On the Bounded Rationality of Gender Stereotyping in Fame Judgments

The false-fame effect is the phenomenon that familiar names are falsely judged famous more often than unfamiliar names. M.R. Banaji and A.G. Greenwald (1995) demonstrated a gender bias in the false-fame effect: In line with existing gender stereotypes, the false-fame effect was larger for male than for female names. A more general explanation for gender biasing in fame judgments is based on cognitive availability. Name gender could be used as an ecologically valid cue when making fame judgments. If the relevant universe of famous names contained more male than female names, a gender bias in fame judgments should be observed, if it contained more female names, the gender bias should be reversed. Indeed, this pattern could be demonstrated experimentally, and we argue that it is not compatible with an account that draws on gender stereotyping but with one based on cognitive availability. Steffens et al. Implicit gender stereotyping in fame judgments 3 On the Bounded Rationality of Gender Stereotyping in Fame Judgments The pervasiveness of stereotyping is an intriguing question. It is all the more intriguing to the extent to which stereotyping seems to occur even in the absence of conscious knowledge and in the presence of deliberate attempts to the contrary. Accordingly, since pioneering studies (Gaertner & McLaughlin, 1983), the emphasis in research on stereotyping has shifted toward their automatic, uncontrollable, or implicit components (e.g., Banaji & Hardin, 1996; Blair & Banaji, 1996; Devine, 1989; Dovidio, Kawakami, Johnson, Johnson, & Howard, 1997; Fazio, Jackson, Dunton, & Williams, 1995; Lepore & Brown, 1997; Rudman, Greenwald, & McGhee, 2001; Wittenbrink, Judd, & Park, 1997). Banaji and Greenwald (1995) developed a unique procedure for measuring implicit gender stereotyping. Their findings have interesting implications for both research on stereotyping and research on memory and thus have appropriately generated a lot of interest. Banaji and Greenwald investigated the so-called false-fame effect that was originally discovered in the data of Neely and Payne (1983) and was later extended into a research paradigm of its own by Jacoby et al. (Dywan & Jacoby, 1990; Jacoby, Kelley, Brown, & Jasechko, 1989; Jacoby, Woloshyn, & Kelley, 1989; Jennings & Jacoby, 1993). A false-fame effect is demonstrated if familiarized names of nonfamous persons are judged famous with a higher probability than the same names when they were not previously familiarized. Apparently, in the absence of explicit knowledge about the fact that the names’ familiarity results from the study phase, familiarity is misattributed to their being the names of famous persons (Steffens, Buchner, Martensen, & Erdfelder, 2000). According to Banaji and Greenwald (1995), due to the fact that there are more famous males than females in society, the gender stereotype implies a closer association between ‘male’ and ‘famous’ than between ‘female’ and ‘famous.’ This, in turn, results in a gender bias in the process of attributing fame: a lower criterion is used when male as opposed to female names are judged. Banaji and Greenwald demonstrated empirically that familiarized nonfamous male names were falsely judged famous more often than familiarized nonfamous female names, whereas there was no such difference for nonfamiliarized names. That is, the Banaji effect took the form of an interaction between name gender and name familiarity on false alarms (i.e., on nonfamous names judged famous). The gender bias in fame judgments was interpreted as an instance of implicit gender stereotyping. As we posit, a stereotyping account of the gender bias implies that it does not simply mirror the proportion of famous males and females in the to-be-judged “universe” in a given context, Steffens et al. Implicit gender stereotyping in fame judgments 4 but rather reflects a more durable association or willingness to call males rather than females famous. That is, the male-famous association should also be found when the proportion of famous males is not higher than the proportion of famous females. Conversely, if a gender bias favoring females instead of males were easily obtained when the proportion of famous females exceeded that of famous males, then the effect would hardly qualify as stereotyping but better be framed as an instance of cognitive availability. The respective mechanism would be adjusting the criterion for judging fame such that the proportion of females judged famous mirrors the proportion of females in the relevant universe of famous people. Such a mechanism would fit the large body of literature on the availability heuristic, showing that probability judgments depend on the availability of the respective instances. The more easily famous females come to mind (see Schwarz & Vaughn, 2002, for distinguishing this rule from: “the more famous females come to mind”), the more liberal will be the criterion for judging females famous. In their seminal paper, Tversky and Kahneman (1973) demonstrated that the proportion of names of females in a previously presented list was estimated to be higher if these females had been more famous, and were thus more easily remembered, than the male names in the list. They described availability as an “ecologically valid clue for the judgment of frequency because, in general, frequent events are easier to recall or imagine than infrequent ones” (Tversky & Kahneman, 1973, p. 209). Empirical Investigation We conducted the present experiment in order to test between a gender-stereotyping and a cognitive availability account of the gender bias in fame judgments. We assumed that participants (not necessarily consciously and purposefully) match their fame judgments with their knowledge of the relative proportions of male versus female persons in the “fame universe.” Consequently, if more males are available in the fame universe, they should judge more familiar male than female names famous. In contrast, they should judge more familiar female than male names famous if more females were available in the fame universe. Normally, participants’ knowledge of famous males and females in general leads to the cognitive availability of more famous males. In the present experiment, we manipulated the proportion of male versus female names in a to-be-learned list of names. Either one quarter (More-Famous-Males Condition) or three quarters (More-Famous-Females Condition) of the names in that list were female names. After participants had learned such a list of names (henceforth referred to as the fame list), we applied the typical false-fame procedure, a study phase in which the pronouncableness of names was judged, and a test phase in which fame judgments were made. The Steffens et al. Implicit gender stereotyping in fame judgments 5 instructions stressed that all famous names in the test phase were from the fame list. In the pronunciation phase as well as in the test phase, participants both in the More-Famous-Males and in the More-Famous-Females Condition received identical lists composed of 50% male and 50% female names, and 50% names from the fame list as well as 50% nonfamous names. We expected an interaction of experimental condition and name gender on false alarms. Specifically, we hypothesized that participants in the More-Famous-Males Condition would falsely assign fame to male names more often than to female names, whereas participants in the More-Famous-Females Condition would show the reverse pattern.