It would be nice if it controlled for various factors like income, and gave it in a more parsimonious form. For example, it claims that while 54% of PC people h

It would be nice if it controlled for various factors like income, and gave it in a more parsimonious form. For example, it claims that while 54% of PC people have completed a college degree, 67% of Mac people have. But perhaps the linkage is through income. Additionally, as someone else mentioned, the "% more likely" isn't really a great measure. Also, what about the significance levels, not to mention the level of bias in the sampling? I know this is just sort of for fun, but if we're expected to make any use of this data, then it needs to be more rigorous (especially the bias).