WPC; (bmo75}՜n"W,e5@K|=`DABB%ڄPwIQc}&N׏{ #gŽl/&i6X̉9tA=)Y%ؑ7*RZY}SV)06@1F`K#tpap͸hIŔf L4YY?~TqJ8k>R)!-NĆž,nlhR ކ'=^aVY2E-^8_Fo~Z-uL' (E-fPɨfruH AKp vV"#+ͅ*#b"qwR=o޷jx2  2b 2 722N&3F(3:n3.4+42535U>I5+5.5>526+P62{6262627+C7+n72727272/8+a828+8+8297F9.}999.9::1:>G:.:2:2:\\CENSUS\PRED-TR2-2-HP8000-3.PRED.DECENNIAL.HQ.CENSUS0(9 Z6Times New Roman RegularX($Ld. jE. :i+003|xbU  WPCD$n$n"S3'!Standard - Wide ,,9du"!5$x,, C}< ??~?~~@???~8  p~<P x~@~p$`~PpÇ8? > ~08  o ? 0?~0p 80`~D @0`A~ @ ~` 0@<~ 0@@x ~ 0@ ~0AAp`߀~ !~   C ~  0 0~ A @! ~ `   @@ ~  ' 0 ~  b8 ~ 0  ~ !@ C ~ BA@  ~ ! @ ~ @@x@?0 ~   |8c ~ ~  A0  B ~ @0 ! 8 |? ~ B  @  ~  @0b >???` ??? <? ? ~ !@D> @ ~ B 8<   ?? 1q ~  @ <0 ????? ? 0 ?  ?x?    < <?  ~ B@ @! ~@` 0` ~ "  <D ~@0   > " ~BP ~  ,``~ @ @~B 6 x @`D  ~0 0@ ~0@@@d <?? ?  ??? 8  y  p0? ? ? x#? x?@p  x! x~" p D@??<~~|8'@1 B?ap ?x8? <?ߌ0~??p?~ @H~@`?0  x?? ~ F  @?? x p; x8 ?8  ?~ @A @ a pp~ ?HH?~8 ???   ?< ?? ?|?~ 0E`~@ @ ~D @B@ ~@~@ @ 3~@0@ ~  @ >0~ A @ @~ "  (@p~@@hA ~ B @~ @"@ ~ ~AA~ ~  0B@p ~ @!H ~ x@$@ H0~"0B @~@@ p8~@@@0 ~ D B @ ` A~ 0!@( ~ ~@0D@ ~   @~@@f~$d $ @~0! @@@~ I B??< ~@ ~  $p !@~@ @ @~$@ 0x@>~ |0bc0$`@~@ @~@ 0c0@Bp~8`@ ~@!`` `~@B0 p~0"8@~  `8O~ s@~@"0~ @p @~B` @@ ~ XH ~;8~ P`s  ~@`@`l!@!~"@ 0@@p~B  l ~ @@# `  ~1 sA~ @   "@~pH@$  ~ ` ~ \ (@ ~ @ @2@A`~@ @ m$ ~ 1  p@@~HB~ @p@'?~  IFgp@ ~@0` !@@~A  d ~ x ~   ~ $@ ~ Hp ~@@8f!!@~    7 ~@ ` @~@ @9<`~ "i` @ ~ C  @ ~ @@`~D> B `8 ~ @@B@0~` %Ȁ6`~ x   ~  C`  ~8  $}r p(`@2~ " @F0~  @pp @~0D( LP ` 4 ~@0 s "@@@~ "  !~@@ `$b @~@@@9 H ~  ~@P p@ ` f~( @ ~   @@ ~?c ~@ c< `~(@D@~ 0 ~H`  00~A`H @~     @ ~@Bp  "/  ~ @L@8 ~h"0(P  0@~  1 ~ !@ p?  0~(  0l ~  "  @~     ~    ~$`0  ~@ !@~  ~  p` @<~@ @ $ ~ i >@ ~   ~@ p$? ~`?8 8@~  |@ A~@  ?0  ~' ?@@ @~     @ ~@@ 0 p`@ ~     B@@~p ` "F  ~ 8  ~  `%0 Dp ~ ? @ ` 0 ~  @ ~@D `p @~<@ D    ~ D@~ 0  ~  2 ~@@@ 0 "@ ~@   2~ p@ ~0@p@~D  ~@$(~ <p`  ~0 p ~"@   ~ @@~BC1 ` ~;`(p@  ~  ~ ` `P$~8x`@~@0 ~@@0`@ @ ~8  @~0@@@~@   ~ @@0@ $~    @~&P A  o03~%@"@ ~I@ !~@p @D ~$  `~@ @~   @?  x? ?~ "@ @ ~D   ~ ` pp  ~@0@@H 8?~` @@ @A ~ X` 0 $~  `~  @H$@~ 0A @p ??~IA   ~A @  `  ? ~0$ "~ @@P@0 I~` 8Dp ? ` ?À? ~@0I@p B  Ç  ? ?  ?|?~ ! ` @@$@~@  $08 p0~ H@ @ Ap? ~ @  $`? ?!~ @@$ 0"D~ A@`0 @ ~  @< p8 8??? q<?|??~0 @~`  0 @D$@ x??Ã |p <? p~ @@  ?p?  8 88?  `?    ? ? ?Á  ~ "A0  `?8$"Small Circle"0  x"Icgkosw{Large BulletLarge Bullet2c0Quick 220 (3$ !  3/((3$ !      0   d  d9999?Heavy ShadowdZd(#$  0  ====7Hairlinedxd)Hairline d@K:\KATHIE'S\PRED2-LI.WMFWPC@@: : !: ƚ: @I  &: &&#TNPP0D & TNPP &&TNPP   :  - "-- !@ -- "-&?  & - &/?  & &z| .  --@    D &^T  &&#TNPP0D & TNPP &&TNPP  D  &`R &`R - "- M- "- - Palatino-  . 2 PRED.--~si-- hPalatino-  . 2 5oP\.Palatino-  .2 5 LANNING, Ndjj+jb .hPalatino-  . 2 5Rf.Palatino-  .2 5ESEARCH,NCNdV[j .-- "--  . 2 AND djc .hPalatino-  . 2 E].Palatino-  .2 t VALUATION \dNddN+ej .hPalatino-  . 2 Dv.Palatino-  .2 (IVISION+\+C+ej.&2.  --D Q ,|&WordMicrosoft Word ,|  -@Times New Roman4-  - "-  f-|- "- v8@Palatino-Y@Palatino--%.&2 %,U.S. Census Bureau/k//KkKkK/wkKJVl.-'-   --'&&< --$H}y+vJrknjgc_,[TN-HWECA?>=8<V<s<=>?A(CQE|HN+T[_cg&jJnmrvy}             --&&`/R -- $R`|@fxkZ_T KbC=8n410j013c6:?gFNWeam {P t      5 P    W  @xqjOe#b`^\k[:YXXWjW4WWXXgZ6[^}c&hov6~_`--&&--#-- pTimes New Roman,-   .%2 Beyond the HorizonPX@@HHH$(H@$hH8(8HHP.&--Times New Roman0-  &TNPP & --'&--"Systemn-&TNPP &?@2K:\KATHIE'S\PRED2-LI.WMFK:\Kathie's Stuff\PRED 2-Line Logo with Tagline.wmfΤ zA X]&JЎ` q8Ȃ(hH  Z6Times New Roman Regulari)rAZ"Arial RegularA-  ۉ )  _  May2002DX XXXD%g % XDX     #%C % %D%g?#O%C % %%CP8XXdd8 @Z  @Z  @Z  @Z  @Z  @Z  @Z  @Z  @Z  @Z  @Z  @Z  @Z  @Z  @Z  @Z  @Z  @Z  @Z  @Z  @Z  @Z  @Z  @Z  @Z  @Z  @Z  @Z  @Z  @Z  @Z  @Z  @Z  @Z @@'#OXX %O%C^#O\XOXPartB:#OXXO\#  8 @@(OEXOXThe 6  @@9 9 %Census2000 4  @@TestingandExperimentation 2  Program#OXXOEs#OiXOX 0  #iOi#G&i ((x3   @  @  @  @  @ 0 @ 0 @ 0 @ #GX XG&I#G.c XGX n$ Q#GX XG.c#G& XGX +|&     #iG&#Oii#OXXOi6#O NXOX # NO N##XX N#Thispageintentionallyleftblank. X%   )# OXXXXO NXOX   TheCensus2000TestingandExperimentationProgram#OXXO N#  t -,X,XX- OlXOX  Introduction  &      4  `      #OXXOl#Asuccessfuldecennialcensus,onethatisresponsivetothenationschangingneeds,cannotbe  L achievedwithoutearlyplanning.ManykeyissuesforCensus2000,suchasdecliningpublic  8 cooperationandtighterfundingrestrictions,werealreadybeingstudiedinthelate1980s.  $ Fundamentaloperationalchanges,suchasthosedesignedtoimprovetheprocessforcapturing   informationonthecensusquestionnaires,camefromthisearlyresearch.Foradecennialcensus, p  muchleadtimeisneededtofirstidentifyandtestpromisingnewprocedures,makeadjustments, \  andretestasneeded.Substantialleadtimeisparticularlynecessaryfortheprocurementand H  testingofmanydifferenttypesofequipmentthatmustbeinplacetoconductthedecennial 4  census.    Earlyin1997,theCensusBureauformedateamtodeveloptheCensus2000programoftesting   andexperimentation.ThetestsandexperimentswereconductedconcurrentlywithCensus2000 p becausethedecennialcensusenvironmentprovidedthebestpossibleconditionstolearnabout \ thevalueofnewordifferentmethodologies. Researchconductedduringthedecennialcensusis H expectedtoguidefuturedecennialcensusdesigns,butalsoprovidevaluableinformationforuse 4 byotherareasoftheCensusBureau.   OlXOX#OXXOl# OlXOXPlannedTestsandExperiments#OXXOl#    SummarydescriptionsofthetestsandexperimentsconductedinCensus2000 dӀareprovidedon  thefollowingpages. ~     .  Thispageintentionallyleftblank.#XXXOX#OXXXX &0!$ 9w,  X,X,X9 Census2000AlternativeQuestionnaireExperiment(AQE2000)  -'+ Їc@4 <DL!T$&)\+- 0d24w,  Xc! X4X!Overview t Thisexperimentwasdesignedtomanipulatethreeindependentquestionnairedesign L components.Thefirstcomponentevaluatestheeffectsoftheamountandpresentationofthe 8 residencerulesontheshortform;thatis,incomparisontothecurrentpresentationwoulda $  brieferandreformattedpresentationoftherulesimprovedataquality?Sincethisisacoverage   issue,a_reinterview_Ԁwasconducted.Thesecondcomponentexaminedthepresentationofthe   racequestiontodeterminewhetherchangesinthewaytheracequestionswereaskedinthe1990  t and2000censusesaffectthequalityandcontentofracedata.Specifically,itevaluatedthe  ` combinedeffectsofvariantquestionwording,format,content,anddesignonracedataquality  L  andcontent.Thethirdcomponentpertainstothelongform,specificallyinregardtothedesign 8  oftheskipinstructionstodeterminewhetherthecurrentformatfacilitatedrespondents $  navigationthroughtheformcorrectly. Skiptoand gotoinstructionvariationswere   examined.InformationlearnedaboutthelongformwilladviseimplementationoftheAmerican p  CommunitySurvey.% 4XX% \  Objectives 4 Theobjectivesofthisexperimentweretocontinueeffortstodevelopauserfriendlymail-out   questionnairethatcanbecompletedaccuratelybyrespondentsandtoevaluatetheeffectsof  questionnairechangesonthedata.Correspondingtothevariablesdescribedabove,thespecific p objectiveswere:1)tocomparetheCensus2000shortform,definedascontainingafullsetof \ residencerules,witharevisedformthatcontainsanalternatepresentationoftherules,2)to H comparetheCensus2000shortformpresentationandsequencingoftheracequestion,including 4 itsprovisionformarkingmultiplecategories,tothatofthe1990presentationandinstructions,   and3)tocomparethestandardskipinstructionontheCensus2000longformwithfourrevised   formats. l    h)#' Thispageintentionallyleftblank.#XXXOX#OXXXX '"% F     AdministrativeRecordsCensus2000Experiment (_AREX_Ԁ2000)  -'+ -w,A  X4X-#XXXOXk#  Overview ` ThePlanforCensus2000explicitlycalledforexperimentationwithanadministrativerecords 8 censusfortworeasons:1)useofadministrativerecordsastheprimarydatacollectionmethod $  hastremendouspotentialforcostsavings,and2)significanttestingofadministrativerecordswas   notdoneaspartofthe1990Censustestingandexperimentationprogramand,asaresult,the   CensusBureauwasnotsufficientlypreparedtoincludeadministrativerecordsintheCensus  t 2000design.Thepotentialbenefitofanadministrativerecordscensusistoreducethecostand  ` responseburdenofdirectdatacollection.  L  The_AREX_Ԁ2000exploredtwomethodsforconductinganadministrativerecordscensus.Inboth $  methods,nationalleveladministrativerecordswereassembled,unduplicatedusingSocial   SecurityNumbers,andassignedblocklevelgeographiccodes.Recordsfortheselectedtestsites p  (approximatelyonemillionhousingunitsinfivecounties)wereextractedandtalliedatthe \  censusblocklevel.ThetwomethodsdifferintheiruseoftheMasterAddressFiletocreatea H  universe(frame)ofhousingunits.ThefirstmethoddidnotusetheMasterAddressFilebut 4 providedonlypopulationcountsattheblocklevel.Thesecondmethodmatchedadministrative   recordstoaddressesontheMasterAddressFileandreconcileddifferencesthroughfield   operations.Thismethodprovidedbothpopulationandhousingunitcountsattheblocklevel.  Theexperimentincludedthefollowingfield/_mailout_Ԁoperations:1)aclerical_geocoding_ \ operationtobeconductedatselectedRegionalCensusCenters,2)afieldaddressverification H operation,and3)a_mailout_ԀtoP.O.Boxandruralstyleaddressestoobtain_geocodable_Ԁhouse 4 number/streetnameinformation.   Objectives l The_AREX_Ԁ2000comparedtwomethodsforconductinganadministrativerecordscensusto D Census2000andevaluatedtheresultsandcosts.Thedataanalysisfortheexperimentincluded 0  comparisonsofsite,censustractandblocklevelpopulationandhousingcountsfrom_AREX_ ! 2000andCensus2000.Theanalysisalsoexaminedthesimilaritiesanddifferencesofpopulation " characteristics(age,gender,race/ethnicity)andsimulatethereplacementofCensus2000 "  _nonresponse_Ԁhouseholdenumerationswithadministrativerecordinformation.Secondary #l! objectivesincludedcollectingrelevantinformationthatwasonlyavailablein2000tobeusedfor $X" ongoingtestingandplanningforadministrativerecordsuseinthe2010Censusandfor %D # comparinganadministrativerecordscensustootherpotential2010methodologies. &0!$ OXXXX  |(#& Thispageintentionallyleftblank. '"%   @+%) 9,X, Xw,A  X9 RROaSocialSecurityNumber,PrivacyAttitudes,andNotificationExperiment (SPAN) t  Overview @ ThepurposeoftheSPANwastoobtainbehavioralandattitudinaldataonseveraltopicsrelated   totheuseofadministrativerecords.Thisincludedhowthepublicrespondedtorequestsfor  x SocialSecurityNumbers(_SSNs_)ondecennialcensusquestionnaires,howthepublicrespondsto  \ differentlywordednotificationsabouttheCensusBureaususeofadministrativerecords,andon  @ whatwerethepublicsattitudesonprivacyandconfidentialitypertainingtothenotionofan  $ administrativerecordscensus. |   9,X, X,X, X9TheSPANconsistedoftwocomponents.Thefirstcomponentcollecteddataonrequestingthe D  _SSN_Ԁandtheuseofdifferentlywordednotifications.Thesecondcomponentinvolveda (  telephonesurveythatmeasuredthepublicsattitudesonprivacyandconfidentialityissues.    ObjectivesforComponent1:Specificobjectivesweretodetermine:1)whateffectarequestfor `  the_SSN_Ԁforeveryhouseholdmemberhasonmailanditemresponse,2)whateffectarequestfor D an_SSN_Ԁforonlythepersoncompletingthequestionnairehasonmailanditemresponse,3)the ( accuracyoftherespondentprovided_SSNs_,and4)whateffectdifferentnotificationsonthe   CensusBureauspossibleuseofadministrativerecordshasonmailanditemresponserates.The d methodologyforachievingtheseobjectivesinvolvedthe_mailout_Ԁofsevenshortformandthree H longformpanelseachpanelcontaininga5,000sampleforatotalof50,000formsduring , Census2000.Thelongformpanelsincludedonlythenotificationtestwithnorequestsfor  _SSNs_.  Thereweretwonotificationsreferredtoas generaland specific.Eachnotificationwas H includedinthecoverletteranddescribedhoworwhytheCensusBureaumayuseadministrative , recordsdatafromotherFederalagencies.A generalnotificationmentionedtheCensus  BureauspossibleuseofstatisticaldatafromotherFederalagencies,whilethe specific h notificationgoesfurthertomentionactualFederalagencies,suchastheInternalRevenue L Service,SocialSecurityAdministration,and agenciesprovidingpublichousingassistance. 0  ObjectivesforComponent2:ThesecondcomponentoftheSPANwasatelephonesurveythat !  measuredthepublicsattitudesonprivacyandconfidentialityissuespertainingtoan "h! administrativerecordscensus.ThissurveyincludedpreandpostmeasurementstoCensus2000 #L" toenableexaminationofthecensusenvironmentseffectonprivacyattitudes.Thepre $0# measurementtookplacebeforethenationalpaidadvertisingandfieldrecruitingcampaigns.The % $ postmeasurementoccurredshortlyafterCensusDay,April1,2000. YEachmeasurementgroup l& % wasanationalsampleof2,000households. Y;ӀSpecificobjectiveswereto:1)determinethe P'!& C'Oa  C publicsopinionoftheFederalgovernmentandtheCensusBureauingeneral,2)assesschange +P&+ inthepublicsattitudesonprivacyrelatedissuesusingresultsfromstudiesconductedin1995 ,4', and1996,and3)determinethepublicsopinionoftheCensusBureaustestingonexpandingthe -(- useofadministrativerecords,possibleinterestincollecting_SSNs_Ԁinthefuture,andthenotionof t an administrativerecordscensus. X       #XXXOX],#OXXXX  $,#  SOYResponseModeandIncentiveExperiment(_RMIE_)  t Overview < #XXXOXp?#TX XXX#XX XTXe@#OXXXXThisexperimentmeasuredtheeffectofanincentiveand/oroptionofalternativeelectronicmodes  ofcollectiononresponsetothecensusshortform.Since1970,responsetothemailedformhas  o declinedandlaborcoststovisit_nonresponding_Ԁhouseholdshavegreatlyincreased.Toaddress  S theseproblems,theCensusBureauexploredothermethodsandtechnologiestocountthe  7 population,suchasincentives,telephones,andtheInternet.Thisexperimentdeterminedwhat   effectanincentivehasongettingrespondentstoanswerthecensususingoneofthreeelectronic s   modesofcollection.Theeffectivenessoftheincentivealsowasmeasuredonhouseholdsnot W  respondingtothe_mailout_ԀofthestandardCensus2000questionnaire. ;  Thealternativemodesofcollectionwere:   Ѐ *(xx("3"  3FE2"3  0    Operatortelephoneinterview.Thisisreferredtoasreversecomputerassistedtelephone W  interview(Reverse_CATI_).Thecoverletteraccompanyingthepaperquestionnaire ; encouragedresponseusingatollfreetelephonenumber.Atelephoneinterviewer  administeredthequestionnaireoverthephone.3FEsE݌w(#(# Ќ   +( xx*"3"  3H2 "3  0    Computertelephoneinterview.Thisisaninteractivevoiceresponsesystemcalledthe ? AutomatedSpokenQuestionnaire(_ASQ_).Thecoverletteraccompanyingthepaper # questionnaireencouragedresponseusingatollfreetelephonenumber.Insteadofanoperator  takingtheinterview,aninteractivevoiceresponse(_IVR_)systempromptedtherespondent w throughtheshortforminstrument.3H2H݌[(#(# Ќ   ,(0xx+"3"  3LK20"3  0    Internet.Thecoverletteraccompanyingthepaperquestionnaireencouragedresponseusing # theInternetandincludedadedicateduniformresourcelocator(URL)forthedatacollection.3LKyK݌{(#(# Ќ  Employingthecollectionmodeslistedabove,thisexperimentincorporatedtwotreatments: C responsemodeandincentive,eachwiththreepanels.Therealsowasacontrolgroupconsisting '  ofthreepanelswhichservedastheuniversefortheresponsephaseoftheexperiment.Thetotal  ! _mailout_Ԁforallpanelswas35,380households.Theincentivewasatelephonecallingcardworth !{  thirtyminutesoffreelongdistanceservicewhichwasactivatedafterresponse.Dataanalysis "_! wasconductedonsevenexperimentalcomponentsthatincluded:1)initial_mailout_/operator #C" assistance,2)_nonresponse_,3)_ASQ_Ԁnamerecognition,4)_ASQ_Ԁcustomersatisfaction,5) $'# InternetUsageSurvey(telephonefollowup),6)InternetCustomerSatisfactionSurvey % $ (administeredontheInternet),and7)Internetadministrativedata.  c& % C+OY  C Objectives +G&+ Thisexperimenthasthefollowingkeyobjectives: -(- Ї '(xx,"3"  3kR2"3  0    Determinetheeffectofincentivesoncooperationrates,householdcooperation,item t _nonresponse_,andonsufficientcompleteness.3kRR݌X(#(# Ќ   ((xx'"3"  3XT2"3  0    Determineeffectofresponsemodeoncooperationrates,householdresponse,anditem   _nonresponse_.3XTT݌ (#(# Ќ   )(xx("3"  3&V2"3  0    Determinetheeffectofincentivesandresponsemodeonthecensus_nonresponding_  X households.3&VSV݌ <(#(# Ќ  "3"  3mW2"3  0    Assesstheoperationalbenefitsofofferingelectronicmodesforresponseanddatacollection x   andcapture.3mWW݌\ (#(# Ќ  @*#XXXOX@#OXXXX  #H" #XXXOXX# Census2000SupplementarySurvey  t Overview 4     h      p  < UXXXXCensus2000includedalongformfor1in6householdsacrossthecountry.Essentiallythesame   processhasbeenusedsince1940tocollectbasicsocioeconomicinformation(suchas  t educational,marital,andveteransstatus;housingcharacteristics;andcommutingpatterns)for  X allgeographicareasoftheUnitedStates,#XXXUXZ#UXXXXrangingfromthenationaldowntothecensustract  < level.   Inspiteoftheefficienciesofusingthedecennialcensustocollectthiscriticalsocioeconomic \  data,thereisstronginterestinmovingawayfromthisapproach#XXXUX[#XX XXXԄ#XX XXX]#UXXXXbothtosimplifythecensus @  processandtoprovidemorecurrentandmoreaccuratedataforfederal,stateandlocalusers.In $  responsetothisinterest,Census2000included,inadditiontoatraditionallongform,a   supplementarysurveydesignedastheoperationalfeasibilitytestofcollectinglongformdata x  throughoutthecountryduringthesametimeframebutinaprocessseparatefromthecensus. \  U&C %XUXN  - ,X, X-  N_  #&C % %U&Cd]#O&C % %&C#OXX %O&C}_#ObjectivesO&C %XOXN_`݌  Ќ  #OXX %O&C`#TheobjectiveoftheCensus2000SupplementarySurveywastodemonstratetheoperationaland : technicalfeasibilityofcollectingthefullrangeofsocioeconomicdatagatheredonthedecennial  censuslongformusingadifferentquestionnaireandestimationmethodology.Toaccomplish  thisobjective,theCensus2000SupplementarySurvey#XXXOXB`#OXXXXwasconductedduringCensus2000using r anexistingquestionnaire#XXXOXb#XX XXXԀ_Ԅ_Ԁ#XX XXXVc#UXXXXthatoftheAmericanCommunitySurvey.Resultswillinformthe V processofremovingthelongformfromthecensus. : #XXXUXc#OXXXX  b-'- Ї Thispageintentionallyleftblank. #H"  YK UseoftheEmployeeReliabilityInventoryFileforNonresponseFollowupEnumerators 0("'  Overview )$) TheOfficeofPersonnelManagement(_OPM_)reportedthatthetestsusedtohiredecennialstaff, +L&+ whilevalid,donotassessanimportantaspectoftheknowledge,skills,andabilitiesneededfor ,0', successfulperformanceinterpersonalskills.Thisexperimentwillhelpdetermineifthe -(- EmployeeReliabilityInventory(_ERI_)meetseachofthreecriteriaforavalidselectiontestandif t itisappropriateforuseinselectingdecennialcensusenumerators.Tobeconsideredavalid X selectionaid,thepersonalitybasedcompetenciesmeasuredbya_noncognitive_Ԁtestshould:1)be < jobrelevant,2)havenobetweengroupdifferences,and3)besubjectrelated.Tomeasurethe   _noncognitive_Ԁcompetenciesofcensusnewhires,analreadyexisting_noncognitive_Ԁinstrument   fromthetestingmarket(_ERI_)wasadministeredtoasampleofpeoplehiredtobe_nonresponse_  t followupenumerators.Theresearchwillanswerthefollowingquestions:  X  K(xx)"3"  3k2"3  0 Z   Doestheuseofa_noncognitive_Ԁtestsignificantlyaddtotheoverallpredictabilityofjob   performanceandtenure?3kk݌x  Z(#Z(# Ќ  "3"  3m2"3  0 Z   Canweidentifywhichtraitsactuallydistinguishthosewhostayfromthosewhodonotand \  thosewhoshowthebestperformancefromthosewhodonot?3mHm݌@ Z(#Z(# Ќ  "3 "  3|n2"3  0 Z   Canwedocumentthatdecennialenumeratorswholeftbeforethecompletionofanoperation $  performeddifferentlyonthe_ERI_Ԁthanthosewhostayed?3|nn݌ Z(#Z(# Ќ  "3 "  3o2"3  0 Z   Canweusethe_ERI_Ԁtoreliablypredictturnoverorjobsuccess?3o&p݌x Z(#Z(# Ќ  Objectives @ Theoverallobjectiveofthisstudywastodetermineifanexisting_noncognitive_Ԁtestprovideda | reliableandvalidmeasureofinterpersonalskillsthatcanbeusedbythecensustomakemore ` preciseemployeehiringdecisions.ThegoalwastodetermineiftheCensusBureaucouldreduce D interviewerturnoverandimproveinterviewerworkperformancebyimprovingenumerator ( selectiontoolsthroughtheuseofacommercial_noncognitive_Ԁtestthe_ERI_.   C(K  C  -(- Ї Thispageintentionallyleftblank.    #XXXOXd#OXXXXEthnographicStudies  "d! Overview $,# TheCensusBureauusedethnographictechniquestostudysurveycoverageasearlyas1971.The h& % NationalAcademyofSciencesPanelonDecennialCensusMethodology,establishedbythe L'!