In 2004, the GSS asked a good proxy question for determining which side of the alpha/beta line respondents fall on. In the following table, those who agreed strongly with the statement that they would rather personally suffer than have a lover suffer are classified as betas, while those who responded otherwise are classified as alphas:
|White (n = 502)||28.0%||72.0%|
|Black (n = 52)||42.9%||57.1%|
|Hispanic (n = 76)||36.1%||63.9%|
Because the question has only been posed once, the non-white samples are small. That said, the results do not surprise me. Black men are fairly close to being evenly split, while there are nearly three white betas for each white alpha. Hispanics fall in between at two betas for each alpha.
Rather than grouping all other responses (from "agree somewhat" to "disagree strongly") together, the issue can also be approached by comparing the average response by race (1 = "agree strongly", 5 = "disagree strongly). The higher the mean, the more alpha the group is:
One standard deviation is .70 points on the index scale, placing the middling black male half a SD higher in alphaness than the average white guy, or at about the 67th percentile among white men.
GSS variables used: AGAPE1, SEX(1), RACE, HISPANIC(2-99)