英文质量管理质量管理.pptx

《英文质量管理质量管理.pptx》由会员分享，可在线阅读，更多相关《英文质量管理质量管理.pptx（39页珍藏版）》请在淘文阁 - 分享文档赚钱的网站上搜索。

1、DefinitionofStatisticsStatisticsisthestudyofvariation,interactionsand,operationaldefinitions.Statisticsisthestudyofdatatoprovideabasisforactiononapopulationorprocess.第1页/共39页TypesofStatisticalStudiesEnumerativestudiesarestatisticalinvestigationsthatleadtoactiononastaticpopulation(thatis,agroupofitem

2、s,people,etc.thatexistinagiventimeperiodand/oratagivenlocation).Analyticstudiesarestatisticalinvestigationsthatleadtoactiononadynamicprocess.Sincethisisatextonqualityimprovement,theprimaryfocuswillbeonanalyticstudies.第2页/共39页EnumerativeStudiesBasicConceptsApopulation(oruniverse)isthetotalityofunits,

3、items,orpeopleofinterestthatexistinagiventimeperiodand/orgivenlocationAdministration.Aframeisalistoftheitemsinapopulation.Thegapisthedifferencebetweentheframeandthepopulation.Asampleisaportionoftheframeunderinvestigation,andisselectedsothatinformationcanbedrawnfromitabouttheframe.第3页/共39页BasicConcep

4、tsNonrandomsamplesareselectedonthebasisofconvenience(conveniencesample),theopinionofanexpert(judgmentsample),oraquotatoensureproportionalrepresentationofcertainclassesofitems,units,orpeopleinthesample(quotasample).Allnonrandomsampleshavethesameshortcomingtheyresubjecttoanunknowndegreeofbiasintheirre

5、sults.Thisbiasiscausedbytheabsenceofaframe.Nonrandomsamplesshouldbeusedonlywhenbetterinformationistoocostlytoobtain.第4页/共39页Randomsamplesareselectedsothateveryelementintheframehasaknownprobabilityofselection.Typesofrandomsamplesincludesimple,stratified,andcluster.Allrandomsamplesallowgeneralizedstat

6、ementstobemadeabouttheframefromthesample.Arandomsampleisselectedbyoperationallydefiningaprocedurethatutilizesrandomnumbersintheselectionofthesampleditemsfromtheframetoeliminatebiasandholduncertaintywithinknownlimits.第5页/共39页Sevenstepsareinvolvedinselectingasimplerandomsample:Step1.Countthenumberofel

7、ementsintheframe,N.Step2.Numbertheelementsintheframefrom1throughN.IfNis25,thentheelementsintheframeshouldbenumberedfrom01through25.Allelementsmustreceiveanidentificationnumberwiththesamenumberofdigits.第6页/共39页Step3.Selectapageinatableofrandomnumbers.Forexample,selectingapageandstartingpointyields.19

8、3040092878311398676069611339043462052856735990548709422934276212493869294093第7页/共39页Step4.Ontheselectedpageofrandomnumbers,selectacolumnofnumbers,selectastartingpointinthatcolumn,anduseasmanydigitsastherearedigitsinN(twodigitsinthecaseofN=25).Forexample,beginningwiththetopofourcolumn,selecting,andus

9、ingthefirsttwodigitsofeachnumberinthatcolumn,thefirstrandomnumberis19.Sample NumberIdentification NumberItems119S209I304D405E513M612L第8页/共39页Step5.Determinethenecessarysamplesize.Step6.Fromthechosencolumnontheselectedpage,selectthefirstsixtwo-digitnumbersbetween01and25,inclusive.Ifanumberisencounter

10、edthatissmallerthan01(e.g.,00)orlargerthan25(e.g.,31),ignorethenumberandcontinuedownthecolumn.Ifanacceptablenumberappearsmorethanonce,ignoreeveryrepetitionandcontinuemovingdownthecolumnuntilsixuniquenumbersbetween01and25havebeenselected.Ifthebottomofthepageisreachedbeforesixuniquerandomnumbersareobt

11、ained,gotothetopofthepageandmovedownthenexttwo-digitcolumn.Step7.Finally,analyzetheinformationasabasisforaction.第9页/共39页Twoimportantpointstorememberare:1.differentmethodsofmeasurementwillalsoyielddifferentresults.2.differentsamplesofsizesixwillyielddifferentresults,andRandomsamples,however,donthaveb

