4 ms·

How often is someone drunk? 30% is absolutely huge conditionally.

• pyre 16y ago
  Really? There is going to be some correlation between the times that people are most likely to commit sexual assaults and the times that people are likely to be drinking. I'm sure people spend a lot of time doing things like commuting to work or sleeping at night, but it's highly unlikely that they will commit sexual assaults during those times. People are just more likely to be drinking at times when they are looking for sex (e.g. at a party, at a bar, etc). You should probably be following something like the following line of reasoning: 1. How many sexual assaults happen on dates? 2. How prevalent is drinking while on a date? (Maybe you'll have to divide the data up regionally or something for it to make sense) 3. How many sexual assaults happen when there was drinking on a date? Then you can start to look at how likely it is that alcohol is really the cause of sexual assault (at least for the dataset that you're looking at).