All projects
Data ScienceJan 2021
Survey Data Exploration & Hypothesis Testing
A statistics course project: cleaning a raw dataset of about 2,000 rows and 600 one-hot columns down to 1,754 rows and 117 columns, then visualising it and testing hypotheses about proportions.
Overview
The raw data arrived one-hot encoded across roughly 600 columns. I merged related columns back into categorical variables, giving 1,754 rows and 117 columns, then visualised comparisons between variables and ran statistical and hypothesis tests on several proportions. The full write-up is in the repo as a paper (PDF).