I now have a sample of an even 50 respondents. I have just e-mailed my collegue about getting the data analyzed. Now, there are are a couple of concepts regarding samples I thought I would mention. Was the sample of our community a random one. No, it was voluntary and there may be reasons why some may not have volunteered. For example, if someone had no symptoms before or after the JW experience they may think their imput didn't matter. Also, some who actualy feel worse now, my not want to admit that. Therefore, the sample may not be representative of the entire mix of posters either. We can only have faith that it is representative, but we can't be sure. Doing this kind of research is like "looking through a glass darkly", I Cor. 13. Can the results of this sample generalize to other groups? I think it could generalize to other groups of internet discussion groups. Could it generalize to all other XJW's? Probably not, if an xJw is extremely ill, or, still psychotic, they are not likely to find the internet. Also, the very poor or poorly educated (formal or informal) may not be represented by our sample. With that said, I think the results will be interesting and I should have the summary statistics this weekend.
Comments on my survey
by larc 8 Replies latest jw friends
-
-
larc
Oh, I almost forgot. Please add your data to the survey, it you have not already.
-
butalbee
Okay, I did. Hope it is useful for yeah.
-
belbab
I started to reply, but couldn't figure out how to paste the questions to the reply box. and
did not have time to figure it out. Does that count as a reply?
belbab
-
Dia
Ummm....what survey?
-
larc
belbab, just put the number and your answer without repeating the questions. That will work. Dia, the thread is called Larc's weekend survey.
-
Prisca
Larc,
Can we email you our replies to your survey, if we don't want the information made public?
-
larc
Prisca, that would be fine. Anyone who does not want the data to be public, can e-mail it to me. Names are not put into the data base and therefore will not appear anywhere.
-
Amazing
Hi Larc: Why not take a second survey by emailing 50 or 100 randomly selected screen names from this board. And have them respond only in private ... and then compare the two surveys and see how close or far apart they are ... a sample of 50 is pretty good, but 100 or more better helps the function of large numbers. I think the findings of a second survey would be most interesting and possibly even more meaningful. - Jim W.