& CensusBureauin1984,recommendedthattheCensusBureauundertakeaseriesofparticipant 0("' observationcoveragestudiesinselectedareas.Exploratoryethnographicresearchwasinitiated )#( inanumberofcommunitiestoidentifyandexplainthecomplexbehavioralprocessesthatleadto )$) _underenumeration_.Basedontheexperienceobtainedinpreliminaryresearch,theCensusBureau *h%* launcheditsmostambitiousphaseofethnographicstudiesassociatedwiththe1990Census. +L&+ Morethan40additionalexploratoryandethnographicstudiesandevaluationswereconductedon ,0', awiderangeofpopulations!suchasthehomeless,migrantworkers,AfricanAmericans,Latinos, -(- AmericanIndians,andAsians!andissuessuchasrespondentsunderstandingofcensuslanguage t andconcepts,andothertypesofcommunications. X Objectives   EthnographicstudiesconductedinassociationwithCensus2000providednewinsightsthatcan  t beusedtoimprovecoverageofselectedsegmentsofthisnationspopulation.Thefollowing  X studiesreflectselectedsocialanddemographicaspectsinAmericansocietythatareimportantto  < explorefromanethnographicperspective.Thisperspective,groundedintheactualbehaviorof   respondents,canofferinsightswhichothermethodsmaynotcapture. x    -(Px x K ProtectingPrivacy:Information,TrustandTechnologyintheDecennialCensusand @  DemographicSurveys: Thegoalofthisprojectwastoconductaqualitativestudyofbelief $  structuresthatinfluencesurveyrespondents'perceptionsof,andreactionsto,surveyinformation    requests,focusingonprivacyconcerns.Thisstudyexploredhowrespondentsassessthe |  consequencesofsurveyparticipationandsurveyresponse,theirsenseofinformationownership, `  theirreactionstoconfidentialitystatements,andtheirreasonsforchoosingtoparticipatein D surveydatacollections. (  Z ComplexHouseholdsandRelationshipsintheDecennialCensusandDemographic d Surveys: Thisethnographicresearchprojecthadthreeobjectives:1)toexploretherangeand H functioningofcomplexhouseholdswithindifferentethnicgroupsintheUnitedStates,2)to 0 examinehowtheresponsecategoriesofthedecennialrelationshipquestioncapturetheemerging  diversityofhouseholdtypes,and3)tocomparethehouseholdcompositionandrelationship  informationcollectedbytheethnographicinterviewstothoseinCensus2000.Thisstudywas h designedtoassesshowwellcensusmethods,questions,relationshipcategories,andhousehold L composition_typologies_Ԁdescribetheemergingdiversityofhouseholdtypesinthiscountry.Six 0 ethnographersorteamseachconducted25ethnographicinterviewswithaselectedethnic/race  group:AfricanAmericans,Hispanics,_Inupiaq_ԀEskimos,Koreans,Navajos,orWhites. l  GenerationXSpeaksOutonCensuses,Surveys,andCivicEngagement:AnEthnographic 4  Approach: Thepurposeofthisnationwideethnographicresearchwastoexaminecivic ! engagement,behaviors,andattitudestowardscensusesandsurveysamongGen_Xers_Ԁ(individuals "  bornduringtheyears19681979)fromvariedsocioeconomicbackgroundsand_ethnicities_, "p! includingindividualsfromhardtoenumeratecategories,suchasyoungminoritymalesand #T" immigrants. $8# Patternsofcivicengagementhaveconsequencesforgovernmentdatacollectioneffortsinterms t&!% ofsurvey_nonresponse_,trustandprivacyconcerns,policyorientedissuesandeffective X'!& educationaloutreachcampaigns.MillennialGenerationindividuals(1418yearsofage)were <("' alsointerviewedinorderthatcomparativelifecycleexperiencesandculturalexplanations  )#( emergewithregardtocensusandsurvey_nonresponse_,governmentengagement,andcivic *$) responsibilityandobligation.d *t%* #XXXOXt#d%C %XXD%g % %%C  ,<', ).AZ*Gill Sans Regular  7$    1    _&C %XXReferto PartC:TheCensus2000EvaluationProgram!Introductionformoreinformationon  theplannedevaluationsandrecentchangesintheprogram.D&g % %&C#DX X %D&gS#D%g % XDX = 5'  _(#(#K(#hEGKOSW[_cAutoList14"""""""")1)a)i) 2+0TxBr_p2     (9 Z 6Times New Roman RegularB-  )USUS.,  _   `     h   May2002     ,  AZ"Arial Regular<  p`*Times New RomanTT8888'dxd(3$ !  <  p`ArialTT i) ^pWVC<<CLevel 1Level 2Level 3Level 4Level 5Level 1Level 2Level 3Level 4Level 5cccc'dxd&0 d d(Vm$0   (O$(vGIOU[agmsAutoList1021.1.1.1.1.1.1.1.(GIOU[agmsAutoList103A:A:A:A:A:A:A:A:0479?EKS[ciC-11.a.i.(1)(a)(i)1)a)a)i)(3$ !  (UEGMSY_ekqAutoList181.1.1.1.1.1.1.1.*+ (_2623  .. EGMSY_ekqAutoList26A.A.A.A.A.A.A.A.(3$ !  _W I&mage <=8C HKKKK'dxd( $ Figure  1  3+0 d d d3+0 d d d(EGMSY_ekqAutoList731.1.1.1.1.1.1.1.'dxd&p d Pd'dxd&p Pd d'dxd'dxd d88888'dxd8'dxd($$   1  (եEGMSY_ekqAutoList781.1.1.1.1.1.1.1.*D+D (_25   ," <DL,23  ..," <DL,   *5+5 (_24  ) <DL)23  ..) <DL)  *2+2 (_23 ` &<<DL&23  ..&<<DL& ` 9999'dxd'dxdP Pd'dxd d'dxd d' dxd'dxd&P d d*/+/ (_22  #DL#23  ..#DL#  *,+, (_21   DL 23  .. DL  *)+) (_20 h DDL23  ..DDL h *&+& (_19  L23  ..L  *#+# (_18   L23  .. L  *>> (_17  2( 4 <DL223  Ԁ2( 4 <DL2  *DD (_16   ," <DL,23  Ԁ," <DL,   *55 (_15  ) <DL)23  Ԁ) <DL)  *22 (_14 ` &<<DL&23  Ԁ&<<DL& ` *// (_13  #DL#23  Ԁ#DL#  *,, (_12   DL 23  Ԁ DL  *)) (_11 h DDL23  ԀDDL h *&& (_10  L23  ԀL  (## &_9   L23  Ԁ L  (>> &_8  2( 4 <DL223  2( 4 <DL2  (DD &_7   ," <DL,23  ," <DL,   (55 &_6  ) <DL)23  ) <DL)  (22 &_5 ` &<<DL&23  &<<DL& ` (// &_4  #DL#23  #DL#  (,, &_3   DL 23   DL  ()) &_2 h DDL23  DDL h (&& &_1  L23  L  &## $_   L23   L  <:Default ParaB-  { )  _   `     h   May2002      i) ? )  _  May2002C-  4(( 9 dC- i)  <<U< 0 7$    2    _ԀAlthoughthescopeoftheCensus2000EvaluationProgramiswelldefined,asmore  databecomeavailableandreviewandanalysisprogresses,theremaybeadditionalprogram  changes.http://www.census.gov/ !!! ((h(  \  `&Times New Roman !!!http://www.census.gov/ 4((   (((~( !!V! ((( ((( ((@( ((h( !!V! !!V! ((Vd( ((( ((( ((@( !!V! ((@( !!V! !!! ((@( -!V!  http://www.census.gov/      http://www.census.gov/          4(d(  http://www.census.gov/ ((( ((( !  _p~ <,(`~0y  `9xE c p (#(#            (#(#      XXCensus2000#XX  "#EXX   #XXEo#  44XXTesting,Experimentation,#XX44#  l  44XXandEvaluation  Program#XX44X#      UpdatedMay2002   p~?<,(`~0( | `1cE( | | (7 p   `     h      p   _c!c!%PP**_ԀForprogramchangessincethereleaseofthisdocument,pleasesee ProgramModifications f# SinceMay2002".PPI R$    (#(#                              -("   ЇBXXXX#XXXBX#Thispageintentionallyleftblank.  |*%) _@pp" lXXTableofContents#XXl#   !!G(# &C %XXPage  Introduction""J(#.i    PartA:TheCensus2000Testing,Experimentation,andEvaluationProgramOverview " "H(#.]A1 8   *XXX*  Introduction""H(#.y y A3 @   8   WhatWeWillLearnfromtheEvaluationProgram""H(#.=A4     WhatWeWillLearnfromtheCensus2000TestingandExperimentationProgram!!G(#.ZA6      PlanningfortheNextDecade""H(#.yy+A7 p     Conclusion""H(#.3 3 A8  H     PartB:TheCensus2000TestingandExperimentationProgram""H(#.GB1 H   0  Introduction#""H(#.y y (#(##B3     Census2000AlternativeQuestionnaireExperiment""H(#.?B5     AdministrativeRecordsCensus2000Experiment""H(#.--<B7 @p  X   SocialSecurityNumber,PrivacyAttitudes,andNotificationExperiment""H(#.UB9 @x  0   ResponseModeandIncentiveExperiment!!G(#.!!5B11 X   Census2000SupplementarySurvey!!G(#.)).B13 @  0   UseoftheEmployeeReliabilityInventoryFileforNonresponseFollowupEnumerators!!G(#.1 1 bB15    EthnographicStudies!!G(#.  #B17   PartC:TheCensus2000EvaluationProgram " "H(#.5C1 @  ? X4 XXX?  Introduction""H(#.y y C5 @   A:0 4 ResponseRatesandBehaviorAnalysis#""H(#.64(#4(##C7    B:0 4 ContentandDataQuality#!!G(#.*4(#4(##C11H@@@$@H@@@$@ @     C:0 4 DataProducts#!!G(#.r r 4(#4(##C17 @h  x   D:0 4 PartnershipandMarketing#!!G(#.*4(#4(##C19 P   E:0 4 SpecialPlacesandGroupQuarters#!!G(#.nn34(#4(##C21 @h  (x   F:0 4 AddressListDevelopment#!!G(#.*4(#4(##C25 @   P    G:0 4 FieldRecruitingandManagement#!!G(#.::14(#4(##C31  (!   H:0 4 FieldOperations#!!G(#.6 6 "4(#4(##C33 !"   I:0 4 CoverageImprovement#!!G(#.$$&4(#4(##C37 "#   J:0 4 EthnographicStudies#!!G(#.&4(#4(##C41 `#$   K:0 4 DataCapture#!!G(#.4 4 4(#4(##C43 8$%   L:0 4 ProcessingSystems#!!G(#.$4(#4(##C45 %` &   M:0 4 QualityAssuranceEvaluations#!!G(#.LL/4(#4(##C49 %8!'   N:0 4 AccuracyandCoverageEvaluationSurveyOperations#!!G(#.D4(#4(##C51 &"(   O:0 4 CoverageEvaluationsoftheCensusandoftheAccuracyand'")4(#4(#   0 4 CoverageEvaluationSurvey#!!G(#.,4(#4(##C57 @  p(#*   P:0 4 AccuracyandCoverageEvaluationSurveyStatisticalDesignandEstimation#!!G(#.[4(#4(##C63 H)$+   Q:0 4 Organization,Budget,andManagementInformationSystem#!!G(#.I4(#4(##C65  *p%,   R:0 4 AutomationofCensusProcesses#!!G(#.04(#4(##C67 *H&- #XX %&Cj#BXXXX + '. Ї8XXdXXd8#XXXBX #E    R F @  @`  @  @  @h  @  @  @p  @ GX XXX @  @`  @  @`  @  @  @h  @  @  @p  @  @  #XX XGXl#@@'\XXPartA:#XX\T#  ` @@~~(EXXThe A  @@  %Census2000 (  @@& & !Testing,Experimentation,   andEvaluationProgram   OverviewGqE i  Ѐ#GX XGq#G& XGX P   ((x3   @  @  @  @ 0 @ 0 @ 0 @ #GX XG&#G.c XGX S'! #GX XG.c)#G& XGX }) $ Ї#GX XG&##XX XGX#Thispageintentionallyleftblank. (D#'    d+%*  @  NXX  @`  @  @  @h  @  @ TheCensus2000 t Testing,#XX N # NXXExperimentation#XX NY!# NXX,andEvaluationProgram#XX N!# 0 q &C %XX?,X, X X4 X?#XX %&C)"#lXXIntroduction#XXl"#  1 IntheirfinalreportonthedesignofCensus2000,theCommerceSecretarys2000Census   AdvisoryCommitteeconcluded: WhateveryonewantsisassimpleasABC...ABetter x  Census. \  Buthowwillweknowifweachieveabettercensusin2000,andhowwillwebuildabetterone $  for2010?AnimportantsourceforansweringthesequestionswillbetheCensus2000Testing,   Experimentation,andEvaluationProgram.BesidesbeingusedtoassessCensus2000,this x  programwillhelpdesigntestingforearly2010Censusplanningandprovideinformationforthe \  AmericanCommunitySurvey,MasterAddressFileUpdatingSystem,andotherCensusBureau @  censusesandsurveys.AsothercountrieslooktotheU.S.CensusBureauasaleaderin $  techniquesandmethodologies,theresultsoftheCensus2000Testing,Experimentation,and | EvaluationProgrammayalsohelptheminmakingmoreinformeddecisionsindesigningtheir ` censuses. D Importantfactorsaffectingthenextdecadeinclude:    +( xx("3"  3w(2 "3  0    TheimplementationoftheAmericanCommunitySurveyinlieuofalongformcensusdata ` collection;3w((݌D(#(# Ќ  "3"  3)2 "3  0    Theabilitytodramaticallychangethe2010Censuswhencollectingonlyshortformcensus   data;3))݌d(#(# Ќ  "3"  3*2 "3  0    Acontinuallymaintainedhousingunitaddressframe;3**݌,(#(# Ќ  "3"  3+2 "3  0    ThemodernizationoftheMasterAddressFileUpdatingSystem;3++݌(#(# Ќ  "3"  3,2 "3  0    Theroleofadministrativerecordsinthe2010Census,theAmericanCommunitySurvey,and !H addresslistdevelopment;3,,݌", (#(# Ќ  "3"  3-2 "3  0    Thechangingcommunityroleinthecensusasmanifestedthroughpartnerships, h$" governmentalactivities,andconstituentgroups;3-.݌L%#(#(# Ќ  "3"  34/2 "3  0    Theimpactofarapidlychangingtechnologicalenvironmentoncensusdatacollection, '!% capture,processing,anddissemination;34/a/݌'"&(#(# Ќ  "3"  3|02 "3  0    Difficultyinelicitingpublicresponsetocensusesandsurveys;and3|00݌)L$((#(#    *0%) "3"  312 "3  0    Theabilitytolimitthepotentialforduplicateresponseswhenalternativewaysofresponding t  tothecensusareoffered.  311݌X(#(# Ќ  &  lXXWhatWeWillLearnfromtheEvaluationProgram#XXl2#    TheCensus2000EvaluationProgramwillmeasuretheeffectivenessandimpactondataquality   oftheCensus2000design,operations,systems,andprocesses.Itwillprovidemeasures' 2ofthe   successofCensus2000anditsoperationswhichareofinteresttointernalandexternal  c stakeholders.Forexample,itwillinformdatausersandstakeholdersaboutdataqualityand  G limitationsofthedata,helpexplainthequalityofcensusdata,andprovideinformationneeded  +  forhistoricalcomparabilityofcensusmethodsandprocedures.Thisalsowillinformplanning   anddevelopmentofthe2010Census,theAmericanCommunitySurvey,andtheMasterAddress g  FileUpdatingSystem.Itwillhelpdeterminewhatsimplificationscanbemadetotheoverall K  2010Censusdesign,assistinoperationalplanning,andinformquestionnairedevelopmentand /  alternativedatacollectionmethodologies.Over100H ;#  1      ׀studiesareplannedinthefollowingareas:   (h(#(#(0  088ResponseRates&BehaviorAnalysisg88 0  088Content&DataQualityK88 0  088DataProducts/88 0  088PartnershipandMarketing88 0  088SpecialPlacesandGroupQuartersk88 0  088AddressListDevelopmentO88 0  088FieldRecruiting&Management388 0  088FieldOperations88 0  088CoverageImprovement88 0  088EthnographicStudiesk88 0  088DataCaptureO88     ProcessingSystems 3 0  QualityAssuranceEvaluationsg(#(# 0  0(#(#  Accuracy&CoverageEvaluationSurvey K Operations/(#(# 0  0(#(#  CoverageEvaluationsoftheCensusand  oftheAccuracyandCoverage k EvaluationSurveyO (#(# 0  0(#(#  Accuracy&CoverageEvaluationSurvey 3! StatisticalDesign&Estimation"(#(# 0  0(#(#  Organization,Budget,andManagement # InformationSystemk$(#(# 0  AutomationofCensusProcesses O%(#(# (h88(Manyoftheissueswearetryingtounderstandwiththeseevaluationstudiesaredescribedbelow. w% Insomecases,wewillbeabletoreachfirmconclusions,whileinothersitwillbemoredifficult [ & todisentangleeffectsofthecensusproceduresfromtheexternalenvironment. ?!' "3 "  3A2 "3  0    TheeffectivenessofthePartnershipandMarketingProgramspaidadvertisinginchanging #) awarenessandmailresponsebehaviorofvariousgroupsandhardtocountpopulations;3AAA݌#w*(#(# Ќ  2  2B"0    WhethernationalandregionalobjectivesoftheexpandedPartnershipProgramwere %? , accomplished;2BB݌&#!-(#(#      {'". "3 "  3C2 "3  0    Theeffectivenessofoperationsusedtobuild,update,andassigngeographiccodestothe t  Census2000addresslist.ThiswillinvolvestudiesoftheMasterAddressFile,theCensus X Bureausgeographicdatabase,thePostalServicesDeliverySequenceFile,fieldoperations, < andpartnershipoperationssuchastheLocalUpdateofCensusAddresses;3CC݌ (#(# Ќ  2  2F"0    Coverageratesforvariousdemographicgroupsandareas,asmeasuredbytheAccuracyand  t CoverageEvaluationSurveyandbydemographicanalysis;2F4F݌ X(#(# Ќ  2  2JG"0    TheeffectivenessofthevariousAccuracyandCoverageEvaluationSurveyoperationsin   measuringerrorsinthecensus;2JGeG݌x  (#(# Ќ   .(x x +2  2H"0    Therelativeeffectivenessofvariousoperationsdesignedtoimproveoverallcoverageor @  reducedifferentialcoverageerrorsforhardtoenumerategroupsandareas;2HH݌$ (#(# Ќ  2  2'J"0    Theuse,effectiveness,anddataqualityofvariousmodesavailableforrespondingtothe x  census(Mail,NonresponseFollowup,Internet,TelephoneQuestionnaireAssistance,Be \  Countedforms);2'JBJ݌@(#(# Ќ  2  2K"0    Thecoverage,content,comparability,andsourcesofinformationusedtoconstructthegroup | quartersframeforthedecennialcensus(andAmericanCommunitySurvey);2KK݌`(#(# Ќ  2  2L"0    TheuseandeffectivenessoflanguageassistanceguidesandnonEnglishlanguage ( questionnaires;2LM݌ (#(# Ќ  2  2M"0    ThesuccessoftheDataCaptureSystem,includingtheOpticalMarkRecognition,Optical ` CharacterRecognition,andoperationalproblems;2MN݌D(#(# Ќ  2  2O"0    Theabilityofvariousfieldandprocessingoperationstoidentifyandunduplicatemultiple   responsesforthesamehouseholdorindividual;2O+O݌d(#(# Ќ  2  2      ResponseModeandIncentiveExperiment "      Census2000SupplementarySurvey    0  UseofEmployeeReliabilityInventoryFileforNonresponseFollowupEnumeratorsv(#(#   0  EthnographicStudiesZ(#(#     Keythingswewilllearnfromthesestudiesinclude: " 2  2\"0    Anassessmentofdifferentquestionnairedesignandcontentoncoverageanddataquality, ^ includingtheeffectsoftheamountandpresentationofresidencerules,instrumentdesign, B andacomparisonofthe1990racequestionwiththatusedin2000;2\\݌&(#(# Ќ  2  2^"0    Anassessment,underdecennialconditions,oftheuseofvarioustypesofadministrative z recordsasaprimarydatacollectiontooltwomajorapproacheswillbestudied;2^^݌^(#(# Ќ  2  2_"0    PublicresponseandtheeffectsonmailanditemresponsetoarequestforSocialSecurity & Numbersonthecensusshortform,andtotwovariationsofanotificationabouttheCensus ~  Bureausproposeduseofadministrativerecordsobtainedfromothergovernmentagencies;2__݌b!(#(# Ќ  2  2a"0    PublicresponsetoalternativemodesofresponsesuchasComputerAssistedTelephone *#! Interviews,interactivevoiceresponse,andtheInternet;2aa݌$"(#(# Ќ  2  2b"0    Theeffectsofofferingalternativeselfadministereddatacollectionmodes,aswellasoffering %b $ anincentivetorespondentswhousethesemodes;2bb݌&F!%(#(# Ќ  2  2c"0    Theoperationalandtechnicalfeasibilityofcollectinglongformdatausingthemethodsofthe (#' AmericanCommunitySurvey,akeyelementinvalidatingtheplantoeliminatethelongform f)#( fromthe2010Census;2c d݌J*$)(#(# Ќ   .+%* 2  2e"0    Thevalidityandfeasibilityofusinganoncognitivetestofpersonalitybasedcompetenciesto t selectinterviewerswithbetterinterpersonalskills,therebyreducingturnoverandimproving X workperformance;and2ee݌<(#(# Ќ  2  2g"0    Qualitativedataaboutresponsebehaviorforhardtoenumeratesubgroupsofthepopulation.2g,g݌ (#(# Ќ  & c lXXPlanningfortheNextDecade#XXl h#  X  ResultsfromtheCensus2000Testing,Experimentation,andEvaluationProgramwillhelp  G informtheCensusBureauseffortstoachievethefollowingobjectivesforthe2010Censusand  +  otherprograms:   'c g2  2i"0    Improvecoverageofthepopulationandreducethedifferentialundercount;2ii݌K (#(# Ќ  2  2qj"0    Improvetheaccuracyofresponsesandlocatingpeoplegeographically;2qjj݌ (#(# Ќ  2  2Ck"0    Increasemailresponseratesandreducefieldactivities;2Ck^k݌g(#(# Ќ  2  2 l"0    Maintainandrefineanopenprocesswithallstakeholdersthroughoutthedecadewhile2 l$l݌/(#(# Ќ    increasingtheconfidenceofourcustomers;and  2  2;m"0    Spreadthecostofdatacollectionandupdatingtheaddresslistmoreevenlythroughoutthe O decadetoreducerisk,simplifylogistics,andimprovemanageability.2;mVm݌3(#(# Ќ  Wewilluseathreeprongedstrategytoachievethesegoals:  2  2n"0    EnhancetheMasterAddressFileandgeographicdatabasethroughmodernizationinitiatives O suchasawebbasedsystem,globalpositioningsystem,andanongoingLocalUpdateof 3 CensusAddress(LUCA)program.Thiswill:2nn݌(#(# Ќ   '(xx .0  !0(#(#enhanceLUCA;S (#(#  .(x x'_p! ! 3/("yx . 3/("yy 3/0  !0(#(#directlyattacktheissuesofacompleteandunduplicatedaddresslist;and7!(#(#  3/("x y3/Xr0  !0(#(#facilitateautomationandelectroniccollection." (#(# 2  2`t"0    ThroughtheAmericanCommunitySurvey,collectandtabulatelongformdataeveryyear. #o" Thiswill:2`t{t݌$S#(#(# Ќ  0  !0(#(#expandourabilitytotarget;&!%(#(# 0  !0(#(#simplifythe2010processallowingustofocusoncoverage;ands'!&(#(# 0  !0(#(#providelongformdataonaflowbasis.W("'(#(#   *$) 2  2v"0    Reengineerthe2010Censusprocessthroughearlyplanning,takingintoaccount t  opportunitiesaffordedbynolongformandanenhancedMasterAddressFile.Using X technologyandashortformonlycensuswill:2vv݌<(#(# Ќ  0  !0(#(#establishaflexiblecosteffectiveinfrastructurethatwillfacilitatecoverageimprovement; (#(# 0  !0(#(#setupadataflowdesigntoallowforefficienciestotheprocess;and t(#(# 0  !0(#(#establishafoundationuponwhichthe perfectcensuscanbebuilt. X(#(# & < TheCensus2000Testing,Experimentation,andEvaluationProgramwillhelpustoaddresskey   planningquestionsforthisdecade: x    -(Px x 3/"3 "  3N{2P"3  0    Doweneedtolengthenorshortentimeperiodsforcensusoperationsforqualityor @  operationalreasons?3N{{{݌$ (#(# Ќ  "3 "  3|2P"3  0    Arethereanyunforeseenoperationaldifficultieswhencollectinglongform'< zdatausingthe x  methodsoftheAmericanCommunitySurvey?3||݌\ (#(# Ќ  "3 "  3}2P"3  0    Whatistheoveralleffectofacontinuouslymaintainedaddressfile?3}~݌$(#(# Ќ  "3"  3~2P"3  0    Howcanwebemosteffectivewithpartnerships,promotion,andadvertising?3~ ݌`(#(# Ќ  "3"  32P"3  0    Whatisthepotentialimpactofusingadministrativerecords?3݌((#(# Ќ  "3"  3Հ2P"3  0    Howaccurateareoursampledesignandproceduresforestimatingtotalanddifferential | undercount?3Հ݌`(#(# Ќ  "3"  32P"3  0    Whatistheimpactonfieldactivitiesandinfrastructureofhiringandtrainingmanymore ( enumeratorsthanareneededfordecennialoperationsinordertocompensateforexpected   turnover?30݌d(#(# Ќ  "3"  32P"3  0    Whichresponseoptionsaremosteffective?3̃݌, (#(# Ќ  lXX Conclusion #XXl# "  ThedesignofCensus2000isbyfarthemostambitiousdecennialcensusinhistory,particularly $" initsuseofanopenplanningprocess,promotion,partnerships,newtechnologies,statistical $x# methodology,andalternativemethodsforhardtocountpopulationsandareas.Yetasournation %\ $ continuestogrowandtheneedforrapidandaccuratedatacontinues,alloftheseapproaches &@!% needtobefurtherrefinedanddevelopedtomeetthechallengesofprovidingdatainthe21st '$"& Centurymoredataneedsatlowerlevelsofgeographyonamoretimelybasis. |(#'   `)#( TheCensus2000Testing,Experimentation,andEvaluationProgramwillassisttheCensus t BureauinevaluatingCensus2000andinexploringnewsurveyproceduresinacensus X environment.Itbuildsthefoundationformakingearlyandinformeddecisionsabout <  theroleandscopeofthe2010Censusinthefederalstatisticalsystemanditsinteractionwiththe   AmericanCommunitySurveyandtheMasterAddressFileUpdatingSystem.Thiswork   providescriticalanalysisandinformationforCensusBureauplanningandimplementationof  t decisionsforthe2010CensusandtheAmericanCommunitySurvey.  X      4  `     h      p     OXXXXE     4  `     h      p          4  `     h      p          4  `     h      p @'O\XOXPartB:#OXXO\#Ԉ \  @(OEXOXThe =   @@9 9 %Census2000 $ @@TestingandExperimentation   Program#OXXOE샍#OiXOX ~ #iOi4#G&i k(" Ї ((xx-   @  @  @  @  @ 0 @ 0 @ 0 @ #GX XG&f#G.c XGX t F#GX XG.c#G& XGX    #iG&ώ#Oii#OXXOiw#O NXOX  a+% Ї # NO NҐ##XX N#Thispageintentionallyleftblank. t     OXXXXO NXOX   TheCensus2000TestingandExperimentationProgram#OXXO N"#  t 9,X,X,X, X9 OlXOX  Introduction        4  `      #OXXOl#Asuccessfuldecennialcensus,onethatisresponsivetothenationschangingneeds,cannotbe  ( achievedwithoutearlyplanning.ManykeyissuesforCensus2000,suchasdecliningpublic   cooperationandtighterfundingrestrictions,werealreadybeingstudiedinthelate1980s. d  Fundamentaloperationalchanges,suchasthosedesignedtoimprovetheprocessforcapturing H  informationonthecensusquestionnaires,camefromthisearlyresearch.Foradecennialcensus, ,  muchleadtimeisneededtofirstidentifyandtestpromisingnewprocedures,makeadjustments,   andretestasneeded.Substantialleadtimeisparticularlynecessaryfortheprocurementand   testingofmanydifferenttypesofequipmentthatmustbeinplacetoconductthedecennial d  census. H  Earlyin1997,theCensusBureauformedateamtodeveloptheCensus2000programoftesting   andexperimentation.ThetestsandexperimentswereconductedconcurrentlywithCensus2000 h  becausethedecennialcensusenvironmentprovidedthebestpossibleconditionstolearnabout L thevalueofnewordifferentmethodologies. Researchconductedduringthedecennialcensusis 0 expectedtoguidefuturedecennialcensusdesigns,butalsoprovidevaluableinformationforuse  byotherareasoftheCensusBureau.  