Adequate Yearly Progress (AYP): Implementation of the No ...

Additional Indicators: In addition to pupil scores on assessments, AYP standards often include one or more supplemental indicators. Examples include high ... Togglenavigation EveryCRSReport.com TopicAreas About Donate AdequateYearlyProgress(AYP):ImplementationoftheNoChildLeftBehind Act July28,2004– August20,2009 RL32495 TitleI,PartAoftheElementaryandSecondaryEducationAct(ESEA),authorizesfinancialaidtolocaleducationalagencies(LEAs)fortheeducationofdisadvantagedchildrenandyouthatthepreschool,elementary,andsecondarylevels.Overthelastseveralyears,theaccountabilityprovisionsofthisprogramhavebeenincreasinglyfocusedonachievementandotheroutcomesforparticipatingpupilsandschools.Since1994,andparticularlyundertheNoChildLeftBehindActof2001(NCLB),akeyconceptembodiedintheserequirementsisthatof“adequateyearlyprogress(AYP)”forschools,LEAs,andstates.AYPisdefinedprimarilyonthebasisofaggregatescoresofvariousgroupsofpupilsonstateassessmentsofacademicachievement.TheprimarypurposeofAYPrequirementsistoserveasthebasisforidentifyingschoolsandLEAswhereperformanceisunsatisfactory,sothatinadequaciesmaybeaddressedfirstthroughprovisionofincreasedsupportand,ultimately,avarietyof“consequences.” UnderNCLB,theTitleI-Arequirementsforstate-developedstandardsofAYPweresubstantiallyexpanded.AYPcalculationsmustbedisaggregated—determinedseparatelyandspecificallyfornotonlyallpupilsbutalsoforseveraldemographicgroupsofpupilswithineachschool,LEA,andstate.Inaddition,whileAYPstandardshadtobeappliedpreviouslyonlytopupils,schools,andLEAsparticipatinginTitleI-A,AYPstandardsunderNCLBmustbeappliedtoallpublicschools,LEAs,andtostatesoverall,ifastatechoosestoreceiveTitleI-Agrants.However,consequencesforfailingtomeetAYPstandardsneedbeappliedunderfederallawonlytoschoolsandLEAsparticipatinginTitleI-A.Anothermajorbreakwiththepre-NCLBperiodisthatstateAYPstandardsmustincorporateconcretemovementtowardmeetinganultimategoalofallpupilsreachingaproficientoradvancedlevelofachievementby2014. TheoverallpercentageofpublicschoolsidentifiedasfailingtomakeAYPforoneormoreyearsonthebasisoftestscoresin2007-2008wasapproximately35%ofallpublicschools.Thepercentageofschoolsforindividualstatesin2007-2008variedfrom7%to80%.Approximately13%ofpublicschoolswereinthe“needsimprovement”status(i.e.,theyhadfailedtomeetAYPstandardsfor2consecutiveyearsormore)onthebasisofAYPdeterminationsfor2007-2008andprecedingschoolyears. TheAYPprovisionsofNCLBarechallengingandcomplex,andtheyhavegeneratedsubstantialinterestanddebate.DebatesregardingNCLBprovisionsonAYPhavefocusedontheprovisionforanultimategoal,useofconfidenceintervalsanddata-averaging,populationdiversityeffects,minimumpupilgroupsize(n),separatefocusonspecificpupilgroups,numberofschoolsidentifiedandstatevariationstherein,the95%participationrule,statevariationsinassessmentsandproficiencystandards,andseveralissuesspecifictotheuseofgrowthmodelstodetermineAYP.TheauthorizationforESEAprogramsexpiredattheendofFY2008,andthe111thCongressmayconsiderwhethertoamendandextendtheESEA.Thisreportwillbeupdatedregularlytoreflectmajorlegislativedevelopmentsandavailableinformation. DownloadPDF DownloadEPUB RevisionHistory Aug.20,2009 HTML · PDF Oct.31,2008 HTML · PDF Aug.2,2007 HTML · PDF Aug.31,2006 HTML · PDF Oct.26,2005 HTML · PDF Jul.28,2004 HTML · PDF Metadata Topicareas EducationPolicy ReportType:CRSReport Source:EveryCRSReport.com,UniversityofNorthTexasLibrariesGovernmentDocumentsDepartment RawMetadata:JSON AdequateYearlyProgress(AYP):ImplementationoftheNoChildLeftBehind Act August20,2009 (RL32495) Contents Background:TitleIOutcomeAccountabilityandtheAYPConcept GeneralElementsofAYPProvisions GenericAYPFactors AYPProvisionsUndertheIASAof1994 ConcernsAbouttheAYPProvisionsoftheIASA AYPUnderNCLBStatute ImplementationoftheAYPProvisionsofNCLBbyED InitialImplementationActions DevelopmentsFollowingInitialImplementationoftheNCLBbyEDGrowth Models StateRevisionsofTheirAccountabilityPlans DataonSchoolsandLEAsIdentifiedasFailingtoMeetAYP SchoolsFailingtoMeetAYPStandardsforOneYear SchoolsFailingtoMeetAYPStandardsforTwoConsecutiveYears(andAnyAdditionalYears) LEAsFailingtoMeetAYPStandards IssuesinStateImplementationofNCLBProvisions Introduction UltimateGoal ConfidenceIntervalsandData-Averaging PopulationDiversityEffects MinimumPupilGroupSize(n) SeparateFocusonSpecificPupilGroups NumberofSchoolsIdentifiedandStateVariationsTherein 95%ParticipationRule IssuesRegardingGrowthModelAlternativestoAYPModelsintheNCLBStatute AreGrowthModelsofAYPMoreFairandAccuratethanStatusorImprovementModels? DoStatesHaveSufficientResourcestoDevelopandImplementGrowthModels? AreGrowthModelsConsistentwithNCLB'sUltimateGoal? Tables Table1.CategoriesofPupilswithDisabilitieswithRespecttoAchievementStandards,Assessments,andAYPDeterminationsUnderESEATitleI-A Table2.ReportedPercentageofPublicSchoolsandLocalEducationalAgencies(LEAs)FailingtoMakeAdequateYearlyProgress(AYP)ontheBasisof2007-2008AssessmentResults Summary TitleI,PartAoftheElementaryandSecondaryEducationAct(ESEA),authorizesfinancialaidtolocaleducationalagencies(LEAs)fortheeducationofdisadvantagedchildrenandyouthatthepreschool,elementary,andsecondarylevels.Overthelastseveralyears,theaccountabilityprovisionsofthisprogramhavebeenincreasinglyfocusedonachievementandotheroutcomesforparticipatingpupilsandschools.Since1994,andparticularlyundertheNoChildLeftBehindActof2001(NCLB),akeyconceptembodiedintheserequirementsisthatof"adequateyearlyprogress(AYP)"forschools,LEAs,andstates.AYPisdefinedprimarilyonthebasisofaggregatescoresofvariousgroupsofpupilsonstateassessmentsofacademicachievement.TheprimarypurposeofAYPrequirementsistoserveasthebasisforidentifyingschoolsandLEAswhereperformanceisunsatisfactory,sothatinadequaciesmaybeaddressedfirstthroughprovisionofincreasedsupportand,ultimately,avarietyof"consequences." UnderNCLB,theTitleI-Arequirementsforstate-developedstandardsofAYPweresubstantiallyexpanded.AYPcalculationsmustbedisaggregated—determinedseparatelyandspecificallyfornotonlyallpupilsbutalsoforseveraldemographicgroupsofpupilswithineachschool,LEA,andstate.Inaddition,whileAYPstandardshadtobeappliedpreviouslyonlytopupils,schools,andLEAsparticipatinginTitleI-A,AYPstandardsunderNCLBmustbeappliedtoallpublicschools,LEAs,andtostatesoverall,ifastatechoosestoreceiveTitleI-Agrants.However,consequencesforfailingtomeetAYPstandardsneedbeappliedunderfederallawonlytoschoolsandLEAsparticipatinginTitleI-A.Anothermajorbreakwiththepre-NCLBperiodisthatstateAYPstandardsmustincorporateconcretemovementtowardmeetinganultimategoalofallpupilsreachingaproficientoradvancedlevelofachievementby2014. TheoverallpercentageofpublicschoolsidentifiedasfailingtomakeAYPforoneormoreyearsonthebasisoftestscoresin2007-2008wasapproximately35%ofallpublicschools.Thepercentageofschoolsforindividualstatesin2007-2008variedfrom7%to80%.Approximately13%ofpublicschoolswereinthe"needsimprovement"status(i.e.,theyhadfailedtomeetAYPstandardsfor2consecutiveyearsormore)onthebasisofAYPdeterminationsfor2007-2008andprecedingschoolyears. TheAYPprovisionsofNCLBarechallengingandcomplex,andtheyhavegeneratedsubstantialinterestanddebate.DebatesregardingNCLBprovisionsonAYPhavefocusedontheprovisionforanultimategoal,useofconfidenceintervalsanddata-averaging,populationdiversityeffects,minimumpupilgroupsize(n),separatefocusonspecificpupilgroups,numberofschoolsidentifiedandstatevariationstherein,the95%participationrule,statevariationsinassessmentsandproficiencystandards,andseveralissuesspecifictotheuseofgrowthmodelstodetermineAYP.TheauthorizationforESEAprogramsexpiredattheendofFY2008,andthe111thCongressmayconsiderwhethertoamendandextendtheESEA.Thisreportwillbeupdatedregularlytoreflectmajorlegislativedevelopmentsandavailableinformation. AdequateYearlyProgress(AYP):ImplementationoftheNoChildLeftBehind Act Background:TitleIOutcomeAccountabilityandtheAYPConcept TitleI,PartAoftheElementaryandSecondaryEducationAct(ESEA),thelargestfederalK-12educationprogram,authorizesfinancialaidtolocaleducationalagencies(LEAs)fortheeducationofdisadvantagedchildrenandyouthatthepreschool,elementary,andsecondarylevels. Sincethe1988reauthorizationoftheESEA(TheAugustusF.Hawkins-RobertT.StaffordElementaryandSecondarySchoolImprovementAmendmentsof1988,or"SchoolImprovementAct,"P.L.100-297),theaccountabilityprovisionsofthisprogramhavebeenincreasinglyfocusedonachievementandotheroutcomesforparticipatingpupilsandschools.SincethesubsequentESEAreauthorizationin1994(theImprovingAmerica'sSchoolsActof1994,P.L.103-382),andparticularlyundertheNoChildLeftBehindActof2001(NCLB,P.L.107-110),akeyconceptembodiedintheseoutcomeaccountabilityrequirementsisthatof"adequateyearlyprogress(AYP)"forschools,LEAs,and(morerecently)statesoverall.TheprimarypurposeofAYPrequirementsistoserveasthebasisforidentifyingschoolsandLEAswhereperformanceisinadequate,sothattheseinadequaciesmaybeaddressed,firstthroughprovisionofincreasedsupportand,ultimately,throughavarietyofconsequences.1 ThisreportisintendedtoprovideanoverviewoftheAYPconceptandseveralrelatedissues,adescriptionoftheAYPprovisionsoftheNoChildLeftBehindAct,andananalysisoftheimplementationoftheseprovisionsbytheU.S.DepartmentofEducation(ED)andthestates.TheauthorizationforESEAprogramsexpiredattheendofFY2008,andthe111thCongressmayconsiderwhethertoamendandextendtheESEA.Thisreportwillbeupdatedregularlytoreflectmajorlegislativedevelopmentsandavailableinformation.GeneralElementsofAYPProvisions ESEATitleI,PartAhasincludedrequirementsforparticipatingLEAsandstatestoadministerassessmentsofacademicachievementtoparticipatingpupils,andtoevaluateLEAprogramsatleasteverytwoyears,sincetheprogramwasinitiatedin1965.However,relativelylittleattentionwaspaidtoschool-orLEA-wideoutcomeaccountabilityuntiladoptionoftheSchoolImprovementActof1988.2UndertheSchoolImprovementAct,requirementsforstatesandLEAstoevaluatetheperformanceofTitleI-Aschoolsandindividualparticipatingpupilswereexpanded.Inaddition,LEAsandstateswereforthefirsttimerequiredtodevelopandimplementimprovementplansforpupilsandschoolswhoseperformancewasnotimproving.However,incomparisontocurrentTitleI-Aoutcomeaccountabilityprovisions,theserequirementswerebroadandvague.UndertheSchoolImprovementActof1988,statesandLEAsweregivenlittledirectionastohowtheyweretodeterminewhetherperformancewassatisfactory,orhowperformancewastobedefined,withonepartialexception. TheexceptionappliedtoschoolsconductingschoolwideprogramsunderTitleI-A.Inschoolwideprograms,TitleI-Afundsmaybeusedtoimproveinstructionforallpupilsintheschool,ratherthanbeingtargetedononlythelowest-achievingindividualpupilsintheschool(asundertheothermajorTitleI-Aservicemodel,targetedassistanceschools).Underthe1988versionoftheESEA,schoolwideprogramswerelimitedtoschoolswhere75%ormoreofthepupilswerefromlow-incomefamilies(currentlythisthresholdhasbeenreducedto40%).TheSchoolImprovementActrequiredschoolwideprograms,inordertomaintaintheirspecialauthority,todemonstratethattheacademicachievementofpupilsintheschoolwashigherthaneitherofthefollowing:(a)theaveragelevelofachievementforpupilsparticipatinginTitleI-AintheLEAoverall;or(b)theaveragelevelofachievementfordisadvantagedpupilsenrolledinthatschoolduringthethreeyearsprecedingschoolwideprogramimplementation. TheembodimentofoutcomeaccountabilityinthespecificconceptofAYPbeganwiththe1994ImprovingAmerica'sSchoolsAct(IASA).UndertheIASA,statesparticipatinginTitleI-AwererequiredtodevelopAYPstandardsasabasisforsystematicallydeterminingwhetherschoolsandLEAsreceivingTitleI-Agrantswereperformingatanacceptablelevel.FailuretomeetthestateAYPstandardswastobecomethebasisfordirectingtechnicalassistance,andultimatelyconsequences,towardschoolsandLEAswhereperformancewasconsistentlyunacceptable.GenericAYPFactors BeforeproceedingtoadescriptionoftheTitleI-AAYPprovisionsundertheIASAof1994andtheNCLBof2001,weoutlinebelowthegeneraltypesofmajorprovisionsfrequentlyfoundinAYPprovisions,actualorproposed. PrimaryBasis:AYPrequirementsarebasedprimarilyonaggregatemeasuresofacademicachievementbypupils.AslongasTitleI-AhascontainedAYPprovisions,ithasprovidedthatthesebebasedultimatelyonstatestandardsofcurriculumcontentandpupilperformance,andassessmentslinkedtothesestandards.Morespecifically,theTitleI-Arequirementshavebeenfocusedonthepercentageofpupilsscoringatthe"proficient"orhigherlevelofachievementonstateassessments,notacommonnationalstandard.However,whenAYPprovisionswerefirstadoptedin1994,statesweregivenanextendedperiodoftimetoadoptandimplementthesestandardsandassessments,andforalengthyperiodafterthe1994amendments,various"transitional"performancestandardsandassessmentswereusedtomeasureacademicachievement.3 UltimateGoal:AYPstandardsmayormaynotincorporateanultimategoal,whichmayberelativelyspecificanddemanding(e.g.,allpupilsshouldreachtheproficientorhigherlevelofachievement,asdefinedbyeachstate,inaspecifiednumberofyears),ormoreambiguousandlessdemanding(e.g.,pupilachievementlevelsmustincreaseinrelationtoeitherLEAorstateaveragesorpastperformance).Ifthereisaspecificultimategoal,theremayalsoberequirementsforspecific,numerical,annualobjectiveseitherforpupilsintheaggregateorforeachofseveralpupilgroups.Theprimarypurposeofsuchagoalistorequirethatlevelsofachievementcontinuouslyincreaseovertimeinordertobeconsideredsatisfactory. SubjectAreas:Withrespecttosubjectareas,AYPstandardsmightfocusonlyonreadingandmathachievement,ortheymightincludeadditionalsubjectareas. AdditionalIndicators:Inadditiontopupilscoresonassessments,AYPstandardsoftenincludeoneormoresupplementalindicators.Examplesincludehighschoolgraduationrates,attendancerates,orassessmentscoresinsubjectsotherthanthosethatarerequired. LevelsatWhichApplied:StatesmayberequiredtodevelopAYPstandardsfor,andapplythemto,schools,LEAs,orstatesoverall.Further,itmayberequiredthatAYPstandardsbeapplicabletoallschoolsandLEAs,oronlytothoseparticipatinginESEATitleI-A. DisaggregationofPupilGroups:AYPstandardsmightbeappliedsimplytoallpupilsinaschool,LEA,orstate,ortheymightalsobeappliedseparatelyandspecificallytoavarietyofdemographicgroupsofpupils—suchaseconomicallydisadvantagedpupils,pupilswithdisabilities,pupilsindifferentethnicorracialgroups,orlimitedEnglishproficientpupils.InaprogramsuchasTitleI-A,thepurposeofwhichistoimproveeducationforthedisadvantaged,itmaybeespeciallyimportanttoconsiderselecteddisadvantagedpupilgroupsseparately,toidentifysituationswhereoverallpupilachievementmaybesatisfactorybuttheperformanceofoneormoredisadvantagedpupilgroupsisnot. BasicStructure:WhileAYPdefinitionsorstandardsmayvaryinamultitudeofrespects,theirbasicstructuregenerallyfallsintooneofthreegeneralcategories.TheNoChildLeftBehindActstatuteplacesprimaryemphasisononeofthesemodels,whileincorporatingasecondmodelasanexplicitlyauthorizedalternative.Inrecentyears,criticsofcurrentpolicyhaveincreasinglyfocusedtheirattentiononathirdmodelofAYP,whichisnowauthorizedthroughregulationsandSecretarialwaivers. ThethreebasicstructuralformsforAYPofschoolsorLEAsarethegroupstatus,successivegroupimprovement,andindividualgrowthmodels.Thekeycharacteristicofthegroupstatusmodelisafixed"annualmeasurableobjective"(AMO),orrequiredthresholdlevelofachievement,thatisthesameforallpupilgroups,schools,andLEAsstatewideinagivensubjectandgradelevel.Underthismodel,performanceatapointintimeiscomparedtoabenchmarkatthattime,withnodirectconsiderationofchangesoverapreviousperiod. ThekeycharacteristicofthesuccessivegroupimprovementmodelisafocusontherateofchangeinachievementinasubjectareafromoneyeartothenextamonggroupsofpupilsinagradelevelataschoolorLEA(e.g.,thepercentageofthisyear's5thgradepupilsinaschoolwhoareataproficientorhigherlevelinmathematicscomparedtothepercentageoflastyear's5thgradepupilswhowereataproficientorhigherlevelofachievement). Finally,thekeycharacteristicoftheindividualgrowthmodelisafocusonthepastorprojectedrateofchangeinthelevelofachievementamongthesamepupils.Suchmodelsmaycomparecurrentperformanceofspecificpupilstopastperformance,ormayprojectfutureperformanceofpupilsbasedonpastchangesintheirperformancelevel.Growthmodelsarelongitudinal,baseduponthetrackingofthesamepupilsastheyprogressthroughtheirK-12educationcareers.Whiletheprogressofpupilsistrackedindividually,resultsaretypicallyaggregatedwhenusedforaccountabilitypurposes.Aggregationmaybebydemographicgroup,byschoolorLEA,orotherrelevantcharacteristics.Ingeneral,growthmodelswouldgivecreditformeetingstepsalongthewaytoproficiencyinwaysthatastatusmodeltypicallydoesnot.4 