Comment Re:Imbalanced Survey? (Score 1) 638
Yeah, I'm not convinced by it. As TapeCutter pointed out they may only be interested in US citizens (doubtful though considering they mention world market share). It's not just that though - they seem to have reversed it in order to draw their conclusions. Sure there may be more "liberal"* Mac users by % than Windows users, but does that mean that liberals are more likely to be Mac users than PC users? Considering the difference in market share (89.2% : 10.8% according to the infographic) the data isn't representative. Also, the breakdown of users surveyed (52% : 25% : 23% - Windows : Mac : "other/those who don't define themselves as either") seem to suggest that the actual breakdown is "liberals" who use Mac is 14.5% (58% of 25%), "liberals" who use Windows is 18.72% (36% of 52%), "others" (non-liberal doesn't necessarily mean conservative) who use Mac is 10.5% (42% of 25%) and "others" who use Windows is 33.28% (64% of 52%) with "non-aligned" (presumably "no preference/both/Linux, BSD etc") people making up the remaining 23%. Assuming my calculations are accurate, that makes the split of those surveyed (who are either Mac or Windows users since they don't give data for the "others") 43.1% "liberal" to 56.9% "non-liberal", which would seem to suggest that those who were surveyed were a particularly liberal bunch considering most Americans seem to describe themselves as conservative or moderate (according to this article the % liberals is actually about 15-25% of the population assuming nothing drastic has happened in the last 2 years).
*"Liberal" is in scare quotes since it is self-reported and what someone claims to be and what they actually are can be wildly different. Different people use different definitions for things.
The same can be said of the "4 year college degree" bit as well. Windows degree = 28.08%, Mac degree = 16.75, Windows !degree = 23.92%, Mac !degree = 8.25%, with the rest, again, being "other". This data is even less useful than the political alignment as it can be pretty much be written off by the fact that Windows is extremely dominant in market share (i.e. the masses generally use Windows) and that, generally speaking, Windows computers are cheaper than the equivalent Macs (degree = higher chance of high payed job -> more likely to be willing to spend a lot of money on a computer).
There also seem to be some other rather suspect figures. They have placed "Win users are more likely to be suburban/rural" opposite "52% of Mac users live in a city". Those are connected but not directly comparable figures (what about urban town dwellers for example, and what are the actual figures for Win users/ non-city Mac users). The fashion/aesthetics clothing section doesn't give any direct comparisons, and under food & drink, I don't see how any of it is relevant other than possibly the vegetarianism bit. The newspaper preference is also not directly comparable (unless by "The Times" they mean "The New York Times" - considering there is a "The Times" paper, in the UK/on the internet at least, that would at best be misleading) and from websites down to magazines they simply list items (which are presumably "preferred" by that group, but without knowing the selection criteria the data is useless; that also applies to most of the food and drink section).
Interestingly, the figures on the bar chart at the bottom which shows preferred computer type has different figures than the main infographic. That's probably down to rounding though as the graph also includes a "skipped" option (when that is removed the numbers almost correspond).
All in all, the whole thing screams of data dredging, whether intentional or not, and is certainly skewed by the fact that all respondents were visitors to the Hunch rather than members of the general public, and that the creator of the infographic is a Mac user (although the same would likely be true in reverse for a Windows user - the only way to be truly unbiased in a situation like this is if the creator is not a computer user).