OlXOX#OXXOl# OlXOXPlannedTestsandExperiments#OXXOl9#  q  SummarydescriptionsofthetestsandexperimentsconductedinCensus2000 Ӏareprovidedon ` thefollowingpages. D    H Thispageintentionallyleftblank.#XXXOX#OXXXX % $ 9w,  X,X,X9 Census2000AlternativeQuestionnaireExperiment(AQE2000)  +L&+ Їc@4 <DL!T$&)\+- 0d24w,  Xc! X4X!Overview t Thisexperimentwasdesignedtomanipulatethreeindependentquestionnairedesign < components.Thefirstcomponentevaluatestheeffectsoftheamountandpresentationofthe   residencerulesontheshortform;thatis,incomparisontothecurrentpresentationwoulda   brieferandreformattedpresentationoftherulesimprovedataquality?Sincethisisacoverage  t issue,areinterviewwasconducted.Thesecondcomponentexaminedthepresentationofthe  X racequestiontodeterminewhetherchangesinthewaytheracequestionswereaskedinthe1990  < and2000censusesaffectthequalityandcontentofracedata.Specifically,itevaluatedthe   combinedeffectsofvariantquestionwording,format,content,anddesignonracedataquality x   andcontent.Thethirdcomponentpertainstothelongform,specificallyinregardtothedesign \  oftheskipinstructionstodeterminewhetherthecurrentformatfacilitatedrespondents @  navigationthroughtheformcorrectly. Skiptoand gotoinstructionvariationswere $  examined.InformationlearnedaboutthelongformwilladviseimplementationoftheAmerican   CommunitySurvey.% 4XX4% x  Objectives @ Theobjectivesofthisexperimentweretocontinueeffortstodevelopauserfriendlymail-out | questionnairethatcanbecompletedaccuratelybyrespondentsandtoevaluatetheeffectsof ` questionnairechangesonthedata.Correspondingtothevariablesdescribedabove,thespecific D objectiveswere:1)tocomparetheCensus2000shortform,definedascontainingafullsetof ( residencerules,witharevisedformthatcontainsanalternatepresentationoftherules,2)to   comparetheCensus2000shortformpresentationandsequencingoftheracequestion,including | itsprovisionformarkingmultiplecategories,tothatofthe1990presentationandinstructions, ` and3)tocomparethestandardskipinstructionontheCensus2000longformwithfourrevised D formats. (    0("' Thispageintentionallyleftblank.#XXXOX#OXXXX h& %     AdministrativeRecordsCensus2000Experiment (AREX2000)  +L&+ -w,A  X4X-#XXXOX#  Overview X ThePlanforCensus2000explicitlycalledforexperimentationwithanadministrativerecords   censusfortworeasons:1)useofadministrativerecordsastheprimarydatacollectionmethod   hastremendouspotentialforcostsavings,and2)significanttestingofadministrativerecordswas  t notdoneaspartofthe1990Censustestingandexperimentationprogramand,asaresult,the  X CensusBureauwasnotsufficientlypreparedtoincludeadministrativerecordsintheCensus  < 2000design.Thepotentialbenefitofanadministrativerecordscensusistoreducethecostand   responseburdenofdirectdatacollection. x   TheAREX2000exploredtwomethodsforconductinganadministrativerecordscensus.Inboth @  methods,nationalleveladministrativerecordswereassembled,unduplicatedusingSocial $  SecurityNumbers,andassignedblocklevelgeographiccodes.Recordsfortheselectedtestsites   (approximatelyonemillionhousingunitsinfivecounties)wereextractedandtalliedatthe x  censusblocklevel.ThetwomethodsdifferintheiruseoftheMasterAddressFiletocreatea \  universe(frame)ofhousingunits.ThefirstmethoddidnotusetheMasterAddressFilebut @ providedonlypopulationcountsattheblocklevel.Thesecondmethodmatchedadministrative $ recordstoaddressesontheMasterAddressFileandreconcileddifferencesthroughfield | operations.Thismethodprovidedbothpopulationandhousingunitcountsattheblocklevel. ` Theexperimentincludedthefollowingfield/mailoutoperations:1)aclericalgeocoding ( operationtobeconductedatselectedRegionalCensusCenters,2)afieldaddressverification   operation,and3)amailouttoP.O.Boxandruralstyleaddressestoobtaingeocodablehouse | number/streetnameinformation. ` Objectives ( TheAREX2000comparedtwomethodsforconductinganadministrativerecordscensusto d Census2000andevaluatedtheresultsandcosts.Thedataanalysisfortheexperimentincluded H comparisonsofsite,censustractandblocklevelpopulationandhousingcountsfromAREX ,  2000andCensus2000.Theanalysisalsoexaminedthesimilaritiesanddifferencesofpopulation ! characteristics(age,gender,race/ethnicity)andsimulatethereplacementofCensus2000 !  nonresponsehouseholdenumerationswithadministrativerecordinformation.Secondary "d! objectivesincludedcollectingrelevantinformationthatwasonlyavailablein2000tobeusedfor #H" ongoingtestingandplanningforadministrativerecordsuseinthe2010Censusandfor $,# comparinganadministrativerecordscensustootherpotential2010methodologies. % $ OXXXX  L'!& Thispageintentionallyleftblank. h& %   )$) 9,X, Xw,A  X9 OaSocialSecurityNumber,PrivacyAttitudes,andNotificationExperiment (SPAN) t  Overview @ ThepurposeoftheSPANwastoobtainbehavioralandattitudinaldataonseveraltopicsrelated   totheuseofadministrativerecords.Thisincludedhowthepublicrespondedtorequestsfor  x SocialSecurityNumbers(SSNs)ondecennialcensusquestionnaires,howthepublicrespondsto  \ differentlywordednotificationsabouttheCensusBureaususeofadministrativerecords,andon  @ whatwerethepublicsattitudesonprivacyandconfidentialitypertainingtothenotionofan  $ administrativerecordscensus. |   9,X, X,X, X9TheSPANconsistedoftwocomponents.Thefirstcomponentcollecteddataonrequestingthe D  SSNandtheuseofdifferentlywordednotifications.Thesecondcomponentinvolveda (  telephonesurveythatmeasuredthepublicsattitudesonprivacyandconfidentialityissues.    ObjectivesforComponent1:Specificobjectivesweretodetermine:1)whateffectarequestfor `  theSSNforeveryhouseholdmemberhasonmailanditemresponse,2)whateffectarequestfor D anSSNforonlythepersoncompletingthequestionnairehasonmailanditemresponse,3)the ( accuracyoftherespondentprovidedSSNs,and4)whateffectdifferentnotificationsonthe   CensusBureauspossibleuseofadministrativerecordshasonmailanditemresponserates.The d methodologyforachievingtheseobjectivesinvolvedthemailoutofsevenshortformandthree H longformpanelseachpanelcontaininga5,000sampleforatotalof50,000formsduring , Census2000.Thelongformpanelsincludedonlythenotificationtestwithnorequestsfor  SSNs.  Thereweretwonotificationsreferredtoas generaland specific.Eachnotificationwas H includedinthecoverletteranddescribedhoworwhytheCensusBureaumayuseadministrative , recordsdatafromotherFederalagencies.A generalnotificationmentionedtheCensus  BureauspossibleuseofstatisticaldatafromotherFederalagencies,whilethe specific h notificationgoesfurthertomentionactualFederalagencies,suchastheInternalRevenue L Service,SocialSecurityAdministration,and agenciesprovidingpublichousingassistance. 0  ObjectivesforComponent2:ThesecondcomponentoftheSPANwasatelephonesurveythat !  measuredthepublicsattitudesonprivacyandconfidentialityissuespertainingtoan "h! administrativerecordscensus.ThissurveyincludedpreandpostmeasurementstoCensus2000 #L" toenableexaminationofthecensusenvironmentseffectonprivacyattitudes.Thepre $0# measurementtookplacebeforethenationalpaidadvertisingandfieldrecruitingcampaigns.The % $ postmeasurementoccurredshortlyafterCensusDay,April1,2000. YEachmeasurementgroup l& % wasanationalsampleof2,000households. YӀSpecificobjectiveswereto:1)determinethe P'!&   *l%* Oa    publicsopinionoftheFederalgovernmentandtheCensusBureauingeneral,2)assesschange t  inthepublicsattitudesonprivacyrelatedissuesusingresultsfromstudiesconductedin1995 X and1996,and3)determinethepublicsopinionoftheCensusBureaustestingonexpandingthe < useofadministrativerecords,possibleinterestincollectingSSNsinthefuture,andthenotionof   an administrativerecordscensus.         #XXXOXj#OXXXX  L'!&  OYResponseModeandIncentiveExperiment(RMIE)  t Overview < #XXXOX#TX XXX#XX XTX#OXXXXThisexperimentmeasuredtheeffectofanincentiveand/oroptionofalternativeelectronicmodes  ofcollectiononresponsetothecensusshortform.Since1970,responsetothemailedformhas  o declinedandlaborcoststovisitnonrespondinghouseholdshavegreatlyincreased.Toaddress  S theseproblems,theCensusBureauexploredothermethodsandtechnologiestocountthe  7 population,suchasincentives,telephones,andtheInternet.Thisexperimentdeterminedwhat   effectanincentivehasongettingrespondentstoanswerthecensususingoneofthreeelectronic s   modesofcollection.Theeffectivenessoftheincentivealsowasmeasuredonhouseholdsnot W  respondingtothemailoutofthestandardCensus2000questionnaire. ;  Thealternativemodesofcollectionwere:   Ѐ *(xx("3"  332"3  0    Operatortelephoneinterview.Thisisreferredtoasreversecomputerassistedtelephone W  interview(ReverseCATI).Thecoverletteraccompanyingthepaperquestionnaire ; encouragedresponseusingatollfreetelephonenumber.Atelephoneinterviewer  administeredthequestionnaireoverthephone.33`݌w(#(# Ќ   +( xx*"3"  32 "3  0    Computertelephoneinterview.Thisisaninteractivevoiceresponsesystemcalledthe ? AutomatedSpokenQuestionnaire(ASQ).Thecoverletteraccompanyingthepaper # questionnaireencouragedresponseusingatollfreetelephonenumber.Insteadofanoperator  takingtheinterview,aninteractivevoiceresponse(IVR)systempromptedtherespondent w throughtheshortforminstrument.3݌[(#(# Ќ   ,(0xx+"3"  320"3  0    Internet.Thecoverletteraccompanyingthepaperquestionnaireencouragedresponseusing # theInternetandincludedadedicateduniformresourcelocator(URL)forthedatacollection.3݌{(#(# Ќ  Employingthecollectionmodeslistedabove,thisexperimentincorporatedtwotreatments: C responsemodeandincentive,eachwiththreepanels.Therealsowasacontrolgroupconsisting '  ofthreepanelswhichservedastheuniversefortheresponsephaseoftheexperiment.Thetotal  ! mailoutforallpanelswas35,380households.Theincentivewasatelephonecallingcardworth !{  thirtyminutesoffreelongdistanceservicewhichwasactivatedafterresponse.Dataanalysis "_! wasconductedonsevenexperimentalcomponentsthatincluded:1)initialmailout/operator #C" assistance,2)nonresponse,3)ASQnamerecognition,4)ASQcustomersatisfaction,5) $'# InternetUsageSurvey(telephonefollowup),6)InternetCustomerSatisfactionSurvey % $ (administeredontheInternet),and7)Internetadministrativedata.  c& % Objectives +G&+ OY    Thisexperimenthasthefollowingkeyobjectives: H  '(xx,"3"  32"3  0    Determinetheeffectofincentivesoncooperationrates,householdcooperation,item   nonresponse,andonsufficientcompleteness.3݌ (#(# Ќ   ((xx'"3"  3O2"3  0    Determineeffectofresponsemodeoncooperationrates,householdresponse,anditem  X nonresponse.3O|݌ <(#(# Ќ   )(xx("3"  32"3  0    Determinetheeffectofincentivesandresponsemodeonthecensusnonresponding h  households.3*݌L (#(# Ќ  "3"  3$2"3  0    Assesstheoperationalbenefitsofofferingelectronicmodesforresponseanddatacollection   andcapture.3$Q݌ (#(# Ќ  @*#XXXOX#OXXXX  X' % #XXXOX# Census2000SupplementarySurvey  d Overview 4     h      p  , UXXXXCensus2000includedalongformfor1in6householdsacrossthecountry.Essentiallythesame   processhasbeenusedsince1940tocollectbasicsocioeconomicinformation(suchas  t educational,marital,andveteransstatus;housingcharacteristics;andcommutingpatterns)for  X allgeographicareasoftheUnitedStates,#XXXUX#UXXXXrangingfromthenationaldowntothecensustract  < level.   Inspiteoftheefficienciesofusingthedecennialcensustocollectthiscriticalsocioeconomic L  data,thereisstronginterestinmovingawayfromthisapproach#XXXUXz#XX XXXԄ#XX XXX#UXXXXbothtosimplifythecensus 0  processandtoprovidemorecurrentandmoreaccuratedataforfederal,stateandlocalusers.In   responsetothisinterest,Census2000included,inadditiontoatraditionallongform,a   supplementarysurveydesignedastheoperationalfeasibilitytestofcollectinglongformdata x  throughoutthecountryduringthesametimeframebutinaprocessseparatefromthecensus. \  U&C %XUXN  - ,X, X-  NT  #&C % %U&C#O&C % %&C#OXX %O&C4#ObjectivesO&C %XOXNT݌ u Ќ  #OXX %O&CK#TheobjectiveoftheCensus2000SupplementarySurveywastodemonstratetheoperationaland * technicalfeasibilityofcollectingthefullrangeofsocioeconomicdatagatheredonthedecennial  censuslongformusingadifferentquestionnaireandestimationmethodology.Toaccomplish  thisobjective,theCensus2000SupplementarySurvey#XXXOX#OXXXXwasconductedduringCensus2000using r anexistingquestionnaire#XXXOXw#XX XXXԀ#XX XXX #UXXXXthatoftheAmericanCommunitySurvey.Resultswillinformthe V processofremovingthelongformfromthecensus. : #XXXUXS#OXXXX +B%*  Thispageintentionallyleftblank. <(!&  +h%*  K UseoftheEmployeeReliabilityInventoryFileforNonresponseFollowupEnumerators d  Overview , TheOfficeofPersonnelManagement(OPM)reportedthatthetestsusedtohiredecennialstaff,   whilevalid,donotassessanimportantaspectoftheknowledge,skills,andabilitiesneededfor  t successfulperformanceinterpersonalskills.Thisexperimentwillhelpdetermineifthe  X EmployeeReliabilityInventory(ERI)meetseachofthreecriteriaforavalidselectiontestandif  < itisappropriateforuseinselectingdecennialcensusenumerators.Tobeconsideredavalid   selectionaid,thepersonalitybasedcompetenciesmeasuredbyanoncognitivetestshould:1)be h  jobrelevant,2)havenobetweengroupdifferences,and3)besubjectrelated.Tomeasurethe L  noncognitivecompetenciesofcensusnewhires,analreadyexistingnoncognitiveinstrument 0  fromthetestingmarket(ERI)wasadministeredtoasampleofpeoplehiredtobenonresponse   followupenumerators.Theresearchwillanswerthefollowingquestions:    K(xx)"3"  32"3  0 Z   Doestheuseofanoncognitivetestsignificantlyaddtotheoverallpredictabilityofjob \  performanceandtenure?3݌@Z(#Z(# Ќ  "3"  32"3  0 Z   Canweidentifywhichtraitsactuallydistinguishthosewhostayfromthosewhodonotand $ thosewhoshowthebestperformancefromthosewhodonot?3݌lZ(#Z(# Ќ  "3 "  3'2"3  0 Z   Canwedocumentthatdecennialenumeratorswholeftbeforethecompletionofanoperation P performeddifferentlyontheERIthanthosewhostayed?3'T݌4Z(#Z(# Ќ  "3 "  32"3  0 Z   CanweusetheERItoreliablypredictturnoverorjobsuccess?3݌Z(#Z(# Ќ  Objectives | Theoverallobjectiveofthisstudywastodetermineifanexistingnoncognitivetestprovideda D reliableandvalidmeasureofinterpersonalskillsthatcanbeusedbythecensustomakemore ( preciseemployeehiringdecisions.ThegoalwastodetermineiftheCensusBureaucouldreduce p  interviewerturnoverandimproveinterviewerworkperformancebyimprovingenumerator T selectiontoolsthroughtheuseofacommercialnoncognitivetesttheERI. 8   +h%* K     Thispageintentionallyleftblank.  <(!& #XXXOX0#OXXXX +h%* ЇEthnographicStudies  d Overview , TheCensusBureauusedethnographictechniquestostudysurveycoverageasearlyas1971.The   NationalAcademyofSciencesPanelonDecennialCensusMethodology,establishedbythe  t CensusBureauin1984,recommendedthattheCensusBureauundertakeaseriesofparticipant  X observationcoveragestudiesinselectedareas.Exploratoryethnographicresearchwasinitiated  < inanumberofcommunitiestoidentifyandexplainthecomplexbehavioralprocessesthatleadto   underenumeration.Basedontheexperienceobtainedinpreliminaryresearch,theCensusBureau h  launcheditsmostambitiousphaseofethnographicstudiesassociatedwiththe1990Census. L  Morethan40additionalexploratoryandethnographicstudiesandevaluationswereconductedon 0  awiderangeofpopulations!suchasthehomeless,migrantworkers,AfricanAmericans,Latinos,   AmericanIndians,andAsians!andissuessuchasrespondentsunderstandingofcensuslanguage   andconcepts,andothertypesofcommunications. x  Objectives @ EthnographicstudiesconductedinassociationwithCensus2000providednewinsightsthatcan l beusedtoimprovecoverageofselectedsegmentsofthisnationspopulation.Thefollowing P studiesreflectselectedsocialanddemographicaspectsinAmericansocietythatareimportantto 4 explorefromanethnographicperspective.Thisperspective,groundedintheactualbehaviorof  respondents,canofferinsightswhichothermethodsmaynotcapture.   -(Px x K ProtectingPrivacy:Information,TrustandTechnologyintheDecennialCensusand ` DemographicSurveys: Thegoalofthisprojectwastoconductaqualitativestudyofbelief D structuresthatinfluencesurveyrespondents'perceptionsof,andreactionsto,surveyinformation , requests,focusingonprivacyconcerns.Thisstudyexploredhowrespondentsassessthe t consequencesofsurveyparticipationandsurveyresponse,theirsenseofinformationownership, X theirreactionstoconfidentialitystatements,andtheirreasonsforchoosingtoparticipatein <  surveydatacollections.  !  Z ComplexHouseholdsandRelationshipsintheDecennialCensusandDemographic "  Surveys: Thisethnographicresearchprojecthadthreeobjectives:1)toexploretherangeand #h! functioningofcomplexhouseholdswithindifferentethnicgroupsintheUnitedStates,2)to $P" examinehowtheresponsecategoriesofthedecennialrelationshipquestioncapturetheemerging %4# diversityofhouseholdtypes,and3)tocomparethehouseholdcompositionandrelationship |& $ informationcollectedbytheethnographicinterviewstothoseinCensus2000.Thisstudywas `' % designedtoassesshowwellcensusmethods,questions,relationshipcategories,andhousehold D(!& compositiontypologiesdescribetheemergingdiversityofhouseholdtypesinthiscountry.Six ()"' ethnographersorteamseachconducted25ethnographicinterviewswithaselectedethnic/race  *#( group:AfricanAmericans,Hispanics,InupiaqEskimos,Koreans,Navajos,orWhites. *$)  +p%*  GenerationXSpeaksOutonCensuses,Surveys,andCivicEngagement:AnEthnographic d Approach: Thepurposeofthisnationwideethnographicresearchwastoexaminecivic H engagement,behaviors,andattitudestowardscensusesandsurveysamongGenXers(individuals 0 bornduringtheyears19681979)fromvariedsocioeconomicbackgroundsandethnicities,   includingindividualsfromhardtoenumeratecategories,suchasyoungminoritymalesand   immigrants.  x Patternsofcivicengagementhaveconsequencesforgovernmentdatacollectioneffortsinterms  @ ofsurveynonresponse,trustandprivacyconcerns,policyorientedissuesandeffective  $ educationaloutreachcampaigns.MillennialGenerationindividuals(1418yearsofage)were l  alsointerviewedinorderthatcomparativelifecycleexperiencesandculturalexplanations P  emergewithregardtocensusandsurveynonresponse,governmentengagement,andcivic 4  responsibilityandobligation.d   #XXXOX#d Z  Z  Z  Z  Z  Z  Z  Z  Z  Z GX XXX Z  Z  Z  Z  Z  Z  Z  Z  Z  Z  Z  Z @EE'#XX XGX#EXXPartC:#XXE#Ԉ  ' %   *#&   EXX@@~~(The d  @@  %Census2000 K  @@& & !EvaluationProgram#XXE&#iXX 2   ((xx -   @  @ 0 @ 0 @ 0 @ #XXi#lXX     Z  Z  Z  Z  Z #XXl#iXX#XXi# %,% ЇThispageintentionallyleftblank. 8   Z  ! H   585 NXXCensus2000EvaluationProgram#XX N# 8 @"TableofContents  h0    Page 0   @  @`  @  @  @h  @  @  @p    Introduction!!F(#.C5   A:ResponseRatesandBehaviorAnalysis!!H(#.2C7   dVm  dVm!0 Z    B:ContentandDataQuality#!!G(#. &Z(#Z(##C11dVm!=݌  P  Ќ  dVm  dVm0 Z    C:DataProducts#!!G(#.8 8 Z(#Z(##C17dVm ݌ P  Ќ  dVm  dVm0 Z    D:PartnershipandMarketing#!!G(#.,,'Z(#Z(##C19dVm݌   Ќ  dVm  dVm0 Z    E:SpecialPlacesandGroupQuarters#!!G(#./Z(#Z(##C21dVm݌   Ќ  dVm  dVm\0 Z    F:AddressListDevelopment#!!G(#.&Z(#Z(##C25dVm\x݌ p Ќ  dVm  dVm+0 Z    G:FieldRecruitingandManagement#!!G(#.-Z(#Z(##C31dVm+G݌ p8 Ќ  dVm  dVm0 Z    H:FieldOperations#!!G(#.  Z(#Z(##C33dVm݌ 8 Ќ  dVm  dVm0 Z    I:CoverageImprovement#!!G(#. "Z(#Z(##C37dVm݌  Ќ  dVm  dVm0 Z    J:EthnographicStudies#!!G(#.N N "Z(#Z(##C41dVm݌  Ќ  dVm  dVm^0 Z    K:DataCapture#!!G(#. Z(#Z(##C43dVm^z݌ X Ќ  dVm  dVm!0 Z    L:ProcessingSystems#!!G(#. Z(#Z(##C45dVm!=݌ X  Ќ  dVm  dVm0 Z    M:QualityAssuranceEvaluations#!!G(#.+Z(#Z(##C49dVm݌  ! Ќ  dVm  dVm0 Z    N:AccuracyandCoverageEvaluationSurveyOperations#!!G(#.@Z(#Z(##C51dVm݌ "! Ќ  dVm  dVm 0 Z    O:CoverageEvaluationsoftheCensusandoftheAccuracyandCoveragedVm  ݌$x#Z(#Z(# Ќ  ݀EvaluationSurvey!!G(#. !C57 %\ $ dVm  dVm!0 Z    P:AccuracyandCoverageEvaluationSurveyStatisticalDesignandEstimation#!!G(#.WZ(#Z(##C63dVm!!݌ \'$"& Ќ  dVm  dVm"0 Z    Q:Organization,Budget,andManagementInformationSystem#!!G(#.EZ(#Z(##C65dVm""݌ $)#( Ќ  dVm  dVm#0 Z    R:AutomationofCensusProcesses#!!G(#.,Z(#Z(##C67dVm##݌ *%* Ќ  v -`(- " <<9   Thispageintentionallyleftblank. % %      &!& 9HlXXIntroduction#XXl%#  %Ԁ ;#  2       U    Foroverhalfacentury,theCensusBureauhasconductedaformalevaluationprogramin D conjunctionwitheachdecennialcensus.ForCensus2000,theEvaluationProgramwillassess (  theeffectivenessofkeyoperations,systems,andactivitiesinorderto evaluatethecurrentcensus   andtofacilitateplanningforthe2010Census,theAmericanCommunitySurvey,andtheMaster   AddressFileUpdatingSystemmodernization.   