sampling
studying part of a group to gain information about the entire group
voluntary response sample
a large group is invited to respond, those who respond are in the sample
experiment
a treatment is imposed on a subject and a response is observed
statistical inference
provides ways to answer questions, guaranteeing that the answers are accurate
population
the group that is sought to be learned about
sample
a subset of the population
convenience sampling
choosing a sample of people easy to reach
bias
a systemic failure of a sample to represent a population
simple random sample
every subject has an equal chance of being picked for the sample next
observational study
observes individuals and measures variables of interest, but doesn’t try to influence results
experimental units
individuals on which an experiment is performed
factors
explanatory variables
placebo effect
response to any treatment with a dummy treatment because an outcome was expected
control
treatments must be compared
randomization
randomly assigning subjects to treatment groups
completely randomized experimental design
all experimental units are randomly allocated among each treatment
statistically significant
an observed effect too large to be attributed to chance
replication
repeating each treatment on a large number of subjects to show systemic effects of the treatment
3 principles of statistical design
control, randomization, replication
hidden bias
not all experimental units are treated the same
double blind
subjects and people with whom they interact shouldn’t know what treatment they got
blocking
dividing subjects into groups that share similar characteristics
matched pairs design
one person gets both treatments, or similar people are matched and each gets a different treatment
stratified random sample
population is seperated into groups, and then people are taken from each group
systematic random sample
every nth subject
multistage sample
cluster sample
groups of people are chosen at a time