Yes, when you report research results to your client, you’re going to want to group the responses into age groupings. Makes sense.
Why, though, are you asking your respondents to enter their answer using those same groupings? Three real issues I have with this approach:
1) When I see groups, I immediately think that some of the groups will qualify to continue, and others won’t. I want that $2, or that sweepstakes entry, or, frankly, I want to kill fifteen minutes; I don’t want to get told I don’t qualify. Therefore, I’m not telling it I’m in the lowest group or the oldest group, and really, probably not the next ones in, either. I’m playing it safe and saying I’m 30-59; that’s probably the safe range.
2) Down the road, if the client wants to see the data broken out in different ways — let’s say the client comes back and asks to see the 25-54 year-olds broken out — there’s no way to provide that data.
3) Why limit yourself to nothing but radio buttons? It’s not difficult to program a text-entry field that’s limited to accepting only digits; it’s not especially difficult to program a slider that lets respondents navigate to their exact age or year of birth, either.
4) Yes, I said three real issues; this is the bonus, fourth issue. This isn’t RDD research — this is web panel work. The panel company already knows my age, as well as my gender, my race, my current mode of internet connectivity, and a thousand other things. Why can’t it use that stored data to fill in the correct responses to these questions without ever even showing them to me? Yes, that requires a lot more work — that’s why I consider this a “bonus” item — but the panel company that figures out how to do this reliably and consistently will reap the benefits of significantly shorter surveys — especially on the qualifications end. Greenfield is particularly bad about this, asking me the same five or ten initial questions before every piece of research and then telling me I’m not qualified for it — so why did they present it to me in the first place?