unit 1 stats

convenience sample

subjects were easy to gather, not a valid sample method

voluntary sample

subjects chose to participate, not a valid sample method

simple random sample

SRS

everyone in the population has an equal chance at being included in the sample

stratified random sample

populations is categorized, subjects are randomly selected proportionally to sample size

cluster sample

subjects are geographically close

multi-stage sample

cluster sampling on multiple stages

systematic sample

every “kth” individual is selected

bias

systematically (not necessarily intentional) misrepresenting the populations

publication bias

researchers only post interesting studies

Hawthorne effect

people act differently when they know they are being monitored

nonresponse bias

subjects selected to be in the sample didn’t participate

response bias

responses are influenced by misworded questions

selection bias

a portion of the population is left out of the sample, often accidental

observational studies

whatever happens, happens, no treatments

retrospective

look into the past to compare similarities and differences

prospective

select individuals and observe them over time

experiments

researchers can administer treatments and observe what happens

control

all non tested variables are constant

randomization

the treatments are assigned to subjects in the experiment randomly to filter out unknown variables

replication

by repeating the experiment, trends will become more apparent

control group

a group that doesn’t receive treatment, determines effectiveness of the treatment, sometimes receives a placebo treatment

completely randomized design

treatments are randomly assigned to subjects, not ideal

blocking

categorize the population, run multiple completely randomized designs

matched pairs

matching 2 subjects that are similar as possible, one gets treatment, one is control

placebo

people that aren’t receiving treatment but think they are may react like they received treatment

blinding

participants don’t know who is receiving treatment

double blind

both subjects and evaluators are blind

single blind

only evaluators are blind or only subjects are blind