PDA

View Full Version : the oldest 100



maria_t
09-01-2010, 05:15 AM
Iím studying activity (number of posts) for the first 100 users on an social network. It is skewness distribution. I am trying to find out if there are activity differences between men and women (differences between posts published by men and by women in this population of the oldest 100 users).
My questions are:

In this situation may we talk about statistical significance? Should I use Mann Whitney to see if there are differences between men and women?

If not, how I can describe the difference between men and women? Should I use median? Or weight mean?

Mean Joe
09-01-2010, 01:05 PM
To make sure your study group is fitting your goal (differences between men and women); why are you studying the first 100 users?
If you find a difference from these first 100 users, then would your conclusion be that men differ from women in general? Your conclusion might only be applicable to early adopters of the social network.

maria_t
09-01-2010, 03:26 PM
To make sure your study group is fitting your goal (differences between men and women); why are you studying the first 100 users?
If you find a difference from these first 100 users, then would your conclusion be that men differ from women in general? Your conclusion might only be applicable to early adopters of the social network.

Thank for your replay. I chose to study the fist 100 users as a case study, as a remember for the beginning of this social media network etc. It's not a sample. I don't want to inference the results to the population.

I measure the total posts published by these 100 first users. I want to know if there are differences between the posts published by men and posts published by women( from these 100).
I've used Mann Whitney test (distribution is not normal), and there are not differences in the publishing activity of men and women. If I did something wrong with this test, please let me know. Thanks again!