TheCensus2000DressRehearsalin1998includedevaluationsofquestionnairedesign,field  G operations,dataprocessing,andestimation.Over40evaluationstudieswereusedtoinformthe +  finalCensus2000design.Asoriginallyplanned,theCensus2000EvaluationProgrammore d  thantripledthiseffortwithnearly150evaluations.Inearly2002,theCensus2000Evaluation H  ProgramwasrefinedandprioritiesreassessedduetoresourceconstraintsattheCensusBureau. ,  Weattemptedtoobtainthebestbalanceofresourcesneededfor:    (Exx("3"  , ,2E3     0 Z   completingandreleasingCensus2000dataproducts,and3 ,F,݌gZ(#Z(# Ќ  "3"  -2E3   K 0 Z   conductingkeyCensus2000evaluations.3-H-݌/Z(#Z(# Ќ  ,8,Toaccomplishthis,wecombinedimportantaspectsofsimilarevaluationsanddroppedthosethat L werelesscriticalfor2010Censusplanning.Wedroppedcomponentsofevaluationsforwhich 0 analyticaldatawerenotavailable.Additionally,someevaluationsplannedfortheAccuracyand  CoverageEvaluationwerenolongerneededwhenthedecisionwasmadenottoadjustthe  Census2000populationcounts.Itwasdeterminedthatsomeofthereportsthatweredeveloped  inanexpeditedmannertoinformtheExecutiveSteeringCommitteeforAccuracyandCoverage k EvaluationPolicy(ESCAP)decisionsweresufficientlycompleteandinformativetoanswerthe O researchquestionsfromFԀP1theearlierplannedevaluationreports.Asaresult,theCensus2000 3 EvaluationProgram,whichpreviouslyincluded149evaluationreports,willnowinclude115,of l which18areESCAPreports.ThesereportsandtheircorrespondingESCAPreportsarenotedin P  thefollowingsectionforevaluationreportdescriptions.ToreviewtheESCAPreportsonthe 4! Internet,goto4 B fO  5  http://www.census.gov/6:3fON3  7 |3Ԁandsearchon ESCAP. " Theevaluationsfallinto18broadcategoriescoveringresponserates,dataquality,partnership #! andmarketing,addresslistdevelopment,fieldoperations,coverageimprovement,datacapture $o" andprocessingsystems,theAccuracyandCoverageEvaluationSurvey,andothers.The %S# evaluationsspeaktoissuesofquality,plausibility,feasibility,accuracy,effectiveness,andvalue, &7 $ andwillprovideacomprehensiveassessmentoftheoperationsandoutcomesofthecensus. p'!% ЇEveryevaluationintheprogramwasapprovedbytheCensus2000EvaluationsExecutive 9 SteeringCommittee.ThesteeringcommitteeincludestheAssociateDirectorforDecennial  Census,theAssociateDirectorforMethodologyandStandards,divisionchiefs,andothercensus   experts.AllevaluationsundergoanextensiveQualityAssuranceprocess.Evaluation   methodologiesandstudyplansarecritiquedbyawideaudienceofcensusexperts.  t Specifications,fieldprocedures,andcomputerprogramsaredocumented,reviewed,and  X approvedbyappropriatecensusstaff.Finally,eachevaluationreportisreviewedforfactual  < accuracyandthensenttotheCensus2000EvaluationsExecutiveSteeringCommitteeand u  CensusBureauExecutiveStafffortheirapproval. Y  Resultsfromtheevaluations,aswellasrelevantresultsfromtestsandexperiments,willbe !  synthesizedintotopicreports thataddressbroadcensussubjectsthatcrosscategories.Current   plansaretopreparereportsforthefollowingtopics:addresslistdevelopment,partnershipand   marketing,coverageimprovement,datacollection,dataprocessing,datacapture,automationof |  censusprocesses,coveragemeasurement(AccuracyandCoverageEvaluationProgram),content `  anddata quality, responseratesandbehavioralanalysis,raceandethnicity,ethnographicstudies, D PuertoRico,specialplacesandgroupquarters,andprivacy.  ,  Foreachofthe18categories,thissectionprovidesan Overviewanda WhatWillWeLearn? I section,followedbyabriefdescriptionofeachplannedevaluation.  -   Z  Z  Z  Z  Z  Z  Z  Z  Z  Z  Z  "  % <U<      5VU5A:ResponseRatesandBehaviorAnalysis  a'? U 4? Z  Z  Z Overview   Theseevaluationsexaminevariousmodesforprovidingresponsestothecensus.Wewillstudy  X theuseofthetelephoneandInternetasresponseoptions.Theeffectivenessofmailingpractices  < andthetargeteddisseminationofformswillalsobeassessed.Theseevaluationsfocuson u  respondentbehaviorandhowthatbehaviorimpactsresponserates(i.e.,mailback,telephone,and Y  Internet).Findingsfromtheseevaluationswillidentifymethodsthatcanbeusedinfuture =  censusestoimprovetheoverallresponserates. !  (#(#K(#WhatWillWeLearn?    Z  Z  Z  Z  Z  Z  Z  Z Thefindingsfromtheseevaluationswillansweranumberofcriticalquestionsabouthow  quicklytheU.S.populationrespondedtoCensus2000.Fromatechnicalstandpoint,theuseof | anInternetQuestionnaireAssistancemodulewilldemonstratetheutilityofemployingthe most ` currenttechnologiesandprovideinsightintorespondentperceptionofusingthismodefor D requestinginformationorcompletingaquestionnaire.Likewise,anenhancedtelephone }( questionnaireassistanceprogramthatisuserfriendlyandcomprehensivewillprovidefurther a  insightintorespondentneedsandpreferences. E Analyzingmailresponse/returnrates(byformtype,demographics,andgeography)andmailing  ! practices,suchastrackingundeliverablequestionnaires,willprovideinsightintoimproving ! overallresponserates.AssessmentoftheBeCountedCampaignwillhelpdetermine "  demographicgroupsthatrespondedviatheBeCountedCampaign.Wealsowillexaminethe #d! frequencyofusinglanguageassistanceguidesandquestionnairesinlanguagesotherthan $H" English,alongwiththenumberofreturnednonEnglishquestionnaires. %,#  +h%*  Z  Z  Z  Z  Z  Z  Z @*  =  @&U@?@ |OYResponseRatesandBehaviorAnalysisEvaluations  U  d(A.1.a)dTelephoneQuestionnaireAssistanceOperationalAnalysis   TheCensus2000TelephoneQuestionnaireAssistancesystemwasdevelopedwithcontractor   supporttoprovidethefollowingservicestorespondents:1)helpingthemcomplete   questionnaires,2)providingquestionnaires(Englishformsonly)andforeignlanguageguides  x uponrequest,and3)conductingshortformquestionnairetelephoneinterviewswhennecessary.  \ Thisoperationalevaluationdassessescallingpatterns,callerbehavior,andsystemperformance.  @ d (A.1.b)TelephoneQuestionnaireAssistanceCustomerSatisfactionSurvey  ]  ThisevaluationfocusesoncustomerreactiontotheCensus2000TelephoneQuestionnaire E  Assistanceprogram.Itincludesanalysesinthefollowingareas:accessibility,easeofuse, )  overallsatisfactionwiththeassistanceandappropriatenessoftheinformationprovided.     (A.2.a)InternetQuestionnaireAssistanceOperationalAnalysis(cancelled)    dThisevaluationwillnotbeconducted.Inearly2002,theCensus2000EvaluationProgramwas d  refinedandprioritiesreassessedduetoresourceconstraintsattheCensusBureau. H d (A.2.b)InternetDataCollectionOperationalAnalysis  e ?+ ` hp x ?XXXXForCensus2000,respondentshadtheopportunitytocompletetheshortformquestionnaireon I theInternet.Thiswasthefirsttimeadecennialcensushadusedthisdatacollectionmode.Since - thereisnobackgrounddata,ageneralevaluationoftheInternetdatacollectionmodeisplanned  onwhatmightbeexpectedintermsoffrequencyofuse.  #XXXXyP# (A.2.c)Census2000InternetWebSiteandQuestionnaire CustomerSatisfactionSurvey  h  CustomersatisfactionsurveysareusedtodmeasurerespondentsatisfactionwithboththeInternet P QuestionnaireAssistanceandtheInternetDataCollectionprograms. 4  (A.3)BeCountedCampaignforCensus2000  Q  TheBeCountedCampaignmadeblankquestionnairesavailableatconvenientlocationsfor 5  peoplewhobelievedtheywereleftoutofCensus2000.Thisevaluationwillexaminepersonand ! housingunitcoveragegainsfromthecampaignalongwiththecharacteristicsofthose ! enumeratedonBeCountedforms.ThisevaluationalsowillassesstheimpactontheMaster "  AddressFilethroughdocumentationofhousingunitaddsresultingfromthisprogram,anditwill #p! evaluateourabilitytogeocodeandprocessBeCountedforms. $T"  (A.4)LanguageProgramUseofNonEnglishQuestionnairesandGuides q& $  Thisstudywilldocumenthowmanyhousingunitsweremailedtheadvanceletterabout U'!% requestinganonEnglishquestionnaire,bystateandtypeofenumerationarea(e.g., 9(!& mailout/mailback,update/leave,etc.);howmanynonEnglishformswererequested,completed, )"' andcheckedin;andthefrequencyofrequestsfornonEnglishshortandlongforms.  Thisstudy *#( alsowilldocumentthenumberoflanguageassistanceguidesrequestedthroughTelephone *$) QuestionnaireAssistance,QuestionnaireAssistanceCenters,andtheInternet,alongwithan +t%* analysisofwhichlanguagesweremostoftenrequested,whethertherequestswereclustered U geographically,andhowmanyrequestsforalanguageassistanceguideresultedinamail 9 returnedform.   (A.5)B.` hp x (#XBӀResponseProcessforSelectedLanguageGroups   E1` hp x (#XE ThisevaluationwillprovideinsightintohowSpanish,Vietnamese,andRussianspeaking  t householdscopedwiththecensusquestionnaireinCensus2000.Specifically,wewilllookat  X howthesenonEnglishspeakinglongformhouseholdswereenumerated.Wewillassesstheir  < useoflanguageguides,QuestionnaireAssistanceCenters,TelephoneQuestionnaireAssistance, u  andtheirexperiencewiththeEnglishform.  Y  (A.6.a)U.S.PostalServiceUndeliverableRatesforCensus2000MailoutQuestionnaires !   ForCensus2000,thequestionnairemailout/mailbacksystemprovidedtheprimarymeansof   enumeration.Thistypeofenumerationwasconductedmainlyinurbanandsuburbanareas,but   alsoinsomeruralareasthatcontainedcity-styleaddress(housenumber/streetname)systems. x  ThisevaluationexaminestheratesatwhichhousingunitswereclassifiedbytheU.S.Postal \  Serviceas undeliverableasaddressed(UAA)forvaryinglevelsofgeography;theoccupancy @ statusofthosehousingunits;demographiccharacteristicsforhousingunitsthatweredeemed y$ undeliverablebuthadafinalstatusofoccupied;theeffectthatundeliverablequestionnaireshad ] onnonresponserates;andthecheck-inpatternofUAAquestionnairesaccordingtodateof A receipt. %  (A.6.b) DetailedReasonsforUndeliverabilityofCensus2000MailoutQuestionnairesby  theUSPS   Thisevaluationfurtherexaminestheissueoftheundeliverabilityofcensusmailout h questionnaires.AftertheU.S.PostalServicedeterminedthatmailpieceswere undeliverableas L addressed(UAA),theCensusBureauattemptedtodeliverthesecasesattheLocalCensus 0 Officelevel.ThisevaluationassessesthequantityofquestionnairesdesignatedasUAAandthe i distributionoftheUAAquestionnairesaccordingtoreasonforundeliverability. M  (A.7.a)Census2000MailbackResponseRates  ! Housingunitsinmailout/mailbackandupdate/leaveenumerationareaswereaskedtoreturn ! questionnairesinpostagepaidenvelopes.Thosequestionnaireswerereceivedandcheckedinat "  DataCaptureCenters.Thisevaluationexaminesmailresponseratesatvaryinglevelsof #l! geographyandquantifiesinformationaboutincomingquestionnairesaccordingtoformtypeand $P" timingwithrespecttocriticaloperationaldates. %4#   (A.7.b)Census2000MailReturnRates  Q' % Housingunitsinmailout/mailbackandupdate/leaveenumerationareaswereaskedtoreturn 5(!& questionnairesinpostagepaidenvelopes,andonceallfollowupoperationswerecomplete,those )"' housingunitswereassignedafinalstatus.Onlythehousingunitsassignedtoreceivean )#( update/leaveormailout/mailbackquestionnairewithafinalstatusofoccupiedonCensusDay *$) (April1,2000)arefactoredintothemailreturnrates.Dataonmailreturnratesprovidemore +p%* accuratemeasuresofcooperationthanmailresponserates,forwhichthedenominatoralso U includesunitsthatturnedouttobevacantornonexistentonCensusDay(April1,2000).This 9 evaluationexaminesmailreturnratesatvaryinglevelsofgeography,quantifiesinformation  aboutincomingquestionnairesfromoccupiedhousingunitsaccordingtoformtypeandtiming   withrespecttocriticaloperationaldates,andprovidesreturnratedataaccordingtocertain   housingunitdemographiccharacteristics.  t  (A.8)PuertoRicoFocusGroupsonWhyHouseholdsDidNotMailBacktheCensus2000  < Questionnaire u   ForCensus2000,theCensusBureauconductedanupdate/leaveenumerationforthefirsttimein Y  PuertoRico.Thatis,censusenumeratorsleftaquestionnaireateachhousingunitwithamailing =  address,andresidentswereaskedtocompletethequestionnaireandreturnittotheCensus !  Bureauinapostagepaidenvelope.Theresponseratewascloseto50percent.Thepurposeof   thisresearchistoobtaininformationonwhynonrespondentsdidnotreturnthequestionnaireby   mailforCensus2000inPuertoRico.Thisinformationwillhelpdevelopstrategiesfor x  improvingtheresponserateforthe2010Census.  \   d. @ ̀ ]     ӀB:ContentandDataQuality ao )#( o Overview +h%* * 5U5   ForCensus2000,thepublichadfivewaysofprovidingcensusdata.Thesemodesincluded 9 mailingbackaquestionnaire,fillingoutacensusshortformontheInternet,pickingupand  returningaBeCountedform,completingashortformcensusinterviewviatelephone   questionnaireassistance,orcompletingapersonalvisitinterviewwithanenumerator.Withthis   inmind,andthelikelihoodthatthe2010Censusmayofferadditionaloptionsforresponse,  t studiesinthiscategorywilldocumentthehundredpercentdataitemnonresponsebyresponse  X mode.Additionally,thedataqualityofeachmodewillbeassessed.Thiscategoryincludesa  < ContentReinterviewSurveystudythatwillmeasureresponsevariance,andaMasterTrace u  Samplestudy.Thelatterwillcreateadatabasecontainingasampleofcensusrecordswith Y  informationpertainingtothemfromtheentirecensusprocess.Otherresearchwillevaluate =  multipleresponsestothenewracequestion. !  WhatWillWeLearn?   Thefindingsfromtheseevaluationswillansweranumberofcriticalquestionsonourprocessto \  definecontent(i.e.,whatquestionstoask)andtheresultingqualityofdataforCensus2000. @ Thesefindings,inturn,canhelpusdoabetterjobforthe2010CensusandtheAmerican y$ CommunitySurvey. ] Wewilllearnaboutthecompletenessofthedatabycalculatingitemimputationratesforseveral % dataitems.Wealsowilllookathundredpercentdataitemnonresponsebydatacollectionmode.   Wewillassessresponsestothenewracequestion.Inparticular,wewillrecontactasampleof  householdswithresponsesoftwoormoreraces,andcollectadditionalinformation,includingan | instructiontochooseasingleracecategory.Thisstudyisneededtomeetthedatarequirements ` ofotheragenciesthatuseonlysingleracecategories,andforcomparisonto1990Censusrace D data. }( Wealsowillgainknowledgeaboutdataqualityincomparisontoexternalbenchmarksby E matchingand/orcomparingcensusdatatodatacollectedbyCensusBureaudemographicsurveys )  includingtheCurrentPopulationSurvey,AmericanCommunitySurvey,andtheSurveyof  ! IncomeandProgramParticipation.Theresultsofthesematchingandcomparisonstudieswill ! alsohelpustoimprovethedesignoffuturesurveysandcensuses. "  SomeofthereportsthatweredevelopedinanexpeditedmannertoinformtheExecutive $H" SteeringCommitteefortheAccuracyandCoverageEvaluation(ESCAP)decisionswere %,# sufficientlycompleteandinformativetoanswerresearchquestionsfromtheplannedevaluation e& $ reports. I' % OneofthesereportsanditscorrespondingESCAPreportisnotedinthefollowingsectionof )"' evaluationreportdescriptions.ToreviewtheESCAPreportontheInternet,goto )#( Ѐ4 fO  5  http://www.census.gov/69~fOM~  7 L {~Ԁandsearchon ESCAP. *$)   +h%* OY+    ContentandDataQualityEvaluations  d  (Exxd(B.1)dB+ ` hp x XBXXXXAnalysisoftheImputationProcessfor100PercentHouseholdPopulationItems , #XXXX# XXXXTodealwithmissingdandinconsistentdata,threecomponentswillcomprisetheimputation   processforCensus2000:assignment,allocation,andsubstitutions.Ratesforeachofthese   componentswillbeproducedforthe100percentcharacteristicsandforthetenureitem.This  t analysiswilldocumentimputationratesandwillserveasasupplementtootherevaluations.#XXXXF#   X  (B.2)DocumentationofCharacteristicsandDataQualitybyResponseType(cancelled)   $ Thisevaluationwillnotbeconducted.Inearly2002,theCensus2000EvaluationProgramwas l  refinedandprioritiesreassessedduetoresourceconstraintsattheCensusBureau.Someofthe P  hundredpercentdataitemnonresponseaspectshavebeenincorporatedintotheB.1evaluation. 4   (B.3)CensusQualitySurveytoEvaluateResponsestotheCensus2000QuestiononRace     OXXXX Databyracefrommostfederalsurveyscurrentlyreflectacollectionmethodologyofasking |  respondentstomarkonlyoneracecategory.UsersoftheCensus2000dataonracewillneedto `  comparetheracedistributionfromCensus2000totheseothersources.Theobjectiveofthe D studyistoproduceadatafilethatwillimproveusersabilitytomakecomparisonbetween ( Census2000dataonrace,thatallowedthereportingofoneormoreraces,anddataonracefrom p  othersourcesthatallowsingleracereporting.Theprimarygoalistoimprovecomparisonsof T 1990andCensus2000racedistributionsatnationalandlowergeographiclevels.Othergoalsare 8 tofacilitatecomparisonsbetweenCensus2000andCensusBureausurveyswhichinstruct  respondentstomarkonerace,andcomparisonswithdatafromthevitalrecordssystem,which  usescensusdata,tocalculatesuchindicatorsasbirthanddeathrates.#XXXOX#OXXXX  #XXXOX8# (B.4)MatchStudyofAccuracyandCoverageEvaluationtoCensustoCompare H ConsistencyofRaceandHispanicOriginResponses(cancelled) ,  Thisevaluationwillnotbeconducted.AcorrespondingExecutiveSteeringCommitteefor t AccuracyandCoverageEvaluationI(ESCAPI)analysisanddocumentationisrelevanttothis X evaluation.RefertoESCAPIreport, A.C.E.ConsistencyofPostStratificationVariables <  (reportB10).  !  (B.5)ContentReinterviewSurvey:AccuracyofDataforSelectedPopulationsandHousing "  CharacteristicsasMeasuredbyReinterview #h!  TheContentReinterviewSurveyutilizesatestretestmethodology,wherebyasampleof $L" householdsdesignatedtoreceivethecensuslongformwerereinterviewedshortlyaftertheyhad %0# beenenumeratedbythecensus.Thesehouseholdsareessentiallyaskedthesamequestions x& $ posedonthelongform.Thentheresponsestothecensusandreinterviewsurveyarecompared. \' % Thissurveyassessesresponsevarianceanderrorthatresultsfromdatacollectionandcapture @(!& operations.d $)"'  d(B.6)MasterTraceSample  +l%* , <hd<   ?+ ` hp x X?XXXXWhilemostevaluationstudieswillprovidedetailedinformationonspecificCensus2000 d operations,theMasterTraceSampledatabasewillprovideinformationthatcanbeusedtostudy#XXXX# XXXX H #XXXXӑ#XXXXvarious#XXXX*#XXXXoperations,alongwithcorrelatesoferroracrossvarioussystems,forarandomlyselected , groupofcensusrecords.Thisdatabasewillcontain,butisnotlimitedto:addresslist   information(e.g.,sourceofaddress),finalvaluesforquestionnaireitemsalongwiththeirvalues   at#XXXXt#XXXXԀ#XXXX #XXXXvarious#XXXXP#XXXXԀstage#XXXX#XXXXs#XXXX#XXXXԀofprocessing,andenumeratorinformation(e.g.,numberofenumeratorattempts  t beforecompletinganinterviewandenumeratorproductionrates).Thisdatabasealsowill  X containinformationaboutthedatacapturesystem(fromrekeyingandreconcilingasubsetof  < MasterTraceSamplequestionnaireimages),theAccuracyandCoverageEvaluation,andthe   ContentReinterviewSurvey.Thisevaluation#XXXX$#XXXXreport#XXXXC#XXXXԀwilldocumenttheprocessofdevelopingthe h  MasterTraceSampledatabase.Itwillincludeinformationonthesourcesofdata,limitations L  withthedata,andsomebasicstatisticsfromthedatabaseitself.Themajorityofresearchand 0  analysisthatwillbeconductedusingtheMasterTraceSampledatabasewillnotbedoneaspart   ofthisevaluation.#XXXX#    (B.7)MatchStudyofCurrentPopulationSurveytoCensus2000 \   UsingtheresultsofapersonlevelmatchofresponsestotheCurrentPopulationSurvey(CPS) @ andCensus2000,thisstudyprovidesadatasetaboutdifferencesbetweentheCensusandSurvey $ estimatesofsocial,demographic,oreconomiccharacteristics.Itsstrengthisitsabilityto l representdifferencesarisingfromnonsamplingvariation.Thestudyfocusesonthedifference P betweenCPSandCensusestimatesofpovertyandlaborforcestatus(whicharemeasured 4 officiallybytheCPS).    h      p      x       ` (B.8)ComparisonsofIncome,Poverty,andUnemploymentEstimatesBetweenCensus | 2000andThreeCensusDemographicSurveys `  ThepurposeofthisevaluationistodeterminetowhatextentCensus2000poverty, D unemployment,andincomeestimatesarecomparablewithestimatesfromtheCurrentPopulation ( Survey,theAmericanCommunitySurvey,andtheSurveyofIncomeandProgramParticipation. p  ThisstudyfocusesonchangesmadetotheCensus2000questionnaireandformsprocessing T systemsthatweredesignedtoimproveunemploymentestimates.Thisevaluationexamines 8  whetherthesechangesbroughttheCensus2000unemploymentestimates(forstates,andfor ! variousdemographicandsocioeconomicgroups)closertotheofficialCurrentPopulation " Surveyestimatesthantheywerein1990.Thisanalysismaybeextendedtocomparedata, "  definitions,andcollectionprocedureswiththeSurveyofIncomeandProgramParticipation. #d! d d(B.9)HousingMeasuresComparedtotheAmericanHousingSurvey(cancelled)  %,#  Thisevaluationwillnotbeconducted.Inearly2002,theCensus2000EvaluationProgramwas t& $ refinedandprioritiesreassessedduetoresourceconstraintsattheCensusBureau.d X' %   d(B.10)HousingMeasuresComparedtotheResidentialFinanceSurvey(cancelled)  *$) |- 5d5    Thisevaluationwillnotbeconducted.Inearly2002,theCensus2000EvaluationProgramwas d  refinedandprioritiesreassessedduetoresourceconstraintsattheCensusBureau. H  (B.11)AmericanCommunitySurveyEvaluationofFollowup,Edits,andImputations    Thisevaluationwillnotbeconductedbecausetheavailabledatacannotanswerthespecified   questionsforthisstudy.  t  (B.12)PuertoRicoRaceandEthnicity  <  Forthefirsttime,inCensus2000,householdsinPuertoRicowereaskedtoanswerquestionson   raceandethnicity.Thesewerethesamequestionsusedonthestatesidequestionnaire.This h  evaluationwillexplorehowthecensusquestionsonraceandHispanicoriginwereansweredby L  peoplelivinginPuertoRico.Specifically,itwillinvestigatewhetherthereareanygeneralizable 0  patternsinhowpeoplerespondedtothesequestionsaccordingtoage,levelofeducation,levelof   income,andresponsemode.Inaddition,itwilllookforpatternsinrespondingtotheHispanic   originquestionbyraceandtotheracequestionbyHispanicorigin.Finally,itwillcomparethe x  patternsofrespondingtothesequestionswiththoseofthegeneralU.S.population.This \  investigationwillbeconductedbypreparingspecialdatatablesfromtheCensus2000datafor @ PuertoRico.Theresultswillbeusedtohelpintheinterpretationofthetabulatedresultsonrace $ andHispanicoriginfortheCommonwealth. l  (B.13) PuertoRicoFocusGroupsontheRaceandEthnicityQuestions 4  ThepurposeofthisresearchistoconductaseriesoffocusgroupsacrossPuertoRicotolearn  moreabouthowpersonsinPuertoRicoviewthesequestions.Theresultsofthefocusgroups  willbeusefulinpreparingforthe2010CensusinPuertoRico.d      `    @* +l%* Ї Thispageintentionallyleftblank.  $H"   5Vd5P fC:DataProducts  d P   `     h      p     Overview , Thefocusof thisevaluation istodetermine theeffectsofdisclosurepreventionmeasureson     Census2000dataproducts. We willexaminethelimitationsandeffectsofdataswappingand  t ourconfidentialityedit!acombinationofstrategiesusedtopreventthedisclosureofdatathat  X canbelinkedtoanindividual!onourdataproducts.  < WhatWillWeLearn? h   Instudyingourdataswappingtechniques,wewillexamineratesfordifferentgeographiclevels 0  andracegroupsanddocumentanyissuesandproblemsthatresultedfrommultipleresponsesto   theracequestion.      `    DataProductsEvaluations  +h%* P 0 <d<   (C.1)EffectsofDisclosureLimitationonDataUtilityinCensus2000 H  ForCensus2000,thedataswappingmethodsfirstusedin1990wererefinedthroughbetter , targetingandexpandedtoincludesampledata.Thisevaluationexaminesvariationsinthe   effectsofswappingdueto:1)aregionsgeographicstructure,2)aregion'sracialdiversity,and   3)thenumberofdimensionsusedintheswapping.  t  (C.2)UsabilityEvaluationofUserInterfaceWithAmericanFactFinder(cancelled)  <  dThisevaluationwillnotbeconducted.Inearly2002,theCensus2000EvaluationProgramwas   refinedandprioritiesreassessedduetoresourceconstraintsattheCensusBureau.Somefocus h  groupswereconductedandtheresultswillbedocumentedinareport, FocusofKeyCustomer L  Segments.Additionally, RevisitingStandardCensusProductsinaNewEnvironment, 0  providespreliminaryresearchfindingsthatwerepresentedatthe2001AmericanStatistical   AssociationConference.Foracopyofthepaper,contacttheCensusBureauon(301)4574218.   Ѐd $'XX$  x   (C.3)DataProductsStrategy(cancelled)  \   dInformationabouttheeffectivenessofthedataproductsstrategywasrolledintoevaluationC.2in @ 2001,whichrecentlywascancelled.Inearly2002,theCensus2000EvaluationProgramwas $ refinedandprioritiesreassessedduetoresourceconstraintsattheCensusBureau.d  l  '    `     h      '    `     h     D:PartnershipandMarketing  *$)  +h%* f1 5Vd5   Overview d  '    `     h   DuringCensus2000,weusednewmethodstopromotecensusawarenessandincreasepublic , cooperation.TheCensus2000PartnershipandMarketingProgramcombinedpublicawareness,   promotion,andoutreachactivitiestogenerateclearandconsistentmessagesaboutthe   importanceofparticipatinginCensus2000.Theplanincorporatedfivecomponents:directmail  t pieces;mediarelations;promotionsandspecialevents(includingCensusinSchools);  X partnershipswithbusinesses,nongovernmentalorganizations,andgovernmententities;andpaid  < advertising.   Theprimarygoalofourcomprehensivemarketingplan,includingthefirsteverpaidadvertising L  campaigncoupledwithanexpandedpartnershipprogram,wastoincreasethemailbackresponse 0  rate,especiallyamonghistoricallyundercountedpopulations.Theadvertisingmarketingstrategy   includedmessagesdeliveredthroughprintmedia,radio,television,andoutofhomemedia   (billboards,busshelters,mobilebillboards).Thepartnershipprogrambuiltpartnershipswith x  state,local,andtribalgovernments,communitybasedorganizations,andtheprivatesector. \  Partnerswereaskedtoassistinthreemajorareas:fielddatacollectionsupport,recruitment,and @ promotion.Inaddition,amajorschoolbasedpublicinformationcampaignwaslaunchedto $ informparentsandguardiansaboutthecensusthroughtheirschoolagechildren.Theplanned l evaluationsforthiscategorywillassesstheeffectivenessoftheseactivitiesaspartofthe P integratedprogramcomponentsoftheCensus2000PartnershipandMarketingProgram. 4 WhatWillWeLearn?  Thesestudieswillhelpusunderstandhowpeoplesattitudes,knowledge,andbehaviorwere ` affectedbytheCensus2000PartnershipandMarketingProgram.Wewillexaminewhich D elementsofthepaidadvertisingmediawerereported/recalledmostoftenbyhardtoenumerate ( groups,andprovidedataforHispanicsandforfiveracecategories:AfricanAmerican,Asian, p  AmericanIndianandAlaskaNative,Hawaiian/PacificIslanders,andWhite.Specifically,we T willlookatwhatimpactthemarketingprogramhadonthelikelihoodofreturningacensusform. 8  Wealsowillcomparethesedatatothe1990Census,whichhadnopaidadvertisingcampaign. ! TheprimarygoalsinstudyingthePartnershipProgramaretomeasurehowwellnationaland " regionalcomponentsaccomplishedtheirobjectivesincommunicatingaconsistentcensus "  messageofprograminitiativesandtodeterminewhichpopulationswerebestservedbythe #d! program.OurevaluationoftheCensusinSchoolsprogramwilltellusabouttheeffectivenessof $H" censuseducationalmaterialsandwhetherteachersreceivingcensusmaterialsincorporatedthem %,# intheircurricula.Overall,wewillanalyzehowwelltheintegratedstrategyoftheCensus2000 t& $ PartnershipandMarketingProgrammetitstwogoals:1)toincreasetheawarenessofthe X' % census,and2)toincreasemailbackresponserates,especiallyamonghistoricallyundercounted <(!& populations.  )"' PartnershipandMarketingEvaluations *#(   (D.1)PartnershipandMarketingProgram +h%* 2 <(~d<   TheCensusBureauhiredtheNationalOpinionResearchCentertoconductanassessmentofthe d marketingandadvertisingcampaignbyfieldingasurveybeforethecampaignbegan,duringthe H educationphaseofthecampaign,andafterthecampaignhadbeenlaunched.Fromthis , evaluation,wewillassessintendedandself-reportedresponsebehaviorandestablishabaseline   andpre-andpost-censusmeasuresofawareness,knowledge,andattitudesofthecensus.We   willobtaintheactualresponsebehaviorforrespondentstooursurvey.Westatisticallymodel  t whateffectself-reportedadvertisingexposurehasonthelikelihoodofrespondingtothecensus.  X Thisevaluationalsoexploresthelinkbetweenraisedawareness,knowledge,attitudes,and  < responsetothecensus.DuetotheintegratedstrategyoftheCensus2000Partnershipand   MarketingProgram,theanalysiswilladdresscomponentsfromtheentireprogram,not h  individuallyforpaidadvertisingandpartnerships. L   (D.2)CensusinSchools/TeacherCustomerSatisfactionSurvey    TheCensusBureauhiredMacroInternationaltoconductapost-censussurveyofschoolteachers   toassesstheeffectivenessofCensuseducationalmaterialsandwhetherteachersreceivingcensus x  materialincorporatedthemintheircurricula. \   (D.3.)SurveyofPartners/PartnershipEvaluation $  WewillassessthehelpfulnessofCensus2000materialstopartners,thetypesandvalueof l servicesrendered,thespecificpartnershipactivitiesconducted,andtheeffectivenessofthe P programinreachingthehardtoenumeratepopulation.Wealsowillobtainfrompartnersthe 4 organizationalcostsincurredtosupportandpromoteCensus2000.Thesampleofpartnerswill  beselectedusingtheContactProfileandUsageManagementSystemdatabase.Westat,an  independentcontractor,washiredtoconductthissurvey.Toimproveresponseamongthe | geographicallydispersedanddiversepartnersinthePartnershipProgram,thisevaluationusesas ` datacollectionmodesamailout/mailbackquestionnairewithanonresponsefollowuptelephone D interview. ( Ѐ  p    5*#5  E:SpecialPlacesandGroupQuarters d *#( TOverview +h%* 3 <d<   dThevastmajorityofU.S.residentsliveasfamiliesorindividuallyinhouses,apartments,mobile H homes,orotherplacescollectivelyknownas housingunits.However,therearemillionsof , peopleintheUnitedStateswholiveingroupsituationssuchascollegedormitories,nursing   homes,convents,grouphomes,migrantworkerdormitories,andemergencyandtransitional   shelters .Ourevaluationswillanalyzetheeffectivenessofprocedurestoenumeratepersons  t livingindifferenttypesofgroupquarters .  X ЀTheCensusBureaudevelopedaspecializedoperationtoenumerateselectedservicelocations   thatservedpeoplewithoutconventionalhousing.Theservicebasedenumerationoperationwas h  conductedfromMarch27toMarch29,2000,atshelters,soupkitchens,mobilefoodvans,and L  targetednonshelteredoutdoorlocations. 0  Somestudieswillfocusonsuchthingsasenumerationat servicebasedlocations(sheltersand   foodfacilitiesforthehomeless;outdoorlocationswherehomelesspeoplesleep).Major x  evaluationsareplannedfortwooperationsdesignedtoenhancetheaddresslistofspecialplaces: \  theSpecialPlaceFacilityQuestionnaireandtheSpecialPlaceLocalUpdateofCensus @ Addresses. $ WhatWillWeLearn? P Thefindingsfromtheseevaluationswillanswerimportantquestionsonhoweffective  enumerationprocedureswereinobtainingthecountforgroupquarters.Wewill comparethe  telephoneandpersonalvisitoperationsoftheFacilityQuestionnaire. Theevaluationswill | includedistributionsof thegroupquarterspopulationsbytypeofgroupquarters,countsof ` personsatgroupquartersonCensusDaywhoindicatedausualhomeelsewhere,andcomparison D ofthepredictedgroupquartersuniversefromtheFacilityQuestionnaireoperationwiththegroup ( quartersuniverseasenumerated.d p   SpecialPlacesandGroupQuartersEvaluations  )"' H.` hp x (#'XH d(E.1.a)SpecialPlace/GroupQuartersFacilityQuestionnaireOperationalAnalysis *$) (cancelled)  +h%*  Thisevaluationwillnotbeconducted.Inearly2002,theCensus2000EvaluationProgramwas d refinedandprioritiesreassessedduetoresourceconstraintsattheCensusBureau. H B.` hp x (#XB (E.1.b)FacilityQuestionnaireCATIandPV    Thisevaluationusedpersonalvisitreinterviewsatasampleofspecialplacestoassessthe   accuracyoftheinformationcollectedfromtheFacilityQuestionnaireviacomputerassisted  t telephoneintervieworpersonalvisit.Thisevaluationwilladdresshowoftenchangesoccurin  X thespecialplacetypecodeandwhetherclassificationdiscrepanciesdifferbytypeofspecial  < place.    X?+ ` hp x (#X?  h    XB.` hp x (#XB (E.2)SpecialPlaceLocalUpdateofCensusAddresses  L  ThisevaluationfocusesonlocalgovernmentsparticipationintheSpecialPlaceLocalUpdateof 0  CensusAddresses.Itwilldocumentchangestotheaddresslistalongwithoperationalissuesthat   wereencountered.   B.` hp x (#XB (E.3)AssesstheInventoryDevelopmentProcessforServiceBasedEnumerations \   Thepurposeofthisstudyistoevaluatethereturnontheeffortsmadetocompiletheinventoryof @ ServiceBasedEnumerationsplacesandaddressesthatwereincludedinCensus2000.Thestudy $ willlookatvarioussourcesthatprovidednamesandaddressesandtheresultsfromenumeration l todeterminewhichsourcesprovedtobemorereliable P  (E.4)DecennialFrameofGroupQuartersandSources   ?+ ` hp x X?XXXXThisstudyevaluatesthecontent,coverage,andsourcesoftheDecennialFrameofGroup  Quartersbycomparingeditionsandrecordstoindependentsources,notablythecontemporary | BusinessRegister.Thisevaluationexaminesthefeasibilityandconstraintstoenrichorintegrate ` theseframes.#XXXX"# D B.` hp x (#XB (E.5)GroupQuartersEnumeration p   Thisstudywilldocumentvariousaspectsofthegroupquartersenumeration.Someofthetopics T coveredbythisstudyincludethetotalcountofthegroupquarterspopulation,thenumberof 8  specialplacesthatwereenumerated,andthenumberofgroupquartersthatwereenumerated. ! Additionally,thenumericaldistributionofgroupquartersperspecialplaceandofresidentsper " groupquarterswillbedocumented. "   (E.6)ServiceBasedEnumeration  $H" ThegoaloftheServiceBasedEnumeration(SBE)wastoenumeratepeoplewithouthousing %,# whomayhavebeenmissedinthetraditionalenumerationofhousingunitsandgroupquarters.A t& $ completeenumerationofshelters , soupkitchens,regularlyscheduledmobilefoodvansand  X' % targetednonshelteredoutdoorlocationswasconductedinMarch2000.Thisevaluationwill d documentdatacollectioncompleteness,partialinterviews,andwhethertheSBEunduplication H processsuccessfullyidentifiedindividualswhowereenumeratedmorethanonce.d ,    `     h      p      x   )"' Thispageintentionallyleftblank. X' %   <(!&  B.` hp x (#XB   @  @` 5d5F:AddressListDevelopment  d Overview , dTheseevaluationscoverabroadspectrumofactivities,bothinternalandexternal,involvedwith   buildingaddressfilesandtherelatedgeographicdatabase,includingfieldoperationsfromwhich  t addressinformationandrelatedmapupdatesaregathered.Theaddresslistdevelopment  X categoryincludesvariousevaluationsoftheCensusBureausMasterAddressFile(MAF),and  < theTopologicallyIntegratedGeographicEncodingandReferencing(TIGER)database.These   includeexaminationofthecompletenessandaccuracyofaddressinformationintheMAF,as h  wellasofthedesignoftheMAFandDMAF.AnevaluationoftheU.S.PostalServices L  DeliverySequenceFileusedintheMAFbuildingprocessisalsoplanned.Avarietyofcensus 0  fieldandlocal/tribalpartneroperationswillbeevaluatedtomeasuretheimpactofeachoperation   ontheMAFandtheTIGERdatabase.Theseinclude,butarenotlimitedto:AddressListing,   BlockCanvassing,Update/Leave,List/Enumerate,andmultiplecyclesoftheAddressList x  Review(alsoreferredtoastheLocalUpdateofCensusAddresses).Combined,thesefield \  operationsoffercomprehensiveaddresschecksinruralandurbanareasandareaprimarysource @ ofaddressinformationusedforMAFandTIGERdatabaseenhancement.Additionalevaluations $ focusonthegeocodingaccuracyofaddressesinthecensus. l WhatWillWeLearn? 4 Thefindingsfromtheaddresslistdevelopmentevaluationswillprovideinsightintothemost  accuratemethodsforupdatingtheMAFandtherelatedTIGERdatabase.Thisincludes | understandingtheindividualcontributionofeachoperationasitisimplemented.Foreach ` operation,wewilllookatthecharacteristicsofaddressesthatwereadded,corrected,orflagged D fordeletion.Wealsowilllookatthegeographicimpactofeachoperation(i.e.,wewillexamine ( howchangestotheMAFaredistributedgeographically).Additionally,wewilllearnsome p  thingsabouttheoverallhousingunitcoverageinthecensus.Finallywewilllearnmoreabout T qualityandcoveragebyexaminingaddressesthatareonthefullMAF,butwerenotincludedin 8  thecensusforvariousreasons.AlloftheseevaluationswillhelpinformcontinuedMAFand ! TIGERdatabaseupdatingthroughthedecadeandalsowillprovideinsightforthe2010Census " andtheAmericanCommunitySurvey. "  d AddressListDevelopmentEvaluations +h%* 8 <d<    '(xx (F.1)ImpactoftheDeliverySequenceFileDeliveriesontheMasterAddressFileThrough H Census2000Operations ,  TheDeliverySequenceFile(DSF)isafileofaddressesproducedandmaintainedbytheU.S.   PostalService.TheCensusBureauusesthisfile,alongwiththe1990Censusaddresslistand   otherinformation,tocreateapermanentnationaladdresslistcalledtheMasterAddressFile  t (MAF).ForCensus2000,theCensusBureauusedtheDSFasaprimarysourcetoenhancethe  X initialMAFformailout/mailbackareasofthecountry.SubsequentDSFswereusedtoupdate  < theaddresslistthroughAprilof2000,inordertomaximizetheinclusionofallexisting   addressesinthecensus.ThisevaluationwillassesstheimpactofeachoftheDSFsthrough h  Census2000operationsbyprofilingthenumberandcharacteristicsofhousingunitsaddedtoand L  deletedfromtheMAFfollowingeachdeliveryoftheDSF. 0   (F.2)AddressListingOperationanditsImpactontheMasterAddressFile    ForCensus2000,anAddressListingOperationwasusedinupdate/leaveareasofthecountryto x  createtheinitialMasterAddressFile(MAF)andprovideacomprehensiveupdateofthe \  streets/roadsandtheirnamesintheTIGERdatabase.Inthisoperation,inareaswherecensus @ questionnairesweresubsequentlyhanddelivered,censusenumeratorswentdoortodoorto $ identifythemailingaddressandphysicallocationofeveryhousingunit.Theyalsoverifiedand l updatedthelocationandnamesofgeographicfeaturessuchasstreets.TheCensusBureauused P thisprocedureinordertocreateafileofgoodlocatableaddressesforCensusBureaufield 4 operationsinCensus2000aswellasitsfuturedemographicsurveys,includingtheAmerican  CommunitySurvey.ThisevaluationwillassesstheimpactoftheCensus2000AddressListing  OperationontheMAFbyprofilingthenumberandcharacteristicsofhousingunitsaddedtothe | MAF. `  (F.3)LocalUpdateofCensusAddresses1998 (  TheLocalUpdateofCensusAddresses(LUCA)operation(alsoknownasAddressListReview) p  forCensus2000includedaLUCA98operationthatfocusedonmailout/mailbackareas.For T thisoperation,localandtribalgovernmententitieswereprovidedaCensusBureauaddresslist 8  containingaddressesderivedfromtheDeliverySequenceFileandthe1990AddressControl ! File.TheobjectiveoftheLUCAoperationswastoprovidelocalentitiestheopportunityto " reviewtheBureausaddressinformationandrelatedmapsandthenprovidefeedbackintheform "  of1)addressadds,deletesandcorrectionsand2)streetandstreetnameadds,deletions,and #d! correctionsonthemaps.TheCensusBureaucomparedtheresultstotheblockcanvassing $H" resultsinmailout/mailbackareas,andalldiscrepancieswerefieldverified.AfterCensusBureau %,# reviewofsubmissions,localandtribalentitiesweregiventheopportunitytoreviewresultsand t& $ toappealsituationsinwhichtheybelievedtheMasterAddressFile(MAF)stillwasincomplete X' % orincorrect.ThisevaluationwillassessthenumberandprofileofhousingunitaddstotheMAF, <(!& theextentofgeographicclusteringoftheseadds,andthetotalnumberandprofileofhousingunit  )"' deletionsandcorrections. *#(   *$) Theevaluationalsowillincludeinformationdocumentingtheparticipationratesoflocaland d  tribalgovernmentsandtheproportionofaddressescoveredbythesegovernments. H  q(U8x' @*d(F.4)EvaluationoftheCensus2000MasterAddressFileUsingEarlierEvaluationData   (cancelled)    Thisevaluationwillnotbeconducted.Inearly2002,theCensus2000EvaluationProgramwas  t refinedandprioritiesreassessedduetoresourceconstraintsattheCensusBureau.  X  (F.5)BlockCanvassingOperation    Forthe1990Census,theCensusBureauconductedanoperationcalledPrecanvasstoimproveits h  addresslistformailout/mailbackareas.ForCensus2000,asimilaroperation,calledBlock L  Canvassing,wasimplemented.Aswiththe1990Precanvass,thisoperationwasconducted 0  primarilyinareaswherecitystyleaddressesareusedformaildelivery;however,forCensus   2000,theBlockCanvassingOperationcoveredalargergeographicareathandidthe1990   PrecanvassOperation,andthescopeoftheoperationwasexpandedtoincludemap(i.e.TIGER x  database)updates.TheobjectiveofthisevaluationistodeterminetheoveralleffectoftheBlock \  CanvassingOperationontheMasterAddressFile(MAF)bymeasuringthenumberand @ characteristicsofhousingunitadds,deletes,andcorrectionstotheMAF. $  (F.6)LocalUpdateofCensusAddresses1999 P  TheLocalUpdateofCensusAddresses(LUCA)operation(alsoknownasAddressListReview) 4 forCensus2000includedaLUCA99operationforUpdate/Leaveareas.ForLUCA99,local  andtribalgovernmententitieswereprovidedwithcensushousingunitblockcountsthatwere  createdusingaddressesobtainedfromtheAddressListingOperation.Participatingentitieswere | askedtoreviewthecountsandprovidefeedbackwhentheybelievedthenumberofhousingunit ` addressesfortheblockshouldhavebeenhigherorlower.Participatinggovernmentscould D challengeblockcounts,butcouldnotprovidespecifichousingunitadds,corrections,ordeletes. ( BlocksthatwerechallengedweresenttoLUCA99FieldVerificationforrelisting,thenreturned p  toparticipatinggovernmentsforanotherreview.Thisevaluationwilldocumenttheparticipation T ratesofthosetribalandlocalgovernmentsthatwereeligibletoparticipate,theproportionof 8  addressescoveredbythosegovernments,thenumberofblocksthatwerechallengedandwentto ! LUCA99FieldVerification,andtheextenttowhichchangesoccurredduringthefield " verification. "   (F.7)CriteriafortheInitialDecennialMasterAddressFileDelivery(cancelled) $H"  Thisevaluationwillnotbeconducted.Inearly2002,theCensus2000EvaluationProgramwas %,# refinedandprioritiesreassessedduetoresourceconstraintsattheCensusBureau. t& $  (F.8)TheDecennialMasterAddressFileUpdateRules(cancelled) <(!&  Thisevaluationwillnotbeconducted.Inearly2002,theCensus2000EvaluationProgramwas  )"' refinedandprioritiesreassessedduetoresourceconstraintsattheCensusBureau. *#(    `     h    +h%*  (F.9)NewConstructionAdds(cancelled) d  Thisevaluationwillnotbeconducted.Inearly2002,theCensus2000EvaluationProgramwas H refinedandprioritiesreassessedduetoresourceconstraintsattheCensusBureau.Someofthe , plannedresearchforthisstudywillbeevidentinevaluationI.4, CoverageImprovement   Followupd.    w(08q d(F.10)Update/Leave  \  TheUpdate/Leaveoperationwasconductedinareaswheremaildeliveryofquestionnaireswas  @ problematic.Fieldstaffdependentlycanvassedtheirassignedarea,updatedtheaddresslistand  $ map,anddistributedaquestionnairetoeachhousingunit.Thisevaluationwilldocument l  addresscorrections,addedunits,andunitsflaggedfordeletionduringtheoperation.Wealso P  willstudyproblemreferralformscompletedbyenumeratorsfordifficultlistingsituations(e.g., 4  unabletoobtainaccess,gateblocked,roadwashedaway,notrespassingsigns),toseehowwell   thesesituationswerefollowedthroughandhowtheymighthavecontributedtocoverageerrors.    (F.11)UrbanUpdate/Leave  `  TheUrbanUpdate/Leavewasanoperationthattargetedwholecensusblocksandwasconducted D inareaswheretheCensusBureauwasnotconfidentthattheaddressedquestionnaireswillbe ( deliveredtothecorrespondinghousingunits.ForCensus2000,eightofthe12RegionalCensus p  Centersidentifiedblocksforthisoperation.TheCharlotte,KansasCity,LosAngeles,andNew T YorkRegionalCensusCentersdecidedtouseotherspecialenumerationmethodologiesinlieuof 8 UrbanUpdate/Leave.Thisevaluationwillassessthenumberofaddressesadded,deleted,  corrected,andmovedasaresultofUrbanUpdate/Leave.Itwillprofilethehousingunit  addressesasfollows:typeofaddress,single/multiunit; drop/nondropdelivery. Delivery  SequenceFilematch/nonmatch.Itwillalsolookattheaddressesintermsofoccupancystatus d andwilldescribethepersonsinUrbanUpdate/Leaveaddressesbysex,age,Hispanicorigin,and H race.  , (F.12)Update/Enumerate X  Update/EnumerateissimilartoUpdate/Leave,exceptthatinterviewersenumeratedtheunitatthe <  timeoftheirvisitratherthanleavingaquestionnairetobecompletedandmailedback.The  ! operationwasconductedincommunitieswithspecialenumerationneedsandwheremost " housingunitsmaynothavehousenumbersandstreetnameaddresses.Theseareasincludesome "  selectedAmericanIndianReservationsandtheColonias.Update/Enumeratewasimplemented #h! inresortareaswithhighconcentrationsofseasonallyvacanthousingunits.Most $L" Update/Enumerateareasweredrawnfromaddresslistedareas,butsomecamefromblock %0# canvassareas.Thisevaluationwilldocumentthenumberandcharacteristicsofhousingunits x& $ added,deleted,corrected,andmovedinUpdate/Enumerateareas. \' %   (F.13)List/Enumerate +l%*  List/Enumeratewasanall-in-oneoperationconductedinsparselypopulatedareasofthecountry. d Theaddresslistwascreatedandthehousingunitsenumeratedconcurrently.Themainobjectives H ofthisevaluationwillbetoprofilealladdressesproducedbytheList/Enumerateoperation,as , wellastospecificallyprofiletheList/EnumerateaddressesthatmatchedtotheDelivery   SequenceFile.d    (F.14)OverallMasterAddressFileBuildingProcessforHousingUnits(cancelled)  X  dThisevaluationwillnotbeconducted.Inearly2002,theCensus2000EvaluationProgramwas  < refinedandprioritiesreassessedduetoresourceconstraintsattheCensusBureau.Objectivesof   thisevaluationwillbemetwiththeAddressListDevelopmenttopicreport,whichwill h  synthesizeresultsfromevaluations,experiments,andincludesomenewanalysisacrossallofthe L  majorsourcesandoperationsthatcontributedtotheMasterAddressFileinCensus2000. 0  d (F.15)QualityoftheGeocodesAssociatedWithCensusAddresses    Theobjectiveofthisevaluationistomeasurethequalityofresidentialaddressgeocodingin x  Census2000andtoidentifythesourceofthegeocode(i.e.,theTIGERdatabase,oneofthe \  severalfieldoperations,LUCA/NewConstructionparticipants,etc.). @  (F.16)BlockSplittingOperationforTabulationPurposes  l BlockSplitoperationsareconductedbytheCensusBureautoprovidefortabulationofdata P wheregovernmentalunitandstatisticalareaboundariesdonotconformtocollectionblock 4 boundaries.Thisevaluationwillmeasuretheaccuracyofblocksplittingoperationsfor  tabulationpurposes.   ff   `     h   @*   `     h     Thispageintentionallyleftblank. h  X' %     G:FieldRecruitingandManagement  +h%* f= 5Vd5   78   8  Overview H Promptedbythedifficultiesinrecruitingapplicantsandhighturnoverofemployeesinthe1990   Census,theCensusBureauredesigneditsrecruitment,staffing,andcompensationprogramsfor   Census2000.Severalnewprogramsweredevelopedtoaddressthe1990issuesandtohelpthe  t CensusBureausuccessfullyrecruitseveralmillionapplicants,hireseveralhundredthousand  X employees,andretainthisstaffthroughthedecennialcensus.Someoftheseprogramsincluded  < frontloading,higherpayrates,andpaidadvertising.   K7X2` hp x (#XKWhatWillWeLearn? L  Thepurposeofthisevaluationistostudytheeffectsofthesenewprogramactivitiesupon   recruitment,staffing,andretention.Acontractor,forexample,determinedthatthe1990District   Office(nowLocalCensusOffice)payrateswerenotadequatelysettoattractandretainstaff x  whencomparedtolocaleconomicconditionsofthatarea.ThemethodologytosettheCensus \  2000payrates,basedonthisknowledge,wasrevisedandsettoaderivativeofthelocal @ prevailingpayrate.Theeffectivenessofthishigherpayratewillbeevaluated,aswellasother $ recruitmentandhiringprograms(suchasfrontloadingandpaidadvertising).  l FieldRecruitingandManagementEvaluations *$)  +h%* > <hd<   (G.1)Census2000StaffingPrograms d  ThisevaluationexaminestheeffectivenessoftheCensus2000hiringprogramsduring H NonresponseFollowup(NRFU).Studyquestionswillfocusupontheeffectivenessofthehigher , payrateprogram,frontloading,paidadvertising,andotherareas.Someofthequestionsare:   1)wastheCensusBureauabletoadequatelyhireandattractstafftoexecuteNRFU,Accuracy   andCoverageEvaluation,andothervariousfieldoperations;2)werethepayrateseffectivein  t attractingandretainingstaffneededforCensus2000NRFU;and3)didrecruitingactivities  X provideanadequatesupplyofapplicantsandreplacements.Aportionofthisstudywillexamine  < theeffectivenessofthehigherpayratesonproductivityandevaluatethepaymodelasapredictor   oflocaleconomicconditions. h   (G.2)OperationControlSystem 0   ThisevaluationhasbeenmovedtoEvaluationCategoryR,AutomationofCensusProcesses;see   R.2.a,OperationsControlSystem2000,SystemRequirementsStudyforitsdescription.     '(x0w  2    `     h    H:FieldOperations edF *#( qFOverview +l%* f? 5Vd5   H( 4 <DLX2XHXXXXThiscategoryincludesstudiesofvariousfieldoperationsandstrategieswhosegoalsweretocurb H questionnairedeliveryandenumerationproblems,andobtaincensusdatafromindividualswho , didnotrespondtothecensusbyaspecifieddate.Forexample,theNonresponseFollowup   operationconsistedofsendinganenumeratortocollectcensusdatafromeveryaddressfrom   whichnomail,telephone,orInternetresponsewasreceived.Evaluationsinthiscategorywill  t analyzewhetherfieldoperationswereconductedasplannedandwillassesstheireffectiveness.  X Additionally,operationalresultswillbedocumentedforeachLCOforhistoricalpurposes.  < Analysesinthiscategoryalsowillexamineoureffortstocountthosecategorizedashardto h  enumerate.Ourtargetingmethodologiesconsistedof1990personandhousingunitcensusdata L  thatareindicatorsofnonresponseandthepotentialtobeundercounted.Thisinformation 0  assistedtheRegionalCensusCentersindeterminingtheplacementofQuestionnaireAssistance   CentersandBeCountedForms.Theinformationalsoassistedparticipantsofourpartnership   program.Studiesinthiscategorywillevaluatehowsuccessfulourtargetingmethodologieswere x  alongwiththeusageofQuestionnaireAssistanceCenters.Inaddition,wewillevaluateour \  targetedenumerationmethodssuchasblitzenumeration(useofagroupofenumeratorsto @ conductenumerationinacompressedtimeframe),teamenumeration(twoenumeratorsworking $ togetherwheresafetyisaconcern),andtheuseoflocalfacilitators(longtimeneighborhood l residentsorchurchleaderswhoassisttheenumeratoringainingentrytotheneighborhood). P Becausesomerespondentswereabletoprovidedatawithoutacensusidentificationnumber  (e.g.,BeCountedandTelephoneQuestionnaireAssistance),itwaspossiblethatrespondents  submittedaddressesthatwerenotonourMasterAddressFile.Weconductedafieldverification | ofthesetypesofaddresses.Ifanenumeratorverifiedthattheaddresswasavalidhousingunit, ` thenitwasaddedtotheDecennialMasterAddressFile.Wealsowillconductanevaluationof D theeffectivenessofthisoperation. (   WhatWillWeLearn? T   Theresultsoftheseevaluationswillgiveusanindicationofhowsuccessfulwewereatobtaining ! datafromnonrespondentsincludingthoselivinginareaswhereweemployedspecial " enumerationmethodologies,andhowtobetterplanthesetypesofoperationsforfuturecensuses. "  TheevaluationofNonresponseFollowupwillreportproxyrates,thenumberofpartial #d! interviews,vacancyrates,andthenumberofunitsenumeratedduringfinalattemptprocedures, $H" whichwillhelpustoassesswhethertheoperationwasconductedasplanned. %,# Otheranalyseswillprovideinformationaboutthequalityofourenumeratortrainingprogram, X' % theutilityofourtargetingmethods,andaprofileofLocalCensusOfficeswhichwillcontain <(!& variousdescriptivestatistics.#XXXXG#   )"' @ <Vdd<   FieldOperationsEvaluations d   (H.1)Useof1990DataforCensus2000Planning(cancelled) ,  dThisevaluationwillnotbeconducted.Inearly2002,theCensus2000EvaluationProgramwas   refinedandprioritiesreassessedduetoresourceconstraintsattheCensusBureau. Usingthe   PlanningDataBasetoPlanDataCollectionforCensus2000,providespreliminaryresearch  t findingsthatwerepresentedatthe2001AmericanStatisticalAssociationConference.Fora  X copyofthepaper,contacttheCensusBureauon(301)4574218.d  <  (H.2)OperationalAnalysisofFieldVerificationOperationforNonIDHousingUnits h   NonIDquestionnairesarethosefromtheBeCountedandTelephoneQuestionnaireAssistance L  operationsorquestionnairesforwhichanenumeratorwasnotabletoverifythattheaddress 0  existed.Duringfieldverification,enumeratorsvisitedthelocationofthesenonIDhousingunits   andverifiedtheirexistenceonthegroundbeforetheywereaddedtotheMasterAddressFile   (MAF)/DecennialMasterAddressFile(DMAF).ForCensus2000,nonIDquestionnairesthat x  weregeocodedtoacensusblock,butdidnotmatchtoanaddressalreadyintheMAFwere \  assignedforfieldverification.Thisoperationalanalysiswillattempttoanswerquestionssuchas @ howmanyunitswereaddedtotheMAF/DMAFafterverificationandifoperationalproblems $ wereencounteredduringtheimplementationoffieldverification. l  (H.3)LocalCensusOfficeDeliveryofCensus2000MailoutQuestionnairesReturnedby 4 U.S.PostalServiceWithUndeliverableasAddressedDesignation(cancelled)   dThisevaluationwillnotbeconducted.Inearly2002,theCensus2000EvaluationProgramwas  refinedandprioritiesreassessedduetoresourceconstraintsattheCensusBureau. | d (H.4)QuestionnaireAssistanceCentersforCensus2000  D TheCensusBureauprovidedwalkinassistancecenterswhererespondentsreceivedassistance ( withcompletingtheirquestionnaire.Languageassistanceguideswereavailableinover40 p  differentlanguages,alongwithBeCountedformsthatwereavailableinEnglishandfiveother T languages.ThisstudywilldocumentvariousdaspectsoftheQuestionnaireAssistanceCenters 8  (QACs)suchaslocation,employees,andtypesofassistanced.Inaddition,thefrequencyofuseof ! theQACswillbeanalyzed. "  (H.5)NonresponseFollowupforCensus2000 #d!  Thisoperationwasconductedforallhousingunitsinthemailout/mailbackandupdate/leave $H" areasforwhichtheCensusBureaudidnotcheckinaquestionnairebyApril11,2000.During %,# NonresponseFollowup(NRFU),enumeratorsvisitedeachnonrespondingunittodeterminethe t& $ occupancystatusoftheunitonCensusDayandtocollecttheappropriatedata(i.e.,longformor X' % shortform)forthehouseholdmembers.Theobjectiveofthisanalysisistodocumentvarious <(!& aspectsoftheNRFUoperations.Someofthetopicscoveredinthisstudyincludedetermination  )"' ofNRFUworkloads,identificationofthedemographicsofthoseenumeratedinNRFU,and *#( documentationofthenumberofdNRFUEnumeratorQuestionnairesthatwerepartialinterviews, *$) refusals,completedviaproxyrespondents,orcompletedduringfinalattemptprocedures.The +h%* A <@d <   percentofNRFUunitsclassifiedasoccupied,vacant,ordeletewillbedocumented. d Additionally,thisevaluationwilldeterminewheneachLocalCensusOffice(LCO)startedand H completedtheirNRFUoperationandthecostoftheoperation.d ,  (H.6)OperationalAnalysisofNonTypeofEnumerationAreaToolKitMethods    Toolkitmethodswerespecialenumerationprocedures(e.g.,blitzenumeration,andtheuseof  t localfacilitators)availableforimprovingcooperationandenumerationinhardtoenumerate  X areas.Forthisevaluation,theCensusBureauwillassessthetoolkitmethodsused,wherethey  < wereused,andtheeffectivenessandfeasibilityofthetoolkitmethods.    (H.7)NonresponseFollowupEnumeratorTraining L   DuringCensus2000,wehiredover500,000peopletofilltemporarypositions.Thelargest 0  numberoftheseworkerswerehiredfortheNonresponseFollowup(NRFU)operation.Adequate   employeetrainingwascriticaltothesuccessofNRFU.Theoverallobjectiveofthisevaluation   istoexaminethequalityoftheNRFUenumeratortrainingprogramaswellastheenumerators x  stateofpreparednessfollowingtraining. \   (H.8)OperationalAnalysisofEnumerationofPuertoRico $  Census2000wasthefirsttimethatanUpdate/Leavemailbackmethodologywasusedtoconduct l theenumerationinPuertoRico.Thisevaluationwilldeterminehowmanyaddresseswere P encompassedbythisenumerationmethodology,aprofileoftheaddresses,andwhatoperational 4 problemswereencounteredinthefieldasaresultofaddresslistcompilationandprocessing  procedures.ThisstudyalsowillmakecomparisonstostatesideUpdate/Leavedata.   (H.9)LocalCensusOfficeProfile `  ThisoperationalsummarywillprovidedescriptivestatisticsattheLocalCensusOffice(LCO) D levelformanycensusoperations.Forexample,totalhousingunits,averagehouseholdsize,and ( mailreturnratewillbeamongthestatisticsreportedforeachLCO. p   (H.10)DateofReferenceforRespondentsofCensus2000  8  TheCensus2000questionnairestatedthattherespondentshouldreportageasofApril1,2000. ! Thisstudywilldocumenttheaveragedateofreferenceusedbycensusrespondentsandthe " averagedateofreferencebymethodofenumeration.Thisstudyalsowilldocumentvarious "  typesofdiscrepanciesbetweendateofbirthandreportedage. #d! Ѐ    `     h      p      $H"   %,#   5d 5  I:CoverageImprovement Rms d s Overview , ThecoverageimprovementevaluationsexaminevariousCensus2000operationsthatare   intendedtoimprovethecoverageofbothhousingunitsandpeopleinthecensus.Followingthe  t mailbackeffortstocompletethecensus,aseriesofoperationswereconductedtoensurethat  X peoplewerecountedattheircorrectCensusDayaddress,toconfirmthestatusofhousingunits  < thatweredeletedorenumeratedasvacant,andtoensuretheinclusionofallpersonsina   householdwhenthereturnedformshoweddiscrepanciesinthenumberofpersonsenumerated. h  WhatWillWeLearn? 0  Fromtheseevaluationswewilllearnabouttheeffectivenessofthesevariousoperationsasthey   attempttoimprovecensuscoverage.FromtheNonresponseFollowupoperation,wewill x  examinethepotentialcoveragegainfromidentifyingmoversandcheckingtoseeiftheywere \  countedattheirCensusDayaddress.Wewillalsoanalyzethesituationswhereentire @ householdswereidentifiedashavinga usualhomeelsewhere.FortheCoverageImprovement $ Followup,wewillexaminethepersonandhousingunitcoveragegainsfromthisoperation, l whichdeterminedtheCensusDaystatusofcertaintypesofhousingunits(mostofwhichwere P identifiedasdeletesorcodedasvacantsinearliercensusoperations).Theevaluationofthe 4 CoverageEditFollowupwillmeasurecoveragegainsfromthisoperation,whichconsistedof  contactinghouseholdswhosecompletedformsshoweddiscrepanciesregardingthenumberof  personsenumerated,orwhosecompletedformindicatedthereweremorethansixpersonsinthat | household.Furthermore,wewillevaluatethecoveragequestionsontheenumerator ` questionnairetodeterminehowwellenumeratorsaskedthesequestionsandusedtheanswersto D obtainanaccuratehouseholdroster. (  CoverageImprovementEvaluations +h%* C <d <    (I.1)CoverageEditFollowupforCensus2000 H  TheCoverageEditFollowup(CEFU)wasdesignedtoincreasewithinhouseholdcoverageand , improvedataqualityintwoways.Astandardquestionnaireonlyhasroomforsixpersons,so   CEFUwasusedtocollectdataonadditionalpersonsinlargehouseholds.Second,itresolved   discrepanciesonmailreturnformsbetweenthereportedhouseholdsizeandtheactualnumberof  t personsforwhichdatawererecordedonthecensusform.Anattemptwasmadetoresolveall  X householdsthatfailededitsforthesesituationsbyusingaComputerAssistedTelephone  < Interview.Thisanalysiswilldocumenttheworkload,operationalaspects,andcoveragegains   fromconductingthisoperation.  h   (I.2)NonresponseFollowupWholeHouseholdUsualHomeElsewhereProbe 0   DuringtheNonresponseFollowup(NRFU),List/Enumerate,andUpdate/Enumerateoperations,   enumeratorsaskedrespondentswhethertheiraddresswasaseasonalorvacationhomeandifthe   wholehouseholdhadanotherplacewheretheylivedmostofthetime.Whenrespondents x  indicatedtheyhadausualhomeelsewhereonCensusDay,enumeratorsrecordedcensus \  informationaboutthisonablankSimplifiedEnumeratorQuestionnaire(SEQaversionofthe @ mailreturnquestionnairethatiseasiertouseforpersonalvisitenumeration)andenumeratedthe $ currentaddressasavacantunitorobtainedinformationaboutthepeoplelivingthereonCensus l Day.ThisevaluationexamineshowoftenSEQswerecompletedasWholeHouseholdUsual P HomeElsewhere(WHUHE);howmanyoftheseaddresseswerematchedtoanaddressonthe 4 DecennialMasterAddressFile(DMAF);howoftenaddressescouldneitherbematchedtothe  DMAForgeocoded;formatchedaddresses,howoftenwastheWHUHEhouseholdalready  includedonthecensusformfortheirususalplaceofresidence;andhowoftendidwefinda | differenthouseholdonthecensusquestionnaire.  `   (I.3)NonresponseFollowupMoverProbe (  InCensus2000,inmovers(householdsthatmovedthereafterCensusDay)wereidentified p  duringtheNonresponseFollowup(NRFU),List/Enumerate,andUpdate/Enumerateoperations T andwereaskediftheywereenumeratedattheirCensusDayaddress.Ifarespondentdidnot 8  recallcompletingacensusformattheirCensusDayaddress,theenumeratorcompleteda ! questionnairefortheinmoverhouseholdusingtheirCensusDayaddress.Thisevaluationlooks " athowmanyofthesecasesoccurred,andhowmanypersonswereaddedtothecensusasaresult "  ofthisprocedure. #d!  (I.4)CoverageImprovementFollowup %,#  TheCoverageImprovementFollowup(CIFU)universeconsistedofunitsclassifiedasvacantor t& $ deletedinNRFU,addsfromthenewconstructionoperation,lateaddsfromUpdate/Leave,blank X' % mailreturns,andlostmailreturns,ifany.DuringCIFU,enumeratorsvisitedtheseunitstoverify <(!& theCensusDaystatusandcollectpersonandhousingunitdataasappropriate.Thisevaluation  )"' willdocumentthepersonandhousingunitcoveragegainfromconductingtheCIFU,including *#(  thenumberofunitsthatchangedstatusfromvacanttooccupiedorfromdeletetoeithervacantor *$) occupied.Thisstudyalsowillexaminethecharacteristicsofpersonsandhousingunitsaddedas d aresultoftheCIFU,start/finishdates,andthecostoftheoperation. H  (I.5)CoverageGainfromCoverageQuestionsonEnumeratorCompletedQuestionnaire    In1990,enumeratorsbegantheirinterviewwithanexplanationofwhoshouldbeincludedas   residentsofthehousehold.ThisprocedurewaschangedforCensus2000.Now,enumerators  t beginbyaskinghowmanypeoplewerelivingorstayinginthehousingunitonCensusDay.  X Aftercollectingappropriatepersonandhousingunitdata,theenumeratoraskedtwocoverage  < questions.Thefirstaskedabouttypicalsituationsinwhichpersonswhoshouldbeincludedas   residentstendtobemissed!babies,fosterchildren,personsawayonbusinessorvacation, h  roomersorhousemates,andtemporaryresidentswithnootherhome.Ifsomeonehadbeen L  missed,thenheorshewasaddedtotheformandtheircensusinformationwascollected.The 0  secondquestionaskedabouttypicalsituationsinwhichpersonswhoshouldnotbeincludedas   residentstendtobeincludedassuch!personsawayatcollege,inthearmedforces,inanursing   home,orinacorrectionalfacility.Ifsomeonewasincludedontheformbutshouldhavebeen x  countedelsewhere,thentheenumeratordeletedthemfromtheformbymarkingthecancelbox \  undertheirname.Thepurposeofthisanalysisistostudytheeffectivenessofthenewcoverage @ questionsintheidentificationofpersonswhowouldhaveotherwisebeenmissedorincludedin $ error. l  (I.6)Coverage,RosteringMethodsandHouseholdComposition:AComparativeStudyof 4 theCurrentPopulationSurveyandCensus2000(cancelled)   dThisevaluationwillnotbeconducted.Inearly2002,theCensus2000EvaluationProgramwas  refinedandprioritiesreassessedduetoresourceconstraintsattheCensusBureau. x  |    `  d +h%* ЇThispageintentionallyleftblank .  %,#  5d 5  f fJ:EthnographicStudies  d    `     h      p   Overview ,    `     h      p      x Theseevaluationswillstudycertainaspectsofcoverageforvariouspopulationsandattemptto   identifyareaswheremethodsofcollectingcensusdataforthesepopulationscanbeimproved.  t Onestudyinthiscategorywillapplysocialnetworkfieldandanalysismethodstoevaluate  X censuscoverageandprocesses.Wealsowillconductethnographicresearchonmobile  < populationsandColonias!areaslackingbasicinfrastructureandservicesalongtheborder   betweentheUnitedStatesandMexico. h  WhatWillWeLearn? 0  Studyresultswillhelpusdeterminewhetherindividualscanbebetteridentifiedfromtheir   positioninsocialnetworks(basedontheirinteractionsandtransactionswithothers)thanby x  comparingsetsofaddressandpersonrecords.Wewillalsolearnhowtoimproveproceduresto \  enumeratemobilepopulationsbytracingCensusDaytravelroutesorstopoversitesforasample @ ofsuchpersonsanddeterminingundercountsormultipleenumerationsoftheminthecensus. $ WealsowilllearnhowtoovercomebarrierstoenumeratingColoniasinfuturecensuses. l ̀  EthnographicStudies +h%* G <d <      `     h      p      x  (J.1)Coverage,RosteringMethodsandHouseholdComposition:AComparativeStudyof H theCurrentPopulationSurveyandCensus2000 ,  ThisevaluationwasreclassifiedundertheCoverageImprovementcategoryasevaluationI.6,   whichwasrecentlycancelled.    (J.2)EthnographicSocialNetworkTracing  X  Thisstudywilluseethnographicandsocialnetworkmethodstostudythefollowingfive  < questions.1)Whatinteractionsinsocialnetworksinfluenceandexplainordeterminethe   durationofindividualsstaysindomiciles(i.e.,households,institutions,orotherplaceswhere h  peoplesleep)andtheirresidentialmobility?2)Howmuchmorelikelyarepeoplewhochange L  domicileonceormorein6monthstobeomittedorerroneouslyenumeratedinCensus2000(and 0  incontemporarydemographicsurveys)thanpeoplewhoremainresidentiallystableovera6   monthperiod?3)Whatcharacteristics(ofpeople,theirnetworks,mobility,housing,household,   occupational,orothersocialoreconomicfactors)arecloselyassociatedwithomissioninthe x  census?4)Canpeoplebemorereliablyidentified(andreidentified)fromtheirpositioninsocial \  networksandfromtheirinteractionswithothers,thanbycomparingsetsofaddressandperson @ records?5)HowwelldoCensusBureaucategoriesfitwiththesociallyrepresented $ characteristicsthatpeopleusetoforminteractingsocialnetworks?d l d (J.3)ComparativeEthnographicResearchonMobilePopulations 4  Inthisstudy,asampleofselectedmobilepeoplewillbetracedtoidentifytheirCensusDay  travelroutesorstopoversites.Theinformationwillbematchedandreconciledwithcensus  results.Coverageerrorsfoundinthecensuswillbeanalyzedtodeveloprecommendationsfor | improvingprocedures. `  (J.4)ColoniasontheU.S./MexicoBorder:BarrierstoEnumerationinCensus2000  ( Coloniasareunincorporated,generallylowincomeresidentialsubdivisionslackingbasic p  infrastructureandservices(e.g.,pavedroadsandpublicwatersystems)alongtheborderbetween T theU.S.andMexico.Inordertodevelopappropriateenumerationproceduresandeffective 8  outreachandpromotionprogramsforColonias,itisnecessarytobetterunderstandtheunique ! situationsandissuesassociatedwithconductingthecensusorotherCensusBureausurveysin " theseareas.ThisresearchwillexaminethepotentialbarrierstocensusenumerationinColonias "  inthecontextofCensus2000throughparticipantobservation,indepthinterviews,andfocus #d! groupswithselectedColoniaresidents.Basedonpreviousresearch,topicsofparticularinterest $H" includeirregularhousing,concernsregardingconfidentiality,complexhouseholdstructure, %,# knowledgeofEnglish,andliteracy.  