KIN330 Lecture 9: Lecture 9 – Sampling.docx

2 major categories in sampling: human research, people, non human research, cell, animal, tissue culture. Why care about sampling: sampling bias: when individuals selected for a sample over represent or under represent certain population related to phenomenon of study, casually asking at random on street, asking people you know only, asking people you can access easily, careful sampling minimizes bias, and therefore maximizes generalizability of your conclusions. Why sample: sampling is more feasible than measurement of a whole population, or even a whole accessible population (obviously, bigger is better in some respects (parameter estimates ie. how prevalent a disease is), but worse in others (measurement quality, always some tension between the two. Overview: probability sampling (best quality sampling you can do, uses random selection of units form the population which means, each unit has equal chance of possessing exposures and characteristics (nutrition, sex, sampling should be free from bias, considered to be representative (many not actually representative, types: