Quote:
However its not like they need to hunt us down, they already have our information.
Firstly, I have no idea how many members they e-mailed, and there is a chance that a lot of the emails ended up in spam filters (including at your ISP). The company which was contracted is as mentioned "cheap" and you need a relatively advanced e-mail system to ensure that the invitations do not get blocked somewhere. Also, doing a very large respondent group is not without cost, there are bounces, the import of databases etc etc. not a huge cost but as mentioned this survey is done on the cheap. Also, the survey system has to be geared for large response numbers (and cheaper online systems are not).
Quote:
I would of thought the more people who complete the survey, the more accurate there results could be... Im sure the person in recieving the information, would have more confidence in there figures, if more people
filled out the surveys.
This is actually not true. I am not a statistics guru, but I work with one.
Normally in research a random 100 responses will give you valid results (depending on the total size of the sample). Actually if you survey 500 people out 30,000 you have about 99% statistical confidence in the results. The only reason you would survey more people is if you want to segment the result (e.g. by age group or gender). So if you want reliable results for 9-sub groups (segments) you would need 100 in each group i.e. at least 900 (depending on how even the spread is in the segments).
I hope this explains it.