Statistics Test 3 Johnson

Double Blind

When both the administrator and the recipient do not know which treatment is being given

SRS / simple random sample

every group of size n has an equal chance of being selected

random assignment

reduces bias

replication

reduces chance results

control

reduces confounding

homogeneous groupings

share a characteristic

SRS, stratified, systematic, cluster

4 types of sampling

completely randomized, block, paired

3 types of experimental design

undercoverage, nonresponse, voluntary response, response

4 types of bias

wording, social desirability

2 types of response bias

replication, randomness, control/comparison

3 characteristics of a "good" experiment