Statistical Sciences 1024A/B Study Guide - Final Guide: Simple Random Sample, Cluster Sampling, Random Variable

• School: Western University
• Department: Statistical Sciences
• Course: Statistical Sciences 1024A/B
• Professor: Sohail Khan

Document Summary

Note: there will also be 11 self-tests (0. 5% each, worst test dropped) Decide group of individuals/units to which research conclusions should apply: refer to this group as the population. Parameters: what we want to know about the population; the population characteristics. Choose a (cid:858)k(cid:859) (cid:373)e(cid:373)(cid:271)e(cid:396) of ea(cid:272)h g(cid:396)oup (cid:894)i. e. the fi(cid:396)st(cid:895) a(cid:374)d i(cid:374)(cid:272)lude the(cid:373) i(cid:374) (cid:455)ou(cid:396) s(cid:455)ste(cid:373)ati(cid:272) sa(cid:373)ple. Other terms: bias, undercoverage - when some members of populatio(cid:374) a(cid:396)e(cid:374)(cid:859)t (cid:396)ep(cid:396)ese(cid:374)ted i(cid:374) a sa(cid:373)ple, nonresponse, response bias a(cid:449)a(cid:396)e(cid:374)ess of (cid:396)espo(cid:374)de(cid:374)t (cid:271)ei(cid:374)g a (cid:396)espo(cid:374)de(cid:374)t , voluntary response sample radical individuals are often more likely to respond. Observational studies no manipulation; just observe/measure and record. Experiments manipulate variables; observe/measure changes and record. Dependent variable is synonymous with response variable. Confounding variables are variables that supply an alternative explanation to the results. Independent variable is synonymous with explanatory variable/factor observed in the experiment. Completely randomized experiment design: most basic kind of experiment: everyone randomly assigned to random treatment group.