If one needs to draw conclusions about the population from the sample results,
the error term 3i in equation (10.1) must satisfy the following properties:
1. the errors have a normal distribution
2. the errors in the model are all independent
3. the same amount of error is found at each level of X. The implication is that the
average difference between the regression line and the observed values is constant
across all values of the independent variables. Having equal variance at each level of
X is known as ‘homoscedasticity’, while having unequal variance is called
‘heteroskedasticity’.