In the spring of 1964, a psychologist named Robert Rosenthal arrived at Spruce Elementary School in San Francisco with a test he called the "Harvard Test of Inflected Acquisition." The name was meaningless — the test was a standard nonverbal IQ instrument, the Flanagan Tests of General Ability — but the name sounded authoritative, and that was precisely the point. Rosenthal administered the test to students across grades one through six and told teachers it was designed to identify "late bloomers": children who were academically unremarkable now but were about to experience a significant intellectual growth spurt.
After testing was complete, Rosenthal gave each teacher a list of students in their class who had purportedly scored in the top range on the bloomer test and could therefore be expected to show unusual intellectual progress over the coming year. The students on those lists had, in fact, been selected at random. They were not distinguished from their classmates in any measurable way. Roughly 20 percent of each class appeared on the lists.
When Rosenthal and his colleague Lenore Jacobson returned eight months later and retested all the children, something striking had happened. The randomly selected "bloomers" had gained significantly more IQ points than the control group. In grades one and two — where the effect was largest — the designated bloomers gained an average of 27.4 IQ points versus 12 points for controls. The teachers had been told these children were special. And in a measurable, quantifiable, psychometrically documented sense, those children had become more special.
Rosenthal and Jacobson published their findings in a 1968 book, Pygmalion in the Classroom, and in doing so gave formal experimental documentation to one of social psychology's most consequential propositions: that what we expect of people influences what they become.
"When teachers expected randomly selected students to show intellectual growth, those students did show significantly greater gains than students in the control group." — Robert Rosenthal & Lenore Jacobson, 1968
What the Pygmalion Effect Is
The Pygmalion Effect is the phenomenon by which higher expectations from an authority figure lead to improved performance in the person being evaluated, operating through the behavioral signals the authority transmits, often without conscious awareness on either side.
Pygmalion Effect vs. Galatea Effect
The Pygmalion Effect is frequently conflated with a related but distinct phenomenon — the Galatea Effect — which concerns the effect of a person's own expectations on their own performance, rather than the expectations of an external authority. Both are species of self-fulfilling prophecy, but they operate through different mechanisms and have different implications for intervention.
| Dimension | Pygmalion Effect | Galatea Effect |
|---|---|---|
| Source of expectations | External authority (teacher, manager, coach) | The individual themselves |
| Primary mechanism | Behavioral signals transmitted by the authority figure | Self-efficacy and internal motivation |
| Theoretical basis | Interpersonal expectancy effects; social communication | Bandura's self-efficacy theory (1977) |
| Intervention target | Authority figure's beliefs and communication style | Individual's belief in their own capacity |
| Empirical context | Classroom studies, military training, management research | Performance psychology, coaching, clinical self-esteem research |
| Reversibility | Depends on changing the authority's expectations or replacing the authority | Depends on changing the individual's self-perception |
| Key limitation | Requires an authority relationship; no authority, no effect | Can amplify existing high self-efficacy but may not rescue severely low self-belief |
The distinction matters practically. A teacher who improves her own expectations of a struggling student is attempting Pygmalion-style intervention. A coach who trains an athlete to develop a more confident self-concept is attempting Galatea-style intervention. A well-designed program typically works both levers simultaneously, which is partly why it is difficult to isolate the two effects empirically.
The Cognitive Science of Expectancy Transmission
The most pressing scientific question raised by Rosenthal and Jacobson's original findings was not whether the effect existed — the data were clear enough on that — but how it worked. If teachers had communicated higher expectations to certain students, what exactly were they doing differently, and through what cognitive channels did those communications translate into IQ gains?
The Four-Factor Model
Rosenthal addressed this directly in his 1973 model, which he called the four-factor or CFTIO theory, later simplified to the "four factors of teacher expectation." The model identified four behavioral channels through which teachers (or any authority figure) communicate expectations:
- Climate: Warmer socio-emotional climate, more frequent smiling, nodding, and leaning toward the student; more sustained eye contact.
- Feedback: More frequent and more differentiated feedback, both positive and corrective; praise that is more specific and more likely to attribute success to the student's ability rather than luck.
- Input: Greater quantity and difficulty of material taught; higher-expectation students receive more intellectually challenging content and more instructional time.
- Output: Greater encouragement of responses; teachers wait longer for high-expectation students to answer, prompt them more when they falter, and follow up their answers more frequently.
