Predictive ability of KRA‐L on third grade

March2014
PredictiveabilityofKRA‐LonthirdgradereadingproficiencyinCleveland
Fischer,Rob,Anthony,Beth,&Kim,Seok‐Joo
CenteronUrbanPovertyandCommunityDevelopment
Overthelasttwodecadestherehasbeenrisinginterestinstrategiestoimprovechildren’s
preparednesstolearnuponkindergartenentry(RAND,2005).Thenotionof‘kindergarten
readiness’reflectsthecompetenciesandskillschildrenneedtopossesstothriveinthe
kindergartenyear(Kagan&Rigby,2003).
InOhio,kindergartenreadinessisdirectlyassessedforallchildrenenteringkindergarten
usingtheKRA‐L(KindergartenReadinessAssessment–Literacy),whichisadministeredby
kindergartenteachersduringthefirstthreeweeksofthefallterm(OhioDepartmentof
Education,2014).TheKRA‐Lwasfirstimplementedinpublicschooldistrictsbeginningin
2007.Ithasavalueof0‐29andhasthreescorebands:Band1=0‐13(Assessbroadlyfor
intenseinstruction),Band2=14‐23(Assessfortargetedinstruction),andBand3=24‐29
(Assessforenrichedinstruction).ChildrenscoringinBand1maybeatseriousriskofbeing
unpreparedforthekindergartenexperience.
ThoughtheKRA‐LhasbeeninuseinOhiosince2007,therehasbeenlittleinformation
availableaboutitsvalidityasameasureofliteracyskills.Oneaspectofvalidityisthe
consistencywithwhichthemeasureispredictiveofotherindicatorsofliteracy,suchasthe
thirdgradereadingassessmentintheOhioAchievementAssessment(OAA).Recently,
researchersatTheOhioStateUniversitycompletedastudyofthepredictiveabilityofthe
KRA‐Lonthirdgradereadingbasedonstudentswhoenteredkindergarteninthe
ColumbusSchoolDistrictbetween2005‐2009(Logan,Justice,&Pentimonti,2014).Overall,
Loganetal.(2014)reportedastrongpositivecorrelationbetweentheKRA‐LandtheOAA
ReadingAssessment(r=.47,p<.001).Specifically,fiveofthesixKRA‐Lsubtestswere
significantlyrelatedtothirdgradereadingscores.Inaddition,theyfoundKRA‐Lbands
werepredictiveofastudent’slikelihoodofpassingtheOAAReadingAssessment.
Thepresentstudywasundertakentoreplicate,inpart,theColumbusstudyusingasample
ofchildrenenrolledintheClevelandMetropolitanSchoolDistrict(CMSD).Thefocushere,
however,isontheoverallpredictiveabilityoftheKRA‐Lscore,ratherthanthatofthe
subtests.PreviousstudiesofhighqualityearlylearningstrategiesinCuyahogaCountyhave
shownthatparticipatingchildrenarriveatkindergartenwithmeanKRA‐Lscores2.5
pointshigherthanunservedchildren(Fischer,Lalich,&Coulton,2013).Thisfindingraises
thequestionaboutwhatincrementsontheKRA‐Lmighttellusaboutstudents’future
success,especiallywhentheincrementdoesnotchangetheKRA‐Lbandinwhichthe
studentfalls.
Center on Urban Poverty & Community Development 1 AspartofastudyfundedbytheOhioEducationResearchCenter,KRA‐Lscoresand3rd
gradeOAAreadingscoreswerelinkedforthreecohortsofkindergartenersinCMSD(2008‐
2010).Thefinalsamplesizeis5,273students.
Overall,KRA‐LbandsstronglyandpositivelypredictOAAreadingassessmentpassage.See
Figure1.AmongchildrenwhoscoredinBand1ontheKRA‐L,61%didnotpasstheOAA
readingassessmentinthirdgrade.ThepatternisreversedforchildrenscoringinBand3of
theKRA‐Linthat83%ofthesestudentsdidpasstheOAAthirdgradereadingassessment.
AmongchildrenscoringinBand2ontheKRA‐L62%laterpassedthethirdgradereading
assessment.
Figure1:PerformanceonThirdGradeReadingAssessmentbyKRA‐LBand
100
Percentage of Students
90
80
39.2
70
61.8
60
82.8
50
40
30
60.8
20
38.2
10
17.2
0
Band 1
Band 2
Band 3
KRA‐L Band
Fail
Pass
UsinglogisticregressiontopredictOAAreadingpassage,theresultsshowthata1point
increaseinKRA‐Lisassociatedwith12.1%increaseintheoddsofpassingOAA3rdgrade
readingassessment(Beta=0.1138,Oddratio=1.121,p<.001).Thisresultdidnotcontrolfor
anycovariates.
ThisfindingsuggeststhatincrementalimprovementsontheKRA‐Lmaywellequateto
academicallysignificantchangesinastudent’sfutureperformance.Forexample,among
studentsintheCuyahogaCountyUPKprogramwherethemeanKRA‐Lis2.5pointshigher
thannon‐UPKstudents,thisequatestoa30%higheroddsofpassingthethirdgrade
readingassessment.Inlightofthis,earlychildhoodpractitioners,providers,andfunders
shouldseetheKRA‐Lnotasanendinitselfbutasamarkeroflateracademicsuccessfor
thechildrentheyserve.
Center on Urban Poverty & Community Development 2 Sources
Fischer,R.,Lalich,N.,&Coulton,C.(2013).CuyahogaCounty’sUniversalPre‐Kindergarten
Pilot:FindingsfromtheFirstFiveYears.Cleveland,OH:CenteronUrbanPoverty&
CommunityDevelopment,CaseWesternReserveUniversity.
Kagan,S.L.,&Rigby,E.(2003).ImprovingtheReadinessofChildrenforSchool:
RecommendationsforStatePolicy.Washington,DC:CenterfortheStudyofSocial
Policy.
Logan,J.,Justice,L.M.,&Pentimonti,J.(2014,winter).ReadytoReadandSchoolSuccess:
KindergartenReadinessandthe“ThirdGradeReadingGuarantee.”Columbus,OH:
CraneCenterforEarlychildhoodResearchandPolicy,TheOhioStateUniversity.
OhioDepartmentofEducation.(2004).KindergartenReadinessAssessment–Literacy
(KRA‐L).Retrievedfromhttp://education.ohio.gov/Topics/Early‐
Learning/Guidance‐About‐Kindergarten/KRAL
RAND(2005).ChildrenatRisk:ConsequencesforSchoolReadinessandBeyond.Santa
Monica,CA:RAND.
Authors
RobFischer,Ph.D.isresearchassociateprofessorandco‐directoroftheCenteronUrban
PovertyandCommunityDevelopmentattheJack,JosephandMortonMandelSchool
ofAppliedSocialSciences,CaseWesternReserveUniversity.
BethAnthony,Ph.D.isseniorresearchassociateattheCenteronUrbanPovertyand
CommunityDevelopmentattheJack,JosephandMortonMandelSchoolofApplied
SocialSciences,CaseWesternReserveUniversity.
Seok‐JooKim,Ph.DisapostdoctoralscholarattheCenteronUrbanPovertyand
CommunityDevelopmentattheJack,JosephandMortonMandelSchoolofApplied
SocialSciences,CaseWesternReserveUniversity.
Center on Urban Poverty & Community Development 3