Adjusted r square
Before starting this please check below r square blog.
Adding a variable to your model increases the r square value. That new variable may or may not be a useful variable to our model. Since r square searches for best fit ,r square doesn’t bother about new variable, it just keeps on increasing whenever you add a new variable.
So question is how can we solve this issue?
Answer: Using adjusted r square
When a useless variable is added
- “k” increases
- In turn overall “denominator” gets reduced.
- We saw Adding a variable to your model increases the r square value, so r square value in equation increases.
- Which in turn reduces the adjusted r square value.
Logically just by thinking on proportionality(directly or inversely proportional) we can understand how adjusted r square reduces on adding a new useless variable.
So Adjusted R-squared is used to determine how reliable the correlation is and how much it is determined by the addition of independent variables.