SWP 538 Lecture 3: Sampling and Statistics
Document Summary
Population: the total number of people from where you draw your sample: denoted by n Elements: individual members of a population/sample i. e. individual human beings (mr. r, mrs. k, etc. ) Rationales for random selection: remove the risk of error. Limitations: less likely to be representative of population: sampling error: error that occurs because only part of the population is directly contacted. Marginalized groups not always accessible through probability sampling. Reactive/obtrusive: participants are aware that they are being studied. Nonreactive/unobtrusive: gathering evidence that participants leave of their social behaviour or actions naturally i. e. observance, content analysis, archives, etc. Rule of parsimony: eliminating all unnecessary questions/data and paring your survey to the absolute minimum: use simple language, avoid: unclear questions. Erosion: observation of selective wearing down of an object of setting. What is measured: frequency: whether or not something occurs, how often, direction: positive, negative, etc. Manifest coding: visible, surface content (e. g. how many times a word appears)