t& $    `     h     K:DataCapture  *$)  +h%* fH 5Vd 5   Overview d    `     h      p      x TheDataCaptureSystemforCensus2000(DCS2000)processedmorethan120millioncensus , formsbycreatingadigitalimageofeachpageandinterpretingtheentriesoneachimageusing   OpticalMarkRecognition(OMR),OpticalCharacterRecognition(OCR),orkeying.These   evaluationsaredesignedtoassesscomponentsofDCS2000,theDataCaptureAuditResolution  t (DCAR)process,andtomeasuretheimpactofthedatacapturesystemondataqualityandon  X subsequentdatacodingoperations.  < WhatWillWeLearn? h  Findingsfromtheseevaluationswilldeterminethelevelofaccuracyatwhichthedatacapture 0  systemperformed.Detailedinformationaboutthesystemwillbecollected,rangingfromthe   numberofformsprocessedbyformtype,date,andprocessingoffice,tomeasuringtheaccuracy   ofeachofthethreecapturemodesOMR,OCR,andKeyFromImage.Operationalproblems x  andtheirresolutionwillbedocumented.EvaluationoftheDCARprocesswillexaminethe \  systemsabilitytoidentifyandresolvecaptureproblemsstemmingfromproblemswithresponse @ entries.Theimpactofdatacaptureerrorsonourabilitytocorrectlyassignindustryand $ occupationcodeswillalsobeassessed. l DataCaptureEvaluations *#(  (Exx' (K.1.a)DataCaptureAuditResolutionProcess +h%* I <@d <   ThisevaluationdocumentstheresultsofDataCaptureAuditResolutionbyfailurereason,form d type,andDataCaptureCenter.Usingthesesamecategories,italsowilldocumentthenumber H andtypesofchangesthatcanbemadebyAuditReviewclerksandtheresultsoftheAuditCount , review.    (K.1.b)QualityoftheDataCaptureSystemandtheImpactofQuestionnaireCaptureand  t ProcessingonDataQuality  X  Thisevaluationexamineshowthedatacapturesystemaffecteddataqualityandwhethertherules  < fordeterminingwherecasesarerouted(e.g.,tokeyfromimage)weresetappropriately.In   addition,thisevaluationwilldocumentandcomparethedataqualityofeachdatacapturemethod h  foreveryfieldonthequestionnaire,aswellasbyformtype,DataCaptureCenter,andracialand L  ethniccategories. 0   (K.1.c)AnalysisofDataCaptureSystem2000KeyingOperations(cancelled)    Thisevaluationwillnotbeconducted.Inearly2002,theCensus2000EvaluationProgramwas x  refinedandprioritiesreassessedduetoresourceconstraintsattheCensusBureau. \  Ѐ (K.1.d)SynthesisofResultsfromK.1.a,K.1.b,andK.1.c $  Thisevaluationwillnotbeconducted.ResultsfromevaluationsK.1.a,K.1.b,andK.1.cwillbe l includedinatopicreportaddressingkeyfindingsforalldatacaptureevaluations. P  (K.2)AnalysisoftheInteractionBetweenAspectsofQuestionnaireDesign,Printing,and  CompletenessWithDataCapture(cancelled)   Thisevaluationwillnotbeconducted.Inearly2002,theCensus2000EvaluationProgramwas | refinedandprioritiesreassessedduetoresourceconstraintsattheCensus.Significantaspectsof ` thisstudywillbeaddressedinevaluationsK.1.aandK.1.b. D  (K.3)ImpactofDataCaptureErrorsonAutocoding,ClericalCodingandAutocoding p  ReferralsinIndustryandOccupationCoding T  Theinformationprovidedbyrespondentstotheindustryandoccupationquestionsonthecensus 8  formwereassigned(coded)toastandardsetofcategories.Thisevaluationexamineshowdata ! captureerrorsaffectedtheabilityoftheautocodingsystemandclericalcoderstoassigncorrect " IndustryandOccupationcodes. "   (K.4)PerformanceoftheDataCaptureSystem2000(cancelled) $H"  Thisevaluationwillnotbeconducted.Inearly2002,theCensus2000EvaluationProgramwas %,# refinedandprioritiesreassessedduetoresourceconstraintsattheCensusBureau. t& $ Ѐ  X' %     L:ProcessingSystems   )"'   Overview *$)  +h%*  fJ  5Vd 5    OncecensusdatafromallsourceswerecapturedbytheDataCaptureSystem2000,theywere d normalized(putinastandardformat,regardlessofinputsource)andstoredinafileknownasthe H DecennialResponseFile,stage1(DRF1).Severalprocessesthenwereappliedbeforethedata , wereusedtoproduceofficialcensuscountsandtabulations.Oneprocesswasappliedtolink   multiplequestionnairesthatwereusedtoenumeratethatsamehousehold.Forexample,alarge   familycouldhaveamailreturnformwithdataonsixmembersofthehouseholdandan  t additionalfollowupComputerAssistedTelephoneInterview(CATI)withdataontherestofthe  X household.Underthesecircumstances,CATIpersonrecordsareappendedtotheoriginalmail  < returnpersonrecords.Anotherprocesswasusedforsituationswheremultiplequestionnaires   involvingdifferenthouseholdswerereceivedforthesameaddress.Forexample,oneformcould h  beforahouseholdthatmovedoutnearCensusDay,andtheotherformcouldbeforthe L  householdthatthenmovedin.AcomputerprogramknownasthePrimarySelectionAlgorithm 0  (PSA)thenwasusedtodecidewhichpersonandhousingunitdatashouldbeusedforcensus   tabulations.TheinputintoPSAistheDRF1,withtheresultingfilebeingtheDecennial   ResponseFile,stage2(DRF2).Followingalltheseprocesses,theDRF2wasmergedwithkey x  elementsoftheDecennialMasterAddressFile(DMAF)tocreatethe100PercentCensus \  UneditedFile(HCUF),whichcontainstheresponsedata(i.e.,the100percentquestionsfrom @ boththeshortandlongformcensusquestionnaires)selectedbyPSAprocessingtorepresenta $ householdinthecensus.TheSampleCensusUneditedFile(SCUF)containstheunedited l responsesforhouseholdswithsample(longform)data. P Avarietyofpostcensusactivitieswereneededtopreparethedatafromtheoriginalresponsesto  releasingtheofficialcountsandtabulations.Theseactivitiesincludeeditingandimputationof  theHCUFandSCUFtocreatethe100PercentCensusEditedFile(HCEF)andtheSample | CensusEditedFile(SCEF)respectively,codingofwriteinresponseitems(suchasrace, ` language,industryandoccupation,andplaceofwork/migration),conversiontotabulation D geography,tabulationrecoding,andapplyingdisclosureavoidancetechniques. ( TheBetaSitewasasoftwaretestingsiteforBureauoftheCensusapplicationdevelopersandis T usedasanintegrationcenterforRegionalCensusCenters(RCC)andLocalCensusOffices 8  (LCO)systems,atestingcenterforallsystems,andasupportcenterforRCC,LCO,andthe ! NationalProcessingCentersystems.Wewillexaminetheeffectivenessofthissoftwaretesting " site.   `  "    $H" K <@d <   5Vd 5WhatWillWeLearn? d AnalysisofareinterviewofmultiplequestionnaireaddresseswilldetermineifthePSA , methodologyandrulesforresolvingthesecasesaccuratelyidentifiedtheCensusDayhousehold   members.TheevaluationoftheDRFcreationandprocesseswillexaminehowwellmultiple   formsforthesamehouseholdwerelinked.AnalysisofCUFcreationwilldocumentthenumber  t oftimeseachspecificDMAF/DRFrulewasapplied.TheBetaSiteanalysiswillinclude  X informationonwhetherthedatacollectionsystemsweresuccessfullyintegrated,andthebenefits  < ofthesoftwaretestingandreleaseprocess.       h  @sV @d@ ProcessingSystemsEvaluations d  (L.1)InvalidReturnDetection ,  ThisevaluationwasnotconductedbecausetheoperationwasnotnecessaryinCensus2000.    (L.2)DecennialResponseFileStage2LinkingandSettingofExpectedHousehold  t Population  X  ThisevaluationwilldocumenthowfrequentlycensusformswerelinkedduringtheDecennial  < ResponseFileprocessingandthetypesoflinkagesthatwereconstructed.Itwillalsoassessthe   accuracyoftheautomatedprocessforsettingtheexpectedhouseholdsizeanditseffectsonthe h  censuspopulation. L   (L.3.a)AnalysisofPrimarySelectionAlgorithmResults(OperationalAssessment)    TheobjectiveofthisevaluationistodocumenttheeffectsofusingthePrimarySelection   Algorithminresolvingsituationswhenmultiplehouseholdquestionnaireswerereceivedforthe x  sameaddress. \   (L.3.b)ResolutionofMultipleCensusReturnsUsingReinterview $  TheobjectiveofthisevaluationistodeterminetheaccuracyofPrimarySelectionAlgorithm l rulesfordeterminingtheCensusDayresidentsforanaddress.Comparisonsweremadebetween P finalCensus2000dataanddatathatwerecollectedusingareinterviewofasampleofaddresses 4 wherethePrimarySelectionAlgorithmwasapplied.   (L.4)CensusUneditedFileCreation |  Thisevaluationdocumentstheresultsoftheprocessofdeterminingthefinalhousingunit ` inventory.Thefinalhousingunitinventoryforthecensuswasdeterminedduringtheprocessof D creatingtheCensusUneditedFile.Thefinalhousingunitinventorywascreatedbymerging ( informationontheprocessedDecennialResponseFilewiththeinformationontheDecennial p  MasterAddressFile. T  (L.5)BetaSite !  ThisevaluationwillanswerquestionsabouthowwelltheBetaSiteintegratedthesoftware " systemssupportingCensus2000anditsoverallutilityforsoftwaretestingandrelease. "   @*   `     h      *#(  Thispageintentionallyleftblank.  <(!& 5d5   @2  @  @`  @  @  @h  @  M:QualityAssuranceEvaluations ho d ! Overview , Census2000involvedmorethan20majorfieldoperationsand,atitspeak,morethan500,000   temporaryworkers.Managingthequalityofthedeliverablesproducedbythislarge,  t decentralized,andtransientworkforcewasamajorchallengefortheCensusBureau.Thequality  X assurance(QA)programsweredesignedtominimizesignificantperformanceerrors,toprevent  < theclusteringofsignificantperformanceerrors,andtopromotecontinuousimprovement.   WhatWillWeLearn? L  ThefirstevaluationwilldeterminetheeffectivenessoftheQAprogramsusedintheaddresslist   developmentandenumerationoperationsandwilldetermineifdifferentQAapproachesshould   beexploredforthenextcensus.Forthesecondstudy,theeffectivenessofvariablesthatwere x  usedtodetectdiscrepancieswillbemeasured,andappropriatevariableswillbeaddedand/or \  deletedfromthedetectionprocess. @    @8d@    "QualityAssuranceEvaluations d  (M.1)EvaluationoftheCensus2000QualityAssurancePhilosophyandApproachUsedfor , theAddressListDevelopmentandEnumerationOperations    Thestudywilldeterminetheeffectivenessofthequalityassurancephilosophyandactivitiesused   tomanagethequalityofthedeliverablesproducedintheaddresslistdevelopmentand  t enumerationoperations.Thisstudywilldocumentoperationalexperienceswiththisapproach,  X measurequalitylevelsachieved,anddetermineifotherapproachesshouldbeexploredforthe  < 2010Census.    (M.2)EffectivenessofExistingVariablesintheModelUsedtoDetectDiscrepanciesDuring L  Reinterview,andtheIdentificationofNewVariables  0  Thereinterviewprogramwasaqualityassurancemeasurewhosemajorobjectivewastodetect   enumeratorswhoseworkindicateddiscrepancies.Thisevaluationexaminesvariablesusedin   thismodeltodetermineiftheywereeffectiveindetectingdiscrepancies;whetherothervariables x  shouldbeaddedtothemodel;andtoprovidesuggestionsonotherwaystoimprovethisprogram. \      N:AccuracyandCoverageEvaluationSurveyOperations ho +h%*  P  5d5    Overview H TheCensusBureauconductedtheAccuracyandCoverageEvaluation(A.C.E.),anationwide   samplesurvey,todeterminethenumberofpeopleandhousingunitsmissedorincorrectly   countedinthecensus.Thebasicapproachwastoindependentlyrelistasampleofblocks, re  t enumeratethemduringtheA.C.E.survey,andthencomparetheresultstothecensusdataforthe  X sameblocks.TheCensusBureaumayusetheresultsoftheA.C.E.tocorrectthecensuscounts  < obtainedthroughtheprecedingenumerationprocedures .   Thestudiesinthiscategory willmeasurehowwelltheCensusBureaucarriedoutdifferent L  componentsoftheA.C.E.Forinstance,analysisprojectsandevaluationswillbeconductedthat 0  measurethecompletenessofthehousingunitlistsusedforA.C.E.interviewing,thequalityof   theA.C.E.personinterviewingprocess,andtheaccuracyoftheproceduresusedtomatch   personscountedduringtheA.C.E.interviewtothosethatwereenumeratedinthecensus.The x  successofeachA.C.E.componentaffectsthequalityofthefinalestimates.  \  ЀWhatWillWeLearn? $ TheresultsoftheseA.C.E.analysisprojectsandevaluationswillhelptheCensusBureauto P documentthiscoveragemeasurementoperationandimproveitsprocedures.Forexample,we 4  willdeterminehowwellwedetectdiscrepantresults,whilealsolookingattheireffectonthe  A.C.E.  TheseoperationalanalysesandevaluationswilldocumenttheA.C.E.processandgivethe ` CensusBureaugreaterinsightintowhatcauseserrorinthemeasurementofcoverageerror. D Moreover,matchingerrorsmayaddtoerrorsintheestimatesofcensuscoverage.One ( evaluationinthiscategorywillexamineasubsampleofrematchedA.C.E.blockstomeasure p  matchingerrors.WealsowillmeasuretheeffectofmatchingerroronDualSystemEstimates T andundercountrates. 8  TheevaluationsinthiscategorywillhelptheCensusBureautoidentifyoperationalcausesof " errorinmeasuringcoverageandwillhelptominimizethemwhenplanningfuturecensuses. "  ManyevaluationsinthiscategorythatwereplannedfortheA.C.E.werenolongerneedwhenthe $H" decisionwasmadenottoadjusttheCensus2000populationcounts.Itwasdeterminedthatsome %,# ofthereportsthatweredevelopedinanexpeditedmannertoinformtheExecutiveSteering t& $ CommitteeforA.C.E.(ESCAP)decisionsweresufficientlycompleteandinformativetoanswer X' % researchquestionsfromtheplannedevaluationreports.Fiveofthesereportsandtheir <(!& correspondingESCAPreportsarenotedinthefollowingsectionofevaluationreport  )"' descriptions.ToreviewtheESCAPreportsontheInternet,goto4 BfO  5  http://www.census.gov/6fO +  7 Ԁand *#( searchon ESCAP. *$)   +h%*  "Q  5Vd5    5  H  R  <@d<    A.C.E.SurveyOperationsEvaluations d  (N.1)ContaminationofCensusDataCollectedinA.C.E.Blocks  , ThisevaluationexamineswhethercensusandA.C.E.operationswerekeptoperationally   independent(akeyrequirementforavoidingbiasinthedualsystemestimatesofcoverageerror)   bycomparingcensusresultsinA.C.E.andnonA.C.E.clustersandthroughdebriefingoffield  t staff.  X  (N.2)AnalysisofListingFutureConstructionandMultiUnitsinSpecialPlaces(cancelled)    Thisevaluationwillnotbeconducted.Inearly2002,theCensus2000EvaluationProgramwas h  refinedandprioritiesreassessedduetoresourceconstraintsattheCensusBureau. L   (N.3)AnalysisofRelistedBlocks(cancelled)    Thisevaluationwillnotbeconducted.Inearly2002,theCensus2000EvaluationProgramwas   refinedandprioritiesreassessedduetoresourceconstraintsattheCensusBureau. x   (N.4)AnalysisofBlocksWithNoHousingUnitMatching(cancelled) @  Thisevaluationwillnotbeconducted.Inearly2002,theCensus2000EvaluationProgramwas $ refinedandprioritiesreassessedduetoresourceconstraintsattheCensusBureau. l Ѐ  P (N.5)AnalysisofBlocksSentDirectlyforHousingUnitFollowup(cancelled) 4  Thisevaluationwillnotbeconducted.Inearly2002,theCensus2000EvaluationProgramwas  refinedandprioritiesreassessedduetoresourceconstraintsattheCensusBureau. x   (N.6)AnalysisofPersonInterviewWithUnresolvedHousingUnitStatus(cancelled) `  Thisevaluationwillnotbeconducted.Inearly2002,theCensus2000EvaluationProgramwas D refinedandprioritiesreassessedduetoresourceconstraintsattheCensusBureau. ( Ѐ  p  (N.7)AnalysisontheEffectsofCensusQuestionnaireDataCaptureinA.C.E.(cancelled) T  Thisevaluationwillnotbeconducted.Inearly2002,theCensus2000EvaluationProgramwas 8  refinedandprioritiesreassessedduetoresourceconstraintsattheCensusBureau.  ! (N.8)AnalysisoftheCensusResidenceQuestionsUsedinA.C.E.(cancelled) "   Thisevaluationwillnotbeconducted.AcorrespondingExecutiveSteeringCommitteefor #d! AccuracyandCoverageEvaluation(ESCAP)analysisanddocumentationisrelevanttothis $H" evaluation.RefertotheESCAPreport, EvaluationResultsforMoversandNonresidentsinthe %,# Census2000AccuracyandCoverageEvaluation(reportB16). t& $  (N.9)AnalysisofthePersonInterviewProcess(cancelled) +h%* S    Thisevaluationwillnotbeconducted.AcorrespondingExecutiveSteeringCommitteefor d AccuracyandCoverageEvaluationI(ESCAPI)analysisanddocumentationisrelevanttothis H evaluation.RefertotheESCAPIreport,A.C.E.PersonInterviewingResults(reportB5). , Additionally, AutomatingtheCensus2000A.C.E.FieldOperationsand ResultsofQuality   AssuranceonthePersonInterviewOperationoftheA.C.E.ofCensus2000"providepreliminary   researchfindingsthatwerepresentedatthe2001AmericanStatisticalAssociationConference.  t Foracopyofthesepapers,contacttheCensusBureauon(301)4574218.   X (N.10)DiscrepantResultsinA.C.E.    Thisevaluationexamineshowwellthequalityassuranceprocessidentifiedinterviewerswho h  entereddiscrepantdataintheA.C.E.interviewandtheimpactofundetecteddiscrepantdataon L  A.C.E.estimates. 0   i(v8x (N.11)ExtendedRosterAnalysis(cancelled)    Thisevaluationwillnotbeconducted.Inearly2002,theCensus2000EvaluationProgramwas x  refinedandprioritiesreassessedduetoresourceconstraintsattheCensusBureau.  \  (N.12)MatchingStagesAnalysis(cancelled) $  Thisevaluationwillnotbeconducted.AcorrespondingExecutiveSteeringCommitteefor l AccuracyandCoverageEvaluationI(ESCAPI)analysisanddocumentationisrelevanttothis P evaluation.RefertotheESCAPIreport, A.C.E.PersonMatchingandFollowupResults 4 (reportB6).Additionally, ResultsofQualityAssuranceontheA.C.E.MatchingOperations,  providespreliminaryresearchfindingsthatwerepresentedatthe2001AmericanStatistical  AssociationConference.dForacopyofthepaper,contacttheCensusBureauon(301)4574218.d |  (N.13)AnalysisofUnresolvedCodesinPersonMatching(cancelled) D  Thisevaluationwillnotbeconducted.AcorrespondingExecutiveSteeringCommitteefor ( AccuracyandCoverageEvaluationI(ESCAPI)analysisanddocumentationisrelevanttothis p  evaluation.RefertotheESCAPIreport, A.C.E.VarianceEstimatesbySizeofGeographic T Area(reportB11). 8   (N.14)EvaluationofMatchingError "  Apotentialsourceoferrorinthecoverageestimatesisthematchingoperationusedtoclassify "  personsasmissedorerroneouslyenumeratedinthecensus.Thisevaluationwilldeterminethe #d! relativeerrorassociatedwiththematchingoperationandhowmatchingerroraffectstheDual $H" SystemEstimates. %,#  (N.15)OutlierAnalysisinthe2000A.C.E.(cancelled) *$)  Thisevaluationwillnotbeconducted.Inearly2002,theCensus2000EvaluationProgramwas +h%* refinedandprioritiesreassessedduetoresourceconstraintsattheCensusBureau. ,L&+ 5T    (N.16)ImpactofTargetedExtendedSearch(cancelled) H   XXThisevaluationwillnotbeconducted.Inearly2002,theCensus2000EvaluationProgramwas , refinedandprioritiesreassessedduetoresourceconstraintsattheCensusBureau.Significant   aspectsofthisstudywillbeaddressedinevaluationN.17.   Ѐ  t  (N.17)TargetedExtendedSearchBlockClusterAnalysis  X  In1990,thesearchareaformatchingwasextendedtosurroundingblocksforallclusters.In  < 2000,thiswasonlydoneforclustersdeemedmostlikelytobenefitfromthisadditional   searching. Thisreportwilldocumentoveralltargetedextendedsearchresultsandidentify h  characteristicsthatmayberelatedtomatchesandcorrectenumerationsfoundinsurrounding P  blocksduetogeocodingerror.  4  (N.18)EffectofLateCensusDataonFinalEstimates(cancelled)    Thisevaluationwillnotbeconducted.AcorrespondingExecutiveSteeringCommitteefor |  AccuracyandCoverageEvaluationII(ESCAPII)analysisanddocumentationisrelevanttothis `  evaluation.RefertotheESCAPIIreport, EffectofExcludingReinstatedCensusPeoplefrom D theA.C.EPersonProcess. (  (N.19)FieldOperationsandInstrumentsforA.C.E. T  Thisanalysisprovidesanoverallassessmentofthequalityofhousingunitandpersoncoverage 8 inA.C.E.operations.SomeofthetopicsaddressedintheanalysisarequalityofA.C.E.listing,  effectofhousingunitfollowupinterviewingontheenhancedlist,effectivenessofhousingunit  andpersonfollowupqualityassurance,andnoninterviewrates.   (N.20)GroupQuartersAnalysis(cancelled) H  Thisevaluationwillnotbeconducted.Inearly2002,theCensus2000EvaluationProgramwas , refinedandprioritiesreassessedduetoresourceconstraintsattheCensusBureau. t   (88i (N.21)AnalysisofMobileHomes(cancelled) <   Thisevaluationwillnotbeconducted.Inearly2002,theCensus2000EvaluationProgramwas  ! refinedandprioritiesreassessedduetoresourceconstraintsattheCensusBureau. "   $L"  Thispageintentionallyleftblank.  X' %   <(!&   5d5  O:CoverageEvaluationsoftheCensusandoftheAccuracyandCoverageEvaluation d Survey d" H S" Overview   ThestudiesinthiscategoryincludeagroupevaluatingA.C.E.coverageandagroupevaluating  t censuscoverage.Thesestudieswillidentifypersonandhousingunitcharacteristicsthatare  X relatedtobeingmissedorerroneouslyenumerated.Analysisinthisareawillalsostudythe  < qualityofdatafromproxyrespondents,andthefrequencyandpatternsofgeocodingerror.   Furthermore,censuscountsanddualsystemestimateswillbecomparedtodemographic h  benchmarkstoevaluateaccuracyandcompleteness. L  WhatWillWeLearn?    ResultsfromtheseevaluationswillallowustodeterminehowcompleteourMasterAddressFile x  wasforCensus2000.Netcoverageratesofhousingunitswillbecomputedatthenationaland \  subnationallevelsalongwithgrossomissionanderroneousenumerationrates.Otherstudieswill @ explainfactorsthatcontributetohousingunitcoverageerror.Forexample,wewilllearn $ whethertypeofaddress(citystyleversusnoncitystyle)hadaneffectonhousingunitcoverage. l Inaddition,therewillbeastudyofhousingunitduplication ,toidentifycharacteristicsof P duplicateunitsandtheiroperationalsource. 4 Similarly,wewillidentifyfactorsthatcontributetopersoncoverageerror.Wewillacquire  knowledgeabouterroneousenumerationsbydeterminingwhichdemographic,housingunittype, | andtypeofenumerationvariableswereassociatedwiththem.Furthermore,wewillconductan ` analysisofmeasurementerror,whichwillhelpusdeterminewhypeoplewereerroneouslylisted D inthecensusandtheAccuracyandCoverageEvaluation. ( ManyevaluationsinthiscategorythatwereplannedfortheA.C.E.werenolongerneedwhenthe T decisionwasmadenottoadjusttheCensus2000populationcounts.Itwasdeterminedthatsome 8  ofthereportsthatweredevelopedinanexpeditedmannertoinformtheExecutiveSteering ! CommitteeforA.C.E.