Preliminary Issues continued assessment

For students are studying English language assessment principle

1. Language Testing & Assessment
Program: TEFL, MA
Session 3
Preliminary Issues (Continued)
INSTRUCTOR: F. BEHZADPOOR
AZARBAIJAN SHAHID MADANI UNIVERSITY
WINTER, 2022

2. Teaching, Assessment, and
Test

Test/Testing&AssessmentareNOTsynonymous

Tests:preparedadministrativeproceduresthatoccuratidentifiabletimesinacurriculum
whenlearnersgatherallfacultiestoindicatetheirmaximumperformance,knowingthat
theirresponsesarebeingmeasuredandevaluated.

Assessment:anongoingprocessthatencompassesamuchwiderdomain

Teachersubconsciouslymakinganassessmentofthestudent’sperformancewhensheis
respondingtoaquestion,offeringacomment,ortryingoutanewwordorstructure

Writtenwork–fromajotted-downphrasetoaformalessay–isperformancethat
ultimatelyisassessedbyself,teacher,andpossiblyotherstudents

Readingandlisteningactivitiesusuallyrequiresomesortofproductiveperformancethat
theteacherimplicitlyjudges

Therefore,mostteachersalwaysassesstheirstudents,whetherthoseassessmentsare
incidentalorintended.

3. Teaching, Assessment, and
Test

Assessmentisthemostinclusiveterm

Assessmentincludesalltypesoftesting:verbaltesting,
non-verbaltesting,paper-and-penciltesting,essay
testing,oraltesting,self-assessment,portfolios,etc.→

Testsareasubsetofassessment (theyarenottheonly
formofassessmentthatateachercanmake)

Allteachinginvolvesassessment: teachersmake
assessmentseverytimetheyteachsomethinginthe
classroom

4. Relationships Among All the Terms
Introduced

5. Informal vs. Formal Assessment

Informalassessment:
takesanumberofforms,startingwithincidental,
unplannedcommentsandresponses,alongwithcoachingandotherimpromptu
feedbacktothestudent

Examples:saying“Nicejob!”“Didyousaycanorcan’t?”orputtinga onsome
homework

Formalassessment:

exercisesorproceduresspecificallydesignedtomeasureandhaveanevaluation
ofavarietyofskillsorknowledge

systematic,plannedsamplingtechniquesconstructedtogivetheteacherand
studentsanappraisalofstudentachievement

6. Informal vs. Formal
Assessment
Question:
Isformalassessmentthesameasatest?

Alltestsareformalassessment.

NOTallformalassessmentistesting.

Astudent’sreflectivejournalorportfolioofmaterialsasaformal
assessmentoftheattainmentofcertaincourseobjectives

Asystematicsetofobservationsofastudent’sfrequencyoforal
participationinclass
►
Tests
areusuallyrelativelytime-constrainedanddrawona limitedsample
ofbehavior.

7. Formative vs. Summative
Assessment

Formative:

Mostclassroomassessmentisformativeassessment

Evaluatingstudentsintheprocessof “forming”theircompetenciesand
skillswiththegoalofhelpingthemtocontinuethatgrowthprocess

Keytoformation:delivery(bytheteacher)and internalization(bythe
student)ofappropriatefeedbackonperformance,withaneyetoward
futurecontinuation(orformation)oflearning

Allkindsofinformalassessmentareformative(focusontheongoing
developmentofthelearner’slanguage).

Whenwegiveastudentacommentorasuggestion,orcallattentionto
anerror,thatfeedbackisofferedinordertoimprovethelearner’s
languageability.

8. Formative vs. Summative
Assessment

Summative

Aimstomeasure,orsummarize,whatastudenthasgrasped

Typicallyoccursattheendofacourseorunitofinstruction

Asummationofwhatastudenthaslearnedimplieslookingbackand
takingstockofhowwellthatstudenthasaccomplishedobjectives, BUTit
doesnotnecessarilypointthewaytofutureprogress.

Examples:

Finalexaminacourse

Generalproficiencyexams`

9. Norm-Referenced vs. Criterion-Referenced Tests

Norm-referencedtest:

Eachtest-taker’sscoreisinterpretedinrelationtomean(averagescore),median
(middlescore),standarddeviation(extentofvarianceinscores),and/or
percentilerank.

Purpose:
placingtest-takersalongamathematicalcontinuuminrankorder.

Scoresareusuallyreportedbacktothetest-takerintheformofanumericalscore
(forexample,90outof100)andapercentilerank(suchas90percent,which
meansthatthetest-taker’sscorewashigherthan90percentofthetotalnumber
oftest-takers,butlowerthan10percentinthatadministration).

