There are many possible reasons. Here are some of the more common ones:
The underlying relationship is not be linear.
The regression has very poor predictive power (coefficient of regression close to zero).
The errors are not independent, identical, normally distributed.
Outliers distorting regression.
Calculation error.
Chat with our AI personalities
Linear regression can be used in statistics in order to create a model out a dependable scalar value and an explanatory variable. Linear regression has applications in finance, economics and environmental science.
I want to develop a regression model for predicting YardsAllowed as a function of Takeaways, and I need to explain the statistical signifance of the model.
The value depends on the slope of the line.
One of the main reasons for doing so is to check that the assumptions of the errors being independent and identically distributed is true. If that is not the case then the simple linear regression is not an appropriate model.
When you use linear regression to model the data, there will typically be some amount of error between the predicted value as calculated from your model, and each data point. These differences are called "residuals". If those residuals appear to be essentially random noise (i.e. they resemble a normal (a.k.a. "Gaussian") distribution), then that offers support that your linear model is a good one for the data. However, if your errors are not normally distributed, then they are likely correlated in some way which indicates that your model is not adequately taking into consideration some factor in your data. It could mean that your data is non-linear and that linear regression is not the appropriate modeling technique.