I've heard some version of one question from an overwhelming number of readers: "I've met lots of Indians and Americans and found the former more racially tolerant than the latter. How can these results possibly be correct?" I'd suggest three possible explanations for this, some combination of which may or may not be true. First, both India and the U.S. are enormous countries; anecdotal interactions are not representative of the whole, particularly given that people who are wealthy enough to travel internationally may be likely to encounter some subsets of these respective populations more than others.
Second, the survey question gets to internal, personal preferences; what the respondents want. One person's experiences hanging out with Americans or Indians, in addition to being anecdotal, only tell you about their outward behavior. Both of those ways of observing racial attitudes might suggest something about racial tolerance, but they're different indicators that measure different things, which could help explain how one might contradict the other.
Third, the survey question is a way of judging racial tolerance but, like many social science metrics, is indirect and imperfect. I cited the hypothetical about Swedes and Finns at the top of this post, noting that perhaps some people are just more honest about their racial tolerance than others. It's entirely possible that we're seeing some version of this effect in the U.S.-India comparison; maybe, for example, Americans are conditioned by their education and media to keep these sorts of racial preferences private, i.e. to lie about them on surveys, in a way that Indians might not be. That difference would be interesting in itself, but alas there is no survey question for honesty.
