Psychology Meets Geography
The origins of this project can best be traced to articles in the popular press, such as those reporting that 20% of 12-year-olds in North Dallas misidentified Brazil as the United States on a world map (Washington Times, 1984), that 39% of high school seniors in Boston could not name all six New England states (Philadelphia Inquirer, 1987), or that about 67% of adults did not know that the Seine was in France (U.S. News & World Report, 1985).
Stories such as these led to a variety of initiatives such as the congressional establishment of 'Geography Awareness Week,' the addition of geography to the 'Goals 2000: Educate America Act,' the incorporation of geography curriculum into 'Sesame Street,' and the popularization of geography television and computer programs. Perhaps the most ambitious responses to reports of geographic ignorance, however, have been those made by NGS. One initiative was the establishment of a network of state 'Geographic Alliances' that help train teachers and provide materials to enhance geography education in the classroom. A second was the establishment in 1989 of an annual 'National Geographic Bee' for children in grades 4-8. To launch the Bee, NGS mailed information to roughly 80,000 schools inviting them to participate. It then sent additional Bee materials (e.g., questions, instructions, prizes) to those 40,000 schools that expressed interest in holding a Bee. Although there has been considerable variation in the way schools conduct preliminary rounds of the contest, each year 10 finalists are eventually identified, and these students then compete in an oral Bee (typically in a school-wide assembly) until a single school winner emerges.
The school winner then takes a 70-item written qualifying test. Roughly half the participating schools return these tests to NGS by the published deadline, and the 100 top-scoring students in each of 57 geographic units (states, the District of Columbia, the territories, and the Department of Defense schools) are then invited to compete in their state-level Bee, held throughout the country on the same date, following identical procedures. Each of the 57 state winners and their teachers are invited, all expenses paid, to Washington, DC, to compete in the National Finals in May. After preliminary rounds eliminating all but 10 students, the final competition, moderated by Alex Trebeck, is held on national television. The stakes are high: The first-, second-, and third-place winners receive, respectively, $25,000, $15,000, and $10,000 college scholarships.
Although the National Geographic Bee has surpassed expectations with respect to enhancing attention to geography (NGS estimates that 6,000,000 school children participate at some level; PBS ratings for the Bee are high), it has simultaneously revealed another problem--a shocking gender disparity among the winners at every level. In 1993, for example, of the 18,000 school winners, about 14,000 were boys. Of the 57 state winners, 55 were boys. Not surprisingly, boys ended up dominating the televised final round of the competition. In most years, all 10 finalists have been boys. Given that my own research has been focused on both the development of spatial cognition and gender, this contest presented an intriguing real-life laboratory in which to test some hypotheses about gender stereotypes, the role of spatial skills for success in a space-related domain, and questions related to changing social and academic skills at puberty. In collaboration with Penn State colleagues with whom I have studied questions related to spatial cognition (Roger Downs, a geographer) and gender stereotypes (Margaret Signorella, a social psychologist), as well as with numerous undergraduate and graduate students, we have been working to identify variables that may account for the wide gender disparity, with a particular eye toward identifying any that may be subject to intervention by NGS. This project is still in progress, and, thus, the description that follows provides an overview of the kinds of variables we are examining, but only some tentative findings.
First, we are attempting to determine if there are factors at the societal level that may lead children to view geography as a masculine domain. Consistent with this possibility, we have found that when asked questions about who 'should be' a geographer or participate in geography-related activities (such as reading maps), there is some tendency for both adults and children to respond in ways that suggest that cultural stereotypes are operating. However, modal responses are neutral, not masculine, and far less extreme than traditionally masculine occupations and activities such as 'being a plumber' or 'fixing cars.' Thus, geography-related stereotypes appear to be weaker than those in math and science, and taken alone, are unlikely to account for the overwhelming difference in boys' and girls' success on the Bee. We are currently exploring the possibility that there are important differences in the artifacts that are presented by our culture to boys versus girls. Preliminary data suggest that this aspect of culture may have considerable explanatory power. For example, we have begun content analyses of 'boys' versus 'girls' popular books (e.g., Jack London versus Beverly Cleary) and find striking differences in the amount of geography-related material contained within them.
Second, we are examining factors related to the Geography Bee itself, that is, factors that relate to the nature of the specific contest. Data suggest that the questions themselves are probably not selected in ways that bias outcomes unfairly for girls. The written qualifying test scores are equally predictive for boys' and girls' ultimate success at later levels of the contest, and independent raters of questions do not find gender bias in content (as they might, for example, if questions had been drawn from traditionally masculine domains such as sports). Consistent with this view are data from other tests (e.g., the International Assessment of Educational Progress) showing that boys surpass girls on other geography tests as well. Another Penn State colleague, Hoben Thomas, is currently using mathematical modeling procedures to explore the extent to which the particular selection procedure used to identify the national finalists may be exaggerating small boy-girl differences at the state level to produce far larger discrepancies at the national level.
Third, we are examining factors related to school procedures that might be affecting students' participation. One potential explanation of the ultimate disparity in outcome was simply that boys and girls might be entering the Bee at strikingly different rates. Data do not support this explanation: At entry, the ratio of boys to girls is almost even (1.04:1). Furthermore, whereas those schools that permit voluntary entry do, as expected, yield more male entrants, the ratio is still quite even (1.10:1). Finally, there is no evidence that the likelihood of producing a male or female winner varies with the entry mechanism used by the schools. Although there were some variations in performance in relation to school type, this, too, appears to be similar for both boys and girls. For example, although it is true that girls tended to perform better in all-girls' schools, the same was true for boys in all-boys' schools.
Finally, we are examining characteristics of the students themselves, both cognitive and social. Again, although only tentative, it appears that there are no striking differences between the way that boys and girls react to the format of the Bee itself (e.g., reported somatic and cognitive anxiety prior to the state bees) and that the relation between performance in other kinds of competition and the Bee is similar (e.g., for both boys and girls, higher performance on the Geography Bee is positively correlated with success in other academic competitions, but negatively associated with performance on nonacademic competitions, suggesting that a competitive spirit per se is not the overriding factor). What does appear to be associated with higher performance and with being male are the variables that are linked to the substantive domain of geography in particular. For example, correlated with better Bee performance and more evident in boys are higher reported 'liking' of geography, greater participation in geography-related leisure activities (e.g., reading maps and atlases), reading certain kinds of books that may be expected to have greater geography content, and higher spatial skills.
Thus, the data so far appear to suggest that the boy-girl difference found in performance on the Geography Bee is actually less likely to be a reflection of something highly specific to the way the Bee is designed by NGS or to the way it is implemented by the schools, than it is to reflect students' knowledge of and interest in geography. If further data and analyses support this generalization, the kinds of interventions likely to be effective are those that encourage girls' greater enjoyment of and exposure to geography, rather than those that are aimed at modifying the format of the National Geography Bee itself. At the same time that these data should prove useful to continuing to plan the educational initiatives of NGS, they are proving useful in examining questions derived from developmental theories concerning, for example, the interaction between self-concept and academic achievement and how these may be moderated by sex, the relation between individuals' gender-stereotypes attitudes toward others and their involvement in gender-stereotyped activities themselves, and how extrinsic and intrinsic motivations for achievement may operate differently for boys and girls. The data base that has evolved reflects both the basic theoretical research interests of the investigators, as well as the very immediate concerns of the staff of the National Geography Bee in their continued quest to promote geographic literacy for all students.