The Pygmalion effect, which is an interpersonal form of a self-fulfilling prophecy, borrows its name from the myth by the Roman poet Ovid and the subsequent play by George Bernard Shaw. Pygmalion was a sculptor who carved a sculpture of a beautiful woman from a piece of flawless ivory. He fell in love with his creation and begged the goddess Aphrodite to intervene on his behalf. She was so moved by his devotion that she brought the statue to life, and they lived happily ever after. In social psychology, Pygmalion effects occur when one person’s beliefs or expectations about another creates or elicits the expected behavior, much as Pygmalion created his perfect woman, or Professor Higgins transformed Eliza Doolittle into a lady in Shaw’s play.
Psychological research on Pygmalion effects began with Robert Rosenthal’s studies that showed that experimenters’ expectations for their participants’ behavior could operate as self-fulfilling prophecies. Self-fulfilling prophecies were first defined by Robert K. Merton (1948) as originally false definitions of the situation that evoke new behaviors that make the false definition come true. In a research context, this means that experimenters may obtain significant results not because their theories are correct, but because they unintentionally bias participants to respond in the hypothesized manner. For example, Rosenthal and Kermit Fode (1963) showed that rats whose experimenters believed them to be specially bred to be "maze bright" actually completed a test maze more quickly than rats believed to be "maze dull," although in reality the "maze bright" and "maze dull" labels had been randomly assigned.
These studies initially generated tremendous controversy because they raised the inevitable question of the extent to which research findings are merely an artifact of experimenter bias. Sufficient studies were conducted to confirm that experimenter expectancy effects can and do occur and that, consequently, steps should be taken in the design of experimental studies to prevent them. The simplest way to do so is to conduct a study in a double-blind manner, such that neither the participant nor the experimenter is aware of the participant’s experimental condition. Today, the double-blind randomized experiment is considered the gold standard of experimental design.
Rosenthal and his colleagues then turned their attention to other domains where Pygmalion effects could be operating. In 1968 Rosenthal and Lenore Jacobson published Pygmalion in the Classroom, describing the results of an experiment in which teachers at an elementary school were informed that certain of their students had performed well on a test of "academic blooming" and could be expected to demonstrate dramatic spurts in academic performance in the upcoming school year. In reality, the students identified as "academic bloomers" had been randomly chosen. Results of IQ tests administered at the end of the school year showed, however, that students labeled "academic bloomers" did in fact demonstrate significantly greater gains in IQ across the school year compared to control students. This finding similarly generated tremendous controversy, owing to the inescapable implication that some children who do not thrive in school may underperform not because of limitations in natural ability but because their teachers do not expect them to do well. This implication is particularly socially problematic when one considers that teachers’ expectations are not random, but are systematically related to the socioeconomic status and race of the students.
The controversy was laid to rest in 1978 when Rosenthal and Donald Rubin published a meta-analytic review of the 345 studies of Pygmalion effects conducted to date. This review showed, first, that the combined significance level for Pygmalion effects was highly statistically significant across all studies and, second, that the average effect size was a Pearson correlation coefficient of r = .33, an effect considered to be moderate in magnitude. The Rosenthal and Rubin review thus provided definitive evidence that Pygmalion effects occur across a wide range of domains and are of practical importance.
Subsequent research on Pygmalion effects has been concerned primarily with documenting their existence in other interpersonal contexts, including everyday interactions and the workplace, and with understanding their underlying mediating processes—in other words, identifying the specific verbal and nonverbal behaviors given off by expecters that elicit the expected behavior from the targets. For example, Mark Snyder, Elizabeth Tanke, and Ellen Berscheid (1977) showed that men spoke in a warmer and friendlier fashion during telephone conversations with women whom they had been led to believe were very attractive, which in turn elicited warmer, friendlier, and more appealing behavior from the women.
Efforts to eliminate Pygmalion effects have been largely ineffective, owing to the fact that the behaviors that cause these effects are often subtle and not under the conscious control of the expecters. Future research needs to focus more on ways in which we can harness the power of positive expectancies and minimize the destructive effects of negative expectancies.
