1
2
Aminimallevelofevidencedoesnot meanthatarecommendationisn’timportant.Itmeansthatmorerigorous researchisneededinthearea.
3
4
5
Poverty &racialprejudice &lack ofpolitical will can be addressed by:sharingthetax wealth (e.g.,FL’s districtsarethecounty,notsmaller
administrativeunits asinmost other states)Incentivize teachers towrok inlow-performing schoolsIntegrate SESgroupsthrough distributed housing plansandeducational
innovations(e.g.,magnet schools,IBprograms;charters)Lack ofrespectforK-12teaching hasled tobrain draintohigher paying professions.Therefore,weaker students going into teacher education.Need toraise beginning
teacher salaries,raise standardsforadmissiontoteacher ed programs,retainteachers byrewarding effectiveness andinnovation.Lack ofscientifically-based SEA/LEAreading policy:state/districtreading plansneed tobe based onthescienceofreading andthere need tobe incentives forthoughtful
andtimely completion andaccountability andrecognitionforsuccess (i.e.,proficiency levels andreduction ofachievement gaps).Personal beliefs trump thescientific method:Reward student achievement gainsinK-12teachers.Universities need tohire faculty toteach thereading certificationcourses
who demonstrate knowledge &skills toteach evidence-based practices.Statereading certificationexamsneed tobe rigorous andbased onthescienceofreading.Instructional materials forteaching reading mustbe based onscientific evidence(Foormanetal.,2004).RELSErubric forevaluating K-5ELAinstructional materials.Systemic approach tomulti-tiered intervention:Generaled andspecial ed teachers,specialists,andadministrators need towork together toarticulate RtI.
SeeNationalImplementationResearchNetworkwebsiteforinformationaboutimplementationscience.
6
RELSEexaminedchangesinteacherknowledgeofearlyliteracyskillsandinratingsofqualityofearlyliteracyskillsinstruction,studentengagementduringearlyliteracyskillsinstruction,andteachingcompetenciesinMSfrom winter2014tofall2015inthetargetschoolsparticipatinginMDE’searlyliteracyinitiative.Teachersintheseschoolsparticipatedinonlineandface-to-facePDonLTRS.RELSEdesignedateacherknowledgesurveythatparticipatingteacherstookinonlinemodulesaftertheLTRSPD.RELSEtrainedstatewidereadingcoachestoconductclassroomobservationsandrateteachingcompetencies.TheRELSEreport,“EducatoroutcomesassociatedwithimplementationofMississippi’sK-3earlyliteracyprofessionaldevelopmentinitiative,”showedthataverageteachers’knowledgeincreased(48th %ile to59th %ile),averageratingsofqualityofinstructionincreased(31st %ile to58%ile),studentengagementincreased(37th%ileto53%ile),andaverageratingofteachingcompetenciesincreased(30th %ile to44th%ile).Theincreaseinteachers’knowledgeandtheincreasesintheaverageratingsofqualityofinstruction,studentengagement,andteachingcompetenciesintargetschoolswereassociatedwithprogressinthePDprogram.
Findings:progressinthePDprogramwasassociatedwithimprovementsinteacherknowledge,qualityofinstruction,studentengagement,andteachercompetencies.
MDE&BarksdaleReadingInstitutehavedoneagreatjobevaluatingMS’steacherpreparationprogramsinreadingandworkingwiththeHigherEdLiteracyCounciltomakeimprovements.
7
8
The Self-study Guide for Implementing Early Literacy Interventions was developed to help district- and school-based practitioners conduct self-studies for planning and implementing early literacy interventions. It is intended to promote reflection about current strengths and challenges in planning for implementation of early literacy interventions, spark conversations among staff, and identify areas for improvement. This guide provides a template for data collection and guiding questions for discussion that may improve the implementation of early literacy interventions and decrease the number of students failing to meet grade-level literacy expectations by the time they enter grade 3.
