Thank you for asking the question, because I went so fast I didn't have a chance to explain very well. I may turn to the Stats Canada folks, because they're the experts in methodology.
I used figures that were comparing average hourly salaries. The reason to do that is that it compares like with like: that is, a person in a job working for one hour gets how much money compared with a person in another job for one hour.
When you do the full year comparisons, you are including things such as the fact that women generally work less in a year, so you're not so much getting a handle on whether pay is differing for someone who has a certain level of education or training. Are they getting the same pay per hour?
Let's take doctors. I don't know what doctors make an hour, but let's say they make $100 an hour. That probably sounds low. If you look at the full year, women doctors on average work fewer hours than men doctors, so if you compare their average annual income, part of the difference is about hours of work. That's just one example.
So if you're trying to get at the question of whether women and men are paid differently for the similar work they're doing at similar levels of education, I think you probably want to look at hourly pay. On the other hand, if you want to get to the issue of hours—whether women are working fewer hours—you'd probably want to use the yearly.
I often read in the literature or in newspapers articles wherein I think people are using the 70% figure as a way of saying women are getting paid less for every hour they work. It doesn't quite mean that.