Random sample - how to get one?

[L.A.]

Hi, I’m doing a research on job satisfaction in lawyers and I would like to use a random sample. My question is: how can I do this?
In the whole country we have 7 regional bars and lawyers must be affiliated to one of them.
Thank you.

 

[interestednew]

Are you doing some sort of experiment.....i.e. show certain lawyers in certan fields/cities/ect are less happy than others? If so you may want to consult a statistician on experiment design. This can be a very involved question depending on what you are trying to accomplish.

 

[L.A.]

Well, I would like to get a general picture of the overall satisfaction in the profession and if there are some differences related to gender or income for example….

 

[Link]

It's kind of hard to get a random sample in that area. In some way, you're going to have bias. Let's ignore for the moment the requirement that lawyers "must be" affiliated with one of the 7 bars, and take for example that you randomly select lawyers from the regional bars. Then, you'd still need to hope that those not responding are completely at random. Its plausible that those not responding could be more dissatisfied with the profession or too busy making money (putting them at higher income).

 

[krytellan]

Yeah, getting a random sample via (I assume) survey mailings is tough, but there is little you can do to work around it. The only other consideration you can make is that the more regional factors you are able to control for, the more you can have some confidence in your results. If you can find a way to control for region and other variables you would consider to be unique across groups, that would be a good idea.

 

[trinker]

ANCOVA could be a good choice here.