9
10
11
Screeningaccuracycanvarydependingontrade-offsbetweentheclassificationindicesofsensitivityandspecificityandpositive/negativepower.Sensitivity istheproportionofindividualswhofailedtheoutcomeandwereidentifiedasatriskonthescreen [A/(A+C)].Specificity istheproportionofindividualswhopasstheoutcometestinthepopulationwhoarenotatriskonthescreen(truenegative;[D/(D+B)]).Bothsensitivityandspecificityarepropertiesofthescreenitself.SensitivityconsideredimportantwithinRtI becauseit’sthepercentageofstudentscorrectlyidentifiedbyscreenorneedingfurtherassessment/intervention.Theoverallcorrectclassificationindex(OCC)is:[(A+D)/(A+B+C+D)].
Incontrast,thesample-basedindicesofpositiveandnegativepredictivepowerdependontheproportionofstudentsinthesamplewhoareatrisk—thatis,thebaserateofrisk.Positivepredictivepoweristheproportionofindividualsidentifiedasatriskonthescreenwhofailtheoutcome([A/(A+B)]),whereasnegativepredictivepoweristheproportionofindividualsidentifiedasnotatriskonthescreenwhopasstheoutcometest[D/(C+D)].
12
#3:Becauseourfocusisonnotunder-identifyingstudents.
13
Justasthevalidityofthescreenistiedtoitsabilitytopredictperformanceonagoldstandardoutcome,thevalidity ofadiagnosticassessmentshouldbetiedtotheabilityofitsdiagnosticprofilestopredicttoagoldstandardoutcometest(Foorman,Petscher,&Stanley,2016),ratherthantounreliabledescriptivenotionsoflearningprofiles(e.g.,Tomlinson,1999)ortoinvalidmetricsfordeterminingreadingfluency(e.g.,Francisetal.,2008).Infact,anefficientapproachistocombinescreeninganddiagnosisintoonesystem,withagatingprocesssothatthediagnosticcomponentisgivenonlytothosestudentspredictedtobeatriskonthescreen.Measuringgrowthalsoimportanttodetermininglearning.
Inadditiontopredictivevalidity,constructvalidityisveryimportanttoscreeninganddiagnosticassessment.Educatorsareresponsibleforteachingstandardsthatareoftenvaguelystatedandnecessarilynarrowedtothosemeasuredinthestateaccountabilitytest.Statetestdesignedtoascertainwhetherstudentsmeetgrade-levelproficiency.Thus,statetestisapoorsubstituteforadiagnosticassessmentbecausedoesn’tmeasureskillsattheloworhighendofthescoredistributioninagrade.Computer-adaptivetestscanaddressperformanceofstudentsoutsideofgrade-levelnorms,suchasELLandspecialed students.
14
15
16
Bothinterventionapproachesweretaughtdailyfrommid-OctobertotheendofMayfor45minutesandconsistedofa25–30minutereadingcomponentanda15minuteorallanguagecomponent.
17
18
19
20
*seeBialystoketal,2003,forbilingualadvantageonPA
21
Togetbettereffects,increase intensitybyreducinggroupsize.Also,mosteffectswereinK,sostartearlier.
22
23
24
Foorman&AlOtaiba(2009)reportthateffectsizesforcommunityvolunteersandcollegestudentsdeliveringreadinginterventionwerelessthanthatofcertifiedteachers.However,well-trainedcommunityvolunteersarecertainlyanoption,butarenotasubstitutionforhighqualityclassroominstruction(Elbaum,Vaughn,Hughes,&Moody,2000).
25
• Districtemployeesareparaprofessionalsorcertifiedhourlyteachers• DistrictemployeesandRELhiredinterventionistswereprovidedthesameinitial
trainingandongoingsupport• Studentdemographics:1060inK(38%EL);1096inG1(34%EL);1309inG2(33%EL).
TotalN=3465.FRLranged from82-86%.
26
Timeline forpreparinginterventionists
27
• For each group, the fall and spring fidelity ratings were then averaged to create separate overall fidelity ratings for the reading and oral language components.
• Instructionineachcomponentofthestand-aloneandembeddedinterventiongroupswasimplementedwithhighfidelity
28
29
30
Fielding,Kerr,&Rosier,2007
31
User-friendlytool(notevaluative);Identifiesresearch-basedpracticesEnhancesliteracyinstructionknowledgeEstablishesconsistentlanguage&expectationsPromotescommunication&collaborationProvidesdatafordecisionmaking(professionallearningneeds,resources)
32
GradeSpecific:CoversFoundationalSkillsReadingComprehensionWritingLanguageSpeaking&Listening
Frequentwalkthroughs5-15minutesperwalkthroughFocusononeormultipleliteracycomponentsCheckboxwhenskill/conceptobserved(beingtaughtbyteacherorappliedbystudents)Recordevidenceintheformofnotes
33
34
35
MTSS =Multi-tieredSystemofSupport
36
37
38
39
Atleasta90-minreadingblock;30minsmall-group interventioninclassroom;30-minsupplementalinterventionforthosepulledout(tier2or3).
40
Theirarticledescribestheknowledge&skillsneededtobeprovidedinpreserviceeducation forspecialeducationteacherstoensuretheycandesignandimplementdata-basedindividualization(DBI).Examplesofthekindsofclinicalpractica invariousspecialed programsandfederally-fundedtraininggrantsisprovided.TheInternationalDyslexiaAssociation’sKnowledge&PracticeStandardsarepraised.
41
Outof422students,80(19%)wereinclass6,performingabout1SDabovethemeanonallmeasures.Class3,with177students(42%),alsotendedtoperformaboveaverageonallmeasuresexceptVocabularyPairs.Class1,with32students(7%),performed1SDormorebelowthemeanonallmeasures.Class4,with97students(23%),andclass5,with28students(7%),wereamirrorimageofeachotherinthatclass4performedatthemeanontheorallanguagemeasuresbutbelowaverageontheprint-relatedmeasuresofPhonologicalAwarenessandLetterSounds.Finally,class2,withonlyeightstudents(2%),hadanunusualprofileofaverageperformanceonVocabularyPairs,FollowingDirections,andLetterSoundsbutaboveaverageperformanceonPhonologicalAwarenessandverylowperformanceonSentenceComprehension(whichtapsreceptivesyntax).
42
43
Butevenamongthelow-performingclassestherewassometimesanotableweaknessorstrength.Forexample,infirstgradeitwastheextremelylowperformanceonFollowingDirectionsthathelpsexplainwhyclass1performedbelowthe25th percentileonreadingcomprehensionwhenthisclasshadsimilarperformancetotwootherclassesonVocabularyPairsandWordReading.Interventionforthe11studentsinClass1wouldneedtotakeintoaccounttheirverylowskillinlisteningtoandrememberingconceptsanddirectionsbyrepeatingdirectionsandprovidingmultiplepracticeopportunitiestolearnnewconcepts.
44
Thevastmajorityofstudents(472,78%)wereinclass3,performingatthemeanonallthreeFRAtasks.Studentsinclasses4and5(representing8%and5%ofstudents,respectively)performedaboveaverage,butclass4was1.5SDshigherontheWordRecognitionTask.Classes1and2(representing5%and4%ofstudents,respectively)performedbelowaverageandweresimilaronSyntacticKnowledgebutmirrorimagesofeachotheronVocabularyKnowledge(whereclass2was1.5SDshigher)andWordRecognition(whereclass1was1.5SDshigher).
ThetwoclassesthatwereaboveaverageonFRAmeasures,class4andclass5,bothperformedabout1SDabovethemeanonreadingcomprehensionandwerenotsignificantlydifferentfromeachotherbutweresignificantlybetterthantheotherthreeclasses.Class3’sandclass2’smeanswereabout0.5SDbelowaverageandwerenotsignificantlydifferentfromeachother.Class2marginallyoutperformedclass1(Hedgesg =-.53).FRAclassesaccountfor24%ofvarianceinSAT-10RC.[butgoesupto61%of
45
varianceinG9.]
45
46
47
48
49
50
51