Single Sample Tests
When you have collected a single sample, it is often divided up in some way. For example, you may have conducted a questionnaire via social media and just want to check if there is a gender bias in the respondents, or that ethnic minorities are not underrepresented. You may know, for example, that your target population is split evenly between males and females but of course even if you are able to take a perfectly random sample, it is unlikely that your sample will be split exactly 50:50 just due to the ‘luck of the draw’. If you were to take many samples, then if they were truly random about half would have more males and about half would have more females. However, conducting a questionnaire via social media certainly does not produce a random sample, because not everyone is equally likely to respond to a request to complete a student’s questionnaire. It would make sense to do a few checks to investigate whether your sample is a bit biased in some way before you go on to analyse the main results. Even if your questionnaire comprises mainly open-ended questions and you intend to use mainly qualitative methods it would make sense to complete this quantitative check.