TohelpillustratethebasicdifferencesamongthesethreeAYPmodels,simplifiedexamplesofbasicaspectsofeacharedescribedbelow.Thereadershouldkeepinmindmanyothervariationsofthesemodeltypesarepossible.Agroupstatusmodel,suchasthecurrentprimarymodelofAYPundertheNCLB(describedfurtherbelow),establishesaseriesofthresholdlevelsorAMOs,whicharepercentagesofpupilsscoringataproficientorhigherlevelofachievementonstatestandards-basedassessmentsofreadingandmathematics.TheseAMOshaveastartingpointandaseriesofincreasestoward(inthecaseoftheNCLB)anultimategoalof100%ofpupilsataproficientorhigherlevelofachievement,coveringamulti-yearperiod(fortheNCLB,theperiodof2001-2002through2013-2014).TheseAMOsarespecifictoeachgradelevelandsubject(readingormathematics)atwhichstateassessmentsareadministered.AkeyfeatureoftheAMOsinthismodelisthattheyarethesameforallpupilgroups—the"allpupil"groupaswellaseachofthedemographicsubgroupsspecifiedundertheNCLB(pupilswithdisabilities,pupilsfromlow-incomefamilies,pupilswithlimitedEnglishproficiency,etc.).Thismodelfocusessolelyoncurrentyearperformanceofthepupilscurrentlyenrolledineachschool/LEAforeverygradelevelatwhichassessmentsareadministered.5ComparisonstopreviousyearperformanceplaynoroleinAYPdeterminations. Anexampleofasuccessivegroupimprovementmodelisthesecondary("safeharbor")modelauthorizedundertheNCLB.Underthismodel,asembodiedintheNCLB,thebasicstructureoftheAYPsystemisthesameasdescribedabove,buttheprimaryfocusshiftstothechangefromthepreviousyearforeachgroup.IfanyspecifieddemographicgroupfailstomeettheprimarygroupstatusAYPcriteriondescribedabove,theschoolorLEAisstilldeemedtomeetAYPstandardsifthepercentageofpupilsscoringbelowtheproficientleveldeclinesby10%incomparisontothepreviousyearforpupilsinthatgradelevel,subjectarea,anddemographicgroup.Thus,theprimaryfocusisonthechangeinachievementfromthepreviousyear,comparing(forexample)theperformanceofthisyear'spupilsfromlow-incomefamiliesintheparticularschool/LEA/gradeleveltolastyear'spupilsfromlow-incomefamiliesenrolledinthatschool/LEA/gradelevel(i.e.,thepupilsareinthesamedemographiccategory,butarenotnecessarilythesamepupils).6 Anindividualgrowthmodelbeginsbytrackingtheperformanceofindividualpupilsovermultiple(atleasttwo)years.Theperformanceofpupilsinthesamegradelevelwhosharerelevantdemographiccharacteristicswithinaschool,LEA,orthestateoverallmaybecombinedintoacohort.Thechangeinscoresforthiscohortiscomparedtoastandardofexpectedgrowth.Theexpectedgrowthmaybeeither"data-driven"(e.g.,thestatewideaveragerateofachievementgrowthforallpupils,orthepredictedrateofgrowthstatewideforpupilswithsimilardemographiccharacteristics),or"policy-driven"(amulti-yeargrowthpathsufficienttomeetanultimategoal,suchastheNCLBrequirementforallpupilstoreachaproficientorhigherlevelofachievementby2013-2014).AschoolorLEAisdeemedtomeetAYPrequirementsiftheachievementgrowthofeachrelevantcohortofpupilsmeetstheexpectedlevelofgrowth.GrowthmodelsasusedforAYPdeterminationsunderNCLBshouldbedistinguishedfromvalue-addedmodels.Value-addedmodelsincorporateavarietyofstatisticalcontrols,adjustmentstoaccountforpupildemographiccharacteristicsorpastachievement,tosharpenthefocusonestimatingtheimpactofspecificteachers,schools,orLEAsonpupilachievementandtomeasurepupilgrowthagainstpredictedgrowthforpupilswithsimilarcharacteristics.TheapplicationofsuchcontrolshasnotbeenallowedbyEDingrowthmodelsusedforAYPdeterminationsunderESEATitleI-A.Proponentsarguethatsuchmodels,withtheircontrolsforbackgroundcharacteristicsandpastlearning,maximizethefocusonfactorsthatareunderthecontrolofteachersandotherschoolstaff.TheTennesseeValue-AddedAssessmentSystem(TVAAS)isonespecificformofgrowthmodelthatusespupilbackgroundcharacteristics,previousperformance,andotherdataasstatisticalcontrolsinordertofocusonestimatingthespecificeffectsofparticularschools,districts,teachers,orprogramsonpupilachievement.7 AssessmentParticipationRate:Itmightberequiredthataspecifiedminimumpercentageofaschool'sorLEA'spupilsparticipateinassessmentsinorderfortheschoolorLEAtobedeemedtohavemetAYPstandards.Theprimarypurposesofsucharequirementaretoassurethatassessmentresultsarebroadlyrepresentativeoftheachievementleveloftheschool'spupils,andtominimizetheincentivesforschoolstafftodiscouragetestparticipationbypupilsdeemedlikelytoperformpoorlyonassessments. ExclusionofCertainPupils:Beyondgeneralparticipationraterequirements(seeabove),statesmaybespecificallyrequiredtoinclude,orallowedtoexclude,certaingroupsofpupilsindeterminingwhetherschoolsorLEAsmeetAYPrequirements.Forexample,statutoryprovisionsmightallowtheexclusionofpupilswhohaveattendedaschoolforlessthanoneyearindeterminingwhetheraschoolmeetsAYPstandards. SpecialProvisionsforPupilswithParticularEducationalNeeds:Beyondrequirementsthatallpupilsbeincludedinassessments,withaccommodationswhereappropriate,theremaybespecialprovisionsforlimitedEnglishproficient(LEP)pupilsorpupilswiththemostsignificantcognitivedisabilities. AveragingorOtherStatisticalManipulationofData:Finally,thereareavarietyofwaysinwhichstatisticalmanipulationofAYP-relateddataorcalculationsmightbeeitherauthorizedorprohibited.Majorpossibilitiesincludeaveragingoftestscoredataoverperiodsoftwoormoreyears,ratherthanuseofthelatestdatainallcases;ortheuseof"confidenceintervals"incalculatingwhethertheaggregateperformanceofaschool'spupilsisatthelevelspecifiedbythestate'sAYPstandards.Thesetechniques,andtheimplicationsoftheiruse,arediscussedfurtherbelow.Ingeneral,theirusetendstoimprovethereliabilityandvalidityofAYPdeterminations,whileoftenreducingthenumberofschoolsorLEAsidentifiedasfailingtomeetAYPstandards.AYPProvisionsUndertheIASAof1994 UndertheIASA,statesweretodevelopandimplementAYPstandardssoonafterenactment.However,statesweregivenseveralyears(generallyuntilthe2000-2001schoolyear)todevelopandimplementcurriculumcontentstandards,pupilperformancestandards,andassessmentslinkedtotheseforatleastthreegradelevelsinmathandreading.8Thus,duringtheperiodbetweenadoptionoftheIASAin1994andofNCLBinearly2002,formoststatestheAYPprovisionswerebasedon"transitional"assessmentsandpupilperformancestandardsthatwerewidelyvaryinginnature.AYPstandardsbasedonsuch"transitional"assessmentswereconsideredtobe"transitional"themselves,with"final"AYPstandardstobebasedonstates'"final"assessments,whenimplemented.ThesubjectareasrequiredtobeincludedinstateAYPstandards(asopposedtorequiredassessments)werenotexplicitlyspecifiedinstatute;EDpolicyguidancerequiredstatestoincludeonlymathandreadingachievementindeterminingAYP,andtheinclusioninAYPstandardsofothermeasureswasoptional. WithrespecttotheultimategoalofthestateAYPstandards,theIASAprovidedbroadlythattheremustbecontinuousandsubstantialprogresstowardagoalofhavingallpupilsmeettheproficientandadvancedlevelsofachievement.However,notimelinewasspecifiedforreachingthisgoal,andmoststatesdidnotincorporateitintotheirAYPplansinanyconcreteway. TheIASA'sAYPstandardsweretobeappliedtoschoolsandLEAs,butnottothestatesoverall.Further,whilestateswereencouragedtoapplytheAYPstandardstoallpublicschoolsandLEAs,statescouldchoosetoapplythemonlytoschoolsandLEAsparticipatinginTitleI-A,andmostdidsolimittheirapplication. TheIASAprovidedthatallrelevantpupils9weretobeincludedinassessmentsandAYPdeterminations,althoughassessmentsweretoincluderesultsforpupilswhohadattendedaschoolforlessthanoneyearonlyintabulatingLEA-wideresults(i.e.,notforindividualschools).LEPpupilsweretobeassessedinthelanguagethatwouldbestreflecttheirknowledgeofsubjectsotherthanEnglish;andaccommodationsweretobeprovidedtopupilswithdisabilities. Importantly,whiletheIASArequiredstateassessmentstoultimately(by2000-2001)providetestresultsthatweredisaggregatedbypupildemographicgroups,itdidnotrequiresuchdisaggregationofdatainAYPstandardsandcalculations.The1994statuteprovidedthatstateAYPstandardsmustconsiderallpupils,"particularly"economicallydisadvantagedandLEPpupils,butdidnotspecifythattheAYPdefinitionmustbebasedoneachofthesepupilgroupsseparately.Finally,thestatutewassilentwithrespecttodata-averagingorotherstatisticaltechniques,aswellasthebasicstructureofstateAYPstandards(i.e.,whethera"groupstatus,""successivegroupimprovement,"or"individualgrowth"modelmustbeemployed).ConcernsAbouttheAYPProvisionsoftheIASA Thus,theIASA'sprovisionsforstateAYPstandardsbrokenewgroundconceptually,butwerecomparativelybroadandambiguous.Althoughstateswererequiredtoadoptandimplementatleast"transitional"AYPstandards,onthebasisof"transitional"stateassessmentresults,soonafterenactmentoftheIASA,theywerenotrequiredtoadopt"final"AYPstandards,inconjunctionwithfinalassessmentsandpupilperformancestandards,untilthe2000-2001schoolyear.Further,stateswerenotallowedtoimplementmostconsequences,suchasreconstitutingschoolstaff,untiltheyadoptedfinalassessments,sotheseprovisionswerenotimplementedbymoststatesuntiltheIASAwasreplacedbyNCLB. AcompilationwaspreparedbytheConsortiumforPolicyResearchinEducation(CPRE)ofthe"transitional"AYPstandardsthatstateswereapplyinginadministeringtheirTitleI-Aprogramsduringthe1999-2000schoolyear.10Overall,accordingtothiscompilation,thestateAYPdefinitionsfor1999-2000werewidelyvariedandsometimescomplex.GeneralpatternsintheseAYPstandards,outlinedbelow,reflectstateinterpretationoftheIASA'sstatutoryrequirements.Mostconsideredonlyachievementtestscores,butsomeconsideredavarietyofadditionalfactors,mostoftendropoutratesorattendancerates. Often,thestateAYPstandardssetathresholdofsomeminimumpercentage,orminimumrateofincreaseinthepercentage,ofpupilsattheproficientorhigherlevelofachievementonacompositeofstatetests.Thesethresholdswereoftenbased,atleastinpart,onperformanceofpupilsinaschoolorLEArelativetostatewideaveragesortotheschool'sortheLEA'sperformanceinthepreviousyear.SeveralstatesidentifiedschoolsasfailingtomakeAYPiftheyfailtomeet"expectedgrowth"inperformanceonthebasisoffactorssuchasinitialachievementlevelsandstatewideaverageachievementtrends.Thesethresholdsalmostneverincorporateda"ladder"ofmovementtowardanultimategoaltobemetbysomespecificdate. WhilesomestateAYPstandardswerebasedonachievementresultsforasingleyear,theywerefrequentlybasedontwo-orthree-yearrollingaverages. TheAYPstandardsgenerallyreferredonlytoallpupilsinaschoolorLEAcombined,withoutaspecificfocusonanypupildemographicgroups.However,theAYPstandardsofsomestatesincludedafocusonasinglecategoryoflow-achievingpupilsseparatelyfromallpupils,andaveryfew(e.g.,Texas)includedaspecificfocusontheperformanceofseveralpupilgroups(AfricanAmerican,Hispanic,White,orEconomicallyDisadvantaged).Onestate(NewMexico)comparedschoolscorestopredictedscoresonthebasisofsuchfactorsaspupildemographics. ThestateAYPstandardsundertheIASAweresometimessubstantiallyadjustedfromyear-to-year(oftenwithconsequentwidevariationsinthepercentageofTitleI-Aschoolsidentifiedasneedingimprovement).AccordingtoCPRE,twostates(IowaandNewHampshire)leftAYPstandardsanddeterminationsalmosttotallytoindividualLEAsin1999-2000.AreportpublishedbyEDin2004,onthebasisofstateAYPpoliciesforthe2001-2002schoolyear,containssimilarconclusionsaboutstateAYPpoliciesintheperiodimmediatelyprecedingimplementationofNCLB.11TherewastremendousvariationamongthestatesintheimpactoftheirAYPpoliciesundertheIASAonthenumberandpercentageofTitleI-AschoolsandLEAsthatwereidentifiedasfailingtomeettheAYPstandards.Insomestates,asubstantialmajorityofTitleI-AschoolswereidentifiedasfailingtomakeAYP,whileinothersalmostnoschoolsweresoidentified.InJuly2002,justbeforetheinitialimplementationofthenewAYPprovisionsofNCLB,EDreleasedacompilationofthenumberofschoolsidentifiedasfailingtomeetAYPstandardsfortwoormoreconsecutiveyears(andthereforeidentifiedasbeinginneedofimprovement)in2001-2002(formoststates)or2000-2001(instateswhere2001-2002datawerenotavailable).12Thenationaltotalnumberoftheseschoolswas8,652;thenumberinindividualstatesrangedfromzeroinArkansasandWyomingto1,513inMichiganand1,009inCalifornia.13Whilethereareobviousdifferencesinthesizeofthesestates,therewerealsowidevariationsinthepercentageofallschoolsparticipatinginTitleI-AthatfailedtomeetAYPforeitheroneyearortwoormoreconsecutiveyears.AYPUnderNCLBStatute NCLBprovisionsregardingAYPmaybeseenasanevolutionof,andtoasubstantialdegreeasareactiontoperceivedweaknessesin,theAYPrequirementsofthe1994IASA.Thelatterwerefrequentlycriticizedasbeinginsufficientlyspecific,detailed,orchallenging.Criticismoftenfocusedspecificallyontheirfailuretofocusonspecificdisadvantagedpupilgroups,failuretorequirecontinuousimprovementtowardanultimategoal,andtheirrequiredapplicabilityonlytoschoolsandLEAsparticipatinginTitleI-A,nottoallpublicschoolsortostatesoverall. UnderNCLB,theTitleI-Arequirementsforstate-developedstandardsofAYPweresubstantiallyexpandedinscopeandspecificity.AsundertheIASA,AYPisdefinedprimarilyonthebasisofaggregatescoresofpupilsonstateassessmentsofacademicachievement.However,underNCLB,stateAYPstandardsmustalsoincludeatleastoneadditionalacademicindicator,whichinthecaseofhighschoolsmustbethegraduationrate.TheadditionalindicatorsmaynotbeemployedinawaythatwouldreducethenumberofschoolsorLEAsidentifiedasfailingtomeetAYPstandards. OneofthemostimportantdifferencesbetweenAYPstandardsunderNCLBandpreviousrequirementsisthatunderNCLB,AYPcalculationsmustbedisaggregated;thatis,theymustbedeterminedseparatelyandspecificallyfornotonlyallpupilsbutalsoforseveraldemographicgroupsofpupilswithineachschool,LEA,andstate.Testscoresforanindividualpupilmaybetakenintoconsiderationmultipletimes,dependingonthenumberofdesignatedgroupsofwhichtheyareamember(e.g.,apupilmightbeconsideredaspartoftheLEPandeconomicallydisadvantagedgroups,aswellasthe"allpupils"group).Thespecifieddemographicgroupsareasfollows:economicallydisadvantagedpupils, LEPpupils, pupilswithdisabilities,and pupilsinmajorracialandethnicgroups,aswellasallpupils.However,asisdiscussedfurtherbelow,therearethreemajorconstraintsontheconsiderationofthesepupilgroupsinAYPcalculations.First,pupilgroupsneednotbeconsideredincaseswheretheirnumberissorelativelysmallthatachievementresultswouldnotbestatisticallysignificantortheidentityofindividualpupilsmightbedivulged.14Asisdiscussedfurtherbelow,theselectionoftheminimumnumber(n)ofpupilsinagroupforthegrouptobeconsideredinAYPdeterminationshasbeenleftlargelytostatediscretion.Statepoliciesregarding"n"havevariedwidely,withimportantimplicationsforthenumberofpupilgroupsactuallyconsideredinmakingAYPdeterminationsformanyschoolsandLEAs,andthenumberofschoolsorLEAspotentiallyidentifiedasfailingtomakeAYP.Second,ithasbeenlefttothestatestodefinethe"majorracialandethnicgroups"onthebasisofwhichAYPmustbecalculated.Andthird,asundertheIASA,pupilswhohavenotattendedthesameschoolforafullyearneednotbeconsideredindeterminingAYPfortheschool,althoughtheyarestilltobeincludedinLEAandstateAYPdeterminations. Incontrasttothepreviousstatute,underwhichAYPstandardshadtobeappliedonlytopupils,schools,andLEAsparticipatinginTitleI-A,AYPstandardsunderNCLBmustbeappliedtoallpublicschools,LEAs,andforthefirsttimetostatesoverall,ifastatechoosestoreceiveTitleI-Agrants.However,consequencesforfailingtomeetAYPstandardsneedonlybeappliedunderfederallawtoschoolsandLEAsparticipatinginTitleI-A. AnothermajorbreakwiththepastisthatstateAYPstandardsmustincorporateconcretemovementtowardmeetinganultimategoalofallpupilsreachingaproficientoradvancedlevelofachievementbytheendofthe2013-2014schoolyear.Thesteps—thatis,requiredlevelsofachievement—towardmeetingthisgoal,knownasAnnualMeasurableObjectives(AMOs),mustincreasein"equalincrements"overtime.Thefirstincreaseinthethresholdsmustoccurafternomorethantwoyears,andremainingincreasesatleastonceeverythreeyears.Asisdiscussedfurtherbelow,severalstateshaveaccommodatedthisrequirementinwaysthatrequiremuchmorerapidprogressinthelateryearsoftheperiodleadingupto2013-2014thanintheearlierperiod. TheNCLBAYPprovisionsincludeanassessmentparticipationraterequirement.InorderforaschooltomeetAYPstandards,atleast95%ofallpupils,aswellasatleast95%ofeachofthedemographicgroupsofpupilsconsideredforAYPdeterminationsfortheschoolorLEA,mustparticipateintheassessmentsthatserveastheprimarybasisforAYPdeterminations.15 TheprimarymodelofAYPundertheNCLBcurrentlyisagroupstatusmodel.Asnotedintheexampleabove,groupstatusmodelssetastheirAMOsthresholdlevelsofperformance,expressedspecificallyintermsofthepercentageofpupilsscoringataproficientorhigher(advanced)levelonstateassessmentsofreadingandmathematics.TheseAMOsmustbemetbyanyschoolorLEA,bothoverallandwithrespecttoallrelevantpupilsubgroups,inordertomakeAYP,whatevertheschool'sorLEA's"startingpoint"(forthemulti-yearperiodcoveredbytheaccountabilitypolicy)orperformanceinthepreviousyear.ThisAMO"uniformbar"isapplicabletoallpupilsubgroupsofsufficientsizetobeconsideredinAYPdeterminations.Thethresholdlevelsofachievementaretobesetseparatelyforreadingandmath,andmaybesetseparatelyforeachlevelofK-12education(elementary,middle,andhighschools).Forexample,itmightberequiredthat65%ormoreofthepupilsinanyofastate'spublicelementaryschoolsscoreattheproficientorhigherlevelofachievementinreadinginorderforaschooltomakeAYP.16 Theinitialminimumstartingpointforthe"uniformbar"wastobethegreaterof(a)thepercentageofpupilsattheproficientoradvancedlevelofachievementforthelowest-achievingpupilsubgroupinthebaseyear(2001-2002),or(b)thepercentageofpupilsattheproficientoradvancedlevelofachievementforthelowest-performingquintile(5th)17ofschoolsstatewideinthebaseyear.18The"uniformbar"mustgenerallyberaisedatleastonceeverythreeyears,althoughintheinitialperioditmustbeincreasedafternomorethantwoyears.Suchgroupstatusmodelsattempttoemphasizetheimportanceofmeetingcertainminimumlevelsofachievementforallpupilgroups,schools,andLEAs,andarguablyapplyconsistentexpectationstoallpupilgroups. ThesecondarymodelofAYPundertheNCLBcurrentlyisthe"safeharbor"provision,anexampleofasuccessivegroupimprovementmodel.ThisisanalternativeprovisionunderwhichschoolsorLEAsthatfailtomeettheusualrequirementsmaystillbedeemedtohavemadeAYPiftheymeetcertainotherconditions.AschoolwhereaggregateachievementisbelowthelevelrequiredunderthegroupstatusmodeldescribedabovewouldstillbedeemedtohavemadeAYP,throughthe"safeharbor"provision,if,amongrelevantpupilgroupswhodidnotmeettheprimaryAYPstandard,thepercentageofpupilswhoarenotattheproficientorhigherlevelintheschooldeclinesbyatleast10%19,andthosepupilgroupsmakeprogressonatleastoneotheracademicindicatorincludedinthestate'sAYPstandards.20Forexample,ifthestandardAMOis65%,andaschoolfailstomeetAYPbecauseoftheperformanceofonepupilgroup(e.g.,themathperformanceofwhitepupils)forwhomthepercentagescoringataproficientorhigherlevelthepreviousyearwas30%,thentheschoolcouldstillmakeAYPifthepercentageofwhitepupilsscoringataproficientorhigherlevelinmathincreasestoatleast37%(the30%fromthepreviousyearplus10%of(100%-30%),orsevenpercentagepoints). AthirdmodelofAYP,individualgrowth,isnotexplicitlyauthorizedbytheNCLB/ESEAstatute.However,asdiscussedlaterinthisreport,ithasbeenallowedthroughwaiversandrevisedprogramregulations.Forthesakeofsimplicity,intheremainderofthisreportwewillrefertothethreeAYPmodelsbytheabbreviatedtitlesof"status,""improvement,"and"growth"models.ImplementationoftheAYPProvisionsofNCLBbyEDInitialImplementationActions StatesbegandeterminingAYPforschools,LEAs,andthestatesoverallonthebasisofNCLBprovisionsbeginningwiththe2002-2003schoolyear.ThedeadlineforstatestosubmittoEDtheirAYPstandardsbasedonNCLBprovisionswasJanuary31,2003,andaccordingtoED,allstatesmetthisdeadline.OnJune10,2003,EDannouncedthataccountabilityplanshadbeenapprovedforallstates.However,manyoftheapprovedplansrequiredstatestotakeadditionalactionsfollowingsubmissionoftheirplan.21 AspectsofstateAYPplansthatapparentlyreceivedspecialattentioninED'sinitialreviewsincluded(1)thepaceatwhichproficiencylevelsareexpectedtoimprove(e.g.,equalincrementsofimprovementovertheentireperiod,ormuchmorerapidimprovementexpectedinlateryearsthanatthebeginning);(2)whetherschoolsorLEAsmustfailtomeetAYPwithrespecttothesamepupilgroup(s),gradelevel(s),orsubjectareastobeidentifiedasneedingimprovement,orwhethertwoconsecutiveyearsoffailuretomeetAYPwithrespecttoanyofthesecategoriesshouldleadtoidentification;22(3)thelengthoftimeoverwhichpupilsshouldbeidentifiedasbeingLEP;(4)theminimumsizeofpupilgroupsinaschoolinorderforthegrouptobeconsideredinAYPdeterminationsorforreportingofscores;(5)whethertoallowschoolscreditforraisingpupilscoresfrombelowbasictobasic(aswellasfrombasicorbelowtoproficientorabove)inmakingAYPdeterminations;and(6)whethertoallowuseofstatisticaltechniquessuchas"confidenceintervals"(i.e.,whetherscoresarebelowtherequiredleveltoastatisticallysignificantextent)inAYPdeterminations.DevelopmentsFollowingInitialImplementationoftheNCLBbyEDGrowth Models InNovember2005,theSecretaryofEducationannouncedagrowthmodelpilotprogramunderwhichinitiallyupto10stateswouldbeallowedtousegrowthmodelstomakeAYPdeterminationsforthe2005-2006orsubsequentschoolyears.23InDecember2007,theSecretaryliftedthecaponthenumberofstatesthatcouldparticipateinthegrowthmodelpilot,andregulationspublishedinOctober200824incorporatethisexpandedpolicy.Themodelsproposedbythestatesmustmeetatleastthefollowingcriteria:theymustincorporateanultimategoalofallpupilsreachingaproficientorhigherlevelofachievementbytheendofthe2013-2014schoolyear; achievementgapsamongpupilgroupsmustdeclineinorderforschoolsorLEAstomeetAYPstandards; annualachievementgoalsforpupilsmustnotbesetonthebasisofpupilbackgroundorschoolcharacteristics; annualachievementgoalsmustbebasedonperformancestandards,notpastor"typical"performancegrowthrates; theassessmentsystemmustproducecomparableresultsfromgrade-to-gradeandyear-to-year;and theprogressofindividualstudentsmustbetrackedwithinastatedatasystem.Inaddition,applicantstatesmusthavetheirannualassessmentsforeachofgrades3-8approvedbyED,andtheseassessmentsmusthavebeeninplaceforatleastoneyearprevioustoimplementationofthegrowthmodels. InJanuary2006,EDpublishedpeerreviewguidanceforgrowthmodelpilotapplications.25Ingeneral,thisguidanceelaboratesupontherequirementsdescribedabove,withspecialemphasisonthefollowing:(a)pupilgrowthtargetsmaynotconsidertheir"race/ethnicity,socioeconomicstatus,schoolAYPstatus,oranyothernon-academic"factor;(b)growthtargetsaretobeestablishedonthebasisofachievementstandards,nottypicalgrowthpatternsorpastachievement;and(c)thestatemusthavealongitudinalpupildatasystem,capableoftrackingindividualpupilsastheymoveamongschoolsandLEAswithinthestate. TherequirementsforgrowthmodelsofAYPunderED'spoliciesarerelativelyrestrictive.Themodelsmustbeconsistentwiththeultimategoalofallpupilsataproficientorhigherlevelby2013-2014,amajorgoalofthestatutoryAYPprovisionsofNCLB.Moresignificantly,theymustincorporatecomparableannualassessments,atleastforeachofgrades3-8plusatleastoneseniorhighschoolyear,andthoseassessmentsmustbeapprovedbyEDandinplaceforatleastoneyearbeforeimplementationofthegrowthmodel.Further,allperformanceexpectationsmustbeindividualized,andthestatemusthaveaninfrastructureofastatewide,longitudinaldatabaseforindividualpupils.Proposedmodelswouldhavetobestructuredaroundexpectationsandperformanceofindividualpupils,notdemographicgroupsofpupilsinaschoolorLEA,althoughindividualresultswouldhavetobeaggregatedforthedemographicgroupsdesignatedinNCLB. Twostates,NorthCarolinaandTennessee,wereapprovedtouseproposedgrowthmodelsinmakingAYPdeterminationsonthebasisofassessmentsadministeredinthe2005-2006schoolyear.26Thirteenadditionalstates—Alaska,Arkansas,Arizona,Colorado,Delaware,Florida,Iowa,Michigan,Minnesota,Missouri,Ohio,Pennsylvania,andTexas—havebeenapprovedtoparticipateinthepilotprogramsubsequently.Thegrowthmodelsforindividualstatesarebrieflydescribedbelow.UnderAlaska'sgrowthmodel,pupilswillbeincludedintheproficientgroupiftheirachievementleveltrajectoryisonagrowthpathtowardproficiencywithinthreeadditionalyearsforpupilsingrades4-9,orwithintwoadditionalyearsforpupilsingrade10.(Alaskacurrentlyhasnostandards-basedassessmentsforgradesbeyond10.)Pupilsinthe3rdgrade(theearliestgradeatwhichstateassessmentsareadministered)willbemeasuredonthebasisofstatusonly,notgrowth.Thegrowthmodelwillnotapplytopupilswithdisabilitieswhotakealternateassessments. InArizona,thegrowthmodelwillbeapplicabletopupilsingrades4-8only.Pupilswillbeincludedintheproficientgroupiftheirachievementleveltrajectoryisonagrowthpathtowardproficiencywithinthreeyearsorby8thgrade,whichevercomesfirst.Pupilsinthe3rdgrade(theearliestgradeatwhichstateassessmentsareadministered)willbemeasuredonthebasisofstatusonly,notgrowth.Unlikesomeotherstatesparticipatinginthegrowthmodelpilot,pupilswithdisabilitieswhotakethestate'salternateassessmentwillbeincludedintheArizonagrowthmodel. UndertheArkansaspolicy,AYPwillbecalculatedeachyearonthebasisofbothstatutoryprovisionsandusingthestate'sgrowthmodel,andaschoolwillmeetAYPstandardsifitqualifiesusingeithermethod.Underthegrowthmodel,pupilsingrades4-8willbedeemedtobeproficientiftheyareonagrowthpathtowardproficiencybytheendof8thgrade.Pupilsalreadyproficientmustbeonapathtocontinuetobeproficientthroughgrade8(i.e.,growthpathcriteriawillbeappliedtoallpupils,proficientandnon-proficient). InColorado,thegrowthmodelwillbeapplicabletopupilsingrades4-10,butwillnotincludepupilswithdisabilitieswhotakealternateassessments.Pupilswillbeincludedintheproficientgroupiftheirachievementleveltrajectoryisonagrowthpathtowardproficiencywithinthreeyearsorgrade10(ifearlier);growthcalculationswillincludecurrently-proficientstudentsonlyiftheyareonatrajectorytomaintainproficiencyoverthenextthreeyearsorgrade10.AYPwillbecalculatedeachyearonthebasisofbothstatutoryprovisionsandusingthestate'sgrowthmodel,andaschoolwillmeetAYPstandardsifitqualifiesusingeithermethod. UndertheDelawaregrowthmodel,AYPwillbecalculatedeachyearonthebasisofboththestatutoryprovisionsandusingthestate'sgrowthmodel,andaschoolwillmeetAYPstandardsifitqualifiesusingeithermethod.Individualpupilperformancewillbetrackedfromoneyeartothenext.Specifiednumbersofpointswillbeawardedonthebasisofchanges(ifany)inpupils'performancelevel;pointswillbeawardedforpartialmovementtowardproficiency,butnotformovementbeyondproficiency.TheaveragegrowthscoresforschoolsandLEAstomeetAYPstandardsincreasesteadilyuntil2013-2014,bywhichtimeallpupilswouldbeexpectedtoachieveataproficientorhigherlevel.27 UndertheFloridamodel,AYPwillbedeterminedseparatelyforeachpupilsubgroupineachschoolorLEA(i.e.,notforschoolsorLEAsasawhole)usingthestatutorymodelsplusagrowthmodel,andtheschoolorLEAwillmeetAYPstandardsifeachpupilsubgroupmakesAYPusinganyofthethreemodels.Florida'sgrowthmodelwillbeessentiallythesameasthecurrentstatusmodelexceptthatproficientpupilswillincludeboththosecurrentlyscoringataproficientorhigherlevelandthosewhoareonanindividualpathtowardproficiencywithinthreeyears.ThemodelwillbeappliedtoAYPdeterminationsforgrades3-10(withsomemodificationsforpupilsingrade3). UndertheIowamodel,pupiltestsscorerangesbelowproficienthavebeendividedintothreecategories:HiMarginal,LoMarginal,andWeak.Astudentwhorisesfromoneoftheselevelstoahigherlevel,andhasnotpreviouslyattainedthehigherlevel,willbedeemedtohavemet"AdequateYearlyGrowth"(AYG).ForschoolsandLEAsthathavenotmetAYPthoughapplicationofthestandardstatusandsafeharbormodels,studentsmakingAYGwillbeaddedtothosescoringproficientorabove,andthiscombinedtotalwillbeusedindeterminingwhethertheschoolorLEAmakesAYPfortheyear.StudentsbeginningattheWeaklevelmustreachproficiencywithinthreeyears,thosebeginningatLoMarginalmustbecomeproficientwithintwoyears,andthosebeginningatHiMarginalmustreachproficiencywithinoneyear.By2014,thegrowthmodelwouldnolongerbeused,andallpupilswillbeexpectedtoachieveataproficientorhigherlevel. InMichigan,studentshavebeendeemedtobeproficientiftheirachievementtestscoresareataproficientoradvancedlevel,orifthescoresofindividualstudentsarewithintwostandarderrorsofmeasurement(ineffect,a95%confidenceinterval)ofthetestscorecutpointforproficiency28(suchstudentsareconsideredtobe"provisionallyproficient").Thegrowthmodeladdsathirdcategoryofstudents"ontrajectory"towardproficiency.Todeterminewhetherstudentsareontrajectorytowardproficiency,eachoftheproficiencylevelsisdividedintothreesub-levels.Similar,butslightlydifferent,proceduresareappliedtoMichigan'sMI-AccessFunctionalIndependencealternateassessment.Thegrowthmodeldoesnotcoverhighschoolstudents.Ifastudent'sperformanceimprovesoverthepreviousyearbyanumberofsub-levelssuchthat,iftheimprovementcontinuedatthesamerateinthefuture,theywouldreachproficiencywithinthreeyears,theyarecountedasbeingontrajectorytowardproficiency.Thus,thetotalofstudentsscoringataproficientlevelplusnon-proficientstudentsonatrajectorytowardproficiencywithinthreeyearsplusthosewhoareprovisionallyproficientwouldbecomparedtothetotalnumberofstudentstestedineachrelevantsubgroup. InMinnesota,boththestatutorymodelsofAYPandagrowthmodelwillbeappliedinallAYPdeterminations,andAYPwillbemadeifeitherofthesecriteriaaremet.TheMinnesotagrowthmodelincorporatesa"valuetable"underwhichvaryingamountsofpartialcreditwillbegivenforgrowthamongsub-levelsofachievementbelowproficient.Thepartialcreditwillbegreater,thegreaterthestudent'sachievementgrowth.TheresultingcalculationswillconvertedtoascaleconsistentwiththestandardAMOsinreadingandmathematicstodetermineiftheAMOshavebeenmet.Pupilsatallgradelevelswillbeincluded,aswellaspupilswithdisabilitiestakingalternateassessments,aslongastwoconsecutiveyearsofassessmentresultsinMinnesotaareavailableforthepupils. InMissouri,ifstudentscurrentlyscoringbelowaproficientlevelareontracktobeproficientwithineitherfouryearsorby8thgrade,whicheveroccursfirst,theywillbeaddedtothenumberofstudentscurrentlyscoringataproficientorhigherlevel.Studentsingrades3and8willbeevaluatedonthebasisofthestatusmodeland"safeharbor"only.Noconfidenceintervalswillbeappliedtogrowthmodelcalculations.OnlythecurrentstatusandsafeharbormodelswillusedforAYPdeterminationsforgrades9-12.Studentswithdisabilities,includingthosetakingthestate'salternateassessmentforstudentswiththemostseverecognitivedisabilities,willbeincludedinthegrowthmodel,applyingtrajectoriesandachievementlevelsassociatedwitheithertheregularoralternateassessments. TheNorthCarolinapolicyaddsaprojectioncomponenttothecurrentgroupstatusmodel.Iftheachievementlevelofanon-proficientpupilisonatrajectorytowardproficiencywithinfouryears,thenthepupilisaddedtotheproficientgroup.Thetrajectorycalculationswillbemadeforpupilsinthe3rdthrough8thgrades. Ohiohasadoptedavariationofthe"projection"or"ontracktoproficiency"approachthatiscommontotheNorthCarolina,Tennessee,Arkansas,andFloridamodels.Afterapplicationofthestandardstatusandsafeharbormodels,ifanypupilgroupfailstomeetAYP,thenadeterminationwillbemadeifasufficientproportionofpupilsinthegroupisontracktowardmeetingtherequiredproficiencythresholdasofa"targetgrade."Inthecaseofelementaryandmiddleschools,thetargetgradewillbeeitherthegradelevelfollowingthehighestgradeofferedbytheschool(i.e.,foraK-5school,the6thgrade),orfourgradesbeyondthepupil'scurrentgrade,whichevercomesfirst.Inthecaseofahighschool,pupilswouldhavetobeontracktowardproficiencybythe11thgrade.Pupilscurrentlyscoringataproficientlevelbutwhoareprojectedtobebelowtheproficientlevelbythetargetgradewillnotbeconsideredtobeproficient. Pennsylvania'sgrowthmodelwillbeappliedincaseswhereAYPisnotmetunderthestatutorymodelsofAYP.Thegrowthmodelwillconsiderwhethereachpupilisprojectedtobeproficientinonetothreeyears(thetimeperiodvariesbygradelevel).Ifacurrentlyproficientpupilisprojectedtoscorebelowproficient,heorshewillbeconsiderednon-proficientunderthegrowthmodel. UndertheTennesseepolicy,schoolsandLEAswillhavetwooptionsformeetingAYP:meetingeithertheAYPstandardsunderthegroupstatusorsuccessivegroupimprovementmodelsofcurrentlaw,ormeetingAYPstandardsaccordingtoa"projectionmodel."Undertheprojectionmodel,pupilsaredeemedtobeataproficientorhigherlevelofachievementiftheirtestscoresareprojectedtobeataproficientorhigherlevelthreeyearsintothefuture,onthebasisofpastachievementlevelsforindividualpupils.Tennessee'sprojectionmodelwillnotbeappliedtohighschools. InTexas,aprojectioncomponentisaddedtothecurrentgroupstatusmodel.Iftheachievementlevelofanon-proficientpupilisprojectedtobeatorabovetheproficientlevelbythenext"highstakesgrade"(5,8,or11),thenthepupilisaddedtotheproficientgroup.Projectionswillbebasedoncurrentyearscoresforindividualpupilsinbothreadingandmathplusmeanscoresinreadingormath(dependingonthesubjectforwhichtheprojectionisbeingcalculated)fortheschooltheyattend.ThistechniquewillbeappliedtopupilstakingthegeneralstateassessmentaswellastheTAKS-MAlternateAssessment;pupilswithmoresubstantialcognitivedisabilitieswhotaketheTAKS-AltAlternateAssessmentwillbeincludedinthegrowthmodelbutusingadifferentmethodbasedontheirrateofimprovementamongsub-levelsofachievement.Overall,mostofthegrowthmodelsapprovedbyEDthusfararebaseduponsupplementingthenumberofpupilsscoringataproficientorhigherlevelwiththosewhoareprojected,ordeemedtobeonatrajectory,tobeataproficientlevelwithinalimitednumberofyears.Elevenofthefifteenapprovedmodelsfollowthisgeneralapproach.Amongthesestates,adistinctionmaybemadebetweeneightstates(NorthCarolina,Arkansas,Florida,Alaska,Arizona,Missouri,Michigan,andTexas)thatcombinecurrentlyproficientpupilswiththosenotproficientwhoare"ontrack"towardproficiency,andfourstates(Ohio,Pennsylvania,Tennessee,andColorado)thatconsideronlyprojectedproficiencylevelsforallpupils(i.e.,currentlyproficientpupilswhoarenotontracktoremainproficientarecountedasnotproficient)whenthegrowthmodelisapplied.Incontrast,themodelsusedbyatleastthreeotherstates—Delaware,Iowa,andMinnesota—focusonawardingcreditformovementofpupilsamongachievementcategoriesuptoproficiency. A2009evaluationreportbyEDfocusesonthetwostatesapprovedtouseagrowthmodelforAYPdeterminationsinthe2005-2006schoolyear,NorthCarolinaandTennessee.29Inthesetwostates,useofthegrowthmodelshadminimalimpactonAYPdeterminationsbasedon2005-2006testresults—noschoolsinNorthCarolinaandonlysevenschoolsinTennesseemadeAYPthroughuseofthegrowthmodelthatwouldnothavemadeAYPthroughthemethodsexplicitlyauthorizedintheESEA. October2008RegulationsonTitleI-AAssessmentsandAccountability SeveralnewfinalregulationsaffectingtheTitleI-Aassessment,AYP,andaccountabilitypolicieswerepublishedintheFederalRegisteronOctober29,2008(pp.64435-64513).MostoftheregulationsdealwithpolicyareasotherthanAYP.Manyoftheregulationsclarifypreviousregulationsorcodifyasregulationspoliciesthathavepreviouslybeenestablishedthroughlessformalmechanisms(suchaspolicyguidanceorpeerreviewerguidance).TheregulationsrelatedtoAYParebrieflydescribedbelow. GroupSize-RelatedProvisionsinStateAYPPolicies Statesmustprovideamoreextensiverationalethanpreviouslyrequiredfortheirselectionofminimumgroupsizes,useofconfidenceintervals,andrelatedaspectsoftheirAYPpolicies.Althoughnospecificlimitsareplacedontheseparameters,statesmustexplainintheirAccountabilityWorkbookshowtheirpoliciesprovidestatisticallyreliableinformationwhileminimizingtheexclusionofdesignatedpupilgroupsinAYPdeterminations,especiallyattheschoollevel.StatesmustalsoreportonthenumberofpupilsindesignatedgroupsthatareexcludedfromseparateconsiderationinAYPdeterminationsduetominimumgroupsizepolicies.Inaddition,theregulationscodifyprovisionsfortheNationalTechnicalAdvisoryCouncilthatwasestablishedinAugust2008toadvisetheSecretaryonavarietyoftechnicalaspectsofstatestandards,assessments,AYP,andaccountabilitypolicies. UndertheregulationsaspublishedinOctober2008,eachstatewouldhavebeenrequiredtosubmititsAccountabilityWorkbook,modifiedinaccordancewiththeregulations,toEDforanewroundoftechnicalassistanceandpeerreview.WorkbooksweretobesubmittedintimetoimplementanyneededchangesbeforemakingAYPdeterminationsbasedonassessmentresultsforthe2009-2010schoolyear.However,inalettertochiefstateschoolofficersdatedApril1,2009,theSecretaryofEducationstatedthatanewroundofpeerreviewsofstateAccountabilityWorkbookswouldnotbeconductedatthistime.30 AssessmentsandAccountabilityPoliciesinGeneral TheregulationsclarifythatassessmentsrequiredunderTitleI-Amayincludemultipleformatsaswellasmultipleacademicassessmentswithineachsubjectarea(reading,mathematics,andscience).Thisdoesnotincludetheconceptof"multiplemeasures,"asthistermhasbeenusedbymanytorefertoproposalstoexpandNCLBthroughinclusionofavarietyofindicatorsotherthanstandards-basedassessmentsinreading,mathematics,andscience.Also,statesarerequiredtoincluderesultsfromthemostrecentNationalAssessmentofEducationalProgress(NAEP)assessmentsontheirstateandLEAperformancereportcards.Further,EDpoliciesregardingprovisionsforstatestorequestwaiversallowingthemtousegrowthmodelsofAYParecodifiedintheOctober2008regulations(previouslytheywerepublishedonlyinpolicyguidanceandpeerreviewerguidancedocuments.) GraduationRates Numerouschangeshavebeenmadetopreviouspoliciesregardinggraduationratesusedasthe"additionalindicator"inAYPdeterminationsforhighschools.Previously,stateswereallowedasubstantialdegreeofflexibilityintheirmethodforcalculatinggraduationratesandwerenotrequiredtodisaggregatetheratesbypupilgroup(exceptforreportingpurposes).Also,althoughstateswererequiredtodeterminealevelof,orrateofimprovementin,graduationratesthatwouldbeadequateforAYPpurposes,theywerenotrequiredtosetanultimategoaltowardwhichtheseratesshouldbeprogressing. UndertheOctober2008regulations,statesmustadoptauniformmethodforcalculatinggraduationrates.Thismethodmustbeusedforschool,LEA,andstatereportcardsshowingresultsofassessmentsadministeredduringthe2010-2011schoolyear,andforpurposesofdeterminingAYPbasedonassessmentsadministeredduringthe2011-2012schoolyear(statesunabletomeetthesedeadlinesmayrequestanextension).ThismethodhasbeenendorsedbytheNationalGovernorsAssociation.Thegraduationrateisdefinedasthenumberofstudentswhograduatefromhighschoolinfouryears31dividedbythenumberofstudentsinthecohortforthestudents'class,adjustedforstudenttransfersamongschools.Statesmayalsoproposeusingasupplementaryextended-yeargraduationrate,inadditiontothefour-yearrate,inordertoaccommodateselectedgroupsofstudents(suchascertainstudentswithdisabilities)whomayneedmorethanfouryearstograduate.Thesegraduationratesmustbedisaggregatedbysubgroup. Statesmustsetanultimategoalforgraduationratesthattheyexpectallhighschoolstomeet.Nofederalstandardisestablished,butthestategoal,aswellasannualtargetstowardmeetingthatgoal,mustbeapprovedbyEDaspartofthestate'saccountabilitypolicy. PupilswithDisabilities SomeofthemostsubstantialofED'sAYPpolicychangesfollowingenactmentoftheNCLBinvolvespupilswithdisabilities.32First,regulationsaddressingtheapplicationoftheTitleI-AstandardsandassessmentrequirementstocertainpupilswithdisabilitieswerepublishedintheFederalRegisteronDecember9,2003(pp.68698-68708).Thepurposeoftheseregulationsistoclarifytheapplicationofstandard,assessment,andaccountabilityprovisionstopupils"withthemostsignificantcognitivedisabilities."Undertheregulations,statesandLEAsmayadoptalternateassessmentsbasedonalternateachievementstandards—alignedwiththestate'sacademiccontentstandardsandreflecting"professionaljudgmentofthehighestachievementstandardspossible"—foralimitedpercentageofpupilswithdisabilities.33ThenumberofpupilswhoseproficientorhigherscoresonthesealternateassessmentsmaybeconsideredasproficientoraboveforAYPpurposesislimitedtoamaximumof1.0%ofalltestedpupils(approximately9%ofallpupilswithdisabilities)atthestateandLEAlevel(thereisnolimitforindividualschools).SEAsmayrequestfromtheU.S.SecretaryofEducationanexceptionallowingthemtoexceedthe1.0%capstatewide,andSEAsmaygrantsuchexceptionstoLEAswithintheirstate.AccordingtoEDstaff,threestatesin2003-2004(Montana,Ohio,andVirginia),andfourstatesin2004-2005(theprecedingthreestatesplusSouthDakota),receivedwaiverstogomarginallyabovethe1.0%limitstatewide.Intheabsenceofawaiver,thenumberofpupilsscoringatthe"proficientorhigher"levelonalternateassessments,basedonalternateachievementstandards,inexcessofthe1.0%limitistobeaddedtothosescoring"belowproficient"inLEAorstate-levelAYPdeterminations. AnewEDpolicyaffectinganadditionalgroupofpupilswithdisabilitieswasannouncedinitiallyinApril2005,withfinalregulationsbasedonitpublishedintheFederalRegisteronApril9,2007.Thenewpolicyisdividedintoshort-termandlong-termphases.Itisfocusedonpupilswithdisabilitieswhoseabilitytoperformacademicallyisassumedtobegreaterthanthatofthepupilswith"themostsignificantcognitivedisabilities"discussedintheaboveparagraph,andwhoarecapableofachievinghighstandards,butmaynotreachgradelevelwithinthesametimeperiodastheirpeers.InED'sterminology,thesepupilswouldbeassessedusingalternateassessmentsbasedonmodifiedachievementstandards. Theshort-termpolicymayapply,withtheapprovaloftheSecretary,tostatesuntiltheydevelopandadministeralternativeassessmentsunderthelong-termpolicy(describedbelow).34Underthisshort-termpolicy,ineligiblestatesthathavenotyetadoptedmodifiedachievementstandards,schoolsmayaddtotheirproficientpupilgroupanumberofpupilswithdisabilitiesequalto2.0%ofallpupilsassessed(ineffect,deemingthescoresofallofthesepupilstobeattheproficientlevel).35ThispolicywouldbeapplicableonlytoschoolsandLEAsthatwouldotherwisefailmeetAYPstandardsduesolelytotheirpupilswithdisabilitiesgroup.Alternatively,ineligiblestatesthathaveadoptedmodifiedachievementstandards,schoolsandLEAsmaycountproficientscoresforpupilswithdisabilitiesontheseassessments,subjecttoa2.0%(ofallassessedpupils)capattheLEAandstatelevels. Thelong-termpolicyisembodiedinfinalregulationspublishedintheFederalRegisteronApril9,2007.Theseregulationsaffectstandards,assessments,andAYPforagroupofpupilswithdisabilitieswhoareunlikelytoachievegradelevelproficiencywithinthecurrentschoolyear,butwhoarenotamongthosepupilswiththemostsignificantcognitivedisabilities(whosesituationwasaddressedbyanearliersetofregulations,discussedabove).Forthissecondgroupofpupilswithdisabilities,stateswouldbeauthorizedtodevelop"modifiedacademicachievementstandards"andalternateassessmentslinkedtothese.Themodifiedachievementstandardsmustbealignedwithgrade-levelcontentstandards,butmayreflectreducedbreadthordepthofgrade-levelcontentincomparisontotheachievementstandardsapplicabletothemajorityofpupils.Thestandardsmustprovideaccesstograde-levelcurriculum,andnotprecludeaffectedpupilsfromearningaregularhighschooldiploma. Aswiththepreviousregulationsregardingpupilswiththemostsignificantcognitivedisabilities,therewouldbenodirectlimitonthenumberofpupilswhotakealternateassessmentsbasedonmodifiedachievementstandards.However,inAYPdeterminations,pupilscoresofproficientoradvancedonalternateassessmentsbasedonmodifiedachievementstandardsmaybecountedonlyaslongastheydonotexceedanumberequalto2.0%ofallpupilstestedatthestateorLEAlevel(i.e.,anestimated20%ofpupilswithdisabilities);suchscoresinexcessofthelimitwouldbeconsidered"non-proficient."Aswiththe1.0%capforpupilswiththemostsignificantcognitivedisabilities,this2.0%capdoesnotapplytoindividualschools.Ingeneral,LEAsorstatescouldexceedthe2.0%caponlyiftheydidnotreachthe1.0%limitwithrespecttopupilswiththemostsignificantcognitivedisabilities.Thus,ingeneral,scoresofproficientoraboveonalternateassessmentsbasedonalternateandmodifiedachievementstandardsmaynotexceedatotalof3.0%ofallpupilstestedatastateorLEAlevel.36Inparticular,statesarenolongerallowedtorequestawaiverofthe1.0%capregardingpupilswiththemostsignificantcognitivedisabilities. TheApril9,2007,regulationsalsoincludeprovisionsthatarewidelyapplicabletoAYPdeterminations.First,statesarenolongerallowedtousevaryingminimumgroupsizes("n")fordifferentdemographicgroupsofpupils.Thisprohibitsthepreviouslycommonpracticeofsettinghigher"n"sizesforpupilswithdisabilitiesorLEPpupilsthanforotherpupilgroups.Second,whenpupilstakestateassessmentsmultipletimes,statesandLEAsmayusethehighestscoreforpupilswhotaketestsmorethanonce.Finally,aswithLEPpupils,statesandLEAsmayincludethetestscoresofformerpupilswithdisabilitiesinthedisabilitysubgroupforuptotwoyearsaftersuchpupilshaveexitedspecialeducation.37 Insummary,therearenowfivegroupsofpupilswithdisabilitieswithrespecttoachievementstandards,assessments,andtheuseofscoresinAYPdeterminations.ThesegroupsaresummarizedbelowinTable1. Table1.CategoriesofPupilswithDisabilitieswithRespecttoAchievementStandards,Assessments,andAYPDeterminationsUnderESEATitleI-A TypeofContentStandards TypeofAchievementStandards TypeofAssessment Capon#ofProficientorAdvancedScoresThatMayBeIncludedinAYPDeterminations Grade-levelcontentstandards Grade-levelacademicachievementstandards Regular(i.e.,thesameasthatapplicabletopupilsgenerally) None Grade-levelcontentstandards Grade-levelacademicachievementstandards Regularwithaccommodations(e.g.,specialassistanceforthosewithsightorhearingdisabilities) None Grade-levelcontentstandards Grade-levelacademicachievementstandards Alternateassessmentsbasedonregular,grade-levelachievementstandards(e.g.,portfoliosorperformanceassessments) None Grade-levelcontentstandards Modifiedacademicachievementstandards Alternateassessmentsbasedonmodifiedacademicachievementstandards Ingeneral,2.0%ofallpupilsassessed Alternatecontentstandards Alternateacademicachievementstandards Alternateassessmentsbasedonalternateachievementstandards Ingeneral,1.0%ofallpupilsassessed ParticipationRates OnMarch29,2004,EDannouncedthatschoolscouldmeettherequirementthat95%ormoreofpupils(allpupilsaswellaspupilsineachdesignateddemographicgroup)participateinassessments(inorderfortheschoolorLEAtomakeAYP)onthebasisofaverageparticipationratesforthelasttwoorthreeyears,ratherthanhavingtoposta95%orhigherparticipationrateeachyear.Inotherwords,ifaparticulardemographicgroupofpupilsinapublicschoolhasa93%testparticipationrateinthemostrecentyear,buthada97%ratetheprecedingyear,the95%participationraterequirementwouldbemet.Inaddition,thenewguidancewouldallowschoolstoexcludepupilswhofailtoparticipateinassessmentsduetoa"significantmedicalemergency"fromtheparticipationratecalculations.Thenewguidancefurtheremphasizestheauthorityforstatestoallowpupilswhomissaprimaryassessmentdatetotakemake-uptests,andtoestablishaminimumsizefordemographicgroupsofpupilstobeconsideredinmakingAYPdeterminations(includingthoserelatedtoparticipationrates).AccordingtoED,insomestates,asmanyas20%oftheschoolsfailingtomakeAYPdidsoonthebasisofassessmentparticipationratesalone.Itisnotknownhowmanyoftheseschoolswouldmeetthenew,somewhatmorerelaxedstandard. LEPPupils InaletterdatedFebruary19,andproposedregulationspublishedonJune24,2004,EDofficialsannouncedtwonewpolicieswithrespecttoLEPpupils.38First,withrespecttoassessments,LEPpupilswhohaveattendedschoolsintheUnitedStates(otherthanPuertoRico)forlessthan10monthsmustparticipateinEnglishlanguageproficiencyandmathematicstests.However,theparticipationofsuchpupilsinreadingtests(inEnglish),aswellastheinclusionofanyofthesepupils'testscoresinAYPcalculations,istobeoptional(i.e.,schoolsandLEAsneednotconsiderthescoresoffirstyearLEPpupilsindeterminingwhetherschoolsorLEAsmeetAYPstandards).Suchpupilsarestillconsideredindeterminingwhetherthe95%testparticipationhasbeenmet. Second,inAYPdeterminations,schoolsandLEAsmaycontinuetoincludepupilsintheLEPdemographiccategoryforuptotwoyearsaftertheyhaveattainedproficiencyinEnglish.However,theseformerlyLEPpupilsneednotbeincludedwhendeterminingwhetheraschoolorLEA'scountofLEPpupilsmeetsthestate'sminimumsizethresholdforinclusionofthegroupinAYPcalculations,andscoresofformerlyLEPpupilsmaynotbeincludedinstate,LEA,orschoolreportcards.Boththeseoptions,ifexercised,shouldincreaseaveragetestscoresforpupilscategorizedasbeingpartoftheLEPgroup,andreducetheextenttowhichschoolsorLEAsfailtomeetAYPonthebasisofLEPpupilgroups. AYPDeterminationsforTargetedAssistanceSchools EDhasreleasedaFebruary4,2004,lettertoastatesuperintendentofeducationprovidingmoreflexibilityinAYPdeterminationsfortargetedassistanceschools.39TitleI-Aservicesareprovidedattheschoollevelviaoneoftwobasicmodels:targetedassistanceschools,whereservicesarefocusedonindividualpupilswiththelowestlevelsofacademicachievement,orschoolwideprograms,inwhichTitleI-Afundsmaybeusedtoimproveacademicinstructionforallpupils.Currently,mostTitleI-Aprogramsareintargetedassistanceschools,althoughthenumberofschoolwideprogramshasgrownrapidlyinrecentyears,andmostpupilsservedbyTitleI-Aareinschoolwideprograms. ThispolicylettergivesschoolsandLEAstheoptionofconsideringonlypupilsassistedbyTitleI-AforpurposesofmakingAYPdeterminationsforindividualschools.LEAandstatelevelAYPdeterminationswouldstillhavetobemadeonthebasisofallpublicschoolpupils.Theimpactofthisauthority,andtheextentofitsutilization,areunclear.Inschoolsusingthisauthority,therewouldbeanincreasedlikelihoodthatpupildemographicgroupswouldbebelowminimumsizetobeconsidered.Atthesametime,ifTitleI-Aparticipantsareindeedthelowest-performingpupilsintargetedassistanceschools,itseemsunlikelythatmanyschoolswouldchoosetobaseAYPdeterminationsonlyonthosepupils. FlexibilityforAreasAffectedbytheGulfCoastHurricanes FollowingthedamagetoschoolsystemsanddispersionofpupilsinthewakeofHurricanesKatrinaandRitainAugustandSeptember2005,interestwasexpressedbyofficialsofstatesandLEAsthatweredamagedbythestorms,orthatenrolledpupilsdisplacedbythesestorms,inthepossibilityofwaivingsomeofNCLB'sassessment,AYP,orotheraccountabilityrequirements.Inaseriesofpolicyletterstochiefstateschoolofficers(CSSOs),theSecretaryofEducationemphasizedformsofflexibilityalreadyavailableundercurrentlawandannouncedanumberofpolicyrevisionsandpotentialwaiversthatmightbegrantedinthefuture. InaSeptember29,2005,lettertoallCSSOs,40theSecretaryofEducationnotedthattheycouldexerciseexistingnaturaldisasterprovisionsofNCLB[§1116(b)(7)(D)and(c)(10)(F)]topostponetheimplementationofschoolorLEAimprovementdesignationsandconsequencesforschoolsorLEAsfailingtomeetAYPstandardsthatarelocatedinthemajordisasterareasinLouisiana,Alabama,Mississippi,Texas,orFlorida,withoutaspecificwaiverbeingrequired.Inaddition,waiversoftheserequirementscouldbeconsideredforotherLEAsorschoolsheavilyaffectedbyenrollinglargenumbersofevacueepupils.Further,allaffectedLEAsandschoolscouldestablishaseparatesubgroupfordisplacedstudentsinAYPdeterminationsonthebasisofassessmentsadministeredduringthe2005-2006schoolyear.Pupilswouldappearonlyintheevacueesubgroup,nototherdemographicsubgroups(e.g.,economicallydisadvantagedorLEP).Waiverscouldberequestedin2006toallowschoolsorLEAstomeetAYPrequirementsifonlythetestscoresoftheevacueesubgroupwouldpreventthemfrommakingAYP.Inanycase,allsuchstudentsmuststillbeassessedandtheassessmentresultsreportedtothepublic.41StateRevisionsofTheirAccountabilityPlans OvertheperiodfollowingtheinitialsubmissionandapprovalofstateaccountabilityplansforAYPandrelatedpoliciesin2003throughthepresent,manystateshaveproposedanumberofrevisionstotheirplans.SometimestheserevisionsseemclearlyintendedtotakeadvantageofnewformsofflexibilityannouncedbyEDofficials,suchasthosediscussedabove,whileinothercasesstatesappeartobeattemptingtotakeadvantageofoptionsorformsofflexibilitythatreportedlybeenapprovedforotherstatespreviously. TheproposedchangesinstateaccountabilityplanshaveapparentlyalmostalwaysbeeninthedirectionofincreasedflexibilityforstatesandLEAs,withreductionsanticipatedinthenumberorpercentageofschoolsorLEAsidentifiedasfailingtomakeAYP.Issuesthathavearisenwithrespecttothesechangesincludealackoftransparency,andpossiblyinconsistencies(especiallyovertime),inthetypesofchangesthatEDofficialshaveapproved;debatesoverwhethertheneteffectofthechangesistomaketheaccountabilityrequirementsmorereasonableortoundesirablyweakenthem;concernthatthechangesmaymakeanalreadycomplicatedaccountabilitysystemevenmorecomplex;andtiming—whetherdecisionsonproposedchangesarebeingmadeinatimelymannerbyED. Themajoraspectsofstateaccountabilityplansforwhichchangeshavebeenproposedandapprovedincludethefollowing:(a)changestotakeadvantageofrevisedfederalregulationsandpolicyguidanceregardingassessmentofpupilswiththemostsignificantcognitivedisabilities,LEPpupils,andtestparticipationrates;(b)limitingidentificationforimprovementtoschoolsthatfailtomeetAYPinthesamesubjectareafortwoormoreconsecutiveyears,andlimitingidentificationofLEAsforimprovementtothosethatfailedtomeetAYPinthesamesubjectareaandacrossallthreegradespansfortwoormoreconsecutiveyears;(c)usingalternativemethodstodetermineAYPforschoolswithverylowenrollment;(d)initiatingorexpandinguseofconfidenceintervalsinAYPdeterminations,including"safeharbor"calculations;(e)changing(usuallyeffectivelyincreasing)minimumgroupsize;and(f)changinggraduationratetargetsforhighschools.AccountabilityplanchangesthathavefrequentlybeenrequestedbutnotapprovedbyEDinclude(a)identificationofschoolsforimprovementonlyiftheyfailedtomeetAYPwithrespecttothesamepupilgroupandsubjectareafortwoormoreconsecutiveyears,and(b)retroactiveapplicationofnewformsofflexibilitytorecalculationofAYPforpreviousyears.42DataonSchoolsandLEAsIdentifiedasFailingtoMeetAYP ThemostrecentavailablecompilationsofstateAYPdataarediscussedbelowintwocategories:reportsfocusingonthenumberandpercentageofschoolsfailingtomeetAYPstandardsforoneormoreyearsversusreportsonthenumberandpercentageofpublicschoolsandLEAsidentifiedforimprovement—thatis,theyhadfailedtomeetAYPstandardsfortwoconsecutiveyearsormore.SchoolsFailingtoMeetAYPStandardsforOneYear Table2providesthepercentageofschoolsandLEAsfailingtomakeadequatelyyearlyprogress,onthebasisof2007-2008assessmentresults,foreachstate,asreportedbyED,basedonConsolidatedStatePerformanceReports.43ThesedataarebasedonallpublicschoolsandLEAsineachstate,notjustthoseparticipatinginTitleI-A.44Thepercentageofpublicschoolsfailingtomakeadequateyearlyprogressfor2007-2008variedwidelyamongthestates,from7%forOklahomaandWisconsinto76%forFlorida,77%fortheDistrictofColumbia,and80%forSouthCarolina.Nationwide,35%ofpublicschoolsfailedtomakeAYPbasedontestscoresforthe2007-2008schoolyear.Thisisanincreasecomparedto29%for2006-2007and27%for2005-2006. AccordingtotheEDreport,"TitleIImplementation—UpdateonRecentEvaluationFindings,"publishedbyEDin2009,ofschoolsfailingtomakeAYPinthe2005-2006schoolyear,35%didsowithrespecttoachievementinreadingormath(orboth)forthe"allpupils"group.Incontrast,24%ofschoolsfailingtomakeAYPdidsoonthebasisofachievementinreadingormath(orboth)foronlyonesubgroupwhilemakingAYPwithrespecttothe"allpupils"group,and20%ofschoolsfailingtomakeAYPdidsoonthebasisofachievementinreadingormath(orboth)fortwoormoresubgroupswhilemakingAYPwithrespecttothe"allpupils"group.Theremaining21%ofschoolsfailingtomakeAYPthatyeardidsowithrespecttotestparticipationratesonly(4%),the"otheracademicindicator"only(6%),orothercombinationsofAYPcriteria(11%).Amongschoolswithnumbersofpupilsineachofthedesignatedcategoriestomeettheminimumgroupsizecriterionfortheirstate,thepercentageofschoolsfailingtomakeAYPwithrespecttomathorreadingachievementin2005-2006wasfoundtovaryfrom2%fortheAsianpupilgroup,3%forWhitepupils,18%forpupilsfromlow-incomefamilies,20%forHispanicpupils,25%forAfrican-Americanpupils,30%forLEPpupils,and43%forpupilswithdisabilities.Withrespecttoeducationlevel,41%ofmiddleschoolsfailedtomakeAYPin2007-2008,comparedto34%ofhighschoolsand19%ofelementaryschools. Table2.ReportedPercentageofPublicSchoolsandLocalEducationalAgencies(LEAs)FailingtoMakeAdequateYearlyProgress(AYP)ontheBasisof2007-2008AssessmentResults State ReportedPercentageofRatedSchoolsNotMakingAYPBasedon2007-2008TestResults ReportedPercentageofLEAsNotMakingAYPBasedon2007-2008TestResults Alabama 16% 1% Alaska 41% 50% Arizona 27% 39% Arkansas 42% 16% California 48% 60% Colorado 43% 58% Connecticut 42% 74% Delaware 29% 32% DistrictofColumbia 77% 81% Florida 76% 97% Georgia 20% 70% Hawaii 58% 100% Idaho 44% 57% Illinois 32% 39% Indiana 46% 16% Iowa 31% 10% Kansas 10% 9% Kentucky 28% 40% Louisiana 19% naa Maine 34% 4% Maryland 17% 67% Massachusetts 63% 78% Michigan 27% 10% Minnesota 49% 58% Mississippi 14% 51% Missouri 57% 74% Montana 28% 32% Nebraska 20% 34% Nevada 40% 6% NewHampshire 62% 44% NewJersey 35% 15% NewMexico 68% 46% NewYork 16% 7% NorthCarolina 68% 92% NorthDakota 37% 39% Ohio 36% 48% Oklahoma 7% 7% Oregon 37% 59% Pennsylvania 28% 8% RhodeIsland 27% 37% SouthCarolina 80% 100% SouthDakota 16% 11% Tennessee 20% 9% Texas 15% 32% Utah 19% 14% Vermont 37% 39% Virginia 25% 57% Washington 62% 72% WestVirginia 19% 89% Wisconsin 7% 1% Wyoming 24% 8% PuertoRico 59% 100% NationalAverage 35% 35% Source:StateConsolidatedPerformanceReports;seehttp://www.ed.gov/admins/lead/account/consolidated/sy07-08part1/index.html.a.NA=Notavailable.Thus,thenationaltotalpercentageforLEAsexcludesthisstate. SchoolsFailingtoMeetAYPStandardsforTwoConsecutiveYears(andAnyAdditionalYears) EDrecentlyposted45datafromtheConsolidatedStatePerformanceReportsonthenumberofschoolsidentifiedforimprovement,correctiveaction,orrestructuringforthe2008-2009schoolyear,onthebasisofassessmentresultsthroughthe2007-2008schoolyear.Atotalof12,599schoolswereidentified,constitutingapproximately13%ofallpublicschools.AswiththepercentageofschoolsfailingtomakeAYP,thepercentageofschoolsidentifiedvariedwidelyamongthestates. AthemereflectedintheseresultsisahighdegreeofstatevariationinthepercentageofschoolsidentifiedasfailingtomeetAYPstandardsorasneedingimprovement.Thesevariationsappeartobebased,atleastinpart,notonlyonunderlyingdifferencesinachievementlevelsbutalsoondifferencesinthedegreeofrigororchallengeinstatepupilperformancestandards,andonvariationsinstate-determinedstandardsfortheminimumsizeofpupildemographicgroupsinorderforthemtobeconsideredinAYPdeterminationsofschoolsorLEAs.