12、ias,andthesamplingerrorcanbeheldtoknownlimitsbyincreasingthesamplesize.第10页/共39页ConductinganEnumerativeStudyThefollowing13stepspresentaguideforconductinganenumerativestudy.Thestepsarethesamewhetherthestudyisbasedonacompletecountofthepopulationorasample.第11页/共39页Step1.Specifythereason(s)youwanttocond

13、uctthestudy(forexample,toestimatetheaveragenumberofsickdaysperemployeeintheXYZCompanyin2002).Ifthisaverageisgreaterthan8.0days,thenanewhealthcareplanwillbeinstituted.Ifitislessthanorequalto8.0days,thecurrentplanwillbemaintained.第12页/共39页Step2.Specifythepopulationtobestudied.Inourexample,thepopulatio

14、nwouldbeallfull-timeemployeesintheXYZCompanyin2002.Anemployeeisconsideredfull-timein2001ifhehadfull-timestatusdesignationatanytimeduringtheyear.第13页/共39页Step3.Constructtheframe(alistofallfull-timeemployees).Everyonewhowillusethestudysresultsasabasisforactionmustagreethattheframerepresentsthepopulati

15、onuponwhichtheywanttotakeaction.第14页/共39页Step4.Performsecondaryresearch(suchastheexaminationofpre-publisheddata)todeterminehowmuchinformationisalreadyavailableabouttheproblemunderinvestigation.Forexample,checktheHumanResourcesDepartmentsrecords.第15页/共39页Step5.Determinethetypeofstudytobeconducted(for

16、example,mailsurvey,personalinterviews,analysisofunits).Inthisexample,wewouldanalyzeemployeeabsenteecardsfor2002.第16页/共39页Step6.Makeitpossibleforrespondentstogiveclear,understandableinformation,and/orfortheresearchertoelicitclear,understandableinformation.Forexample,themethodforanalyzingabsenteecards

17、shouldbeclearandstraightforward.Considertheproblemofnon-response.Inthisexample,besurenoabsenteecardsaremissing,andmakesurethedatagatherersknowhowtointerprettheabsenteecards.第17页/共39页Step7.Establishthesamplingplantobeused,determinetheamountofallowableerrorintheresults,andcalculatethecostofthesampling

18、plan.Atthisstage,Steps1and2mayneedrevisionsduetocostconsiderations.Forexample,wemaydecidetodrawasimplerandomsampleofemployeeabsenteecardsusingrandomnumbers,atacostof$1percard,assuminganallowableerrorofonequarterofadayintheestimate.第18页/共39页Step7requiresarandomsample.Theresultofanonrandomsampleinanen

19、umerativestudyisworthnomorethanthereputationofthepersonwhosignsthereport.第19页/共39页Step8.Establishprocedurestodealwiththenon-responsesanddifferencesbetweeninterviews,testers,inspectors,andsoon.Forexample,beabletoassessdifferencesincollecteddataduetodifferencesinthedatagatherersabilities.第20页/共39页Step

20、9.Prepareunambiguousinstructionsforthedatagatherersthatcoverallphasesofdatacollection.第21页/共39页Step10.Establishplansfordatahandling,includingformatoftables,headings,andnumberofclasses.第22页/共39页Step11.Pretestthedata-gatheringinstrumentanddata-gatheringinstructions.第23页/共39页Step12.Conductthestudyandth

21、etabulations.Itiscriticalthatthestudybecarriedoutaccordingtoplans.Fromthegathereddata,calculatethesamplingerrorsofinterest.Thisisimportantsothatthestudysuserscanunderstandthedegreeofuncertaintypresentinthestudyresults.Inourexample,wewouldcalculatethestandarderrorfortheaveragenumberofdaysabsentperemp

22、loyee.第24页/共39页Step13.Interpretandpublishtheresultssothatdecisionmakerscantakeappropriateaction.Forexample,iftheaveragenumberofdaysabsentperemployeeisgreaterthaneight,thenestablishthenewhealthcareprogram.第25页/共39页AnalyticStudiesBasicConceptsAnalyticstudiesarestatisticalinvestigationsthatleadtoaction

23、sthatwillimproveorinnovateaprocessovertime.第26页/共39页Theconceptofapopulationdoesnotexistforananalyticstudybecausefutureprocessoutput,whichdoesnotyetexist,cannotbepartofthepopulation.Aframecannotexistwithoutapopulation.Lackofapopulationandframemakesitimpossibletodrawarandomsampletostudyaprocesssbehavi

