How the (sprout)(mince)(turkey) do I work out a response rate of an online survey that went out in an organisation?

I don't think I can say it went out to EVERYONE (30,000 employees) in that organisation.

I started with sending it out to 50 people say and then got it snowballed. Final sample is N=240, but I have no idea how many people SAW the questionnaire and didn't fill it out.

(please mods can we have some more smilies)

Do you know if it went to any lists like @all_staff of some such similar thing? You might be able to ask admin for the numbers of recipients on any lists used.

no, I spent 3 months trying to get it on a @staff type email list, but you had to put a 'business case' forward - which involved a 30 page document, meeting with senior partners, getting your department to budget it (its an F***ing email!!!) etc. and as a 'student' rather than an employee I just couldn't do it.

In the end I just emailed it to people who had been interested in previous studies and got them to forward it on. It did go out to a network of people in the oganisation, but that was to make sure I got enough of that specific sample e.g. green haired people, I still need to say it adequatly (or not) samples the WHOLE organisation - I know it adequatly samples the green haired people.

oh and i've found a study that explains how you can compare the organisations stats against stats for the questionanire - so I can test sampling adequacy.

Its just the actual response rate I'm confused about!

In previous work I've been advised to state that whilst X number of people work at organisation Y where the study was advertised it is not possible to state the response rate as snowballing was used.

It's pretty common, I've seen it stated in loads of academic papers.

I've also stated blah blah blah lowest possible response rate is Z% (where Z=X/Y*100) but I personally think it's pretty useless and normally makes your response rate look horrendous!

Good luck

A