Sampling People, Networks and Records

Good samples are the foundation of good data collection. There are many options for choosing the right samples. Although it is possible to randomly or conveniently select people, records, networks or other units from the population, one should question their quality and what these selection methods are for drawing accurate conclusions after analysis. Although samples can be selected more carefully based on the researcher's judgement, one must ask if that judgment is biased by personal factors. You can draw samples in statistically rigorous and precise ways. This includes using control and random selection methods. These will provide sound representations and cost control. These last types of samples will be covered in this course. Simple random sampling can be used to collect records or people, cluster sampling can be used to gather records or networks of people, stratification can be used to stratify simple random or cluster samples, systematic selection and stratified multistage sample. This course will conclude with a short overview on how to estimate and summarize uncertainty in random sampling.