chapter 4 review - stat

simple random sample

subjects drawn from pop without any respect to grouping, equal subset equal chances

systematic random sample

randomly selective starting point then pick in fixed intervals

stratified random sample

divide population into groups and select certain number from each group

cluster smapling

divide pop into groups randomly select gorups then sample all fo the selected group

experiments

attempt to manipulate/influence subject/situations for data random to control and treatment

obserational study

measure characterisitcs without manipulation, try to generalize characteristics of sample to population

non random sampling bias

mainfest convience sample and voluntary response samples

convience sample

most easily avaliable subjects chosen

voluntary response

subjects elect themselves

undercoverage bias

some parts of population are systematically excluded from sample

nonresponse bias

individuals refusing to reply

response bias

measured response diff from true, confusion on subjects part, sensitive or leading questions, introduced by surveryer

experimental unit

subject, person/object which treatment is applied

explanatory variable

independent, influences/explains changes in response variable

response variable

dependent, variable of interest what we measure after treatment

binding

who knows what treatments are given to subjects

single bind

only researcher knows

double blind

neither researcher or patients know

unforeseen variables

confounding variable or lurking variables

random assignment

treatments randomly assigned even out effect of unforeseen so changes can be attributed to explanatory variable

random sampling

select representative sample from population so results can be generalized

completely randomized design

experimental units randomly assigned without concern for anything

matched pairs design

experimental units are related or matched before test