Done the survey.
TBH the lad’s got plenty of time to sit down and do a second survey if he feels it would help.
I assume the questions are chosen to provide data which is comparable to existing stuff, but IMO they won’t – use the same entire set of questions, except ones that really aren’t applicable, survey only “elite” mountain bikers (people who actually race, or that bloke with the new Ragley who never stops), or include a question on whether the respondent races, and you’ll be closer. My motivations to nip out on the bike and Steve Peat or Lance Armstrong’s will be totally different most of the time.
Or take an entirely different tack and ask a set of questions that are more suited to mountain biking, include more possiblities (ie, nature / outdoors, etc), and recognise that you’re not going to produce comparable data. Ask different questions of a different group and try to compare it with the research that’s gone before and you’ll struggle.
I’m sure his tutor will be putting the fear of god into him at present, but to be honest, an undergrad dissertation is 10000 words. The guy’s got 5 weeks to write it up, with some spare time to proof and print it. By dissertation time you should know what’s a realistic per day word count, and know if you can realistically take more time to research. Online surveys have a huge advantage over traditional data collection in terms of speed of data collection and collation for one. They’re also the hallmark of someone writing in a hurry for that reason – maybe go to a race or trail centre for an afternoon and question a few people…
I’d personally stop now and do more research – starting with a decent Research Methods text like Bryman.
Having done a couple of dissertations myself, I know this is hardly music to anyone’s ears, but I reckon you’ve got the time to make some big improvements to the research, even a quick face to face survey would make a fair impact.
Edit: I’ve made the assumption that this is the main piece of primary research. If it’s an aside to something else / a dataset someone else has given you, then fair enough.