Fichas
STATS FINAL TERMS
Fichas del conjunto (58)
- 1Parameternumber which summarizes something about the population (often an average or proportion)
- 2Statisticnumber which summarizes something about the sample
- 3Biased Samplemore likely to produce some outcomes than others
- 4Conveniencesamples that are too easy to take
- 5Volunteer ResponseSelf-selected sample of people who responded to a general appeal (table tents, online votes, etc.)
- 6Simple Random Sampletaken in a way that every set of n items has an equal chance of being chosen
- 7Stratified Random Samplespopulation divided into groups (strata) and random samples are taken within those groups
- 8Cluster samplespopulation divided into naturally occuring groups called clusters
- 9Systematic Samplesselect every kth item from sampling frame
- 10Multi-Stage Samplescombination of sampling techniques
- 11Bad Sampling Framecommon type of sampling error
- 12Under-coveragesampling frame doesn't include all population
- 13Data Entry Errorsrecords answers wrong
- 14Non-responsesome part of population doesn't response
- 15Response Ratepercent of contacted individuals who gave a response
- 16Response BiasGiven different from the truth
- 17Skewed RightLong tail to the right
- 18Skewed LeftLong tail to the left
- 19Symmetrictails approximately equal in both directions
- 20Mean of skewed rightgreater than the median
- 21Mean of skewed leftless than the median
- 22Mean of symmetricabout the same as the median
- 23Outlierseffect the mean by making it unrepresentative, but usually does not change the median
- 241st Quartilehas 25% of data below
- 253rd Quartile25% of data above
- 26Variancedistance between all individuals and the mean
- 27Margin of Errornumeric indication of distance a statistic may be from true parameter
- 28Z-Scoreconfidence coefficient
- 29Lurking Variablesmay influence the response but often aren't studied explicitly
- 30Experimentsimpose a difference in explanatory variables and try to determine difference in the outcome variable
- 31Subjectsindividuals being measure
- 32Response VariableVariable that measures the outcome of interest in a study (y or dependent variable)
- 33Explanatory VariableVariable we think explains changes in the response variable (x or independent variable)
- 34Treatmentspecific regimen or procedure assigned to subjects
- 35Random Assignmentsubjects are assigned to different levels of explanatory variable by random mechanism
- 36ControlAbsence of treatment; some baseline or reference condition that is used for comparison
- 37Placebofalse drug or treatment that the subjects believe is real
- 38Hawthorne Effectpeople act different when they know they are in an experiment
- 39Blindingsomeone is not aware of what group assigned to
- 40Double blindingneither know which group a subject is
- 41Non-response/dropoutsome subjects do not complete the study
- 42Non-adherencesome subjects don't follow the study protocol
- 43Generalization/realismcan we say the results will hold true for whole population
- 44Completely Randomized Designeach subject receives one treatment without taking other variables into consideration.
- 45Matched pairs designEach unit receives two treatments. The units could be a single subject or 2 subjects matched together
- 46Statistical SignificanceNot likely to have been just by random chance
- 47Hypothesis Testallow us to establish statistical significance
- 48Null Hypothesisbeginning claim
- 49Alternative HypothesisAnother theory
- 50Test statisticnumeric measure of distance from sample value to eat is expected under null hypothesis
- 51Null DistributionDistribution of test statistic assuming null hypothesis is true
- 52p-valueproportion of null distribution that is more extreme than the observed value
- 53Least Squares LineLine that relates x and y that minimizes the sum of squared errors.
- 54slopethe rate of change in y as x changes
- 55interceptjust a place holder
- 56correlationnumerical measure of association (r and ranges from -1 to 1)
- 57Square of correlation(r^2 from 0 to 100%, closer to 100 smaller variability along the line. Is the percentage of variability in y accounted for by straight line relationship with x.
- 58assumptions required for a test of sloperandomization in collecting data, straight line is the correct model for the data, spread of points around the line have same SD for all x, points are normally distributed around the line
Información
- Descripción
- Fichas sobre STATS FINAL TERMS, creado por mkbeards el 07/12/2015.
- Creado por
- mkbeards
- Visitas
- 1
- Valoración