Do the same trends emerge among non-whites when it comes to IQ and political orientations? The following tables show IQ scores, converted from wordsum scores on the assumption that the mean white wordsum score represents an IQ of 100 with a standard deviation of 15, by race and political orientation. For contemporary relevance, data are from 2000 onward:
| Blacks (n = 479) | IQ |
| Liberal | 89.3 |
| Moderate | 89.9 |
| Conservative | 88.7 |
| Hispanics (n = 247) | IQ |
| Liberal | 90.0 |
| Moderate | 90.3 |
| Conservative | 87.8 |
| Asians (n = 140) | IQ |
| Liberal | 91.8 |
| Moderate | 93.9 |
| Conservative | 94.3 |
Conservative NAM IQs are a bit more modest than moderate and liberal NAM IQs are. The opposite is true regarding Asians. While the differences are too marginal to put much stake in, this generally meshes with my personal observations. I am surprised to see how moderate scores are not conspicuously lower than liberals and conservatives among non-whites like they are among whites, however. I'm not sure why this is the case.
Relative to actual measures of intelligence, Asian wordsum scores are low and should not be considered an accurate measure of IQ in comparison to non-Asians. East Asians (and in the case of the GSS, we are primarily talking about East and Southeast Asians rather than Indians) perform worse on verbal tests of intelligence than whites do, but more than compensate for that in mathematical and visuo-spatial testing. The GSS wordsum test is one based entirely on vocabulary and consequently puts Asians at a disadvantage. Still, for intra-Asian comparisons, wordsum is presumably a reasonable proxy measure to use in gauging differences in intelligence.
GSS variables used: RACECEN1(2)(4-10)(15-16), WORDSUM, POLVIEWS(1-3)(4)(5-7), YEAR(2000-2008)
8 comments:
Can you pull the Jews out of the whites and then check?
I think PARTYID is a better variable to use here. As you know, even though large numbers of minorities self-identify as conservative, they vote reliably Democratic, so it doesn't necessarily mean the same thing it does for whites. If you use PARTYID instead, the "moderate drop" appears for Hispanics and Asians but not for blacks - it may be that they are so reliably Democrat that the smarter blacks go along with the popular opinion rather than risk ostracism from the community.
Anon,
Jews identifying as politically conservative are almost non-existent among whites (8 of the 761 respondents so classified), and they score essentially the same as white gentiles do, so it makes no difference on the conservative side. Among liberals, they are 25 of 458, with an IQ advantage of about 10 points over liberal white goyim. Thus Jews inflate the white liberal score by half an IQ point, narrowing the gap a little but still leaving it at almost 3 points between white liberals and white conservatives.
Jokah,
Interesting. Black independents actually come off looking better than Democratic (barely) and Republican (by a lot) blacks do. But as you say, the moderate = modest IQ relationship holds among Asians and Hispanics. You're explanation seems plausible to me.
Sample sizes are too small to make anything out of this.
If anything, it shows that political orientation is orthogonal to IQ.
I think Asian Wordsums are only low on GSS due to a high proportion of immigrants (who have relatively weak language skills).
If you control with Born(1) - American born -, I don't think lower Asian Wordsum exists and they actually have the strongest score on this variable.
Anon,
Wow, great insight. I didn't realize the foreign-born percentage of Asians was so high. You're exactly right--among those born in the US, wordsum scores are on par with whites (actually slightly higher).
Can you do a significance test on this data to see if this spread is significant or just as likely as random error? I really doubt that there is a significant difference on the liberal data. That is too close, but is the 5 point spread between high and low significant. I also give a 10 point margin of error when I write up an IQ score on a psychological evaluation I have given.
The spreads are not statistically significant for any of the intra-race scores, though the difference between blacks and Asians and Hispanics and Asians are (at 95% confidence).
Post a Comment