Doing an masters by research at UCC and my main data source was an online survey via surveymonkey sent to a sample group of multi-national company managers. The survey was well received I think, but I believe company firewalls etc may have prevented the email invite and link to the survey from reaching many. Still, I'll end up with a response rate of about 30%,out of roughly 220 invites sent. Not analysed the data in detail yet. My question is : on the face of it, is a 30% response rate broadly acceptable to examiners? Asking my supervisor the same question, but interested in a wider 'take' on this.
FYI time constraints meant this was the most expedient method to gather data, but I think next time, I'd have back-ups.
Depends greatly on the type of questions and your intended analysis. Were they largely open ended? Multi-choice? Are you building scales?
If you're going to be doing serious quants on them, then you will have to pay very close attention to statistical tests relevant to your sample characteristics. If your questions are largely qualitative, you will have to be just as careful with your postcoding.
Can you elaborate on the nature of the survey (the instrument itself)? Details of methodology?
Thanks for your response. Clearly I'm going to have to bring myself up to speed quickly on this topic as its been years since I studied research methods.
To answer your questions. Survey was largely multi-choice, about twenty questions seeking to answer three tightly defined research questions. Answers had options of comment boxes varying between a line and a para as appropriate. Definitely not doing 'serious quants' cause I've no idea where to start with that. Questions involve a performance management methodology.....does your company use it.....does it align with strategy....is it reviewed frequently.......thats the gist of what I'm seeking to find out via the survey's twenty odd questions. Used survey monkey to automate it and collect it. Collection finishes next week, then I have three months to analyse and write up conclusions etc and submit. I know this is very basic stuff compared to scientific/peer reviewed research and the way collection was done hasn't turned out to be ideal, but times a big issue. Must submit by August.
Any additional helpful input, most welcome.
Would really need to have a look at your proposal and questionnaire to be sure.
There are ways to manage this within your thesis though - In terms of analysis, if you were aiming to measure frequencies amongst the respondents on individual variables, your results are not likely to be generalizable to the sample with your response rate. This in turn depends on how much went into drawing your sample - if you explicity sought out respondents, weighted your sample by business type (sorry I dont know business), gender, or location then you may run into problems.
As in, you couldn't say '30% of x agree with y' with reasonable confidence. Will you be doing anything to your data besides frequencies?
There are ways to manage this within the context of your thesis though
You could approach your analysis inductively depending on the nature of your questions, but again, I would need to see your questionnaire - pm the surveymonkey link if you want me to have a look.
There is always scope to adjust your methodology, and there are many ways to analyse - I wouldn't worry too much about the response rate. Seems as if you may be aiming for an elusive sample - try finding some articles by researchers who may have run into the same difficulties.
hi, I'm doing a masters at the moment as well, we were told response rate of about 40% is ok!! I'm sure 30% is fine and I wouldnt worry about it too much