WPC AJ5v|9f?͝hTl˦/O\6v.APzW ,F7kˢl+ K.kgz(Cb}-1Jm%p5Tԡ%i>95@>q q ✀ڰ- :qPlόIfӰp}FD'),I fi?ybdn|T~u[pU 8sblٜk~ҝf6q>nr)kQ:CU)w[QBoFCBX{u2!3%eG1UT qNR^ܮ:擃U")k~OŕP3h`RpXvihV lDU=6Gկ5N3; 領K}*^jj3^'ZrI]oݥz4fuw#UN % 0: 1u! 72 0-c 0yw+^ w4 m\\BIGMOO\CEZANNET(,,,,T(0&d9 Z6Times New Roman RegularX($USUS.,3#37=CIQYag1.a.i.(1)(a)(i)1)a)i)(;3$2#  0  .3  0  (O;$0  2#  a  .3  0` (#(#%, ,:i+003|xU !USUS.,  _Sociology_357_ԀSummer2002_Study_ԀGuidePartII  A.SamplingTheory(Willbeatleast20%ofexam)   (#33"3"  32#  1  .3  0    Definitionofaprobabilitysample:everyelementofthepopulationhasa knownnonzero  ` probabilityofselection.3݌P (#(# Ќ  ";"  ;0  2#  a  .3  0` (#(#  Recognizeobviousexamplesofprobabilityandnonprobabilitysamples; ݌< ` (#` (# Ќ  ";"  ;0  2#  b  .3  0` (#(#  NOTE:Probabilitysamplesarealsoknownasrandomsamples;-݌( x` (#` (# Ќ  "3"  32#  2  .3  0    Understanddifferencebetweenpopulationoftheoreticalinterestandtheactualpopulation  d towhichgeneralizationsmaybemade(thepopulationlistedbythesamplingframe)3D݌ P (#(# Ќ  "3"  32#  3  .3  0    Samplingframe3݌ < (#(# Ќ  ";"  ;x0  2#  a  .3  0` (#(#  Thelistorruledefiningthepopulation;x݌( ` (#` (# Ќ  ";"  ;}0  2#  b  .3  0` (#(#  Recognizeexamples;}݌ ` (#` (# Ќ  ";"  ;l 0  2#  c  .3  0` (#(#  Understandthecrucialroleinsampling:generalizationsfromprobabilitysamples   canbemadeonlytothepopulationlistedbythesamplingframe;l ݌ ` (#` (# Ќ  ";"  ; 0  2#  d  .3  0` (#(#  Understanddifferencebetweenthesampleandthesamplingframe;  ݌ ` (#` (# Ќ  ";"  ; 0  2#  e  .3  0` (#(#  Recognizethatthesamplingframeitselfisnotrandomlychosenandthatitmustbe t evaluatedforitsfittothepurposesofresearchandabilitytotapthepopulationof ` theoreticalinterest; 7 ݌L` (#` (# Ќ  "3"  3 2#  4  .3  0    Recognizeanddistinguishamongexamplesofthetypesofrandom(probability)samples:3 ݌8(#(# Ќ  ";"  ;0  2#  a  .3  0` (#(#  Simple;݌$t` (#` (# Ќ  ";"  ;0  2#  b  .3  0` (#(#  Systematic;݌`` (#` (# Ќ  ";"  ;0  2#  c  .3  0` (#(#  Stratified;݌L` (#` (# Ќ  ";"  ;0  2#  d  .3  0` (#(#  Cluster;݌8` (#` (# Ќ  ";"  ;0  2#  e  .3  0` (#(#  StratifiedCluster;݌$` (#` (# Ќ  "3"  3q2#  5  .3  0    Ideaofsamplingerrorastheerrorinanestimateyougetfromasample.Notthetechnical  definition,buttheheuristicideaofawiderornarrowerrangewithinwhichtheestimate  falls,andtheriskofasamplewhoseaverageismarkedlydifferentfromthepopulation3q݌(#(# Ќ  "3"  3p2#  6  .3  0    SystematicSample3p݌(#(# Ќ  ";"  ;J0  2#  a  .3  0` (#(#  Knowthatitiscomparabletosimplerandomifthelistisrandom;Jw݌p` (#` (# Ќ  ";"  ;h0  2#  b  .3  0` (#(#  Knowthatthereisimplicitstratificationifthelistisordered(firstonegroup,then \  another,etc.);h݌H!` (#` (# Ќ  ";"  ;0  2#  c  .3  0` (#(#  Knowthatthereisariskofhigherrorifthelisthasaperiodicitythatcoincides 4"  withthesamplinginterval;݌ #p!` (#` (# Ќ  ";"  ; 0  2#  d  .3  0` (#(#  Recognizeexamplesofthese; M݌ $\"` (#` (# Ќ  "3"  32#  7  .3  0    Clustervs.stratifiedvs.stratifiedcluster3E݌$H #(#(# Ќ  ";"  ;0  2#  a  .3  0` (#(#  Recognizeexamplesofeach;;݌%4!$` (#` (# Ќ  ";"  ;0  2#  b  .3  0` (#(#  Knowthatstratificationsamplessomefromeachgroup,whileclustersamples & "% someofthegroups.;2݌' #&` (#` (# Ќ  ";"  ;X0  2#  c  .3  0` (#(#  Knowthat,comparedtosimplerandomsample,clustersampleshavehighererrors (#' whilestratifiedsampleshavelowererrors.Havesomebasicunderstandingofwhy )$( thisistrue.;X݌*%)` (#` (# Ќ  ";"  ; 0  2#  d  .3  0` (#(#  Recognizethedifferencebetweenarandomclustersampleanda_nonprobability_ l+&* sample.; 9 ݌X,'+` (#` (# Ќ  ";"  ;r!0  2#  e  .3  0` (#(#  Knowthatstratificationisworthwhilewhenthegroupsaredifferentfromeach D-(, other,anditisnottoocostlytoperformthestratification.;r!!