Example:

Standardizedtests(TOEFL,IELTSorSAT)intendedtobeadministeredtolarge
audiences,withresultsefficientlydisseminatedtotest-takers(Suchtestsmusthave
fixedpredeterminedresponsesinaformatthatcanbescoredquicklyat
minimumexpense)

10. Norm-Referenced vs. Criterion-Referenced
Tests

Criterion-referencedtest:

Designedtogivetest-takersfeedback,usuallyintheformofgrades,on
specificcourseorlessonobjectives

Example:
Classroomtestsinvolvingthestudentsinonlyoneclass,and
connectedtoacurriculum

Muchtimeandeffortonthepartoftheteacher(testadministrator)are
sometimesrequiredinordertodeliveruseful,appropriatefeedbackto
students,orwhatOller(1979)called“instructionalvalue.”

Distributionofstudents’scoresacrossacontinuummaybeoflittleconcern.

11. Norm-Referenced vs. Criterion-Referenced
Tests

NRTsmeasuregloballanguageabilities(forinstance,overallEnglishlanguageproficiency,
academiclisteningability,readingcomprehensionandsoon).

InNRTsstudents’arecomparedwitheachotherintermsoftheirperformanceonthetests.
Infact,eachstudent’sscoreonanNRTisinterpreted relativeto(incomparisonwith)the
scoresofallotherstudentswhotookthetest.Suchcomparisonsareusuallydonewith
referencetotheconceptofthenormaldistribution.

ThepurposeofanNRTistospreadstudentsoutalongacontinuumofscoressothatthose
with“low”abilitiesinageneralareasuchasreadingcomprehensionareatoneendof
thenormaldistribution,whilethosewith“high”abilitiesareattheotherend.Infactthis
normaldistributionranksthestudentsbasedontheirabilitiesandthisisinlinewiththe
purposeofanNRT.

InNRTs,whilestudentsmayknowthegeneralformatofthequestionsonanNRT(for
example,multiple-choice,true-false,dictation,oressay),theywilltypicallynotknowwhat
specificcontentorskillswillbetestedbythosequestions.

AnidealNRTtestisatestthatdifferentiatesandranksthestudentsbasedontheirscores
andwecanlistthestudentsfromthemostproficienttotheleastproficientones.

12. Norm-Referenced vs. Criterion-Referenced
Tests

ACRTisdesignedtomeasurewell-definedandfairlyspecificobjectives.Oftenthese
objectivesarespecifictoaparticularcourse,program,schooldistrict,orstate.

TheinterpretationofscoresonaCRTisconsidered absoluteinthesensethateachstudent’s
scoreismeaningfulwithoutreferencetootherstudents’scores.Inotherwords,astudent’s
scoreonaparticularobjectiveindicatesthepercentofknowledgeorskillinthatobjective
thatthestudenthaslearned.

ThedistributionofscoresonaCRTneednotnecessarilybenormal.Ifallthestudentsknow
100%ofthematerialonalltheobjectives,thenallthestudentsshouldreceivethesamescore
withnovariationatall.

ThepurposeofaCRTistomeasuretheamountoflearningthatastudenthasaccomplished
oneachobjective.

Inmostcases,thestudentswouldknowinadvancewhattypeofquestions,tasks,andcontent
toexpectforeachobjectivebecausethequestioncontentwouldbeimplied(ifnotexplicitly
stated)intheobjectivesofthecourse.

AnidealCRTisatestthatattheendofthecourse,allthestudentsgetthehighestscore.
BecauseinCRT,theteacherissupposedtoteachandtotestbasedonthesyllabusthe
idealizationisthatattheendofthecoursetheteachercomesupwiththehighestscores

13. Norm-Referenced vs. Criterion-Referenced
Tests
Characteristic Norm-Referenced Tests Criterion-Referenced Tests
Type of InterpretationRelative(Astudent’sperformanceis
comparedtothatofallotherstudents
inpercentileterms.)
Absolute(Astudent’sperformanceis
comparedonlytotheamount,or
percentage,ofmateriallearned.)
Type of Measurement Tomeasuregenerallanguageabilities
orproficiencies
Tomeasurespecificobjectives-based
languagepoints
Purpose of Testing Tospreadstudentsoutalonga
continuumofgeneralabilitiesor
proficiencies
Toassesstheamountofmaterial
known,orlearned,byeachstudent
Distribution of ScoresNormaldistributionofscoresarounda
mean
Varies,usuallynon-normal(students
whoknowallofthematerialshouldall
score100%)
Test Structure Afewrelativelylongsubtestswitha
varietyofquestioncontents
Aseriesofshort,well-definedsubtests
withsimilarquestioncontents
Knowledge of
Questions
Studentshavelittleornoideawhat
contenttoexpectinquestions
Studentsknowexactlywhatcontent
toexpectintestquestions

14. High-stakes vs. Low-stakes
Decisions

High-stakes:
decisionsthatdeterminesomeone’snextphase
oflifelikeuniversityentranceexamsorjobinterviews

Low-stakes:
decisionsthatdonotinfluencesomeone’slifea
lotlikequizzes,midtermsandclassroomfinals