Tuesday, October 20, 2015

Facebook Experiment Update and Global Data Competition



Post Type
11-Oct
18-Oct
Friend/family
2
2
Friend/Political
2
0
friend/share
2
2
Post from group
1
3
Post from page
2
1
promoted post
1
2

I looked  at the types of posts from the first 10 posts I saw in my news feed on Facebook.  Though some of the counts are different from one week to the next, there was no significant difference according to Fisher's exact test for the ungrouped table seen above (p=0.7336) and the grouped table seen below (p=0.6307).  This could be due to a low sample size but, as a pilot study, this was a method of estimating differences without knowing the result.  A large sample size would be needed to reach the universal p-value for significance of 0.05.


post
11-Oct
18-Oct
friend
6
4
group
1
3
page
2
1
promoted
1
2

That was a fun exercise. Now I am working on the Global Data Competition to showcase my skills.   I will be on a team working on global climate change.  I will report on the results.

**Related Posts** 

Changes in the Algorithms of Research and Facebook

 

Twitter Primary: Following and Presidential Candidate Support

 

It's All About The Likes


Global Warming, Wikileaks, and Statistics: What Barry Sanders Can Teach Us.