These four channels operate largely below conscious awareness. In a 1985 meta-analysis published in the Journal of Personality and Social Psychology, Mary Harris and Robert Rosenthal examined 136 studies of expectancy communication and found that the "input" factor — the sheer amount of material taught — showed the most robust relationship with student outcomes, followed by the socio-emotional climate factor. Feedback and output showed weaker but consistent effects. The meta-analysis covered effect sizes ranging from modest to substantial, with a weighted mean effect size of approximately r = 0.40 across the literature — a medium-to-large effect by conventional standards in psychology.
Behavioral Confirmation
A parallel line of research addressed expectancy effects in social interaction more broadly. Mark Snyder, Elizabeth Decker Tanke, and Ellen Berscheid published a now-classic study in the Journal of Personality and Social Psychology in 1977 titled "Social Perception and Interpersonal Behavior: On the Self-Fulfilling Nature of Social Stereotypes." Male participants were shown photographs (manipulated to show either an attractive or unattractive woman) and then conducted telephone conversations with a female participant they believed corresponded to the photograph. The men who believed they were speaking with an attractive woman behaved more warmly and engagingly — and independent raters, listening only to the woman's side of the conversation, rated her as warmer, more confident, and more socially skilled. The woman's behavior had been shaped by the man's expectations, transmitted through subtle differences in his vocal tone and conversational engagement, even though she had no knowledge of the photograph manipulation.
This study established that expectancy effects do not require prolonged relationships, hierarchical authority, or conscious transmission. A few minutes of interaction, mediated by a manipulated belief, was sufficient to alter observable behavior. The implication is that Pygmalion-type dynamics are not confined to classrooms and management chains — they are a routine feature of social interaction.
Neurocognitive Mechanisms
More recent work has begun to situate expectancy effects within neurobiological frameworks. Alia Crum and Ellen Langer published a study in Psychological Science in 2007 showing that hotel housekeepers who were informed that their daily work constituted significant physical exercise (and thus fulfilled the Surgeon General's recommendations for active lifestyle) showed measurable changes in weight, body mass index, blood pressure, and body fat percentage relative to a control group — with no change in actual physical activity. The belief that they were exercising appears to have triggered physiological responses that exercise itself would have triggered. While this is not strictly a Pygmalion effect (the source was information provided to participants rather than expectations held by an authority), it illustrates the magnitude of cognitive-to-physiological pathways that expectancy research has documented.
Dopaminergic reward circuits are implicated in several models of expectancy effects: anticipating a positive outcome activates reward circuitry in ways that can improve attention, motivation, and cognitive persistence. If a student perceives — through the behavioral cues described by Rosenthal's four-factor model — that her teacher expects her to succeed, this may prime reward circuitry in ways that functionally improve her performance on cognitively demanding tasks.
Four Case Studies Across Domains
Case Study 1: Military Training — Eden's Israeli Defense Forces Research
Dov Eden, a management psychologist at Tel Aviv University, conducted a series of experiments in Israeli Defense Forces settings in the 1980s and 1990s that constitute some of the strongest evidence for Pygmalion effects outside educational contexts. In his foundational 1984 study, published in the Journal of Applied Psychology, Eden and Uri Shani randomly assigned trainees in an IDF combat command course to one of three conditions: some instructors were told their trainees had high command potential; others were told potential was uncertain; a control group received no information. Trainees whose instructors believed they had high potential outperformed controls on objective outcome measures including a written examination and ratings of performance in simulated command exercises.
In a 1990 follow-up, Eden extended the design to an IDF platoon setting, manipulating expectations held by commanding officers about their entire platoons rather than individual soldiers. Platoons whose commanders received high-expectation information about platoon capability showed significantly higher performance on combat exercises than control platoons. The unit-level effect was comparable in magnitude to individual-level effects, suggesting that Pygmalion dynamics can operate simultaneously at multiple levels of an organizational hierarchy.
Eden also documented a "Golem Effect" in this research — the mirror image of Pygmalion, in which low expectations depress performance. Platoons assigned to commanders who had received negative information about their units showed performance decrements relative to controls. This asymmetry is important: low expectations are not merely the absence of high expectations but an active force that suppresses performance through the same behavioral channels in reverse.
