Can you please tell us a little bit about the data/population that served as the basis of this analysis (ex: how big is total population of data set, how big is the subset of users who went to both gym and park, etc). It would help us better provide feedback and understand the your results.
The restaurant data is the only one that is at odds with intuition, and strongly so. Why would Thursday be 3.5x more popular than Wednesday? Without an explanation, I'm more willing to believe that there was a data anomaly.
Am i infer that half-way between wednesday and thursday (midnight) that the check-in rate for restaurants is half way between the two (~450)?<p>Silly choice of graph type.
I'm amazed so few people are at the parks in the mornings. I always thought that they were crammed with people walking their animals, exercising, etc... from around 5:30 to 8:00 am.<p>I guess people there like to sleep in.
Has any study been done about the likelihood of people to check-in with one of these social apps? If someone checks-in at the gym on Monday, are they going to be less likely to check in again on Tuesday?
People I know seem to check-in when they are subtly bragging about what they're doing, so is it likely they'll check-in at a non-trendy restaurant on a Tuesday evening?
Updated graph on gym usage by hour: <a href="http://simplyryan.files.wordpress.com/2010/07/gymhour.png" rel="nofollow">http://simplyryan.files.wordpress.com/2010/07/gymhour.png</a>