An interesting usability problem: my bank offered me 3 sets of security questions. I was to choose one from each set. If they had offered me all of them, from which to choose three, it would have been OK, but I had problems picking one from each set, because of the assumptions behind the questions:
- Several questions assumed the respondent was married (spouse’s middle name, where you got engaged, city in which you met your spouse) — I’m not. (And I’m certainly not going to answer about my ex.)
- Some were about kids (date of birth of your first child) – I have none.
- Some were about popular culture –name of your favorite actor/actress: I don’t have one, and, even if I did, this would be likely to change.
- Name of high school mascot: I went to a girls’ high school, and though we did have a basketball team, we had no mascot.
- Some expect you to remember things that I don’t, so I may not come with the same answer some possibly-distant time in the future: first state that you visited outside of your home state (we did road trips when I was a kid and I don’t know where we went when), name of your first boyfriend/girlfriend (does the boy I kissed in 8th grade count?).
The people who thought up these questions would no doubt say, We don’t expect them all to apply, but you only need one from each list. But if MANY of them don’t apply to me, and since they offer only a subset of them for each question, I had to pick at least one question to which I’m not sure I’ll provide a consistent answer.
I can easily imagine someone from another culture having considerably more trouble.
And no, writing your own question was not an option.
This is a symptom of a larger usability problem: we assume that the people for whom we are designing are relatively homogeneous, and while they may be, to a large degree, what happens to those who are different? One would think that in the Bay Area, of all places, there would be more awareness that such assumptions about people’s life experiences would be better thought through.