easier to defer gratification, too. Children who distracted themselves performed better. Hot and cool perception interlace with abstraction and diversion to create the person who, down the line, when you've seen their Sats scores and healthy internal organs, you'd call more "rounded".

Mischel references fMRI scans without sufficient caveat; where there is activity in the "stranger" area rather than the "self" area of the brain, that doesn't necessarily mean the subject is thinking of his or her future self as a stranger. The process isn't that fine-tuned. In its current stage, fMRI scanning has been likened to trying to discover what's going on in an office by looking at whether or not it has its lights on. We are a long way from being able to tell what a person is thinking about. The argument about self-perception, however – how different you perceive your future self to be from your current self, and what this says about your ability to save for the future or go to the gym – is interesting. If you can conceive of the future as a real place, this not only engages your "willpower", but also your ethical centre – respondents who identified themselves closely with their future selves were less likely to condone fraud and theft. Illustratively, Mischel suggests a technique for decision-making that involves bringing that future point, whatever it is, to now: how would you feel about x future job if you were to do it today? Or, in the opposite way, how do you feel about x argument or y love affair if you regarded it as a fly on the wall would, rather than immersing yourself in it?

These are relatively simple exercises, yet enlightening: trying to consciously embody the future makes you realise how distantly you really conceive of it, while stepping out of the self enables the cool appraisal that can dissipate heightened emotion. Though I think I'd already read the first advice in a Nancy Mitford novel (when you get an invitation to cocktails, you must always think: would I say yes if it were about to happen today, in half an hour?).

One of the core strategies in this thinking is the "if-then", a framework of encouraging self-talk that even young children concoct for themselves. "If x temptation occurs, then I'll embark on y strategy … It sounds simple, and it is. By forming and practising implementation plans, you can make your hot system reflexively trigger the desired response, whenever the cue occurs. Over time, a new association or habit is formed." This is a strategy of cognitive behavioural therapy; its close sibling is "thought-stopping" (used for anxiety disorders, where the problematic thought is replaced with a pleasurable one, until the two become automatic). A prison psychologist told me once that he'd tried it on a heroin addict; it hadn't worked, because the prisoner had replaced his temptation, heroin, with a pleasurable thought that was also heroin.

At the centre of all this is a sincerely held belief that the answer is within you – that if you can modify your hot animal with your cool reason, if you can strategise against any temptation, if you can conceive of yourself as a striver rather than a finished product, then yours is the world and everything that's in it. (Except, that is, for the things that are addictive or may negatively impact on you in some other way.) It's rather an atomised view of us all – batting away our demons in our own personal video game. But it's fascinating and often horribly recognisable.

• To order The Marshmallow Test for £16, go to bookshop.theguardian.com or call 0330 333 6846.