1. An increase in the observed sample size from 100 customers to 400 customers in a study of promotion response using simple regression produces predictions that are twice as accurate.
2. Doubling the sample size used to ft a regression can be expected to reduce the standard error of the slope by about 30%.
3. The simple regression model presumes, for example, that you have appropriately used logs or other transformation to obtain a linear relationship between the response and the explanatory variable.
4. Prediction intervals get wider as you extrapolate outside the range of the data.
5. The assumption of a normal distribution for the errors in a regression model is critical for the confidence interval for the slope.