FC Connect 2020

The Power of Relative Importance Analysis, As Demonstrated on a COVID-19 Dataset

Relative Importance Analysis is a modern extension of regression analysis. It’s ideal for survey data, which often includes highly correlated driver variables that can make traditional approaches to regression invalid.

We’ll use it to explore a dataset collected by Qualtrics pertaining to perceptions of COVID-19, including the following:
– How has reduced social contact affected people’s lives?
– How many people would intentionally infect themselves with the virus if that gave them greater access to certain jobs?
– How sick do respondents believe they would be if they contracted COVID-19?
– Most importantly for this presentation, what are the demographic and attitudinal drivers of responses to the above?

Fill out this form to watch the session: