The Pygmalion effect is the claim that a teacher's expectations about a student can become true partly because the teacher holds them, a self-fulfilling prophecy in which belief shapes behaviour, behaviour shapes outcomes, and the outcome confirms the belief. The name comes from the sculptor in Greek myth whose statue came to life, by way of Robert Rosenthal and Lenore Jacobson's 1968 book Pygmalion in the Classroom.
Rosenthal and Jacobson gave children at a Californian elementary school a test they described to teachers as identifying pupils about to "bloom" intellectually. In fact the named pupils were chosen at random; the only thing that changed was what teachers believed about them. The researchers reported that these randomly designated "bloomers" gained more on IQ measures over the year than their classmates, especially in the youngest grades. The interpretation was that teacher belief alone, with no change in the children themselves, had altered their measured ability. The book landed at a moment of intense argument about whether intelligence was fixed, and it spread quickly into teacher training as evidence that low expectations could quietly hold disadvantaged children back.
If expectations do influence outcomes, the mechanism is not telepathy but ordinary differences in how a teacher treats a child they expect much from. Rosenthal later organised these into four factors. Climate: warmer, more attentive emotional treatment of favoured pupils. Input: teaching them more material and harder content. Output: giving them more and longer opportunities to respond in class. Feedback: more, and more differentiated, responses to their answers, including praise. A child on the receiving end of richer instruction, more airtime, and more useful feedback may well learn more, which closes the loop and confirms the teacher's original expectation. The factors matter for practice because each is observable and adjustable, unlike a belief held privately.
The study drew sharp methodological criticism almost at once. Robert Thorndike attacked the IQ instrument, noting that some of the reported pre-test scores were so low as to be implausible and that the test was unreliable at the ages used, making the gains hard to trust. Richard Elashoff and Richard Snow's 1971 reanalysis questioned the statistical handling and found the dramatic effects concentrated in a few grades and shaky elsewhere. Replication attempts through the 1970s produced a mixed and often null record, and the original IQ-jump finding has not held up as a robust, general phenomenon.
Lee Jussim and Kent Harber's 2005 review in Personality and Social Psychology Review settled much of the debate by separating the strong claim from the defensible one. Self-fulfilling prophecies in classrooms are real, but the effects are typically small, they do not accumulate much across teachers or over time, and they are more likely to fade than to compound. The widely held image of expectations snowballing year on year into large achievement gaps is not what the evidence shows. The important qualification is that effects are larger and more durable for some groups, notably students from stigmatised or lower-achieving backgrounds, where a teacher's expectation has more room to move an outcome that is not already locked in by prior success. The honest summary is a modest, conditional effect, not the sweeping force the original book implied.
The Pygmalion literature is a case study in how a striking finding can outrun its evidence. The 1968 result was overgeneralised in textbooks and training long after the original IQ effect had failed to replicate, and the gap between the popular "teacher expectations create reality" slogan and the careful "small, often fading, larger for some groups" conclusion remains wide. A second caution runs the other way: because naturally occurring teacher expectations are usually accurate, formed from real prior performance, most of the correlation between what teachers expect and what students achieve reflects teachers reading reality correctly, not creating it. Treating every expectation-achievement correlation as a self-fulfilling prophecy overstates how much teacher belief is bending outcomes.
The defensible lesson is to watch the four transmittable behaviours rather than to police private beliefs, which is neither feasible nor mainly where the action is. A teacher can audit whether they are spreading challenging content, response opportunities, wait time, and substantive feedback evenly, or quietly concentrating them on students they have already judged promising. The largest leverage is with students whose trajectory is not yet fixed, where a withheld expectation can still cost an outcome that better treatment would have reached. The point is not to inflate praise or fake optimism, which children read accurately, but to make sure low expectations are not silently rationing the inputs that produce learning.