# Regression or Correlation for this RQ?

Our little group at uni is investigating if there is a relationship between 3 measures of social well-being (social anxiety, social connectedness and self esteem) and usage time (on-screen time in hours) on several social media platforms. Would you recommend that we use correlation or regression in this case?

You need to provide us with more information on your variables, are they categorical or numerical? If usage time is numerical then you can build a linear regression model and the get the coefficients and associated p-values out of it.

If all variables are numerical you could calculate the correlation between them, but the linear model gives more information / is superior.

• Hi! Thanks for the feedback. All variables are numerical - that is usage time and social well-being measures (questionnaire scores). I guess this means we can build a regression model then? Which one should be the IV and which would be the DV? If we have usage time for multiple social media platforms (9 platforms) and 3 measures for social well-being, will this be multiple regression? – Cafka Sep 14 '18 at 9:25
• Linear model would be my choice. (FYI Linear models work with both categorical and numerical independent variables and a numerical dependent variable) – user2974951 Sep 14 '18 at 9:28
• usage time is you DV, the other 3 are your IV's, this is multiple regression. In R lm(usage~v1+v2+v3) – user2974951 Sep 14 '18 at 9:47

If you want to predict things that can't be precisely measured (such as social anxiety, social connectedness, and self esteem) I recommend using your numerical screen time data to perform principal component analysis.

In short, principal component analysis "uses an orthogonal transformation to convert a set of observations of possibly correlated variables into a set of values of linearly uncorrelated variables called principal components."

The principal components CAN be interpreted as measurements of something you don't have a measurement for, in this case self esteem, etc... This is a pretty big stretch so if you are doing this in a rigorous academic environment then you are going to need something more to back up your claims.

Read paragraph 2.5 in this paper to see what I mean: http://www.math.montana.edu/jobo/writing/documents/barbour.pdf