Studied by 1 person
response variable
values of this variable record the results of each trial with respect to what we were interested in
trial
the sequence of several components representing events that we are pretending will take place
population
the entire group of individuals or instances about whom we hope to learn
sample
a representative subset of a population, examined in hope of learning about the population
bias
when sampling methods either over or under emphasize some characteristic of the population
voluntary response
bias where we allow members of our population to respond at will
undercoverage
some portion of the intended population is either unrepresented or under-represented
nonresponse bias
bias created when people who are randomly selected refuse to particiapte in the sample
response bias
anything in the survey design that influences the responses from the sample
randomization
the random selection of subjects used to reduce bias in either surveys or experiments
matched design
when our data values are not independent of one another and are related somehow to one subject - it is a form of blocking
census
attempting to survey the entire population
population parameter
mathematical measures taken from the entire population like the population mean and standard deviation
sample statistics
mathematical measures taken from a sample of the population of interest like the sample mean and sample standard deviation
simple random sample
abbreviated SRS, this requires that every item in the population has an equal chance to be chosen and that every possible combination of items has an equal chance to exist. No grouping can be involved.
stratified random sample
population is divided into homogenous groups and then a random sample is drawn from each group
cluster sample
population is split into parts or clusters usually based on geography and then entire clusters are selected randomly and sampled
systematic sample
a sample is drawn from the population by following a systematic plan to give every object in the population an equal chance to be chosen but all possible groups can not be formed
multistage sampling
sampling schemes that combine several different sampling methods at different stages
observational study
a study based on data where no manipulation of factors has been employed, we simply observe what has or will happen
experiment
a study where there is a definite random assignment of subjects to treatments
factor
a variable whose levels are controlled by the experimenter to discover its effects on the subjects
response variable
the variable that will be measured in an experiment
level
the experimental values that are chosen for each factor
treatment
the process, intervention, etc that is applied to each subject in an experiment
principles of experimental design
control, randomize, replicate, block
control group
a group created the provide a basis for comparison usually either an old or placebo treatment
single blind
when the subjects in an experiment do not know if they are in the treatment or control group
double blind
when the subjects as well as those evaluating the result of an experiment do not know which subjects are in the treatment or control groups
placebo
a fake treatment often given to the control group to help prevent the placebo effect (subjects improving by the power of suggestion)
blocking
subjects or experimental units are grouped together homogeneously because of some charcteristic they have in common that is believed might effect the experiment
confounding
when the results of an experiment can be determined because a variable has not be accounted for and has interferred with your results
convenience sampling
a poor way to collect a sample - do what is convenient for the surveyor, data is usually very biased
prospective study
a study where we collect data as it happens
retrospective study
a study where we collect data from events that have happened in the past
Note 
Flashcard (22)