Outliers can cause your model to be biased because they affect the values of the estimated regression coefficients. The change in this one point has a dramatic effect on the regression model chosen to fit the data. With the outlier present, the regression model changes – its gradient is reduced the line becomes flatter and the intercept increases – the new line will cross the Y axis at a higher point. It is important to try to detect outliers to see whether the model is basied in this way. These are detected by looking for larger differences.