(ESCAP)decisionsweresufficientlycompleteandinformativetoanswer " researchquestionsfromtheplannedevaluationreports.Tenofthesereportsandtheir "  correspondingESCAPreportsarenotedinthefollowingsectionofevaluationreport #d! descriptions.ToreviewtheESCAPreportsontheInternet,goto4BfO  5  http://www.census.gov/6-fO- +  7 .Ԁand $H" searchon ESCAP.  %,# @d"@  5CoverageEvaluationsoftheCensusandoftheA.C.E.  d    `     h      p      x (O.1)TypeofEnumerationAreaSummary(cancelled) ,  Thisevaluationwillnotbeconducted.Inearly2002,theCensus2000EvaluationProgramwas   refinedandprioritiesreassessedduetoresourceconstraintsattheCensusBureau.    (O.2)CoverageofHousingUnitsintheEarlyDecennialMasterAddressFile(cancelled)  X  Thisevaluationwillnotbeconducted.Inearly2002,theCensus2000EvaluationProgramwas  < refinedandprioritiesreassessedduetoresourceconstraintsattheCensusBureau.     (O.3)Census2000HousingUnitCoverageStudy L   Thisevaluationassesses1)thenetcoveragerateofhousingunits,2)thegrossomissionrateof 0  housingunits,and3)theerroneousenumerationrateofhousingunits.Theseassessmentsare   madeatthenationallevel,smallergeographiclevels,andforeachpost-strata.Thisevaluation   alsoexaminesthepotentialimpactonhousingunitcoveragehadweexcludedspecificMaster x  AddressFilebuildingoperations.ThisstudyissimilartotheHousingUnitCoverageStudy \  conductedin1990.Italsohas acorrespondingExecutiveSteeringCommitteeforAccuracyand @ CoverageEvaluationII(ESCAPII)analysisanddocumentation.RefertotheESCAPIIreport, $  Census2000HousingUnitCoverageStudy(report17). l  (O.4)AnalysisofConflictingHouseholds 4  DuringA.C.E.housingunitmatching,situationswerefoundwherethecensusandA.C.E.listed  twoentirelydifferentfamilies.Thisstudywilldocumentthefollow-upinterviewingresultsfor  thesehouseholdstodetermineifthecensuswasinerror,theA.C.E.wasinerror,ifthetwo | familiesbothliveattheaddress,iftherewasmisdeliveryofthecensusform,andsoon.   ` (O.5)AnalysisofProxyDataintheA.C.E. (  BoththecensusandA.C.E.sometimesmustcollectdatafromproxyrespondentspersonswho p  arenotmembersofthehouseholdwheredataareneeded.Thisstudywillexaminematchrates T anderroneousenumerationratesforsuchcasesintheA.C.E.  8   (O.6)PSampleNonmatchesAnalysis(cancelled) "  Thisevaluationwillnotbeconducted.AcorrespondingExecutiveSteeringCommitteefor "  AccuracyandCoverageEvaluationII(ESCAPII)analysisanddocumentationisrelevanttothis #d! evaluation.RefertotheESCAPIIreport, PsampleNonmatchAnalysis(report18). $H" Additionally, ConsistencyofCensus2000PostStratificationVariables,providespreliminary %,#   researchfindingsthatwerepresentedatthe2001AmericanStatisticalAssociationConference. t& $ Foracopyofthepaper,contacttheCensusBureauon(301)4574218.   x  X' %    `     (O.7)AnalysisofPersonCoverageinPuertoRico(cancelled) +h%* 5X    Thisevaluationwillnotbeconducted.Inearly2002,theCensus2000EvaluationProgramwas d  refinedandprioritiesreassessedduetoresourceconstraintsattheCensusBureau.  H (O.8)AnalysisofHousingUnitCoverageinPuertoRico(cancelled)    Thisevaluationwillnotbeconducted.Inearly2002,theCensus2000EvaluationProgramwas   refinedandprioritiesreassessedduetoresourceconstraintsattheCensusBureau.  t  (O.9)GeocodingErrorAnalysis(cancelled)  <  Thisevaluationwillnotbeconducted.AcorrespondingExecutiveSteeringCommitteefor   AccuracyandCoverageEvaluationII(ESCAPII)analysisanddocumentationisrelevanttothis h  evaluation.RefertotheESCAPIIreports,  EsampleErroneousEnumerationAnalysis(report L  5)and AnalysisofNonmatchesandErroneousEnumerationsUsingLogisticRegression 0  (report12).    (O.10)HousingUnitDuplicationinCensus2000 x   Duplicationinthecensuswasonetypeoferroneousenumeration. Thisanalysiswillidentify \  duplicatehousingunitsinCensus2000andtheircharacteristics.Thestudywillalsodetermineif @ duplicationwasmorelikelyforonegrouporanother(e.g.,ownersvs.renters).Thecensus $ operationsmostlikelytoproducehousingunitduplicationwillbeidentified,alongwiththemost l plausiblesourcesofduplication. P  (O.11)ESampleErroneousEnumerationAnalysis(cancelled)   Thisevaluationwillnotbeconducted.AcorrespondingExecutiveSteeringCommitteefor  AccuracyandCoverageEvaluationII(ESCAPII)analysisanddocumentationthatisrelevantto | thisevaluation.RefertotheESCAPIIreport, ESampleErroneousEnumerationAnalysis ` (report5).Additionally, Census2000ESampleErroneousEnumerations,provides D preliminary  researchfindingsthatwerepresentedatthe2001AmericanStatisticalAssociation ( Conference.dForacopyofthepaper,contacttheCensusBureauon(301)4574218.d p    (O.12)AnalysisofNonmatchesandErroneousEnumerationsUsingLogisticRegression 8  (cancelled) !  Thisevaluationwillnotbeconducted.AcorrespondingExecutiveSteeringCommitteefor " AccuracyandCoverageEvaluationII(ESCAPII)analysisanddocumentationisrelevanttothis "  evaluation.RefertotheESCAPIIreport, LogisticRegression(report19).Additionally, #d!  ModelingA.C.E.NonmatchesintheCensus2000,providespreliminary  researchfindingsthat $H" werepresentedatthe2001AmericanStatisticalAssociationConference.dForacopyofthe %,# paper,contacttheCensusBureauon(301)4574218.d t& $    `     h      p   (O.13)AnalysisofVariousHouseholdTypesandLongFormVariables <(!&  Thisstudycombinesthehousingunitdataandthepersondatatostudycoverage.A newlink  )"' betweenA.C.E.housingunitsandcensushousingunitsinthesamplewillbecreatedinthe *#( combineddatabasedonthepersonmatchingresult.Thenthecombineddatawillbeusedto *$) examinewhethercoveragesareaffectedbyvariablessuchasaddressstyle,income,education, d propertyvalueorrent,typeoffamily,andhouseholdcomplexity.  H (O.14)MeasurementErrorReinterviewAnalysis(cancelled)    Thisevaluationwillnotbeconducted.AcorrespondingExecutiveSteeringCommitteefor   AccuracyandCoverageEvaluationII(ESCAPII)analysisanddocumentationisrelevanttothis  t evaluation.RefertotheESCAPIIreports, EvaluationResultsforChangesinA.C.E.  X EnumerationStatus(report3)and FollowupReview(report24).  <  (O.15)ImpactofHousingUnitCoverageonPersonCoverageAnalysis h   dThisevaluationwillnotbeconductedbecausetheavailabledatacannotanswerthespecified L  questionsforthisstudy.d 0   (O.16)PersonDuplicationinCensus2000    Peoplewereduplicatedinthecensusformanydifferentreasons.Thisanalysiswillidentifythe x  numberandcharacteristicsofduplicatepersonsinCensus2000.Thestudywillalsodetermineif \  duplicationwasmorelikelyforonegrouporanother(e.g.,owners/renters).Thecensus @ operationsmostlikelytocauseduplicationwillbeidentified,alongwiththemostplausible $ sourcesoftheduplication. l  (O.17)AnalysisofHouseholdsRemovedBecauseEveryoneintheHouseholdisUnder16 4 YearsofAge(cancelled)   Thisevaluationwillnotbeconducted.Inearly2002,theCensus2000EvaluationProgramwas  refinedandprioritiesreassessedduetoresourceconstraintsattheCensusBureau. |  (O.18)SynthesisofWhatWeKnowAboutMissedCensusPeople(cancelled) D  Thisevaluationwillnotbeconducted.Inearly2002,theCensus2000EvaluationProgramwas ( refinedandprioritiesreassessedduetoresourceconstraintsattheCensusBureau. p   (O.19)AnalysisofDeletedandAddedHousingUnitsInCensus2000Measuredbythe 8  AccuracyandCoverageEvaluation  ! Thegoalofthisstudyistoassessthecompletenessofhousingunitcoverageontheearly " DecennialMasterAddressFile(DMAF).Wewilldeterminewhichcensusoperations "  contributedtoundercoveragebydeletingunitsthatshouldhavenotbeendeleted,andwhich #d! operationsimprovedcoveragebyaddingunitsnotpreviouslyaccountedfor.Wealsowill $H" identifywhichcensusoperationsreducedhousingunitduplication  %,# (O.20)ConsistencyofCensusEstimateswithDemographicBenchmarks X' %  JThisstudyusesindependentdemographicbenchmarkstoevaluatetheaccuracyoftheCensus <(!& 2000countsandthecompletenessofcoverageinCensus2000.Whilethisapproachcannot  )"' produceestimatesforasmanydemographicgroupsandgeographicareasasA.C.E.,resultscan *#( becomparedtoA.C.E.ataggregatelevels.  *$)   (O.21)ImplicationsofNetCensusUndercountonDemographicMeasuresandProgram H Uses(cancelled) ,  Thisevaluationwillnotbeconducted.ItwascancelledbeforeFebruary2001 .   (O.22)EvaluationofHousingUnitsCodedasErroneousEnumerations(cancelled)  x  Thisevaluation,whichwasaddedafterFebruary2001,willnotbeconducted.Acorresponding  \ ExecutiveSteeringCommitteeforAccuracyandCoverageEvaluationII(ESCAPII)analysisand  @ documentationisrelevanttothisevaluation.RefertotheESCAPIIreport, EvaluationofLack  $ ofBalanceandGeographicErrorsAffectingA.C.E.PersonEstimates(report2).  l   (O.23)AnalysisofInsufficientInformationforMatchingandFollowup(cancelled) 4   Thisevaluation,whichwasaddedafterFebruary2001,willnotbeconducted.Acorresponding   ExecutiveSteeringCommitteeforAccuracyandCoverageEvaluationII(ESCAPII)analysisand   documentationisrelevanttothisevaluation.RefertotheESCAPIIreport, ESample |  ErroneousEnumerationAnalysis(report5). `   (O.24)EvaluationofLackofBalanceandGeographicErrorsAffectingPersonEstimates ( (cancelled) p   Thisevaluation,whichwasaddedafterFebruary2001,willnotbeconducted.Acorresponding T ExecutiveSteeringCommitteeforAccuracyandCoverageEvaluationII(ESCAPII)analysisand 8 documentationisrelevanttothisevaluation.RefertotheESCAPIIreport, EvaluationofLack  ofBalanceandGeographicErrorsAffectingA.C.E.PersonEstimates(report2).   (O.25)MoverAnalysis(cancelled)  d Thisevaluation,whichwasaddedafterFebruary2001,willnotbeconducted.Acorresponding H ExecutiveSteeringCommitteeforAccuracyandCoverageEvaluationII(ESCAPII)analysisand , documentationisrelevanttothisevaluation.RefertotheESCAPIIreport, Analysisof t Movers(report15). X  (O.26)AnalysisofBalancingintheTargetedExtendedSearch(cancelled)  !  Thisevaluation,whichwasaddedafterFebruary2001,willnotbeconducted.Acorresponding " ExecutiveSteeringCommitteeforAccuracyandCoverageEvaluationI(ESCAPI)analysisand "  documentationisrelevanttothisevaluation.RefertotheESCAPIreport, A.C.E.Dataand #h! AnalysistoInformtheESCAPReport(reportB1). $L"    x& $ 5   5Vd5P.AccuracyandCoverageEvaluationSurveyStatisticalDesignandEstimation d  AfOverview  ,  TheevaluationsinthiscategoryweredesignedtoexaminethequalityofAccuracyandCoverage   Evaluation(A.C.E.)estimates. BecauseofresourcereallocationandExecutiveSteering  t CommitteeforAccuracyandCoverageEvaluation(ESCAP)analysesanddocumentationthat  X informsevaluations,theevaluationsforthiscategorywerenotconducted.Refertospecific  < evaluationsformoreinformation.   SomeofthereportsthatweredevelopedinanexpeditedmannertoinformESCAPdecisions L  weresufficientlycompleteandinformativetoanswerresearchquestionsfromtheplanned 0  evaluationreports.TwoofthesereportsandtheircorrespondingESCAPreportsarenotedinthe   followingsectionofevaluationreportdescriptions.ToreviewtheESCAPreportsonthe   Internet,goto4B fO  5  http://www.census.gov/6jfO k +  7;kԀandsearchon ESCAP. x   A.C.E.SurveyStatisticalDesignandEstimationEvaluations  *$) \ <dd<   (P.1)MeasurementofBiasandUncertaintyAssociatedWithApplicationoftheMissing H DataProcedures(cancelled) , B.` hp x (#XB  Thisevaluationwillnotbeconducted.AcorrespondingExecutiveSteeringCommitteefor   AccuracyandCoverageEvaluation(ESCAP)analysisanddocumentationisrelevanttothis   evaluation.RefertotheESCAPIIreport, AnalysisofMissingDataAlternativesfortheA.C.E  t (report12) .  X (P.2)SyntheticDesignResearch/CorrelationBias(cancelled)  $  Thisevaluationwillnotbeconducted.Inearly2002,theCensus2000EvaluationProgramwas l  refinedandprioritiesreassessedduetoresourceconstraintsattheCensusBureau. P  Ѐ  4  (P.3)VarianceofDualSystemEstimatesandAdjustmentFactors(cancelled)    Thisevaluationwillnotbeconducted.AcorrespondingExecutiveSteeringCommitteefor   AccuracyandCoverageEvaluationI(ESCAPI)analysisanddocumentationthatisrelevantto |  thisevaluation.RefertotheESCAPIreport, A.C.E.:VarianceEstimatesbySizeofGeographic `  Area(reportB11).d D d (P.4)OverallMeasuresofA.C.E.Quality(cancelled) p   B.` hp x (#XBThisevaluationwillnotbeconducted.Asummaryreportofvariousqualitymeasureswas T previouslyplanned.However,thisinformationwillbeincludedinatopicreportassessingkey 8 findingsforallA.C.E.evaluations.   (P.5)TotalErrorAnalysis(cancelled)   Thisevaluationwillnotbeconducted.Inearly2002,theCensus2000EvaluationProgramwas d refinedandprioritiesreassessedduetoresourceconstraintsattheCensusBureau. H   X ]     5d5  Q:Organization,Budget,andManagementInformationSystem 20u d vOverview , Researchinthiscategorywilldocumentheadquartersdecisionmakingprocessesandtheimpact   ofheadquartersorganizationalstructureonthedecennialcensus.Weplantostudythe  t managementapproach,structure,processes,andmanagementtools.  X Ѐ  < WhatWillWeLearn?   ThefindingsfromthisstudywillhelptheCensusBureautobettermanagefuturecensusesand L  similarprojects.ThisstudywilldocumenthowwelltheManagementInformationSystem 0  workedinhelpingusmanageCensus2000.Wewillcomparetheactivitiesand   recommendationsoftheCensus2000researchanddevelopmentprogramtowhatwasactually   implementedforCensus2000todeterminewhichprojectsweremostbeneficial.Inaddition,we x  willexaminetherolesandinfluencesofbothexternalandinternalentitiesonplanningand \  implementingthecensus. @ ЀOrganization,Budget,andManagementInformationSystemEvaluation +h%*  ,L&+  5^^  <d<    (Q.1)ManagementProcessesandSystemsofthe2000DecennialCensus  d Thepurposeofthisstudyisto evaluatethemanagementmodelforCensus2000includingthe H organizationalstructure,thedecisionmakingprocess,andthemanagementinformationtools. , Thestudywillalsoassessthestaffing,theuseandmanagementofcontracts,andtheimpactof   externalinfluencessuchastheCensusMonitoringBoard,theCongress,thefundinghistory,the   GeneralAccountingOffice,andotherstakeholders.  t    `   0    5d5  R:AutomationofCensusProcesses 20~ d 0 Overview , ForCensus2000,theCensusBureauimplementedaseriesofautomatedsystemstoaidthe   conductofthedecennialcensus.Thesesystemsincluded,butwerenotlimitedto,datacollection  t andcapture,costandprogressreporting,managementcontrols,customerreaction,quality  X assuranceandanalysis,andtheInternet.Manyofthesesystemswereimplementedforthefirst  < timeinCensus2000.Thereareatotaloftwelvesystemsthatwewillevaluate.Ingeneral,we   willassesswhetherthecorrectrequirementsandproperfunctionalitywerespecifiedforeachof h  thesetwelvesystems,whetherthesystemsperformedadequatelyintermsofeitherimpacton L  dataqualityorinprovidingusefulmanagementinformation,andwhetherwespecifiedour 0  requirementsinatimelymanner.Wealsowillexamineanycontractmanagementissues,as   applicable.ԀÃThetwelvesystemstobeevaluatedareasfollows:TelephoneQuestionnaire   Assistance;CoverageEditFollowup;InternetQuestionnaireAssistance;InternetData x  Collection;OperationsControlSystem2000;LaptopComputersfortheAccuracyandCoverage \  Evaluation;AccuracyandCoverageEvaluation2000ControlSystem;MatchingandReview @ CodingSystemfortheAccuracyandCoverageEvaluation;PreAppointmentManagement $ System/AutomatedDecennialAdministrativeManagementSystem;AmericanFactFinder; l ManagementInformationSystem2000;andCensus2000DataCapture. P ddWhatWillWeLearn?  Evaluationreportswillbegeneratedusinginformationcollectedfromdebriefingswithprogram | managers,systemsusers,andothersaffiliatedwiththesystems.Questionnaireswillbe ` developedforeachsystemthatwilladdressgeneralissuesconcerningthesystemsfunctionality D andthecorrectandtimelyspecificationofitsrequirementsalongwithquestionsthatareunique ( toeachsystem.Weexpecttogaininsight,asappropriate,inareassuchasmaintenanceand p  securityneeds,respondentacceptance,initialinvestmentrequired,ease/difficultyofsetup, T reliability,leveloftrainingrequired,effectsoncoverageandresponserates,additionalcostsor 8  savings,andtechnologylifecycleissues. !    `     h      p      x  AutomationofCensusProcessesEvaluations +h%* ` <d<   Alistofkeyquestionstobeansweredforeachsystemfollows. H  (Ex8"3"  32E3  0    Didwespecifytherightrequirementsandfunctionality?3݌ (#(# Ќ  "3"  3|2E3  0    Didthesystemdowhatweneededittointermsofeitheritsimpactondataqualityorin   providingusefulmanagementinformation?3|݌ t(#(# Ќ  "3"  3Ɏ2E3  0    Didwedefineourrequirementsinatimelymanner?3Ɏ݌ X(#(# Ќ  "3"  32E3  0    Ifacontractorwashiredtoworkonthesystem,didthecontractoreffectivelycomplete  < therequiredtasks?3ޏ݌ (#(# Ќ  (R.1.a)TelephoneQuestionnaireAssistance  L  Telephonequestionnaireassistance(TQA)wasatollfreeserviceprovidedbyacommercial 0  phonecentertoanswerquestionsaboutCensus2000orthecensusquestionnaire.Thissystem   alsoincludedareverseCATI(computerassistedtelephoneinterview)operation.ForCensus   2000,TQAwasoperatedoutof22phonecentersnationwidefromMarchthroughJune2000. x   (R.1.b)CoverageEditFollowup @  CoverageEditFollowup(CEFU)wasanoutboundserviceoperatingoutof13callcentersto $ resolvecountdiscrepancies(coverageeditfailures)andtoobtainmissinginformationforlarge l households.TheCEFUwasconductedfromMaytoAugust2000. P  (R.1.c)InternetQuestionnaireAssistance   Internetquestionnaireassistance(IQA)wasanoperationthatallowedrespondentstousethe  CensusBureausInternetsitetoaskquestionsandreceiveanswersaboutthecensus | questionnaire,jobopportunities,orgeneralquestionsaboutthepurposeofthecensus.This ` servicewasoperativefromMarchthroughJune2000. D  (R.1.d)InternetDataCollection  p  FromMarchthroughApril2000,respondentstotheCensus2000shortformhadtheoptionof T completingtheircensusformelectronicallybyaccessingtheCensusBureausInternetsiteand 8  providinga22digitIDnumberfoundontheirformreceivedinthemail. !  (R.2.a)OperationsControlSystem2000  "  TheOperationsControlSystem(OCS)2000wasadecennialfieldinterfacesystemandwasused #d! forcontrol,tracking,andprogressreportingforallfieldoperationsconductedforthecensus, $H" includingproductionofmaterialsusedbyfieldstafftodotheirwork.Thissystemwasoperative %,# fromOctober1997throughOctober2000fortheprecensusanddecennialoperationalphases. t& $   d d   )"' (R.2.b)LaptopComputersforAccuracyandCoverageEvaluation  d TheAccuracyandCoverageEvaluationwasacoveragemeasurementmethodologythatwasused H todeterminethenumberofpeopleandhousingunitsmissedorcountedmorethanoncein , Census2000.Thelaptopcomputerswereusedtoconductpersonalandtelephoneinterviews.   Thisevaluationexaminesandassessestheuseoflaptopcomputersindeterminingcoverage   error.  t  (R.2.c)AccuracyandCoverageEvaluation2000ControlSystem   < TheAccuracyandCoverageEvaluation2000ControlSystemwasadecennialsystemtoaid   managementintrackingandcontrollingtheAccuracyandCoverageEvaluationfieldoperations. h   (R.2.d)MatchingandReviewCodingSystemfortheAccuracyandCoverageEvaluation  0  TheMatchingandReviewCodingSystemfortheAccuracyandCoverageEvaluationwasalso   referredtoastheAccuracyandCoverageEvaluationsurveymatchingsystem.Thesystem   providedforacomputermatchingofhousingunitsandpersonsfollowedbyaclericalreviewof |  unmatchedrecords.ThissystemwasusedattheCensusBureausNationalProcessingCenterin `  Jeffersonville,IN. D  (R.3.a)PreAppointmentManagementSystem/AutomatedDecennialAdministrative p  ManagementSystem  T ThePreAppointmentManagementSystem/AutomatedDecennialAdministrativeManagement 8 Systemwasanintegratedstructureofadministrativemanagementprogramsthatsupport  applicanttrackingandprocessing,backgroundchecks,selectionrecords,recruitingreports,  personnelandpayrollprocessing,andarchivingofhistoricaldata.Thissystemwasusedinthe  hiringoftemporaryworkersforthecensus. d  (R.3.b)AmericanFactFinder  , TheAmericanFactFinderisageneralizedelectronicsystemforaccessanddisseminationof t CensusBureaudata.ThesystemisavailablethroughtheInternetandoffersprepackageddata X productsandtheabilitytobuildcustomproducts.Thesystemwillserveasthevehiclefor <  accessinganddisseminatingdatafromCensus2000(aswellasthe1997EconomicCensusesand  ! theAmericanCommunitySurvey).ThesystemwasformerlyknownastheDataAccessand " DisseminationSystem.Census2000dataproductswillbeavailablethroughtheAmerican "  FactFinderbeganinJanuary2001. h  #h!  (R.3.c)ManagementInformationSystem 2000  %0# TheManagementInformationSystem(MIS)providesdecisionsupportfunctions,suchas, |& $ criticalpathanalysisandwhatifanalysis.Italsoprovidesinformationondates,theresponsible `' % organization,budget,costtodate,andcurrentprogressofCensus2000operations.TheMIS D(!& includesthemasteractivityschedule,theexecutiveinformationsystem,andthecostand ()"' progresssystem.DesignedasatoolforCensus2000,theMIShasanongoingfunction.  *#(  (R.3.d)Census2000DataCapture  *$) Thedatacaptureprocessisafullelectronicdatacaptureandprocessingsystemforimaging d  Census2000questionnaires.Thisprocessinvolves:1)thecheckinofpaperforms;2)the H scanningandimagingofthoseforms;and3)theuseofopticalmarkandopticalcharacter , recognitionstocapturedatafromcensusquestionnaireimagesthatconvertittoacomputer   readableformat.TheCensusBureauworkedwithprivatesectorcompaniestooperatefourdata   capturecentersthatwereoperativefromMarchthroughOctober2000.  t    `     h      p