(Ingeneral,largerminimumsizesforpupildemographicgroupsreducethelikelihoodthatmanydisadvantagedgroups,suchasLEPpupilsorpupilswithdisabilities,willbeconsideredindeterminingwhetheraschoolorLEAmeetsAYP.)LEAsFailingtoMeetAYPStandards Althoughmostattention,inboththestatuteandimplementationactivities,thusfarhasbeenfocusedonapplicationoftheAYPconcepttoschools,alimitedamountofinformationisbecomingavailableaboutLEAsthatfailtomeetAYPrequirements,andtheconsequencesforthem.AsshowninTable2,accordingtotheConsolidatedStatePerformanceReportsreferredtoabove,approximately35%ofallLEAsfailedtomeetAYPstandardsonthebasisofassessmentresultsforthe2007-2008schoolyear.46Amongthestates,therewasevengreatervariationforLEAsthanforschools.Twostates—AlabamaandWisconsin—reportedthat1%oftheirLEAsfailedtomakeadequateyearlyprogress,while100%oftheLEAsinSouthCarolina,plusthesingle,statewideLEAinHawaii,failedtomeetAYPstandards.IssuesinStateImplementationofNCLBProvisionsIntroduction TheprimarychallengeassociatedwiththeAYPconceptistodevelopandimplementschool,LEA,andstateperformancemeasuresthatare:(a)challenging,(b)providemeaningfulincentivestoworktowardcontinuousimprovement,(c)areatleastminimallyconsistentacrossLEAsandstates,and(d)focusattentionespeciallyondisadvantagedpupilgroups.Atthesametime,itisgenerallydeemeddesirablethatAYPstandardsshouldallowflexibilitytoaccommodatemyriadvariationsinstateandlocalconditions,demographics,andpolicies,andavoidtheidentificationofsomanyschoolsandLEAsasfailingtomeetthestandardsthatmoraledeclinessignificantlysystemwideanditbecomesextremelydifficulttotargettechnicalassistanceandconsequencesonlow-performingschools.TheAYPprovisionsofNCLBarechallengingandcomplex,andhavegeneratedsubstantialcriticismfromseveralstates,LEAs,andinterestgroups.Manycriticsareespeciallyconcernedthateffortstodirectresourcesandapplyconsequencestolow-performingschoolswouldlikelybeineffectiveifresourcesandattentionaredispersedamongarelativelylargeproportionofpublicschools.OthersdefendNCLB'srequirementsasbeingameasuredresponsetotheweaknessesofthepre-NCLBAYPprovisions,whichweremuchmoreflexiblebut,asdiscussedabove,hadseveralweaknesses. TheremainderofthisreportprovidesadiscussionandanalysisofseveralspecificaspectsofNCLB'sAYPprovisionsthathaveattractedsignificantattentionanddebate.Theseincludetheprovisionforanultimategoal,useofconfidenceintervalsanddata-averaging,populationdiversityeffects,minimumpupilgroupsize(n),separatefocusonspecificpupilgroups,numberofschoolsidentifiedandstatevariationstherein,the95%participationrule,statevariationsinassessmentsandproficiencystandards,andseveralissuesspecifictotheuseofgrowthmodelstodetermineAYP. ItshouldbenotedthatthisreportfocusesonissuesthathavearisenintheimplementationofNCLBprovisionsonAYP.Assuch,itgenerallydoesnotfocusonalternativestothecurrentstatutoryprovisionsofNCLB.UltimateGoal Therequiredincorporationofanultimategoal—ofallpupilsataproficientorhigherlevelofachievementwithin12yearsofenactment—isoneofthemostsignificantdifferencesbetweentheAYPprovisionsofNCLBandthoseunderpreviouslegislation.SettingsuchadateisperhapstheprimarymechanismrequiringstateAYPstandardstoincorporateannualincreasesinexpectedachievementlevels,asopposedtotherelativelystaticexpectationsembodiedinmoststateAYPstandardsunderthepreviousIASA.Withoutanultimategoalofhavingallpupilsreachtheproficientlevelofachievementbyaspecificdate,statesmightsimplyestablishrelativegoals(e.g.,performancemustbeashighasthestateaverage)thatprovidenorealmovementtoward,orincentivesfor,significantimprovement,especiallyamongdisadvantagedpupilgroups. Nevertheless,agoalofhavingallpupilsataproficientorhigherlevelofachievement,within12yearsoranyotherspecifiedperiodoftime,maybeeasilycriticizedasbeing"unrealistic,"ifoneassumesthat"proficiency"hasbeenestablishedatachallenginglevel.ProponentsofsuchademandingultimategoalarguethatschoolsandLEAsfrequentlymeetthegoalsestablishedforthem,evenratherchallenginggoals,ifthegoalsareveryclearlyidentified,defined,andestablished,iftheyareattainable,andifitismadevisiblyclearthattheywillbeexpectedtomeetthem.Thisisincontrasttoapre-NCLBsystemunderwhichperformancegoalswereoftenvague,undemanding,andpoorlycommunicated,withfew,ifany,consequencesforfailingtomeetthem.Ademandinggoalmightmaximizeeffortstowardimprovementbystatepublicschoolsystems,evenifthegoalisnotmet.Further,ifalessambitiousgoalweretobeadopted,whatlowerlevelofpupilperformancemightbeacceptable,andforwhichpupils? Atthesametime,bysettingdeadlinesbywhichallpupilsmustachieveattheproficientorhigherlevel,theAYPprovisionsofNCLBcreateanincentiveforstatestoweakentheirpupilperformancestandardstomakethemeasiertomeet.Inmanystates,onlyaminorityofpupilsarecurrentlyachievingattheproficientorhigherlevelonstatereadingandmathematicsassessments.Eveninstateswherethepercentageofallpupilsscoringattheproficientorhigherlevelissubstantiallyhigher,thepercentageofthoseinmanyofthepupilgroupsidentifiedunderNCLB'sAYPprovisionsissubstantiallylower.Itwouldbeextremelydifficultforsuchstatestoreachagoalof100%oftheirpupilsattheproficientlevelwithoutreducingtheirperformancestandards. Therehasthusfarbeensomeapparentmovementtowardloweringproficiencystandardsinasmallnumberofstates.Reportedly,afewstateshaveredesignatedlowerstandards(e.g.,"basic"or"partiallyproficient")asconstitutinga"proficient"levelofperformanceforTitleI-Apurposes,orestablishednew"proficient"levelsofperformancethatarebelowlevelspreviouslyunderstoodtoconstitutethatlevelofperformance,andotherstateshaveconsideredsuchactions.47Forexample,insubmittingitsaccountabilityplan(whichwasapprovedbyED),Coloradostatedthatitwoulddeemstudentsperformingatbothits"proficient"and"partiallyproficient"levels,asdefinedbythatstate,asbeing"proficient"forNCLBpurposes.48Initssubmission,thestatearguedthat"Colorado'sstandardsforallstudentsremainhighincomparisontomoststates.Colorado'sbasicproficiencylevelonCSAPisalsohighincomparisontomoststates."Similarly,Louisianadecidedtoidentifyits"basic"levelofachievementasthe"proficient"levelforNCLBpurposes,statingthat"[t]hesestandardshavebeenshowntobehigh;forexample,equipercentileequatingofthestandardshasshownthatLouisiana's'Basic'issomewhatmorerigorousthanNAEP's'Basic.'Inaddition,representativesfromLouisiana'sbusinesscommunityandhighereducationhavevalidatedtheuseof'Basic'asthestate'sproficiencygoal."49 ThisisanaspectofNCLB'sAYPprovisionsonwhichtherewilllikelybecontinuingdebate.Itisunlikelythatanystate,andfewschoolsorLEAsofsubstantialsizeandaheterogeneouspupilpopulation,willmeetNCLB'sultimateAYPgoal,unlessstatestandardsofproficientperformancearesignificantlyloweredorstatesaggressivelypursuetheuseofsuchstatisticaltechniquesassettinghighminimumgroupsizesandconfidenceintervals(describedbelow)tosubstantiallyreducetherangeofpupilgroupsconsideredinAYPdeterminationsoreffectivelylowerrequiredachievementlevelthresholds. Somestateshaveaddressedthissituation,atleastintheshortrun,by"backloading"theirAYPstandards,requiringmuchmorerapidimprovementsinperformanceattheendofthe12-yearperiodthanatthebeginning.Thesestateshavefollowedtheletterofthestatutorylanguagethatrequiresincreasesof"equalincrements"inlevelsofperformanceafterthefirsttwoyears,andatleastonceeverythreeyearsthereafter.50However,theyhave"backloaded"thisprocessby,forexample,requiringincreasesonlyonceeverytwo-threeyearsatthebeginning,thenrequiringincreasesofthesamedegreeeveryyearforthefinalyearsoftheperiodleadingupto2013-2014.Forexample,bothIndianaandOhioestablishedincrementalincreasesinthethresholdlevelofperformanceforschoolsandLEAsthatareequalinsize,andthataretotakeeffectintheschoolyearsbeginningin2004,2007,2010,2011,2012,and2013.Asaresult,therequiredincreasesperyeararethreetimesgreaterduring2010-2013thaninthe2004-2009period.ThesestatesmaybetryingtopostponerequiredincreasesinperformancelevelsuntilNCLBprovisionsarereconsidered,andpossiblyrevised,byCongress.ConfidenceIntervalsandData-Averaging ManystateshaveusedoneorbothofapairofstatisticaltechniquestoattempttoimprovethevalidityandreliabilityofAYPdeterminations.Useofthesetechniquesalsotendstohaveaneffect,whetherintentionalornot,ofreducingthenumberofschoolsorLEAsidentifiedasfailingtomeetAYPstandards. Theaveragingoftestscoreresultsforvariouspupilgroupsovertwo-orthree-yearperiodsisexplicitlyauthorizedunderNCLB,andthisauthorityisusedbymanystates.Insomecases,schoolsorLEAsareallowedtoselectwhethertoaveragetestscoredata,andforwhatperiod(twoyearsorthree),whicheverismostfavorableforthem.Asdiscussedabove,recentpolicyguidancealsoexplicitlyallowstheuseofaveragingforparticipationrates. TheuseofanotherstatisticaltechniquewasnotexplicitlyenvisionedinthedraftingofNCLB'sAYPprovisions,butitsinclusionintheaccountabilityplansofseveralstateshasbeenapprovedbyED.Thisistheuseof"confidenceintervals,"usuallywithrespecttotestscores,butinacoupleofstatesalsotothedeterminationofminimumgroupsize(seebelow).Thisconceptisbasedontheassumptionthatanytestadministrationrepresentsa"samplesurvey"ofpupils'educationalachievementlevel.Aswithallsamplesurveys,thereisadegreeofuncertaintyregardinghowwellthesampleresults—averagetestscoresforthepupilgroup—reflectpupils'actuallevelofachievement.Aswithsurveys,thelargerthenumberofpupilsinthegroupbeingtested,thegreatertheprobabilitythatthegroup'saveragetestscorewillrepresenttheirtruelevelofachievement,allelsebeingequal.Putanotherway,confidenceintervalsareusedtoevaluatewhetherachievementscoresarebelowtherequiredthresholdtoastatisticallysignificantextent. "Confidenceintervals"maybeseenas"windows"surroundingathresholdtestscorelevel(i.e.,thepercentageofpupilsattheproficientorhigherlevelrequiredunderthestate'sAYPstandards).51Thesizeofthewindowvarieswithrespecttothenumberofpupilsintherelevantgroupwhoaretested,andwiththedesireddegreeofprobabilitythatthegroup'saveragescorerepresentstheirtruelevelofachievement.Thisisanalogoustothe"marginoferror"commonlyreportedalongwithopinionpolls.Whiletestresultsarenotbasedonasmallsampleoftherelevantpopulation,asareopinionpollresults,sincethetestsaretobeadministeredtothefull"universe"ofpupils,theresultsfromanyparticulartestadministrationareconsideredtobeonlyestimatesofpupils'truelevelofachievement,oroftheeffectivenessofaschoolorLEAineducatingspecifiedpupilgroups,andthusthe"marginoferror"or"confidenceinterval"conceptsaredeemedbymanytoberelevanttothesetestscores.Theprobability,orlevelofconfidence,ismostoftensetat95%,butinsomecasesmaybeaslowas90%orashighas99%—thatis,itis95%(or90%or99%)certainthatthetrueachievementlevelforagroupofpupilsiswithintherelevantconfidenceintervaloftestscoresaboveandbelowtheaveragescoreforthegroup.Allotherrelevantfactorsbeingequal,thesmallerthepupilgroup,andthehigherthedesireddegreeofprobability,thelargeristhewindowsurroundingthethresholdpercentage. Forexample,considerasituationwherethethresholdpercentageofpupilsattheproficientorhigherlevelofachievementinreadingforelementaryschoolsrequiredunderastate'sAYPstandardsis40%.Withoutapplyingconfidenceintervals,aschoolwouldsimplyfailtomakeAYPiftheaveragescoresofallofitspupils,orofanyofitsrelevantpupilgroupsmeetingminimumsizethresholds,isbelow40%.Incontrast,ifconfidenceintervalsareapplied,windowsareestablishedaboveandbelowthe40%threshold,turningthethresholdfromasinglepointtoavariablerangeofscores.Thesizeofthisscorerangeorwindowwillvarydependingonthesizeofthepupilgroupwhoseaveragescoresarebeingconsidered,andthedesireddegreeofprobability(95%or99%)thattheaverageachievementlevelsforpupilsineachgrouparebeingcorrectlycategorizedasbeing"truly"belowtherequiredthreshold.Inthiscase,aschoolwouldfailtomakeAYPwithrespecttoapupilgrouponlyiftheaveragescoreforthegroupisbelowthelowestscoreinthatrange.52 TheuseofconfidenceintervalstodeterminewhethergrouptestscoresfallbelowrequiredthresholdstoastatisticallysignificantdegreeimprovesthevalidityofAYPdeterminations,andaddressesthefactthattestscoresforanygroupofpupilswillvaryfromonetestadministrationtoanother,andthesevariationsmaybeespeciallylargeforarelativelysmallgroupofpupils.Atthesametime,theuseofconfidenceintervalsreducesthelikelihoodthatschoolsor(toalesserextent)LEAswillbeidentifiedasfailingtomakeAYP.Also,forrelativelysmallpupilgroupsandhighlevelsofdesiredaccuracy(especiallya99%probability),thesizeofconfidenceintervalsmayberelativelylarge.Ultimately,theuseofthistechniquemaymeanthattheaverageachievementlevelsofpupilgroupsinmanyschoolswillbewellbelow100%proficiencyby2013-2014,yettheschoolswouldstillmeetAYPstandardsbecausethegroups'scoresarewithintherelevantconfidenceinterval.PopulationDiversityEffectsMinimumPupilGroupSize(n) AnotherimportanttechnicalfactorinstateAYPstandardsistheestablishmentoftheminimumsize(n)forpupilgroupstobeconsideredinAYPcalculations.NCLBrecognizesthatinthedisaggregationofpupildataforschoolsandLEAs,theremightbepupilgroupsthataresosmallthataveragetestscoreswouldnotbestatisticallyreliable,orthedisseminationofaveragescoresforthegroupmightriskviolationofpupils'privacyrights. BoththestatuteandEDregulationsandotherpolicyguidancehavelefttheselectionofthisminimumnumbertostatediscretion.Whilemoststateshavereportedlyselectedaminimumgroupsizebetween30and50pupils,therangeofselectedvaluesfor"n"isratherlarge,varyingfromasfewasfivetoasmanyas200pupils53undercertaincircumstances.Onestate(NorthDakota)hassetnospecificlevelfor"n,"relyingonlyontheuseofconfidenceintervals(seeabove)toestablishreliabilityoftestresults.Althoughmoststateshavealwayssetastandardminimumsizeforallpupilgroups,somestatesuntilrecentlyestablishedhigherlevelsof"n"forpupilswithdisabilitiesorLEPpupils.54 Ingeneral,thehighertheminimumgroupsize,thelesslikelythatmanypupilgroupswillactuallybeseparatelyconsideredinAYPdeterminations.(Pupilswillstillbeconsidered,butonlyaspartofthe"allpupils"group,orpossiblyotherspecifiedgroups.)ThisgivesschoolsandLEAsfewerthresholdstomeet,andreducesthelikelihoodthattheywillbefoundtohavefailedtomeetAYPstandards.Inmanycases,ifapupilgroupfallsbelowtheminimumgroupsizeattheschoollevel,itisstillconsideredattheLEAlevel(whereitismorelikelytomeetthethreshold).Inaddition,sinceminimumgroupsizesforreportingachievementdataaretypicallylowerthanthoseusedforAYPpurposes,55scoresareoftenreportedforpupilgroupswhoarenotseparatelyconsideredinAYPcalculations.Atthesametime,relativelyhighlevelsfor"n"weakenNCLB'sspecificfocusonavarietyofpupilgroups,manyofthemdisadvantaged,suchasLEPpupils,pupilswithdisabilities,oreconomicallydisadvantagedpupils.SeparateFocusonSpecificPupilGroups ThereareseveralongoingissuesregardingNCLB'srequirementfordisaggregationofpupilachievementresultsinAYPstandards,namelytherequirementthatavarietyofpupilgroupsbeseparatelyconsideredinAYPcalculations.Thefirstofthesewasdiscussedimmediatelyabove:theestablishmentofminimumgroupsize,withthepossibleresultthatrelativelysmallpupilgroupswillnotbeconsideredintheschoolsandLEAsofstatesthatset"n"atacomparativelyhighlevel,especiallyinstatesthatsetahigherlevelforcertaingroups(e.g.,pupilswithdisabilities)thanothers. Asecondissuearisesfromthefactthatthedefinitionofthespecifiedpupilgroupshasbeenleftessentiallytostatediscretion.Thisisnoteworthyparticularlywithrespecttotwogroupsofpupils:LEPpupilsandpupilsinmajorracialandethnicgroups.RegardingLEPpupils,manyhavebeenconcernedaboutthedifficultyofdemonstratingthatthesepupilsareperformingataproficientlevelifthispupilgroupisdefinednarrowlytoincludeonlypupilsunabletoperforminregularEnglish-languageclassroomsettings.Inotherwords,ifpupilswhonolongerneedspeciallanguageservicesarenolongeridentifiedasbeingLEP,howwillitbepossibletobringthosewhoareidentifiedasLEPuptoaproficientlevelofachievement? IndevelopingtheirAYPstandards,somestatesaddressedthisconcernbyincludingpupilsintheLEPcategoryforoneormoreyearsaftertheynolongerneedspeciallanguageservices.Aswasdiscussedabove,EDhasrecentlypublishedpolicyguidanceencouragingallstatestofollowthisapproach,allowingthemtocontinuetoincludepupilsintheLEPgroupforuptotwoyearsafterbeingmainstreamedintoregularEnglishlanguageinstruction,andfurtherallowingthescoresofLEPpupilstobeexcludedfromAYPcalculationsforthefirstyearofpupils'enrollmentinUnitedStatesschools.Ifwidelyadopted,thesepoliciesshouldreducetheextentthatschoolsorLEAsareidentifiedasfailingtomeetAYPstandardsonthebasisoftheLEPpupilgroup. Anotheraspectofthisissuearisesfromthediscretiongiventostatesindefining"majorracialandethnicgroups."NeitherthestatutenorEDhasdefinedthisterm.Somestatesdefinedthetermrelativelycomprehensively(e.g.,MarylandincludesAmericanIndian,AfricanAmerican,Asian,White,andHispanicpupilgroups)andsomemorenarrowly(e.g.,Texasidentifiesonlythreegroups—White,AfricanAmerican,andHispanic).Amorenarrowinterpretationmayreducetheattentionfocusedonexcludedpupilgroups.ItwouldalsoreducethenumberofdifferentthresholdssomeschoolsandLEAswouldhavetomeetinordertomakeAYP. Afinal,overarchingissuearisesfromtherelationshipbetweenpupildiversityinschoolsandLEAsandthelikelihoodofbeingidentifiedasfailingtomeetAYPstandards.Allotherrelevantfactorsbeingequal(especiallytheminimumgroupsizecriteria),themorediversethepupilpopulation,themorethresholdsaschoolorLEAmustmeetinordertomakeAYP.Whileinasensethiswasanintendedresultoflegislationdesignedtofocus(withinlimits)onallpupilgroups,theimpactofmakingitmoredifficultforschoolsandLEAsservingdiversepopulationstomeetAYPstandardsmayalsobeseenasanunintendedconsequenceofNCLB.ThisissuehasbeenanalyzedinarecentstudybyThomasJ.KaneandDouglasO.Staiger,whoconcludedthatsuch"subgrouptargetscauselargenumbersofschoolstofail...arbitrarilysingleoutschoolswithlargeminoritysubgroupsforsanctions...orstatisticallydisadvantagediverseschoolsthatarelikelytobeattendedbyminoritystudents....Moreover,whilethecostsofthesubgrouptargetsareclear,thebenefitsarenot.Althoughthesetargetsaremeanttoencourageschoolstofocusmoreontheachievementofminorityyouth,wefindnoassociationbetweentheapplicationofsubgrouptargetsandtestscoreperformanceamongminorityyouth."56AccordingtotheEDreport,"TitleIImplementation—UpdateonRecentEvaluationFindings,"publishedin2009,thepercentageofschoolsfailingtomakeAYPrangedfrom7%forthosewithonly1subgroupto20%forthosewith2subgroups,37%forthosewith3subgroups,and43-51%forthosewith4-8subgroups. However,withoutspecificrequirementsforachievementgainsbyeachofthemajorpupilgroups,itispossiblethatinsufficientattentionwouldbepaidtotheperformanceofthedisadvantagedpupilgroupsamongwhomimprovementsaremostneeded,andforwhosebenefittheTitleI-Aprogramwasestablished.Underpreviouslaw,withoutanexplicit,specificrequirementthatAYPstandardsfocusonthesedisadvantagedpupilgroups,moststateAYPdefinitionsconsideredonlytheperformanceofallpupilscombined.AnditistheoreticallypossibleformanyschoolsandLEAstodemonstratesubstantialimprovementsinachievementbytheirpupilsoverallwhiletheachievementoftheirdisadvantagedpupilsdoesnotimprovesignificantly,atleastuntiltheultimategoalofallpupilsattheproficientorhigherlevelofachievementisapproached.Thisisespeciallytrueundera"status"modelofAYPsuchastheoneinNCLB,underwhichadvantagedpupilgroupsmayhaveachievementlevelswellabovewhatisrequired,andanoverallachievementlevelcouldeasilymaskachievementwellbelowtherequiredthresholdbyvariousgroupsofdisadvantagedpupils. Onepossiblealternativetocurrentpolicywouldbetoallowstatestocounteachstudentonlyonce,innet,inAYPcalculations,withequalfractionsforeachrelevantdemographiccategory(e.g.,aHispanicLEPpupilfromalow-incomefamilywouldcountasone-thirdofapupilineachgroup).NumberofSchoolsIdentifiedandStateVariationsTherein