SOAN 3120 Chapter Notes - Chapter 18: Statistical Inference, Inference, Confidence Interval

• School: University of Guelph
• Department: Sociology and Anthropology
• Course: SOAN 3120
• Professor: Andrew Hathaway

Document Summary

Any confidence interval or significance test can be trusted only under specific conditions. If (cid:455)ou(cid:396) data do(cid:374)"t (cid:272)o(cid:373)e f(cid:396)o(cid:373) a (cid:396)andom sample or a randomized comparative experiment, your conclusions may be challenged: to answer the challenge, you must usually rely on subject-matter knowledge, not statistics. It is common to apply statistical inference to data that are not produced by random selection. There is no simple rule for deciding when you can act as if a sample is an srs. Pay attention to these cautions: practical problems such as nonresponse in samples or dropouts from an experiment can hinder inference even from a well-designed study. If a survey has a high-response rate, data is quite trustworthy: different methods are needed for different designs. The z p(cid:396)o(cid:272)edu(cid:396)es a(cid:396)e(cid:374)"t (cid:272)o(cid:396)(cid:396)e(cid:272)t fo(cid:396) (cid:396)a(cid:374)do(cid:373) sa(cid:373)pli(cid:374)g desig(cid:374)s (cid:373)o(cid:396)e (cid:272)o(cid:373)ple(cid:454) tha(cid:374) a(cid:374) r : there is no cure for fundamental flaws like voluntary response surveys or uncontrolled experiments. The z procedures are designed for normal populations.