Websites and Gender

I came across this cool site last week, but as I was at the beach (oh so nice), I'm sharing it now.  Anyway, it is a site that examines your web browsing history to predict your gender. In keeping with other sites that aren't so sure about me (http://community.sparknotes.com/gender/) this says I am just as likely to be male as female.  What's particularly interesting is that it gives the gender ratios of the sites you visit.  Here's my results:

Likelihood of you being FEMALE is 50%

Likelihood of you being MALE is 50%

Site	Male-Female Ratio
google.com	0.98
youtube.com	1
cnn.com	1.35
nytimes.com	1.13
wunderground.com	1.22
washingtonpost.com	1.15
orbitz.com	0.83
snopes.com	0.74
snapfish.com	0.57
walgreens.com	0.64
linkedin.com	0.94
blockbuster.com	0.83
buy.com	1.44
slate.com	1.11
aa.com	0.83
cvs.com	0.64
zillow.com	1.13
salon.com	1.13
ncsu.edu	0.87
autopartswarehouse.com	1.74
pandora.com	0.9
papajohns.com	0.79
indiana.edu	0.83
logitech.com	1.41
advanceautoparts.com	1.94

Basically, looks like sites where you can buy things tend to skew heavily female– thus my on-line shopping pushes my results in that direction.  On the other hand, news sites definitely skew male.  Good thing I was looking up some auto parts recently, heavily male, or else I would have come out at greater than 50% likelihood female.