Aswasdiscussedearlier,concernhasbeenexpressedbysomeanalystssinceearlydebatesonNCLBthatarelativelyhighproportionofschoolswouldfailtomeetAYPstandards.Onthebasisofassessmentresultsfor2007-2008,35%ofallpublicschoolsnationwidefailedtomakeAYP.Further,approximately13%ofallpublicschoolswereidentifiedasneedingimprovement(i.e.,failedtomeetAYPstandardsfortwoormoreconsecutiveyears)for2007-2008.Futureincreasesinperformancethresholds,astheultimategoalofallpupilsattheproficientorhigherlevelofachievementisapproached,mayresultinhigherpercentagesofschoolsfailingtomakeAYP. Inresponsetotheseconcerns,EDofficialshaveemphasizedtheimportanceoftakingactiontoidentifyandmovetoimproveunderperformingschools,nomatterhownumerous.TheyhavealsoemphasizedthepossibilitiesforflexibilityandvariationintakingconsequenceswithrespecttoschoolsthatfailtomeetAYP,dependingontheextenttowhichtheyfailtomeetthosestandards.Itshouldalsobere-emphasizedthatmanyoftheschoolsreportedashavingfailedtomeetAYPstandardshavefailedtomeetAYPforoneyearonly,whileNCLBrequiresthataseriesofactionsbetakenonlywithrespecttoschoolsorLEAsparticipatinginESEATitleI-AthatfailtomeetAYPfortwoconsecutiveyearsormore. Further,someanalystsarguethatasetofAYPstandardsthatone-thirdormoreofpublicschoolsfailtomeetmayaccuratelyreflectpervasiveweaknessesinpublicschoolsystems,especiallywithrespecttotheperformanceofdisadvantagedpupilgroups.Totheseanalysts,theidentificationoflargepercentagesofschoolsisapositivesignoftherigorandchallengeembodiedinNCLB'sAYPrequirements,andislikelytoprovideneededmotivationforsignificantimprovement(andultimatelyareductioninthepercentageofschoolssoidentified). Othershaveconsistentlyexpressedconcernabouttheaccuracyandefficacyofanaccountabilitysystemunderwhichsuchahighpercentageofschoolsisidentifiedasfailingtomakeadequateprogress,withconsequentstrainonfinancialandotherresourcesnecessarytoprovidetechnicalassistance,publicschoolchoiceandsupplementalservicesoptions,aswellasotherconsequences.Inaddition,somehaveexpressedconcernthatschoolsmightbemorelikelytofailtomeetAYPsimplybecausetheyhavediverseenrollments,andthereforemoregroupsofpupilstobeseparatelyconsideredindeterminingwhethertheschoolmeetsAYPstandards.Theyalsoarguethattheapplicationoftechnicalassistanceand,ultimately,consequencestosuchahighpercentageofschoolswilldiluteavailableresourcestosuchadegreethattheseresponsestoinadequateperformancewouldbeinsufficienttomarkedlyimproveperformance. TheproportionofpublicschoolsidentifiedasfailingtomeetAYPstandardsisnotonlyrelativelylargeintheaggregate,butalsovarieswidelyamongthestates.Aswasdiscussedabove,thepercentageofpublicschoolsidentifiedasfailingtomakeAYPonthebasisofassessmentresultsfor2007-2008rangedfrom7%to80%amongthestates.Thisresultissomewhatironic,giventhatoneofthemajorcriticismsofthepre-NCLBprovisionsforAYPwasthattheyresultedinasimilarlywidedegreeofstatevariationintheproportionofschoolsidentified,andthemoreconsistentstructurerequiredunderNCLBwaswidelyexpectedtoleadtogreaterconsistencyamongstatesintheproportionofschoolsidentified. ItislikelythatstatevariationsinthepercentageofschoolsfailingtomeetAYPstandardsarebasednotonlyonunderlyingdifferencesinachievementlevels,aswellasavarietyoftechnicalfactorsinstateAYPprovisions,butalsoondifferencesinthedegreeofrigororchallengeinstatepupilperformancestandardsandassessments.ParticularlynowthatallstatesreceivingTitleI-Agrantsmustalsoparticipateinstate-leveladministrationofNAEPtestsin4thand8thgradereadingandmatheverytwoyears,thisvariationcanbeillustratedforallstatesbycomparingthepercentageofpupilsscoringattheproficientleveloraboveonNAEPversusstateassessments.Suchacomparisonwasconductedbyaprivateorganization,Achieve,Inc.,basedon8thgradereadingandmathassessmentsadministeredinthespringof2003.57Foravarietyofreasons,theanalysisexcludedseveralstates;29stateswereincludedinthecomparisonforreading,and32statesformath.Accordingtothisanalysis,thepercentageofpupilsstatewidewhoscoreataproficientorhigherlevelonstateassessments,usingstate-specificpupilperformancestandards,wasgenerallymuchhigherthanthepercentagedeemedtobeattheproficientorhigherlevelontheNAEPtests,andemployingNAEP'spupilperformancestandards.Ofthestatesconsidered,thepercentageofpupilsscoringataproficientorhigherlevelonthestateassessmentwaslowerthanonNAEP(implyingamorerigorousstatestandard)forfivestates58(outof32)inmathandonlytwostates(outof29)inreading.Further,amongthemajorityofstateswherethepercentageofpupilsattheproficientlevelorabovewasfoundtobehigheronstateassessmentsthanonNAEP,therelationshipbetweenthesizeofthetwogroupsvariedwidely—insomecasesonlymarginallyhigheronthestateassessment,andinothersthepercentageattheproficientlevelwasmorethantwiceashighonthestateassessmentasonNAEP. Morerecently,areportbytheNationalCenterforEducationStatisticsmappedeachstate'sstandardforaproficientlevelofperformanceinreadingandmathematicsatthe4thand8thgradelevelsforthe2004-2005schoolyearontotheequivalentNAEPscales.59Thepurposewastocomparethelevelofperformancedeemedtobeproficientundereachstate'sassessmentprogramwiththeproficientlevelofperformanceontheequivalentNAEPtest.Thereport'sauthorsconcludedthatincomparisontothecommonstandardembodiedinNAEP,statestandardsofproficiencyvariedwidely,andinalmostallcaseswereloweronstateteststhanunderNAEP.60Infact,theproficientlevelofperformanceinmanystateswasfoundtobelowerthanthebasiclevelofperformanceonNAEP.61 Asecondissueiswhethersomestatesmightchoosetolowertheirstandardsof"proficient"performance,inordertoreducethenumberofschoolsidentifiedasfailingtomeetAYPandmakeiteasiertomeettheultimateNCLBgoalofallpupilsattheproficientorhigherlevelbytheendofthe2013-2014schoolyear.Intheaffectedstates,thiswouldincreasethepercentageofpupilsdeemedtobeachievingata"proficient"level,andreducethenumberofschoolsfailingtomeetAYPstandards. Itseemslikelythatthepre-NCLBvariationsintheproportionofschoolsfailingtomeetAYPreflectedlargedifferencesinthenatureandstructureofstateAYPstandards,aswellasmajordifferencesinthenatureandrigorofstatepupilperformancestandardsandassessments.WhilethebasicstructureofAYPdefinitionsisnowsubstantiallymoreconsistentacrossstates,significantvariationsremainwithrespecttothefactorsdiscussedinthissectionofthereport(suchasminimumgroupsizeoruseofconfidenceintervals),andsubstantialdifferencesinthedegreeofchallengeembodiedinstatestandardsandassessmentsremain.Overall,itseemslikelythatthekeyinfluencesdeterminingthepercentageofastate'sschoolsthatfailstomakeAYPinclude(innoparticularorder):(1)degreeofrigorinstatecontentandpupilperformancestandards;(2)minimumpupilgroupsize(n)inAYPdeterminations;(3)useofconfidenceintervalsinAYPdeterminations(andwhetherata95%or99%levelofconfidence);(4)extentofdiversityinpupilpopulation;(5)extentofcommunicationabout,andunderstandingof,the95%testparticipationrule;and(6)possibleactualdifferencesineducationalquality.95%ParticipationRule Itappearsthatinmanycases,schoolsorLEAshavefailedtomeetAYPsolelybecauseoflowparticipationratesinassessments,meaningthatfewerthan95%ofallpupils,orofpupilsinrelevantdemographicgroupsmeetingtheminimumsizethreshold,tooktheassessments.While,asdiscussedabove,EDrecentlypublishedpolicyguidancethatrelaxestheparticipationraterequirementsomewhat—allowinguseofaverageratesovertwo-tothree-yearperiods,andexcusingcertainpupilsformedicalreasons—thehighrateofassessmentparticipationthatisrequiredinorderforschoolsorLEAstomeetAYPstandardsislikelytoremainanongoingfocusofdebate. Althoughfewargueagainsthavinganyparticipationraterequirement,itmaybequestionedwhetheritneedstobeashighas95%.Inrecentyears,theoverallpercentageofenrolledpupilswhoattendpublicschoolseachdayhasbeenapproximately93.5%,anditisgenerallyagreedthatattendanceratesarelowerinschoolsservingrelativelyhighproportionsofdisadvantagedpupils.Eventhoughschoolsareexplicitlyallowedtoadministerassessmentsonmake-updaysfollowingtheprimarydateoftestadministration,anditisprobablethatmoreschoolsandLEAswillmeetthisrequirementastheybecomemorefullyawareofitssignificance,itislikelytocontinuetobeverydifficultformanyschoolsandLEAstomeeta95%testparticipationrequirement.IssuesRegardingGrowthModelAlternativestoAYPModelsintheNCLBStatute WhyisthereincreasedinterestingrowthmodelsfordeterminingAYPunderNCLB?WhatmightbethemajoradvantagesanddisadvantagesofgrowthmodelsofAYP,incomparisontostatusorimprovementmodels?Thesequestionsareaddressedinthefollowingpages.AreGrowthModelsofAYPMoreFairandAccuratethanStatusorImprovementModels? Manyproponentsofgrowthmodelsforschool/LEAAYPseethemasbeingmorefair—tobothpupilsandschoolstaff—andaccuratethanstatusorimprovementmodels,primarilybecausetheycanbedesignedtotakeintoconsiderationthecurrentlywidelyvaryinglevelsofachievementofdifferentpupilgroups.Growthmodelsgenerallyrecognizetherealitythatdifferentschoolsandpupilshaveverydifferentstartingpointsintheirachievementlevelsandrecognizeprogressbeingmadeatalllevels(e.g.,frombelowbasictobasic,orfromproficienttoadvanced),givingcreditforallimprovementsoverpreviousperformance. Growthmodelswouldlikelyincreasetheabilitytoattributepupilachievementtotheircurrentschool,asopposedtotheirpastschoolsorbackgroundcharacteristics,especially(butnotonly)ifcontrols(and/orpredictedgrowthelements)areincludedinthemodel.Theymoredirectlymeasuretheeffectofschoolsonthespecificpupilstheyserveoveraperiodofyears,attemptingtotrackthemovementofpupilsbetweenschoolsandLEAs,ratherthanapplyingasinglestandardtoallpupilsineachstate.Theyhavetheabilitytofocusonthespecificeffectivenessofschoolsandteacherswithpupilswhomtheyhaveactuallytaughtformultipleyears,ratherthanthechangeinperformanceofpupilgroupsamongwhomtherehasusuallybeenasubstantialamountofmobility.Theycandirectly(aswellasindirectly)adjustfornon-schoolinfluencesonachievement,comparingthesamestudentsacrossyearsandreducingerrorsduetostudentmobility. ProponentsofgrowthmodelsoftenarguethatstatusmodelsofAYPinparticularmakeschoolsandLEAsaccountableforfactorsoverwhichtheyhavelittlecontrol,andthatstatusmodelsfocusinsufficientlyonpupilachievementgains,especiallyifthosegainsarebelowthethresholdforproficientperformance,orgainsfromaproficienttoanadvancedlevel.Statusmodels,suchasthecurrentprimarymodelofAYPunderNCLB,mightevencreateanundesirableincentiveforteachersandschoolstofocustheirattention,atleastintheshortrun,onpupilswhoareonlymarginallybelowaproficientlevelofachievement,inhopesofbringingthemabovethatsolekeythreshold,ratherthanfocusingonthemostdisadvantagedpupilswhoseachievementiswellbelowtheproficientlevel.ThecurrentstatusmodelofAYPalsoconfersnocreditforachievementincreasesabovetheproficientlevel,thatis,bringingpupilsfromtheproficienttotheadvancedlevel. Atthesametime,growthmodelsofAYPhavethesignificantdisadvantageofimplicitlysettinglowerthresholdsorexpectationsforsomepupilgroupsand/orschools.AlthoughanygrowthmodeldeemedconsistentwithNCLBwouldlikelyneedtoincorporatethatact'sultimategoalofallpupilsataproficientorhigherlevelofachievementby2013-2014(seebelow),themajorityofsuchmodelsusedcurrentlyorinthepastdonotincludesuchgoals,andtendtoallowdisadvantagedschoolsandpupilstoremainatrelativelylowlevelsofachievementforconsiderableperiodsoftime. GrowthmodelsofAYPmaybequitecomplicated,andmayaddresstheaccountabilitypurposesofNCLBlessdirectlyandclearlythanstatusor(toalesserextent)improvementmodels.IftheprimarypurposeofAYPistodeterminewhetherschoolsandLEAsaresucceedingatraisingtheachievementoftheircurrentpupilstochallenginglevels,withthosegoalsandexpectationsappliedconsistentlytoallpupilgroups,thenthecurrentprovisionsofNCLBmightmoresimplyanddirectlymeetthatpurposethangrowthmodelalternatives. PupilmobilityamongschoolsandLEAsissubstantial,andhasimportantimplicationsforallmodelsofAYP.However,itsimplicationsaremultifaceted,anddonotnecessarilyfavoraparticularAYPmodel.Growthmodelshavetheadvantageofattemptingtotrackpupilsthroughlongitudinaldatasystems.ButiftheytherebyattributetheachievementofhighlymobilepupilsamongavarietyofschoolsandLEAs,accountabilityisdispersed.Atthesametime,thepresenceofhighlymobilepupilsinthegroupsconsideredindeterminingAYPunderstatusandimprovementmodelsmayseemunfairtoschoolstaff.However,theimpactofsuchpupilsinschool-levelAYPdeterminationsislimitedbyNCLB'sprovisionthatpupilswhohaveattendedaparticularschoolforlessthanoneyearneednotbeconsideredinsuchdeterminations.DoStatesHaveSufficientResourcestoDevelopandImplementGrowthModels? ItisgenerallyagreedthatgrowthmodelsofAYParemoredemandingthanstatusorimprovementmodelsinseveralrespects,especiallyintermsofdatarequirementsandanalyticalcapacity.Foralongitudinaldatasystemsufficienttosupportagrowthmodel,itislikelythatstateswouldneedtohavepupildatasystemsincorporatingatleastthefollowing: 1.auniquestatewidestudentidentifier; 2.theabilitytoproducecomparableresultsfromgrade-to-gradeandfromyear-to-year(vertically-scaledassessments); 3.student-levelenrollment,demographicandprogramparticipationinformation; 4.informationonuntestedstudents; 5.studentlevelgraduationanddropoutdata;and 6.astatewideauditsystem.62 