Thank you, Mr. Chair.
I'll ask Mr. Swales to give the details. In paragraph 3.50, we explain what we did. At the end, yes, we took a sample of vehicles to identify whether the information had been collected or not, but as we say in paragraph 3.50, we started with 19 million records, and based on those we identified 511,000 that looked like anomalies. Then, we took the sample from those 511,000 to determine whether they really were anomalies or not. The first pass of the complete population indicated to us that there could be as many as 511,000. From the sample, we were able to reduce that down to about 300,000. It wasn't simply a matter of taking a certain number or a small sample out of the 19 million. We did start by looking at the whole population.
I'll ask Mr. Swales if he needs to correct anything I just said, or if he has any more details to add to that.