hi, i'm about to launch a postal survey with 500 questionnaires ......i have thought of putting two envelopes an two stamps as a return envelope, so firms can return it free, but A. I'm concerned with cost B. i'm not sure the uni would cover it...

has anyone got any idea how to reduce costs in this process?

thanks! :-(

What are you hoping to do with such a big sample size? A factor analysis? Anyway, I'd only expect about a 40% response rate, maybe less with businesses. Have you thought about improving your response rates by calling them before hand, so you can set up some kind of dialogue with them beforehand? Then they're expecting it and are more likely to return it. There's a Cochrane review on how to increase response rates - you'll find it in the Cochrane Collaboration Library online. Have you thought of posting the survey online, maybe using survey monkey, so that they can have more than one way of completing the same thing? Maybe even e-mailing them the survey to complete and return it to you in the form of an electronic document - Adobe LiveCycle is excellent for this.
At the very least, make sure that you use an invitation letter with your universitie's logo to make it look officious. Also, if your university will help you with the costs, see if you can get them to mail out your surveys, as that way it will have the university's frank on it. My university wouldn't help me with my costs, but I got an 80% response rate by ringing up my respondents and reminding them to send me my flippin' forms back.

hi1 yes i am going to try all those methods......I also have a sponsor outside the uni to add to my questionnaire as a reference too..


I am aiming for 100 firms and above out 500-600 so 40% is a great response rate. I need them to run a regression after....

how exactly do you send the envelopes and get firms to send them back? you send two of them ? two envelopes enclosed in one?

Hey Billy. I have been sending out questionnaires to participants- and firstly, I'd agree with the post below. I got about a 95% return rate by speaking to the participants over the phone and introducing myself and the project before sending the questionnaire out. With respect to the postage I enclosed a freepost return envelope- the university were able to give me some stickers with the return address and a freepost code on them, which I put on the envelopes. So basically when the envelopes are returned the postage is charged to my funding account, but the advantage of this is that if the questionnaires aren't returned, you haven't paid for the postage in advance like you would if you put a stamp on the envelope, so you're only charged for the ones that actually get returned. It's worth asking your uni if they can arrange something like this for you, it has worked well for me. Best, KB

thanks! did you put a code on each envelope in case you want to send it back to the firm due to errors etc?

No, I didn't code the envelopes as the questionnaires inside were already coded anyway so that I would be able to trace them back to the participant if I needed to. Best, KB

thanks! oh boy that would take ages.......matching each code witha firm!

what do you consider to be a good number of questionnaires to send for a pilot study? 10-20 or more?

thanks!

Is your questionnaire ready for piloting? That is, have you got some other people, other students are a good bet as they have fellow sympathy :-) so that you have the wording right? What I did was to then test it out on some people who were leaving my specific area of study, so they had the knowledge necessary to understand the questions from a professional point of view, but as they were leaving, they didn't receive the final version. This was quite important for my study, but might not be for yours. However, one thing mentioned at one of our meetings here was the need to make any necessary changes and then pilot the changed questionnaire too, something worth thinking about. I suppose the actual number depends upon how many people are available. The entire number of people available to me was much smaller than your proposed number of questionnaires, but my pilot was with 15 people who were leaving. Also the group knows what I am doing, knows me (in a virtual way) and the process was conducted via the internet. This made it swift, I had a high return rate (only 2 people didn't manage to fill it in, and that was through a technical problem which we were unable to solve, and they have subsequently joined in other bits of discussion anyway. I still produced a covering letter though, I think it is quite important, especially as it allows you to not only introduce yourself, but provide the recipient with information, like how they can withdraw, how you are going to treat the information etc. This would probably reassure the recipient and make them more likely to fill it in. 

yes it's ready for piloting and i send it to colleagues to get ideas too.....i'm just concerned at the size.

although it's according to the literature i'm not sure if 6 pages looks ok.....it has only 15 questions but is 6 pages (one side only) including coversheet.

what kind of size do you consider ok?

If you're sending them into organisations, you should really call the HR people or someone appropriate otherwise you may get into problems later down the line with ethics etc. My uni are really stressy about organisational research.

I would say 40% is very high, I was lucky to get 10% - although my sample is within an organisation and a specific group, so pretty hard to get.

Have you thought about online? - I know most people will sit in their lunch hour and fill in a questionnaire online, but few will bother to post it back, if they can be bothered to find the bit of paper they were given.

well i already passed the ethics check with my uni....so i'm ok there.

I thought about online initially but was told postal gets higher results , since people can fill it in at any time and anywhere (eg. airplane, train, tube, )....also sending many emails can look like spam whilst envelopes look more formal.
of course it costs more and i need to add return envelopes.........also, my sample is all the same, so more than one person can answer my questionnaire it doesnt matter provided they have specific info, assuming they do answer!
I am planing to launch an online one in parallel though.


how big should i keep it?

an online parrallel is a good idea - you can then snowball it.

I would keep a postal questionnaire down to 2 sides of A4, max. And an online down to 3 pages e.g. 1st page = demographics, 2nd page = questionnaire, 3rd page = questionnaire and thank you.

wwo....2 sides of A4? that would mean squeezing the questions quite abit.....not sure if it's ok for the attractiveness of it all.
I read 4-6 pages is the average and am trying to make it 4-5 pages or 3 sides to keep it short.

I guess it depends, mine was a likert scale questionnaire, so by the end of 2 sides of A4 participants had had enough. When I did extend it to 3 sides (I included another questionnaire), people either didn't fill in the last page, or they just put 'neutral' for every question :-s