How can I get a more diverse range of people to take my survey?

Try Survey Monkey. Not only is it a great way to create surveys but it is an easy/fun way for people to respond to surveys and will help get a more diverse group to respond. At least that has been my experience. One downfall, though, is that you might have to subscribe. I think you can join for free but access may be limited. At the least it might give you some ideas.

Ah… one of the hardest bits for anyone sending out a survey is trying to make people fill the damn thing in.

You could look at your survey and ask yourself why the majority of respondents were all of the same demographic. It could be that unconsciously you’ve created a questionnaire that appeals to that particular group. That might be something to consider when you write up your work for class.

As for actually improving the number of responses you have to be careful that your methods don’t actually influence what people put in their replies. If you hassle people too much they might reply but do so without really answering your questions properly there by screwing up your results. Having said that a gentle reminder may improve responses. You say you sent it out to the whole school so perhaps you could liase with the teachers to remind their classes that there is a survey to fill in.

As for your follow on question. I don’t think it would get your question flagged but, would providing a link to the survey make answering your first question any easier?

Ava asked for us to fill out a survey for her on fluther. Or maybe it was andrew asking for her. So I would ask away.

Why does it matter if you get more demographic variation? It’s not like you’ve got a generalizable survey. It’s a convenience sample, not a random sample, so you won’t be able to say anything more about the answers even if you do get more demographic variation.

Did a professor tell you to do this? Is this just for practice? Or fun?

The best way to improve response rate is to contact people personally. You probably don’t have the money to do that. You can send out the survey again. Word the cover letter differently. Say only females or people over the age of 80 should respond. You can call people. Beg people. Another way people try to improve response rate is by offering a reward. You could buy an iPOD or something, and say that everyone who replies will be entered into a drawing for the iPOD, then hold the drawing and send out the gift to the winner.

Since you don’t know who is responding, or rather, you don’t know if they are telling the truth about themselves, your survey isn’t going to have much validity in the real world. I wouldn’t think it worth while to do all this work. But perhaps you could explain? Of course, if it’s just for fun, that’s a whole different situation.

Well I’ve written the survey and sent it to all the students at my school (which is an all-boys school) which is why I have a majority of responses that are males born between 1990–1996! The idea of the survey is to find if people have changed their attitudes to traditional media like newspapers, magazines, TV shows and have started to use more online media like blogs and social networking sites.

This would be fine, except that I also posted a link on Facebook so now I have a majority male but also about 20 females, all of which are over the age of 40 (my uncle sent it on to everyone I think).

I’m not really fussed if I can’t get a more diverse range, I’ll just mention that as I analyse the data. The point is that I might as well try to get as diverse as possible audience for my survey (It’s also been distributed in paper format) so I can compare media use across age-group, gender, country etc.

The thing is that I just sent the link in my Facebook status and via email and got over 80 responses, so I don’t think I particularly need to tempt people into doing it.

There doesn’t seem to be much enthusiasm on Fluther so I think i’ll do another push through Facebook and send out a full address-book email.