ap-statistics-样本分类公开课获奖课件.pptx

APStatisticsLuHao

2.3Anexperimentisaplannedinterventionundertakentoobservetheeffectsofoneormoreexplanatoryvariables.

Explanatoryvariable(factor)ResponsevariableExperimentalcondition/treatment:(combinationofvaluesforexplanatoryvariable)Designofaexperiment:overallplan

ExtraneousfactorDirectlycontrol(holdingsomeextraneousfactorsconstant)Blocking(creategroups/blocksthataresimilar,tryalltreatmentineachblock)Randomization(randomassignment)Replication(makingmultipleobservationforeachexperimentalcondition)

DefinitionofBlockingUsingextraneousfactorstocreategroupsthataresimilar.Allexperimentalconditonsarethentriedineachblock.

ExampleforBlockingAnexperimentisdesignedtotestanewdrugonpatients.Therearetwolevelsofthetreatment,?drug,and?placebo,administeredto?male?and?female?patientsina?doubleblind?trial.Thesexofthepatientisa?blocking?factoraccountingfortreatment?variability?between?males?and?females.Thisreducessourcesofvariabilityandthusleadstogreaterprecision.

Theidealsituationwouldbetohavebothrandomselectionofsubjectsandrandomassignmentofsubjectstoexperimentalconditons

HomeworkP40,19a,b,c,dP49,32P6762Readsection2.5and2.6

3.1 ComparativeBarChartForcategoricaldataUserelativefrequency

PiechartCategoricaldataSmallnumberofpossiblecategoriesForillustratingproportionssegmentedbarchart(P81)

3.2Stem-and-leafdisplaysForunivariatenumericaldataDoesnotworkwellforverylargedatasetsDef:Outlieranunusuallysmallorlargedatavalue

Stem-and-leafdisplaysWhattolookfor:gaps,symmetry,peaks,outliersrepresentativeortypicalvalue

RepeatedstemstostretchadisplayComparativestem-and-leafdisplay(92)

3.3Frequencydistributionandhistogram

fornumericaldataDiscretevariableContinuousvariable

Histogram

fordiscretenumericaldataWorkswellevenforlargedatasetExamp