Case Study 2: Education — Special Education Labeling Studies
One of the most practically significant extensions of Rosenthal and Jacobson's classroom work concerns the effects of special education labeling. Robert Rosenthal collaborated with Elisha Babad and Jacinto Inbar to examine what they called the "Pygmalion-in-the-Classroom" effect in authentic school settings rather than experimental manipulations. Their 1982 study, published in the Journal of Educational Psychology, found that teachers who scored high on a measure of susceptibility to expectancy effects showed larger performance differences between students they perceived as bright versus those they perceived as slow — not because the susceptible teachers were worse educators in any general sense, but because they were more responsive to expectancy cues and thus amplified naturally occurring expectation differences.
The labeling literature documents a variant of this dynamic at institutional scale. When students are classified as having learning disabilities or behavioral disorders, these labels carry expectational freight that affects every subsequent teacher's behavior, peer interactions, and the student's own self-concept. Jere Brophy and Thomas Good documented in their 1970 study in the Journal of Educational Psychology that teachers behave demonstrably differently toward students they perceive as low-achieving even within the first days of a school year — before they have any direct evidence of performance — based on the child's appearance, socioeconomic indicators, and verbal style. The expectation precedes the evidence and then shapes the evidence that accumulates.
Case Study 3: Organizational Management — Corporate Performance Reviews
Research on performance appraisal processes in corporate settings has documented Pygmalion dynamics that have significant implications for personnel management. A 2003 study by Brian McNatt and Timothy Judge, published in the Academy of Management Journal, examined the effects of supervisor expectations on new employee performance in a financial services firm. New employees randomly assigned to supervisors who had been given positive (fabricated) information about the new hire's prior performance showed measurably higher performance ratings at their 90-day review than controls, controlling for actual prior performance.
The mechanism appeared to operate primarily through differential coaching behavior: supervisors who expected more gave more developmental feedback, assigned more challenging initial tasks, and provided more frequent informal guidance. This mirrors Rosenthal's input and feedback factors. The study also found that the effect was larger for employees who rated themselves as higher in need for achievement — individuals who were already motivated to succeed responded more strongly to environmental signals of expected competence.
McNatt extended this line of inquiry in a 2000 meta-analysis in the Academy of Management Journal examining Pygmalion effects in work settings across 17 studies, finding a mean effect size of d = 0.81 — a large effect by Cohen's convention. The effect was stronger in military contexts than in civilian organizations, which McNatt attributed to the more explicit and consequential nature of authority hierarchies in military settings.
Case Study 4: Clinical Psychology — Therapeutic Alliance and Patient Outcomes
Psychotherapy research offers a domain in which expectancy effects are both well-documented and practically significant. The therapeutic alliance — the quality of the working relationship between therapist and client — is one of the most robust predictors of psychotherapy outcome across treatment modalities, accounting for approximately 30 percent of outcome variance in some meta-analyses. A substantial component of the therapeutic alliance appears to be expectancy transmission: therapists who communicate confident expectations of improvement produce better outcomes than those who do not, controlling for technical competence.
Irving Kirsch has documented expectancy effects in pharmacological contexts that parallel Pygmalion dynamics. In analyses published across multiple peer-reviewed journals between 1998 and 2008, Kirsch and colleagues found that a substantial portion of antidepressant drug effects in clinical trials are mediated by patient expectancies — specifically, the expectation of improvement induced by taking a medication one believes to be active. The implication is not that the drugs are inert but that the expectancy transmitted by the prescribing physician and the treatment context amplifies drug effects and may account for a larger share of clinical benefit than is typically acknowledged.
Michael Lambert's longitudinal research at Brigham Young University, published in a 2001 paper in the Journal of Consulting and Clinical Psychology, found that therapists who received ongoing feedback about patient progress produced significantly better outcomes than those who did not — in part because the feedback altered therapist expectations and behavior. Therapists who learned that a patient was not progressing as expected adjusted their approach; those without feedback continued as before, even when patients were deteriorating.
Intellectual Lineage: From Merton to Rosenthal
The Pygmalion Effect did not emerge in a theoretical vacuum. Its intellectual prehistory runs through several distinct traditions.
The most direct precursor is the concept of the self-fulfilling prophecy, named and formalized by sociologist Robert K. Merton in a 1948 essay in The Antioch Review. Merton defined a self-fulfilling prophecy as "a false definition of the situation evoking a new behavior which makes the originally false conception come true." His canonical example was bank runs: if enough depositors believe a bank is about to fail and withdraw their funds, the bank does fail — not because the initial belief was accurate but because the behavior it generated made it accurate. Merton's contribution was to identify the circular causality at the core of the phenomenon: the belief causes the behavior that confirms the belief.
