If using PMCC why do the variables have to be random
A random sample enables proper inference about the population to be undertaken
A random sample is unbiased .
For the hypothesis test (using PMCC) to valid it is necessary to assume that the sample is random
A is independent as values are not subject to random variation
Because this would be extrapolation and it is possible that the relationship is different for young children
Because the values of a are non random so it makes no sense to try predict them
Emphasis on the “valid”
“Underlying distribute may not be normally distributed “