Althoughtheavailabilityofinformationonstatedatasystemsisinsufficienttoenableonetodeterminewithprecisionhowmanystatescouldorcouldnotcurrentlyimplementsuchmodelsiftheychosetodoso,itisverylikelythatgrowthmodelsgenerallyrequireresourcesanddatasystemsthatsomestatescurrentlylack. ThisconcernisbeingaddressedthroughanEDprogramintendedtohelpstatesdesign,develop,andimplementstatewide,longitudinaldatasystems.Aninitialappropriationof$24.8millionwasprovidedforthisprogram,administeredbyED'sInstituteofEducationSciences(IES),63forFY2005.Subsequently,$24.6millionwasappropriatedforeachofFY2006andFY2007,$48.3millionforFY2008,and$65millionforFY2009.Inaddition,$250millionwasappropriatedforthisprogramforFY2009underP.L.111-5,theAmericanRecoveryandReinvestmentAct(ARRA).Further,theestablishmentoflongitudinaldatasystemsforeducationisapriorityforstateparticipationintheStateFiscalStabilizationFundandthe"RacetotheTop"discretionarygrantcompetitionundertheARRA.64 Thusfar,atleast41stateshavereceivedawardsthroughthreeroundsofcompetition.65Underthisprogram,aidisprovidedtostateeducationalagencies(SEAs)viacooperativeagreements,notgrants,toallowincreasedfederalinvolvementinthesupportedactivities.AccordingtotheannouncementintheApril15,2005,FederalRegister,theprogramisintended"toenableSEAstodesign,develop,andimplementstatewide,longitudinaldatasystemstoefficientlyandaccuratelymanage,analyze,disaggregate,anduseindividualstudentdata....Applicationsfromstateswiththemostlimitedabilitytocollect,analyze,andreportindividualstudentachievementdatawillhaveapriority...."AccordingtoED,theprogramisdesignedtohelpSEAsmeettheAYPandreportingrequirementsofNCLB,aswellastoconductvalue-addedorachievementgrowthresearch,including"meaningfullongitudinalanalysesofstudentacademicgrowthwithinallsubgroupsspecifiedbytheNoChildLeftBehindActof2001."Therewillalsobeanemphasisonencouragingdatasharingamongstates,whileatthesametimeprotectingthesecurityandprivacyofdata.AreGrowthModelsConsistentwithNCLB'sUltimateGoal? MostgrowthmodelsusedbeforeinitiationofED'sgrowthmodelpilot,orstillusedaspartofstate-specificaccountabilitysystems,havenotincorporatedanultimategoalsuchastheoneunderNCLB—thatallpupilsreachaproficientorhigherlevelofachievementby2013-2014.Non-NCLBgrowthmodelshavegenerallyincorporatedoneoftwotypesofgrowthtarget,the"howmuchimprovementisenough"aspectofthemodel:(a)datadriven/predictedgrowth,or(b)policydriven/requiredgrowthtargets.Thefirsttypeofgrowthtargethasbeenmostcommon,whileNCLB'sultimategoalwouldrepresentagrowthtargetofthesecondvariety,withseparatepaths(withpresumablyseparatestartingpoints)foreachrelevantpupilcohort.ThemodelsapprovedthusfarunderED'sgrowthmodelpilotarguablymeettheultimategoalrequirement.However,undersomeofthesemodels,pupilsneedonlybeproficientorontracktowardproficiencywithinalimitednumberofyearsasof2013-2014. Acknowledgments [authornamescrubbed],formerSpecialistinEducationPolicy,wastheoriginalauthorofthisreport. Footnotes 1. Theseconsequences,aswellaspossibleperformance-basedawards,arenotdiscussedindetailinthisreport.Forinformationonthem,seeCRSReportRL33731,EducationfortheDisadvantaged:ReauthorizationIssuesforESEATitleI-AUndertheNoChildLeftBehindAct,by[authornamescrubbed]. 2. Foradditionalinformationonthislegislation,seeCRSReport89-7,EducationforDisadvantagedChildren:MajorThemesinthe1988ReauthorizationofChapter1,by[authornamescrubbed](outofprint,availableuponrequest). 3. ForadditionalinformationonthestandardandassessmentrequirementsunderESEAtitleI-A,seeCRSReportRL31407,EducationalTesting:ImplementationofESEATitleI-ARequirementsUndertheNoChildLeftBehindAct,by[authornamescrubbed]. 4. Thereisavariantofthegroupstatusmodel,sometimescalledan"indexmodel,"underwhichpartialcreditwouldbeattributedtoperformanceimprovementsbelowtheproficientlevel—e.g.,frombelowbasictobasic. 5. Scoresaretypicallycombinedforpupilsinallassessedgradelevelsinaschool. 6. Onestate,Massachusetts,hasinjectedapartialgrowthelementintoitssafeharborprovision.Inthatstate,aschoolorLEAthatfailstomeetthestandardAYPrequirementsstillmakesAYPifthenumberofpupilsinrelevantgroupsandsubjectsscoringbelowtheproficientleveldeclinesby10%ormorefromthepreviousyearordeclinessufficientlytoputthemontracktowardproficiencybytheendofthe2013-2014schoolyear. 7. See,forexample,IssuesintheDesignofAccountabilitySystems,byRobertL.Linn,CSETechnicalReport650,NationalCenterforResearchonEvaluation,Standards,andStudentTesting,April2005. 8.  FormoreinformationonallaspectsoftheESEATitleI-Aassessmentrequirements,seeCRSReportRL31407,EducationalTesting:ImplementationofESEATitleI-ARequirementsUndertheNoChildLeftBehindAct,by[authornamescrubbed]. 9. AllpupilsinstateswhereAYPdeterminationsweremadeforallpublicschools,orallpupilsservedbyESEATitleI-AinstateswhereAYPdeterminationsweremadeonlyforsuchschoolsandpupils. 10. Seehttp://www.cpre.org/Publications/Publications_Accountability.htm. 11. U.S.DepartmentofEducation,OfficeoftheUndersecretary,PolicyandProgramStudiesService,EvaluationofTitleIAccountabilitySystemsandSchoolImprovementEfforts(TASSIE):First-YearFindings,2004.HereafterreferredtoastheTASSIEFirst-YearReport. 12. SeetheU.S.DepartmentofEducation,"PaigeReleasesNumberofSchoolsinSchoolImprovementinEachState,"pressrelease,July1,2002athttp://www.ed.gov/news/pressreleases/2002/07/07012002a.html. 13. AnotherreportpublishedbyEDin2004(theTASSIEFirst-YearReport—seefootnote11)statedthat8,078publicschoolshadbeenidentifiedasfailingtomeetAYPstandardsfortwoormoreconsecutiveyearsinthe2001-2002schoolyear. 14. Programregulationspublishedin2002didnotrequiregraduationratesandotheradditionalacademicindicatorstobedisaggregatedindeterminingwhetherschoolsorLEAsmeetAYPstandards.However,regulationspublishedsubsequentlyinOctober2008(discussedlaterinthisreport)requiregraduationratestobedisaggregatedinAYPdeterminations. 15. IfthenumberofpupilsinaspecifieddemographicgroupistoosmalltomeettheminimumgroupsizerequirementsforconsiderationinAYPdeterminations,thentheparticipationraterequirementdoesnotapply. 16. IthasoccasionallybeensaidthattheAYPsystemsapprovedbyEDforafewstatesbeforeinitiationofthegrowthmodelpilotannouncedinNovember2005incorporate"growth"elements.However,suchclaimsappeartobebasedprimarilyontheinclusionintheAYPsystemsof"pupilachievementindexes"thatgivepartialcreditforachievementgainsbelowtheproficientlevel,comparingthisyear'spupilgroupswithlastyear's.Theydonotmeetthedefinitionofgrowthmodelasusedinthisreport. 17.  Thisisdeterminedbyrankingallpublicschools(oftherelevantgradelevel)statewideaccordingtotheirpercentageofpupilsattheproficientorhigherlevelofachievement(basedonallpupilsineachschool),andsettingthethresholdatthepointwhereone-fifthoftheschools(weightedbyenrollment)havebeencounted,startingwiththeschoolsatthelowestlevelofachievement. 18. Underprogramregulations[34C.F.R.§200.16(c)(2)],thestartingpointmayvarybygradespan(e.g.,elementary,middle,etc.)andsubject. 19. Asnotedearlier,undertheaccountabilitypolicyapprovedforuseinMassachusetts,aschoolorLEAalsomeetsthesafeharborrequirementifthenumberofpupilsinrelevantgroupsandsubjectsscoringbelowtheproficientleveldeclinessufficientlytoputthemontracktowardproficiencybytheendofthe2013-2014schoolyear. 20. UnderNCLB,stateAYPsystemsmustincludeatleastoneindicator,otherthanachievementtestscores.Forseniorhighschools,theadditionalindicatormustbethegraduationrate.Atypicaladditionalindicatorforelementaryandmiddleschoolsistheattendancerate. 21. Theplans,asupdatedovertime,havebeenpostedonlinebyEDathttp://www.ed.gov/admins/lead/account/stateplans03/index.html. 22. EDhasapprovedstateaccountabilityplansunderwhichschoolsorLEAswouldbeidentifiedasfailingtomeetAYPonlyiftheyfailedtomeettherequiredlevelofperformanceinthesamesubjectfortwoormoreconsecutiveyears,buthasnotapprovedproposalsunderwhichaschoolwouldbeidentifiedonlyifitfailedtomeetAYPinthesamesubjectandpupilgroupfortwoormoreconsecutiveyears. 23. Seehttp://www.ed.gov/news/pressreleases/2005/11/11182005.html. 24. SeetheFederalRegisterforOctober29,2008(pages64435-64513). 25. Seehttp://www.ed.gov/policy/elsec/guid/growthmodelguidance.pdf. 26. Oneotherstate,Massachusetts,incorporatesapartialgrowthelementintoitssafeharborprovision.Inthatstate,aschoolorLEAthatfailstomeetthestandardAYPrequirementsstillmakesAYPifthenumberofpupilsinrelevantgroupsandsubjectsscoringbelowtheproficientleveldeclinesby10%ormorefromthepreviousyearordeclinessufficientlytoputthemontracktowardproficiencybytheendofthe2013-2014schoolyear. 27. Delaware'sproposalincludeduseofconfidenceintervalsatanunspecifiedlevelinimplementingthegrowthmodel;however,EDapproveduseofthemodelwithoutconfidenceintervals. 28. MoststatesuseconfidenceintervalsintheirAYPdeterminations.However,inmostcases,theconfidenceintervalsareappliedtogroupaveragepercentagesofstudentsscoringproficientorabove,notindividualstudentscores. 29. "Evaluationofthe2005-2006GrowthModelPilotProgram,"January15,2009,U.S.DepartmentofEducation,availableathttp://www.ed.gov/admins/lead/account/growthmodel/index.html. 30. Seehttp://www.ed.gov/policy/elsec/guid/secletter/090401.html. 31. Thisincludesstudentswhograduatefollowingasummerprogramaftertheirfourthyear. 32. Foramorecompletediscussionandanalysisofthistopic,seeCRSReportR40701,AlternateAssessmentsforStudentswithDisabilities,byErinD.Caffrey. 33. Thislimitationdoesnotapplytotheadministrationofalternateassessmentsbasedonthesamestandardsapplicabletoallstudents,forotherpupilswith(non-cognitiveorlessseverecognitive)disabilities. 34. Undercurrentregulations,theshort-termpolicycannotbeextendedbeyondthe2008-2009schoolyear. 35. Thiswouldbecalculatedonthebasisofstatewidedemographicdata,withtheresultingpercentageappliedtoeachaffectedschoolandLEAinthestate.InmakingtheAYPdeterminationusingtheadjusteddata,nofurtherusemaybemadeofconfidenceintervalsorotherstatisticaltechniques.(Theactual,notjusttheadjusted,percentageofpupilswhoareproficientmustalsobereportedtoparentsandthepublic.) 36. The3.0%limitmightbeexceededforLEAs,butonlyif—andtotheextentthat—theSEAwaivesthe1.0%capapplicabletoscoresonalternateassessmentsbasedonalternateachievementstandards. 37. Insuchcases,theformerpupilswithdisabilitieswouldnothavetobecountedindeterminingwhethertheminimumgroupsizewasmetforthedisabilitysubgroup. 38. SeeFederalRegister,June24,2004,pp.35462-35465;andhttp://www.ed.gov/nclb/accountability/schools/factsheet-english.html. 39. Seehttp://www.ed.gov/policy/elsec/guid/stateletters/asaypnc.html. 40. Seehttp://www.ed.gov/policy/elsec/guid/secletter/050929.html. 41. Foradditionalinformationonthistopic,seeCRSReportRL33236,Education-RelatedHurricaneRelief:LegislativeAction,by[authornamescrubbed]etal. 42. SeeCenteronEducationPolicy,RuleChangesCouldHelpMoreSchoolsMeetTestScoreTargetsfortheNoChildLeftBehindAct,October22,2004,availableathttp://www.cep-dc.org/nclb/StateAccountabilityPlanAmendmentsReportOct2004.pdf;TitleIMonitor,ChangesinAccountabilityPlansDiluteStandards,CriticsSay,November2004;CouncilofChiefStateSchoolOfficers,RevisitingStatewideEducationalAccountabilityUnderNCLB,September2004,availableathttp://www.ccsso.org;and"RequestsWinMoreLeewayUnderNCLB,"EducationWeek,July13,2005,p.1. 43. Seehttp://www.ed.gov/admins/lead/account/consolidated/index.html. 44. DataarealsoavailableonTitleI-A-recipientschoolsandLEAsthatfailtomakeadequateyearlyprogress.However,intheaggregate,theresultsarequitesimilar.For2007-2008,35%ofallpublicschoolsand36%ofallTitleI-AschoolswerereportedasfailingtomakeAYP.ForLEAs,35%ofallLEAsand38%ofallTitleI-ALEAsfailedtomakeAYP. 45. Seehttp://www.ed.gov/programs/statestabilization/schooldata.pdf. 46. ThiscalculationwasbasedondataforallstatesexceptLouisiana. 47. See,forexample,"StatesRevisetheMeaningof'Proficient',"EducationWeek,October9,2002. 48. Seehttp://www.ed.gov/admins/lead/account/stateplans03/cocsa.pdf,p.7. 49. Seehttp://www.ed.gov/admins/lead/account/stateplans03/lacsa.doc,p12. 50. AccordingtoSection1111(b)(2)(H),"EachStateshallestablishintermediategoalsformeetingtherequirements,...ofthisparagraphandthatshall—(i)increaseinequalincrementsovertheperiodcoveredbytheState'stimeline...."Theprogramregulationsalsowouldseemtorequireincreasesinequalincrements:"EachStatemustestablishintermediategoalsthatincreaseinequalincrementsovertheperiodcoveredbythetimeline...."(34C.F.R.§ 200.17). 51. Alternatively,theconfidenceinterval"window"maybeappliedtoaveragetestscoresforeachrelevantpupilgroup,thatwouldbecomparedtoafixedthresholdscoreleveltodeterminewhetherAYPhasbeenmet. 52. ThetextabovedescribesthewayinwhichconfidenceintervalshavebeenusedbystatesforAYPdeterminations.Theconceptcouldbeappliedinadifferentway,requiringscorestobeatorabovethehighestscoreinthe"window"inordertodemonstratethatapupilgrouphadmeetAYPstandardstoastatisticallysignificantdegree.Thiswouldreflectconfidence(atthedesignatedlevelofprobability)thataschoolorLEAhadmetAYPstandards,whereasthecurrentusagereflectsconfidencethattheschoolorLEAhadfailedtomeetAYPstandards. 53. InTexas,theminimumgroupsizeforpupilgroups(otherthanthe"allpupils"group,wheretheminimumis40)isthegreaterof50studentsor10%ofallstudentsinaschoolorLEA(uptoamaximumof200).InCalifornia,theminimumgroupsizeisthegreaterof50studentsor15%ofallstudentsintheschoolorLEA(uptoamaximumof100). 54. UnderregulationspublishedonApril9,2007,thispracticeisnolongerallowed. 55. MinimumgroupsizesforAYPpurposesaretypicallyintherangeof30to40pupils,whilethoseforreportingaretypicallyintherangeoffiveto20pupils. 56. ThomasJ.KaneandDouglasO.Staiger,"UnintendedConsequencesofRacialSubgroupRules,"inPaulPetersonandMartinWest,eds.,NoChildLeftBehind?ThePoliticsandPracticeofSchoolAccountability(Washington:BrookingsInstitutionPress,2003),pp.152-176. 57. CenteronEducationPolicy,FromtheCapitaltotheClassroom,Year2oftheNoChildLeftBehindAct(January2004),p.61. 58. Intwoadditionalstates,thepercentageswereessentiallythesame. 59. NationalCenterforEducationStatistics,U.S.DepartmentofEducation,Mapping2005StateProficiencyStandardsOntotheNAEPScales,Washington,DC,June2007,http://nces.ed.gov/nationsreportcard/pdf/studies/2007482.pdf. 60. Ofthefoursubjectandgradelevelcomparisons,thestateproficiencystandardwasfoundtobeequivalenttoorhigherthantheNAEPproficiencystandardfornostatesin4thor8thgradereading,fortwostatesin4thgrademath,andforthreestatesin8thgrademath. 61. Ofthefoursubjectandgradelevelcomparisons,thestateproficiencystandardwasfoundtobelowerthantheNAEPbasicstandardfor24statesin4thgradereading,for12statesin8thgradereading,forsixstatesin4thgrademath,andforeightstatesin8thgrademath. 62. AimeeGuidera,directoroftheDataQualityCampaign,asquotedin:CommissiononNoChildLeftBehind,CommissionStaffResearchReport,"GrowthModels:AnexaminationwithinthecontextofNCLB,"August2006,availableathttp://www.aspeninstitute.org/atf/cf/{DEB6F227-659B-4EC8-8F84-8DF23CA704F5}/Growth%20Models%20and%20NCLB%20Report.pdf,visitedonSeptember6,2006. 63. ThisprogramisauthorizedbySection208oftheEducationSciencesReformActof2002,P.L.107-279.Theauthorizedfundinglevelis$80millionforFY2003and"suchsumsasmaybenecessary"foreachofthesucceedingfivefiscalyears. 64. SeeCRSReportR40151,FundingforEducationintheAmericanRecoveryandReinvestmentActof 2009 (P.L. 111-5),by[authornamescrubbed]etal.fordetails. 65. Foradditionalinformation,seehttp://nces.ed.gov/Programs/SLDS/.