A couple of caveats:
First, it’s entirely likely that some people lied when answering this question; it would be surprising if they hadn’t. But the operative question, unanswerable, is whether people in certain countries were more or less likely to answer the question honestly. The willingness to state such a preference out loud, though, might be an indicator of racial attitudes in itself. Second, the survey is not conducted every year; some of the results are very recent and some are several years old, so we’re assuming the results are static, which might not be the case.