SOCA01H3 Lecture Notes - Symbolic Culture, Global Warming, Abstract Window Toolkit
Document Summary
Reliability: do the research again & if you obtain the same result then its reliable. Validity: does your research variable measure what it claims to measure. Sample: (part of group) selected from sampling frame (list of people in population of interest) but its too costly & time consuming to get in touch with entire population. Hopefully it will represent the entire population/ universe so, we can generalize. Best way is to do random sampling. Bias: oversample certain ethnic groups e. g. people whose surname begins with s. Select a specific group & take a sample. If you are writing about something but doing something else then it lacks validity. Undercounting: occurs in some categories of population because of an imperfect sampling frame (e. g. under representativeness of sample when internet surveys are used). Non- response: the refusal to participate to participate. Wording effects: the way questions are phrased or ordered can influence & invalidate response.