PSYB01H3 Chapter Notes - Chapter 5: Nth Metal, Sampling Frame, Statistical Parameter
Conducting a census studying the entire population of interest which avoids the problem of sampling (where only a limited number of people represent the entire population) If we want to assess the cross-population generaliziability of our findings, we need to compare results obtained from samples of different populations. That"s 5 out of 15 people which is 33% that are happy. You choose a sample of 6 out of this 15. A representative sample of this would be if you took 2 happy people out of 6. That would also be 33% that is happy. But if you take 4 happy people out of the original population out of the 6 then it will show that 66% are happy which would become a unrepresentative sample. Of course, representation in a sample is never perfect, but it is important to provide information about how representative a given sample is. Random sampling error may or may not result in an unrepresentative sample.