݌` (#` (# Ќ  ";"  ;"0  2#  f  .3  0` (#(#  Knowthatclustersamplingisworthwhilewhentheclustersarecomparableto  eachotherandtherearesubstantialcostsavingsfromclustering,sothatalarger  sampleispossible;"#݌t` (#` (# Ќ  ";"  ;$0  2#  g  .3  0` (#(#  Recognizeexamplesofreasonableandunreasonablestratificationorclustering.;$$݌`` (#` (# Ќ  ";"  ;%0  2#  h  .3  0` (#(#  Understandtheideaofcreatingstrataofclusters,andthensamplingclustersfrom L  eachstratumofclusters.;%&݌8 ` (#` (# Ќ  ";"  ;2'0  2#  i  .3  0` (#(#  Recognizeexamplesofstratifiedclustersampling,anddistinguishitfromother $ t typesofsamples.;2'_'݌ `` (#` (# Ќ  "; "  ;(0  2#  j  .3  0` (#(#  Complexmultistageprobabilitysamplesofteninvolveseveralnestedstepsof  L  stratifiedclustersampling.;((݌ 8 ` (#` (# Ќ  "3 "  3)2#  8  .3  0    Responseratesandnonresponsebias3) *݌$ (#(# Ќ  ";"  ;*0  2#  a  .3  0` (#(#  Needforcallbackstobringupresponserate;**݌ ` (#` (# Ќ  ";"  ;+0  2#  b  .3  0` (#(#  Sampleswithoutcallbacksmaybe,inpractice,indistinguishablefromcontrolled   quotasamples.;+,݌ ` (#` (# Ќ  "3"  3'-2#  9  .3  0    Conveniencevspurposivesampling3'-T-݌ (#(# Ќ  "; "  ;.0  2#  a  .3  0` (#(#  Neitherisprobabilitysample;.>.݌p` (#` (# Ќ  "; "  ; /0  2#  b  .3  0` (#(#  Convenienceispureconvenience,notrelatedtopurposesofresearch; /8/݌\` (#` (# Ќ  "3 "  3,02#  10  .3  0    Justificationsforpurposivesampling:3,0Y0݌H(#(# Ќ  "; "  ;10  2#  a  .3  0` (#(#  Smallsamples<30,toosmallforrandomsamplingtobeadequate:choosesubjects 4 appropriatelyforpurposesofresearch;1I1݌ p` (#` (# Ќ  "; "  ;20  2#  b  .3  0` (#(#  Hard-to-getpopulationsthatcannotbefoundthroughscreeninggeneral  \ population;22݌H` (#` (# Ќ  "3 "  332#  11  .3  0    Quotasampling333݌4(#(# Ќ  "; "  ;40  2#  a  .3  0` (#(#  Recognizeexampleswhendescribed;44݌ ` (#` (# Ќ  "; "  ;50  2#  b  .3  0` (#(#  Recognizethisisessentiallyconveniencesampling;55݌ ` (#` (# Ќ  "; "  ;60  2#  c  .3  0` (#(#  Controlledquotasamplesuseprobabilitysamplingdowntothelastlevel,then  quota.;66݌` (#` (# Ќ  "3 "  372#  12  .3  0    Factorsaffectingneededsamplesize:Pairedexamplesmayaskyouwhichonewouldneed  largersamplesize,oryoumaybeaskedtopicktheseoutofalistincludingother l (irrelevant)factors378݌X (#(# Ќ  "; "  ;90  2#  a  .3  0` (#(#  Recognizethataccuracyofsampledependsuponsamplesize,notratioofsample D! topopulation;99݌0" ` (#` (# Ќ  "; "  ;:0  2#  b  .3  0` (#(#  Heterogeneity:needlargersampletostudymorediversepopulation;:;݌#l!` (#` (# Ќ  "; "  ;<0  2#  c  .3  0` (#(#  Desiredprecision:needlargersampletogetsmallererror;<=<݌$X"` (#` (# Ќ  "; "  ;'=0  2#  d  .3  0` (#(#  Samplingdesign:smallerifstratified,largerifcluster;'=T=݌$D #` (#` (# Ќ  "; "  ;=>0  2#  e  .3  0` (#(#  Natureofanalysis:complex_multivariate_Ԁstatisticsneedlargersamples;=>j>݌%0!$` (#` (# Ќ  B.Other '#& "3 "  3?2#  13  .3  0    Understandtheideathatdifferentmeasuresofthesameconceptwillbecorrelatedwith (#' (havebivariateassociationwith)eachother.3??݌)$((#(# Ќ  "3 "  3A2#  14  .3  0    Possiblyhavesimpleexamplesofreliabilityanalysisanddiscusswhattheyimplyaboutthe |*%) measurementoftheconcept.3A0A݌h+&*(#(# Ќ  "3"  3XB2#  15  .3  0    Generalideaofstatisticalcontrolforextraneousvariablesinisolatingcauseeffect T,'+ relations