Average vs Outlier Differences and from Variable to

Average vs. Outlier Differences and from Variable to Explanation

Average vs. Outlier Differences Question: What explains seemingly systematic gender, racial, and ethnic performance differences in sports, academics, career professions, etc? 494 of the top 500 times in the 100 -meters are held by sprinters of West African ancestry. • • 10, 000 meters: 15 of the top 20 fastest times were by Africans 5, 000 meters: 18 of 20 1, 500 meters: 13 of 20 Sprints (100 meters or less): 20 of 20 Elite endurance runners are primarily East African (Kenyans, Tanzanians), but rarely, if ever, West African: New York and Boston Marathons No white, Asian, or East African has ever broken 10 seconds in the 100 meters, while dozens and dozens of West Africans have.

Average vs. Outlier Differences 14% of U. S. population is black: NFL has 65% of its roster filled by black athletes, NBA 80%, WNBA 70%. So what do people often assume? KEY POINT: Variability Differences vs. Average Differences

Average vs. Outlier Differences • Males are much likelier than females to be found on both of the tail ends of the bell curve, among the superhigh scorers and the very bottom performers: For example, among college-bound seniors who took the math SAT’s in 2001, for example, nearly twice as many boys as girls scored over 700, and the ratio skews ever more male the closer one gets to the top tally of 800 (there is a 1. 3 boys-to-girls ratio in the top 10% of SAT math scores, 1. 5 ratio in the top 5%, and a whopping 7 -to-1 ratio in the top 1% of scores). *Yet boys are also more likely than girls to get nearly all the answers wrong. • The overwhelmingly male tails of the bell curve may be telling. Such results, taken together with assorted other neuro-curiosities like the comparatively greater number of boys with learning disorders, autism, and ADHD, suggest that the male brain is a delicate object, inherently prone to extremes, both of incompetence and genius. This may explain how and why girls’ average cumulative GPAs in college are higher than boys’. The SAT turns out to underpredict female performance and overpredict male performance in college. (e. g. , Pi Sigma Alpha candidates in the Political Science Dept. )

Other Possible Factors for Performance Differences Environment? Top-scoring girls are only about 60% as likely as top-scoring boys to pursue science or engineering careers, for reasons that remain unclear. - Surveying a representative population of working scientists and engineers, Dr. Catherine Weinberger (economist, UC Santa Barbara) discovered that women were more likely than men to have very high tests scores. “Women are more cautious about entering these professions unless they have very high test scores to begin with, ” she said. Discrimination? Recent experiment in which Princeton students were asked to evaluate two highly qualified candidates for an engineering job – one with more education, the other with more experience. The students picked the more educated candidate 75% of the time. BUT when the candidates were designated as male or female, and the educated candidate bore a female name, suddenly she was preferred only 48% of the time (NYTimes, 1/24/2005).

Possible Environmental Factors for Performance Differences • “ability attribution” – success is due to ability/talent, while failure is someone or something else’s fault (e. g. , boys and math) • “learned helplessness” – the state in which failure is perceived as insurmountable (e. g. , girls and math) -----------------------------------------Substitute the words “sports” for “math” and “whites” for “girls”: - many top white athletes have been saddled with the athletic equivalent of “learned helplessness” - many top female math students have been saddled with the academic equivalent of “learned helplessness” * Wild-care factor desire: the great intangible in performance (baseball and Dominicans or Puerto Ricans; hockey and Canadians; chess and Russians; socioeconomic class and elite athletes, etc. )

“Tipping Point” Case Studies diffusion models: hybrid corn seed in Greene County, Iowa - Innovators & Early Adopters then Early Majority, Late Majority, and finally the Laggards (introduced in Iowa) diffusion models and the distortion process. . . leveling: all but a few details are dropped sharpening: certain details are sharpened assimilated: a rumor or a product is changed so that it makes more sense e. g. , anatomy of a rumor and the Chinese teacher in Maine in 1945 Johns Hopkins needle-exchange program and public health: role of “super-exchangers” (Connectors) in starting a counter-drug epidemic

From Variable to Explanation Why? questions -- Variables: “Why do boys (independent variable 1), on average, get higher combined SAT scores (dependent variable 1), but girls (independent variable 2) get higher high school and college cumulative GPAs (dependent variable 2)? Or, again, why do SAT scores (independent variable) seem to underpredict girls’ aggregate GPA’s (dependent variable 1) and overpredict boys’ aggregate GPA’s (dependent variable 2). Pollack, pp. 32 -33: Why did Hush Puppies “tip” and go on to sell over 1. 5 million pairs in 1996? Gladwell links a dependent variable (the success or failure of contagion), to an independent variable (the presence or absence of connectors).