Regression analysis generates lots of numerical values in the outputs. It would appear to be quite overwhelming at first. So, we are going to break down by focusing on a few questions at a time.
The 4 questions
At this stage, we will first work on finding the evidence to answer the following 4 questions:
- Is the relationship between Y and X significant or random?
- How much is the impact of X on Y?
- What’s the regression model for prediction?
- How good is the prediction?
These 4 questions are be grouped based on two objectives of regression analysis: 1) to understand the relationship and 2) use the relationship to predict. While most people are so very interested in predicting, we need to keep in mind that understanding the relationship is the first step and a very important objective.