Hi All, Judging by the 4 downvotes, It's pretty clear that there's something wrong with my question skeptics.stackexchange.com/questions/35616/…

I think I know what it is. Since the image already has a source, i just misunderstood it, I was asking a question about data that wasn't really questionable in the first place. Is this correct?

I could just delete this question and walk away, but I want to participate here and having a bad first experience (even if it's my own fault) is discouraging. As such, I'd really like to understand why I got this wrong so that in the future when I have another question I don't make the same mistake.

@stanri I've tried to motivate my concerns in a comment. Unfortunately this came with putting the question on hold.

The problem is that all we can deduce from the image itself is that at some point, during this electoral campaign, someone counted the subset of votes for trump and clinton based on skin color and gender, then put out a map of who of the two candidates would have more votes.

The problem is that people interpret it as a prediction, or anything resembling a vote or a voting poll. But this is a methodological error, not really a data error.

i.e. you can't ask people who they would vote for in a normal election and pretend that the same answers would apply if the election were completely different.

A site where this question might have better fortune is Cross Validated, but it's very unlikely that it can ever work here unless we reduce it to the bare minimum (is the data correct) removing any implied claim, but even then I can see why people would downvote the question as promoting a bad use of stats.

@Sklivvz Thanks. I understand now. My mistake was (1) not being very clear on what these maps mean and how they are made and (2) not understanding properly what qualifies as a good question here (particularly notableness and a claim being written). I've read a couple of the meta posts you linked and have learnt a lot on both points. Thanks for your help!

When I do have a more appropriate question to ask in the future, I'll give it a better shot at getting it right :)

I'm tempted to turn this into a question, but not sure how

techcrunch.com/2016/10/25/…