top of page
Data Preparation
As the datasets contain an enormous number of questions,
we have selected only relevant columns and kept only those columns that are of use
for our analysis. We kept 55 columns for the year 2014, and 78 columns for the
year 2019 to continue with.

Data Conversion

As the responses for these questions were categorical format, we transformed them into numeric format in the following manner:
​


However, in both years, we had two questions in the well-being section with a negative tone. Responses to these 2 questions have been altered in a way to balance the tone in the entire data frame as if a participant chose 1, that implies satisfaction in all columns.

bottom of page