Researcher bias - selecting people who you think will produce a certain result, complete selection control, avoid whoever they like therefore low internal validity
Unrepresentative of target population - drawn from particular area therefore difficult to generalise to target population