Key Driver Analysis explains overarching employee outcomes like engagement, satisfaction, or intent to leave. Here we apply multivariate regression to identify the largest contributors to overall employee satisfaction for the 2015 Federal Employee Viewpoint Survey. To help select the questions with the largest effects on overall satisfaction and to help reduce potential issues of multicollinearity, we employ LASSO regression.

The key point of this analysis is that the items with the lowest relative responses aren’t necessary the ones that management should focus on to improve overall satisfaction. If lower scoring items aren't associated with higher satisfaction, choosing to focus on them may yield little to no gains.

The results of the Key Driver Analysis are illustrated in the plot below, which is divided into four quadrants based on the average question score and average importance:

# Question
Average
Importance
Successes
Critical Improvements
Basics
Low Yields
Remaining Questions