When the sample is random it enables proper inference about the population undertaken because is less likely to be biased
Outliers
Because it is not representative
How is the critical value affected by sample size?
The larger the sample size, the smaller the critical value and the smaller the significance level