Expectancy effect
Also known as: Expectation effect, Observer-expectancy effect
What we expect to happen quietly helps make it happen.
What it means
An expectancy effect is any case in which holding an expectation about an outcome helps produce that outcome, because the expectation changes behavior, perception, or physiology before the result is settled. It runs by two routes. The interpersonal route works through the expecter: believing something about another person, we adjust our warmth, patience, attention, and feedback, and they respond in kind — behavioral confirmation, of which the Pygmalion effect is the best-known case. The intrapersonal route works through one's own body and attention, as in the placebo response. It shows up in classrooms, management, medicine, and the lab, where a researcher's hopes leak into how data are gathered and coded. The limits matter: expectancy moves ambiguous, malleable outcomes rather than fixed ones, and estimates vary widely by setting and shrink under stronger designs. It matters because good intentions are no protection — only blinding and structure are.
The four channels of behavioral confirmation
The interpersonal route is not one act but a bundle. Reviewing his experiments, Rosenthal proposed four channels through which an expecter shapes a target: a warmer emotional climate toward those expected to succeed; more material taught to them; more chances to respond; and more frequent, differentiated feedback. Each is small and largely unconscious — a held gaze, an extra second before moving on, a follow-up question rather than the answer handed over. The target reads these signals as information about their standing and adjusts effort and self-image to match. Because the channels compound quietly and the expecter rarely notices them, the eventual result — the star performer, the confirmed diagnosis — arrives looking like independent proof of the belief that produced it.
What the classroom evidence shows
Rosenthal and Jacobson's 1968 Pygmalion in the Classroom reported that randomly labelled 'bloomers' gained more IQ than classmates, but the effect was carried almost entirely by first and second graders and rested on a questionable test; older grades showed little. Later syntheses shrank the claim. Raudenbush's 1984 meta-analysis of eighteen experiments found an average effect near d = 0.11, and close to zero once teachers had known their pupils for about two weeks — the belief bites only before firsthand knowledge arrives. Jussim and Harber's 2005 review concluded that classroom self-fulfilling prophecies are real but small, do not accumulate over time, and are frequently accuracy masquerading as causation: teachers predict outcomes largely because their expectations are correct, not because they create them.
The version that corrupts research
The same mechanism turns on the people who study it. In Rosenthal's demonstration, assistants told their rats were 'maze-bright' recorded faster runs than those told the rats were 'dull,' though the animals were assigned at random — the experimenters' hopes had leaked into handling and scoring. Rosenthal and Rubin's 1978 synthesis of the first 345 studies put the average interpersonal expectancy effect around d = 0.70, across reaction-time tasks, animal learning, and person perception. This is why the double-blind trial exists: not because scientists are dishonest, but because an honest observer who knows which group is which will, without meaning to, gather and code data in the expected direction. Blinding removes the knowledge, not the temptation.
Where it stops
Expectancy moves outcomes that are ambiguous, effortful, or self-reported; it does not rewrite fixed facts. Estimates fall sharply under the designs that matter — randomization, blinding, prior acquaintance — so the strongest studies tend to find the weakest effects. The subjective end is real but bounded: open-label placebos, where patients are openly told the pill is inert, still ease irritable-bowel and pain symptoms, showing that a persuasive narrative rather than deception does the work — yet the gains sit in symptom report, not in the underlying disease. The practical lesson is defensive. Because good intentions offer no protection, the only dependable guards are structural: randomize assignment, blind whoever measures, and treat any outcome that passed through a hopeful human's hands as provisional.
Examples
A manager told a new hire is 'a rising star' hands them the harder brief, more airtime, and quicker feedback — and six months on, the strong rating looks like proof.
A runner given what they are told is a performance drink, actually water, reports the last mile as easier and pushes harder — expectation doing work the drink never could.
Rosenthal's assistants, told their rats were bred 'maze-bright' or 'maze-dull,' handled the bright ones more gently and recorded faster runs. The rats had been assigned at random.
A physician who reads 'anxious' at the top of a chart hears ordinary chest tightness as panic, orders fewer cardiac tests, and the reassuring workup that follows seems to confirm the first hunch.
An interviewer who has decided a candidate is weak keeps the questions clipped, leans back, and cuts answers short; the candidate reads the chill, grows flustered, and the halting performance seems to confirm the judgment that produced it.
First described in Robert Rosenthal (1963, 1966).
Key references
- de Boer, H., Timmermans, A. C., & van der Werf, M. P. C. (2018). The effects of teacher expectation interventions on teachers' expectations and student achievement: narrative review and meta-analysis. Educational Research and Evaluation, 24(3-5), 180-200. doi.org/10.1080/13803611.2018.1550834
- 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. doi.org/10.1207/s15327957pspr0902_3
- Raudenbush, S. W. (1984). Magnitude of teacher expectancy effects on pupil IQ as a function of the credibility of expectancy induction: A synthesis of findings from 18 experiments. Journal of Educational Psychology, 76(1), 85-97. doi.org/10.1037/0022-0663.76.1.85
- Kaptchuk, T. J., Friedlander, E., Kelley, J. M., Sanchez, M. N., Kokkotou, E., Singer, J. P., Kowalczykowski, M., Miller, F. G., Kirsch, I., & Lembo, A. J. (2010). Placebos without deception: A randomized controlled trial in irritable bowel syndrome. PLoS ONE, 5(12), e15591. doi.org/10.1371/journal.pone.0015591
- Rosenthal, R., & Rubin, D. B. (1978). Interpersonal expectancy effects: The first 345 studies. Behavioral and Brain Sciences, 1(3), 377-386. www.cambridge.org/core/journals/behavioral-and-brain-sciences/article/interpersonal-expectancy-effects-the-first-345-studies/44A0027CD2B1D8CB47FE77D952770C7F