simple random sample
label the population 1-n, use random number table or calculator to generate random numbers (ignoring repeats), select those individuals
stratified random sample
grouping individuals into strata based on a homogenous characteristic, take an SRS from each strata
cluster sampling
make heterogeneous groups, choose a few of these groups, every individual in selected group included
systematic sampling
choose random starting point, choose every nth individual from there to sample
bias
systematic tendency to favor certain responses over others; leads to inaccurate data
undercoverage bias
when part of the population has a reduced chance of being included in a sample
nonresponse bias
when individuals chosen for a sample do not respond- leads to bias if these individuals differ from respondents
identify population and sample
explain how selected individuals might differ from general pop.
explain how this leads to an over/underestimate
when writing about bias
voluntary response bias
occurs when an invitation is sent to all individuals in a pop. to participate; those who choose to participate may differ from individuals who don’t
question wording bias
when survey questions are confusing or leading
self-reported response bias
when individuals inaccurately report their own traits
control
comparison
randomization
replication
4 principles of experimental design
control
keep other variables constant
comparison
using 2 or more treatments, such as a placebo
replication
using large number of experimental units
blind experiment
subjects are unaware of which treatment they are recieving
double-blind experiment
subjects and those administering treatment don’t know which treatment subjects are receiving
observational study
observing and recording, no variables changed
experiment
imposing a treatment, explanatory and response variables, control group
randomized experimental design
experimental units assigned to the treatments completely at random
blocking experimental design
individuals put into homogenous “blocks,” random assignment of experimental units is carried out separately within each block
matched pairs experimental design (similar groups)
two very similar experimental units are paired and the two treatments are randomly assigned within each pair
matched pairs experimental design (own control)
each experimental unit receives both treatments in a random order
flip a coin, random number generator, roll dice, names in hat, etc.
how to randomly assign
cause and effect
random assignment can conclude…
inferring results to population
random selection can conclude…
convenience sample
selects individuals from the pop. who are easy to reach, leads to biased and inaccurate data
confounding
2 variables are associated in such a way that their effects on a response variable cannot be distinguished from each other
treatment
specific condition applied to individuals in an experiment (combo of levels)
experimental unit
the object to which a treatment is randomly assigned
factor
explanatory variable that is manipulated and may cause a change in the response variable
level
different values of a factor
sampling variability
data from a sample unlikely to be the exact same as data from entire population
(number of individuals in population) / (number of individuals in sample)
choosing a starting point for systematic random sample