|
|
|
|
|
|
by Shank
2899 days ago
|
|
|
This is a really cool process, but the author stops just short of completing the analysis. > It's impossible to determine any correlation at all based on this graph. I know my happiness is influenced by a lot of factors, but so far I cannot tell if sleep is one of them. Yeah, this is true. You can't just look at a graph and say that the job is done. This is the entire point for the statistics field. You need to do an actual statistical test for correlation -- not just plot a graph. Obvious trends are definitely visible in graphs, but more subtle correlations are often obscured and not visible. I'd be much more interested to see what statistical analysis on the data looks like, rather than just a graph at the end. With a huge focus on methods, I'm flabbergasted that some form of regression wasn't tried at all. It's probably worth chopping out obvious problem areas, like relationship stress, though. (You could also just add categorical traits to the model, though, and let relationship stress act as a factor on happiness directly.) ---- A tangent: it's pretty well known that sleep deprivation can actually improve symptoms in people who are depressed -- especially at extremes. I don't know if this has anything to do with a large scale analysis like this, though. |
|
|
OP: start with a multiple regression, then do a PCA, to see which factors may have commonalities.
If you don't know how, upload your data somewhere. Me or anyone else here could do that in 5 min. The conclusions would certainly be more interesting than the wall of text - even if it covers just 1 person, you have many datapoints!