Exploratory Data Analysis
Exploratory data analysis is a structured, systematic method of initially viewing a dataset aimed at understanding the many facets of the data, features, and relationships contained in a given dataset.
- point is to determine the main characteristic of the data set
- e.g.,
- determine how much of your data is qualitative (categorical) versus quantitative (countable/measurable)
- provides you with up-front information about the data
- easiest type of analysis to visualize
- benefits
- Informs what type of statistical testing may be appropriate based on the data collected
- Helps you to investigate potential relationships between data early on
- Allows you to identify correlation between data points
- Informs whether your research questions can be potentially answered with the data
- all data should go through exploratory analysis before it is analyzed any further
- popularized by John Tukey in 1970s
- to help determine better methods of data collection and discover potential research questions
- type of exploratory analysis performed will vary depending on your needs