There is a reason motivational content aimed at girls rarely changes anything measurable. It usually targets the wrong construct.
Psychology distinguishes between global self-esteem — how positively a person regards herself overall — and self-efficacy, which is the expectation of being able to accomplish a specific task in a specific domain. The two are correlated but distinct, and their predictive power is not remotely equal.
What each one predicts
Global self-esteem correlates with wellbeing. It has a weaker and more contested relationship with performance and behaviour. Decades of research on self-esteem interventions found that raising self-regard, on its own, does not reliably raise achievement, and in some designs produced no effect at all.
Self-efficacy behaves differently. A student's belief that she can solve this type of problem, given effort, predicts whether she attempts it, how long she persists when it becomes difficult, and whether she chooses similar tasks in future. This relationship holds across academic domains, sport and health behaviour, and it holds after controlling for prior attainment.
The distinction is not academic pedantry. It determines what an intervention should try to do.
Why the difference matters for girls specifically
Where self-assessment gaps appear between girls and boys, they are largely domain-specific rather than global. Girls' overall self-esteem is not uniformly lower in most population studies. What differs is expectation of success in particular fields — most consistently in mathematics, physics and computing.
That is a self-efficacy pattern, not a self-esteem pattern. And self-efficacy has a well-described set of sources, which means it has a well-described set of levers.
Where self-efficacy comes from
The standard account identifies four sources, in rough order of strength.
Mastery experience is the strongest. Having actually done the thing — solved the problem, finished the race, shipped the code — builds expectation of being able to do it again more powerfully than anything else. Nothing else in the list comes close.
Vicarious experience is watching someone similar to you succeed. The similarity matters; this is the mechanism by which role models work, and it explains why the effect is stronger when the model resembles the observer in relevant ways.
Verbal persuasion — being told you can do it — is third, and considerably weaker than the first two. It has some effect, particularly from a credible source, and it can be undone quickly by a contradicting experience.
Physiological state is fourth. How your body feels during a task gets interpreted as evidence about the task. A racing heart read as fear lowers expectation; the same racing heart read as readiness does not.
Most interventions target the weakest lever
Assemblies, posters, hashtags and inspirational speakers are almost entirely verbal persuasion. They are the third item on a four-item list, and they are being asked to do the work of the first.
This is why a well-received talk produces enthusiasm on the day and nothing detectable a term later. Nothing in the student's actual experience changed. She was told something, and then she went back to a maths class where the same things happened as before.
Programmes that do produce measurable change tend to be structured around mastery. A coding club that ships a working project. A sports programme that runs long enough for skills to visibly improve. A research placement where a student produces a result. These are expensive and slow, which is precisely why they are less popular than assemblies.
The feedback problem
Mastery experience only builds self-efficacy if it is recognised as mastery. This is where feedback style becomes load-bearing.
Research on attribution shows that the explanation attached to a success matters. A result attributed to effort and strategy supports the expectation of repeating it. The same result attributed to luck, to the task being easy, or to the student being "naturally good" is much weaker — the first two because they are not repeatable, the third because it becomes fragile the moment something is hard.
There is a well-documented tendency for feedback given to girls to emphasise neatness, effort and compliance, and feedback given to boys to emphasise ideas and ability. Both patterns have costs. The relevant one here is that praise for diligence does not build expectation of succeeding at a harder problem; it builds expectation of being praised for working hard.
What to do with this
For anyone designing a programme: budget for repetition and completion, not for inspiration. The question to ask is not "will this be motivating?" but "will a participant end this having actually done something difficult?"
For teachers and coaches: attach explanations to outcomes. "You got that because you tried the second method when the first stalled" does work that "well done" does not.
For a student reading this: the finding that should interest you is that the strongest source of self-belief is evidence, and evidence is something you can go and manufacture. Doing a hard thing badly and then less badly is the mechanism. It is unglamorous and it is the one that works.
None of this means encouragement is worthless. It means encouragement is a supplement to experience rather than a substitute for it, and that programmes built entirely from the supplement should not be surprised when they do not nourish anyone.