Merton's analysis drew on the earlier work of sociologist W.I. Thomas, whose 1928 observation — now known as the Thomas Theorem — holds that "if men define situations as real, they are real in their consequences." Thomas was describing how subjective definitions of reality shape objective social outcomes, which is the same logical structure Merton later elaborated and that Rosenthal would eventually test experimentally.
The name "Pygmalion Effect" connects the phenomenon to the Greek myth of Pygmalion, a sculptor who carved an ivory statue of a woman so perfect that he fell in love with it. In Ovid's Metamorphoses, the goddess Aphrodite, moved by his devotion, brought the statue to life. George Bernard Shaw adapted the myth in his 1913 play Pygmalion, in which phonetician Henry Higgins transforms Eliza Doolittle — a flower girl from the London streets — into a convincing aristocratic lady through sustained belief in the project's possibility and intensive instruction. Shaw's play dramatizes the same mechanism Rosenthal would later document empirically: the transformation is accomplished partly through Higgins's relentless expectation of success, which shapes both his behavior and Eliza's self-concept.
Rosenthal's own earlier research on experimenter effects formed the immediate empirical background for the classroom study. Through a series of experiments in the late 1950s and early 1960s, Rosenthal documented that researchers' expectations about experimental outcomes systematically influenced those outcomes — not through fraud but through subtle, unconscious differences in how they ran experiments, interacted with research subjects, and recorded ambiguous data. This line of research, summarized in Experimenter Effects in Behavioral Research (1966), established that expectancy transmission was a real and measurable phenomenon in controlled laboratory settings. Pygmalion in the Classroom was the natural extension of that finding to the educational field.
Empirical Research: Replication, Scaling, and Boundary Conditions
Rosenthal and Jacobson's 1968 findings were immediately controversial. Several psychologists questioned the statistical analyses, the reliance on IQ as an outcome measure, and the size of the effect relative to what could plausibly be explained by behavioral transmission alone. Thorndike (1968) published a pointed critique in the American Educational Research Journal arguing that the IQ gains in the youngest grades were implausibly large given what was known about IQ test measurement error.
These critiques prompted a substantial empirical literature over the following decades. By the mid-1980s, enough studies had accumulated for quantitative synthesis. The Harris and Rosenthal 1985 meta-analysis, examining 136 studies of interpersonal expectancy effects (not all in educational settings), found a mean effect size of r = 0.40. However, this aggregate conceals substantial heterogeneity. Some studies found no effect; others found very large effects. The four-factor model predicted that the effect would be stronger when authorities held extreme expectations, when they had more contact with the people they expected something of, and when the targets had less prior performance history against which the expectation could be disconfirmed.
Lee Jussim and Kent Harber published a comprehensive and critical review of the teacher expectation literature in 2005 in the Personality and Social Psychology Review, titled "Teacher Expectations and Self-Fulfilling Prophecies: Knowns and Unknowns, Resolved and Unresolved Controversies." Their analysis identified several important qualifications that the field had established over the 37 years since Rosenthal and Jacobson's book:
First, teacher expectations are generally accurate — they reflect real differences in student ability more often than they create ability differences from nothing. Jussim and Harber estimated that naturally occurring teacher expectations account for somewhere between 5 and 10 percent of variation in student outcomes, which is meaningful but substantially smaller than the 1968 original suggested. The experimental paradigm, in which expectations were induced by false information, may overestimate the size of the effect because it forces teachers to hold expectations that are strongly discrepant from all available evidence.
Second, the Pygmalion effect appears to be larger and more durable for students from stigmatized groups. Jussim, Eccles, and Madon (1996) published a longitudinal study in Journal of Personality and Social Psychology finding that teacher expectations in sixth grade predicted student performance in seventh grade, and that the predictive relationship was stronger for students from lower socioeconomic backgrounds and for African American students than for white students from higher-income families. The interpretation is that students who lack strong alternative sources of expectational information — mentors outside school, family members with educational credentials, cultural messages of academic belonging — are more susceptible to school-based expectancy effects, for better and for worse.
Third, accuracy and self-fulfilling prophecy are not mutually exclusive. A teacher may have accurate expectations in aggregate — correctly perceiving that student A reads better than student B — while still generating Pygmalion effects at the margin by communicating those expectations in ways that alter the developmental trajectories of both students.
