That's a wonderful question.
When we developed the victimization surveys, we took a lot of care and concern about respondents. Two strategies were used. On the individual's computer, they would follow a link to Statistics Canada and the questions would come up. There was a help button, so if someone was concerned about their safety—perhaps their partner was in the room—they could click that button and it would pop out and go to some other home page within Statistics Canada about cow purchases. I can't remember what it was. Secondary to that, we also added, at the very end, a list of victim services and call centres, if anyone was in need of help.
For the telephone interviews, we trained interviewers to detect if there was a change in patterns between the interviewer and the respondent. Perhaps they were now being reluctant to share stuff. The interviewer would say, “Is this a good time or a bad time?” In addition, they offered respondents links to different victim services. We took a lot of care.
We would never follow up in a secondary survey to ask additional questions. Everything was anonymized. Once you responded to the survey, there was no way for us to contact you again, because your first and last name would have been stripped from the data file.