24、or.第27页/共39页Modelscanbeusedtostudythesecause-and-effectsystems.Thesemodelsincludeflowchartsofaprocess,simulationsoftheprocess,andcause-and-effectdiagrams.Inananalyticstudy(unlikeanenumerativestudy),thereisnounderlyingstatisticaltheorythatenablesquantificationofuncertainty.第28页/共39页Expertopinionisinv

25、aluableinunderstandingthemagnitudeoftheuncertaintycausedbychangestoaprocess.Generally,thisuncertaintyisbestexplainedbyanexpertwhoisinvolvedwiththeprocessunderstudy.第29页/共39页ImprovementorinnovationofaprocessisaccomplishedusingthePlan-Do-Study-Act(PDSA)cycle.ThePDSAcycleisusedtonarrowthedifferencebetw

26、eenprocessperformanceandcustomer(eitherinternalorexternal)needsandwants.第30页/共39页TheaimofthePLANstageistodetermineanimprovementorinnovationtoaprocessthatwillnarrowthedifferencebetweenprocessperformanceandcustomerneedsandwants.(i.e.oldflowcharttonewflowchart)TheDOstageofthePDSAcyclerequiresthatplanne

27、dexperimentsbeconductedtodeterminehowbesttoimplementthePLANestablishedinthepreviousstage.Experimentsshouldbeconductedonasmallscale,usingalaboratory,anofficeorplantsite,or,withhisconsent,evenacustomerslocation.TheaimoftheSTUDYstageistodetermineifthePLANhasbeeneffectiveindecreasingthedifferencebetween

28、processperformanceandcustomerneedsandwants.IntheACTstage,thePLANshouldbeintegratedintotheprocess.第31页/共39页ErrorsinAnalyticStudiesAtypeoneerroroccurswhenactionistakenonaprocesswhenitshouldhavebeenleftalone.Atypetwoerroroccurswhenwefailtotakeactiononaprocesswhenactionisappropriate.第32页/共39页DesignofAna

29、lyticStudiesWecanincreaseourdegreeofbeliefinapredictionfromananalyticstudybyconsidering:BuildingofknowledgeinasequentialfashionTestingoverawiderangeofconditionsSelectionofunitsforthestudy第33页/共39页Building of Knowledge in a Sequential FashionExperimentsperformedintheiterationsofthePDSAcycle,incombina

30、tionwiththeoryabouttheprocessfromsubjectmatterexperts,mayincreasethedegreeofbeliefinpredictionsaboutthefuturebehavioroftheprocessunderstudy.Thisdegreeofbeliefisincreasedassequentialpredictionsabouttheprocesssfuturebehaviorcomeclosertotheactualperformanceoftheprocess.第34页/共39页Testing over a Wide Rang

31、e of ConditionsThedegreeofbeliefinthepredictivevalueoftheknowledgegainedfromananalyticstudyisincreasediftheanalyticstudyyieldsthesameresultsoverawiderangeofconditions.Onlyanexpertinthesubjectmatterunderstudycananswerquestionssuchashowwidearangeofconditionsisadequatetohaveadegreeofbeliefsufficienttom

32、akeaprediction,orhowclosetoactualconditionsmusttheexperimentalconditionsbetohaveadegreeofbeliefhighenoughtomakepredictions.第35页/共39页Selection of the Units for the StudySincethereisnoframeinananalyticstudy,therecanbenorandomsampleandquantifiabledegreeofuncertaintyinapredictionaboutthefuture.Judgments

33、amplesareusedtoconductanalyticstudies.第36页/共39页AnalysisofDatafromAnalyticStudiesA Stable ProcessAstableprocessprovidesaforumforasubjectmatterexperttoconductexperimentstogainknowledgetopredictthefuturebehavioroftheprocess.Graphical AnalysisThegeneralapproachforconductinganalyticstudiesreliesongraphic

34、altechniques,suchascontrolcharts,whichutilizebothstatisticalknowledgeandsubjectmatterknowledgetolearnabouttheprocesstopredictitsfuturebehavior.第37页/共39页DistinguishingBetweenEnumerativeandAnalyticStudiesIfa100percentsampleoftheframeanswersthequestionunderinvestigation,thestudyisenumerative;ifnot,thestudyisanalytic.第38页/共39页感谢您的观看！第39页/共39页