Limits and Nuances
When Expectations Fail to Translate
Not all expectations become self-fulfilling prophecies. Several conditions appear necessary for the Pygmalion effect to operate, and their absence substantially attenuates or eliminates the effect.
Prior performance history: The effect is weakest when the target of expectations has extensive prior performance history that contradicts the expectation. A student who has been assessed by many teachers over many years acquires a stable self-concept and a documented performance record that buffers against any single teacher's expectation. This is why Rosenthal and Jacobson found the largest effects in the youngest grades, where students had the least accumulated performance history. It also explains why the effect is more robust in laboratory studies (where participants have no prior history with the experimenter) than in naturalistic classroom settings.
Duration of the relationship: A brief interaction may produce a Behavioral Confirmation effect (as in Snyder et al., 1977), but producing durable changes in ability requires sustained contact. One semester of elevated expectations may be insufficient to produce lasting IQ gains if the child then encounters a teacher who holds lower expectations.
Credibility of the expectation: If the expectation being communicated is radically discrepant from what the student already knows about themselves, the student may reject or discount the authority's signals rather than incorporate them. Pygmalion effects are most reliable when the expectation is somewhat elevated above current performance — a stretch goal rather than a fantasy.
Individual differences in susceptibility: Not all students respond equally to teacher expectations. Students with low self-efficacy (Bandura, 1977) may be more susceptible to downward expectancy effects (Golem effects) and less responsive to upward ones, because negative expectational information confirms existing self-beliefs while positive information is discrepant with them. Conversely, students with high implicit theories of intelligence as fixed (Dweck's "entity theorists") may be less susceptible to either direction of expectancy effect, because they have already decided that their capacity is fixed.
The Ethics of Expectancy Manipulation
Rosenthal and Jacobson's study raises an ethical question that the field has never fully resolved: is it permissible to induce false beliefs in authority figures — telling teachers their students have "high potential" when they do not — to produce real performance gains? The pragmatic answer from subsequent research is that the gains from artificially induced high expectations may be real but modest, and they are likely to be temporary once feedback from the environment corrects the false expectation. The more durable strategy is to change the institutional conditions that generate low expectations in the first place — tracking systems, racially segregated schools, poverty — rather than manipulating individual teachers' beliefs.
Distinguishing Accuracy from Bias
One of the most important conceptual clarifications in the post-2000 literature is the distinction between accurate expectations (which improve outcomes by providing appropriate challenge and support) and biased expectations (which are systematically distorted by race, class, gender, or other demographic characteristics of the student). Jussim and Harber (2005) argue that much of what is attributed to Pygmalion effects in real classrooms is better described as accurate teacher perception: teachers give more attention to students who are already performing well because those students are easier to teach and more rewarding to interact with. The self-fulfilling component — the expectation causing the outcome rather than merely tracking it — is real but smaller than the popular account suggests.
This distinction has practical consequences. If the primary problem is biased expectations toward students from stigmatized groups, the intervention is to correct the bias. If the primary problem is accurate-but-self-reinforcing expectations that trap low-performing students in low-performance trajectories, the intervention is more complex — it requires disrupting the feedback loop that makes accurate expectations become determinative.
Domain-Specificity
The magnitude of the Pygmalion effect varies substantially across domains. Eden's military research produced large, reliable effects; corporate management research produced medium effects; classroom research produces effects that are real but modest when carefully measured. The explanation may involve the density and speed of performance feedback in different settings. In military training, performance is assessed frequently and explicitly; a trainee's response to an expectation can compound rapidly. In classrooms, performance feedback is slower and more diffuse; the expectation has more time to be diluted by other influences.
References
Rosenthal, R., & Jacobson, L. (1968). Pygmalion in the Classroom: Teacher Expectation and Pupils' Intellectual Development. Holt, Rinehart & Winston.
Rosenthal, R. (1973). The Pygmalion effect lives. Psychology Today, 7(4), 56–63.
Harris, M. J., & Rosenthal, R. (1985). Mediation of interpersonal expectancy effects: 31 meta-analyses. Psychological Bulletin, 97(3), 363–386.
Jussim, L., & Harber, K. D. (2005). Teacher expectations and self-fulfilling prophecies: Knowns and unknowns, resolved and unresolved controversies. Personality and Social Psychology Review, 9(2), 131–155.
