In this tutorial you will learn the basics of verifying segmentation, analyzing the data, and creating segments. When reviewing survey data, you will typically be handed likert questions (ex: answers with a scale of 1 to 5 with 1 being bad and 5 being good). Using a few techniques, you can verify the quality of the survey and start grouping respondents into populations. The steps we will be following are listed below:
- Analyzing our data set for scale.
- Using Principle Component Analysis (PCA) to verify that the survey is sound and grouping data.
- Checking for correlated questions.
- Setting up the Exploratory Factor Analysis (EFA) to create the final segments.
This tutorial is written as a Python Jupyter notebook. All instructions and code is self-contained within the notebook file itself. That file can be found in the top directory of the repository: Survey Segmentation.ipynb.
To run the notebook yourself, install Anaconda to get all of the required software. It will come with a Python environment, Jupyter Notebook, and an IDE (Spyder) to run your Python code. If you just want to view the notebook, click on the ipynb file and GitHub will display the contents within the browser for you.
When running the code on your own system, make sure to point the CSV file to the correct directory.
- Jason Wittenauer - All work - Jason's Repos
This project is licensed under the GNU General Public License v3.0 - see the license.md file for details.