56 terms

Chapter 4 Test (AP STATS)

STUDY
PLAY
population
the entire group of individuals that we want information about
census
an attempt to gather information about every individual member of the population
sample
part of the population that we actually examine in order to gather information
sampling frame
list of individuals from which the sample is drawn (attempts to be a list of the entire population)
design
refers tot he method used to choose the sample from the population
sampling without replacement
once an individual is selected for inclusion, it may NOT be selected again in the sampling process
explanatory variable
the variable which separates the cases to be considered; same as the independent variable (there doesn't always have to be one)
response variable
the variable/attribute that is being measured; same as the dependent variable
surveys (observational studies)
used to estimate parameters of a fixed, well defined population
observational studies
in this case, subjects are observed and measurements are made of variables of interest, but there is no attempt to influence the respond; goal is to draw conclusions, but no causation can be determined
experiments
used to establish cause and effect relationships by comparing treatments that are actively imposed on the subject
experimental units
individuals on which the experiment is done
lurking variable
a variable that is not of interest in the current study, but is thought to affect the response variable
confounding variable
related both to group membership and to the response variable
simple random sample
each possible grouping of individuals is equally likely to be selected using a type of randomization *best method*
probability sample
sample is chosen by chance (such as flipping a coin)
stratified random sample
divide the population into strata of similar individuals and then choose an SRS in each strata and combine these SRSs to form the full sample; purpose is to insure that the sample is more representative of a population; accepted method
systematic sample with random start
samples a certain percentage of the population by taking every nth individual starting with one randomly selected from the first n individuals; as long as it is random, it is a reasonable method
cluster sample
randomly select a group of individuals that are heterogenous and take a census of the selected groups; reasonable method with randomization
multi stage sample
select successively smaller groups within the population in stages resulting in a sample consisting of cluster of individuals; as long as it is random, it is a reasonable method
voluntary response sample
consists of people who chose themselves by responding to general appeal; has high bias as people with strong opinions are most likely to respond; not an acceptable method
convenience sample
consists of individuals easiest to reach; often have something in common and therefore bias; never an acceptable method
judgement sampling
form of convenience sampling where an expert selects a sample he/she considers representative; not an acceptable method
quota sampling
type of convenience sampling using clusters and data; not an acceptable method
sampling error
difference between a sample result and the true population result; such an error results from chance sample fluctuations
non-sampling error
occurs when the sample data are incorrectly collected, recorded, or analyzed; can result from selecting a nonrandom and biased sample, using a defective measuring instrument, using a biased survey question; obtaining a large number of refusals; copying the data incorrectly
within treatment variability
represents the differences in the response variable that is due to the chance and natural differences in the population
between treatment variability
difference in the response variable in the two factors in an experiment
bias
study or sampling design where it systematically favors certain outcomes
selection bias
method of selection systematically excludes some part of the population
size bias
tendency to choose more of a certain group in the population with higher probability than their proportion due to sampling method
volunteer/convenience bias
due to convenience sampling
judgement bias
due to having an "expert" pick the sampling
incomplete sampling frame
forgetting to include homeless people as part of a town's population
undercoverage/nonresponse bias
sample differs from the population because data are obtained from all individuals selected for inclusion due to individual refusal to participate or inability to contact selected individuals
measurement/response bias
sample differs due to behavior of the interviewer or the respondent or from misleading due to poorly worded questions (wording bias, incorrect response, measurement)
factor
an experiment needs to manipulate this
response
an experiment needs to measure this
levels
specific values that the experimenter chooses
treatment
combination of levels from all of the factors that an experimental unit receives
randomized comparative treatment
used to avoid confounding because the placebo effect and other influences affect both groups equally
completely randomized design
treatments get assigned completely at random without restriction
control
for the effects of lurking variables by comparing several random treatments in the same environment and sources of variation other than factors by making conditions as similar as possible for all groups
control group
placebo group or standard treatment comparison
blocking
refers to a deliberate grouping of subjects in an experiment based on a characteristic that you suspect will affect responses to treatments in a systematic way to reduce variability; need to randomly assign subjects to treatments within blocks
randomized block design
treatments are randomly assigned within blocks of similar units
matched pairs
pairs of similar units are randomly assigned different treatments to compare the treatments
repeated measures
subjects can be their own matched pairs; randomize order of treatments; subject serves as his/her own control
replication
using a large enough number of experimental units to reduce chance variation in a study
blinding
any individual associated with an experiment who is not aware of how subjects have been allocated to treatment groups
single blind
when every individual in either of these classes ins blinded
double blind
when everyone in both classes is blinded
confounding
two groups you want to compare differing in some important way other than the relevant response variable; makes it possible to determine whether the treatment or something else caused the response
placebo effect
occurs when an untreated subject incorrectly believes that he/she is receiving a real treatment and reports an improvement in systems
hawthorne effect
occurs when treated subjects somehow respond differently because they are part of an experiment
experimenter effect
occurs when the researcher/experimenter unintentionally influences subjects through such factors as facial expression, tone of voice, or attitude
YOU MIGHT ALSO LIKE...