Eden, D., & Shani, A. B. (1982). Pygmalion goes to boot camp: Expectancy, leadership, and trainee performance. Journal of Applied Psychology, 67(2), 194–199.
Eden, D. (1990). Pygmalion without interpersonal contrast effects: Whole groups gain from raising manager expectations. Journal of Applied Psychology, 75(4), 394–398.
Snyder, M., Tanke, E. D., & Berscheid, E. (1977). Social perception and interpersonal behavior: On the self-fulfilling nature of social stereotypes. Journal of Personality and Social Psychology, 35(9), 656–666.
Jussim, L., Eccles, J., & Madon, S. (1996). Social perception, social stereotypes, and teacher expectations: Accuracy and the quest for the powerful self-fulfilling prophecy. Advances in Experimental Social Psychology, 28, 281–388.
Merton, R. K. (1948). The self-fulfilling prophecy. The Antioch Review, 8(2), 193–210.
McNatt, D. B. (2000). Ancient Pygmalion joins contemporary management: A meta-analysis of the result. Journal of Applied Psychology, 85(2), 314–322.
Brophy, J. E., & Good, T. L. (1970). Teachers' communication of differential expectations for children's classroom performance: Some behavioral data. Journal of Educational Psychology, 61(5), 365–374.
Bandura, A. (1977). Self-efficacy: Toward a unifying theory of behavioral change. Psychological Review, 84(2), 191–215.
Frequently Asked Questions
What is the Pygmalion effect?
The Pygmalion effect is the phenomenon whereby higher expectations from authority figures lead to improved performance in those being evaluated. Named after the Greek myth of Pygmalion — and George Bernard Shaw's play in which a phonetics professor transforms a flower girl by believing she can become a duchess — the effect was empirically established by Robert Rosenthal and Lenore Jacobson's 1968 Pygmalion in the Classroom study. Teachers who expected certain students to bloom intellectually behaved differently toward those students in ways that produced measurable IQ gains, even though the students were randomly assigned to the 'bloomer' category.
What did the Pygmalion in the Classroom study find?
Rosenthal and Jacobson (1968) administered a real IQ test to students at Spruce Elementary School in San Francisco, then told teachers that approximately 20% of students — actually selected at random — had scored high on a 'Harvard Test of Inflected Acquisition' that predicted intellectual blooming. When retested eight months later, first-grade 'bloomers' had gained an average of 27.4 IQ points compared to 12 points for controls. Second-grade bloomers gained 16.5 points versus 7 for controls. The effect was strongest in younger grades and diminished in older grades, where prior performance records constrained teacher expectations.
How do teacher expectations translate into student outcomes?
Robert Rosenthal's 1973 four-factor model identified the behavioral mechanisms. Climate: teachers create warmer socio-emotional environments for high-expectancy students — more eye contact, more encouragement, closer proximity. Input: teachers teach more material and more difficult material to students they expect to succeed. Feedback: high-expectancy students receive more detailed and more accurate feedback on their performance. Output: teachers give high-expectancy students more opportunities to respond and more time to formulate answers. Harris and Rosenthal's 1985 meta-analysis of 136 studies confirmed all four factors, with input and output showing the strongest effect sizes (r = 0.40 for overall expectancy effects).
Does the Pygmalion effect work in organizational settings?
Dov Eden's research with the Israeli Defense Forces produced some of the strongest organizational replications. In Eden's 1984 study, IDF military instructors were told that certain trainees had high command potential (actually randomly assigned). Trainees in the high-expectancy condition subsequently outperformed controls on objective performance measures. Eden's 1990 field experiment with IDF combat soldiers produced similar results. McNatt's 2000 meta-analysis of 17 organizational studies found a mean effect size of d = 0.81, suggesting the Pygmalion effect is stronger in workplace settings than in classrooms, possibly because adult performance is more responsive to supervisor behavior than children's test scores are to teacher behavior.
How robust is the Pygmalion effect?
Jussim and Harber's 2005 Personality and Social Psychology Review reassessment found that while the effect is real, its size in naturalistic settings is modest — explaining perhaps 5-10% of variance in student outcomes. The effect is strongest for students from stigmatized groups (who may be subject to negative default expectations that positive expectancy manipulation counteracts), for younger students with limited prior performance records, and for short time horizons. Long-term, students' actual performance eventually overrides expectancy-based teacher perceptions. The Pygmalion effect is real but smaller and more conditional than early enthusiasm suggested.