Reliability and Validity in Testing

Reliabilityand Validity in Testing

Author’sName

Reliabilityand Validity in Testing

Reliabilityandvalidityarecloselyrelatedin testing.Theyare usedto measurea test’saccuracyandconsistencyin a trial.In reliability,thesamescoreshould be attainedwhenadministeringthesametesteverytime.However,evenifthescoreof thetestis thesameeverytimebutthetestis not testingonthesubjectitis supposedto assess.Thenthetestis givingthewronginformation(Cohen, Swerdlik &amp Sturman, 2013). From thisthenarisestheconceptof validity.Validityin testingshowsifa testis accuratelytestingwhatitpurportsto prove.Itwould be verywrongandcompletelyinvalidto usean arithmetic testin assessingverbalskills(Makkonen, 2013).

Intesting,reliabilitymeansan extremelypositivecorrelationbetween theresultsobtainedfrom theassessment.In this,a testshould be ableto producestableandconsistentresultsifadministeredto thesamegroupeverytime.Fortheoutcomeof a testto be reliable,theyare to be approximatelythesameeachtimethetestisconducted(Geisinger, 2013). Thetestcannot be dependableifitproducesdifferentresultseverytimeas itwould not showtheactualrelationbetween thevariousintervalsitwasadministered.An exampleforthiswould be like whenan individualweighsthemselves in theday.Thepersonwould expectthesameresultsin theeveningifnot thentheassessmentwasnot satisfactory.Asthiswould implythattheweighingmachineisbrokenfornot beingconsistentsince theindividual’sweightcannot fluctuateverymuchin a day.Thistoois expectedwhenadministeringa test.Theresultof an analysisshould be similarandconsistentwhenbeingappliedto thesamegrouprepeatedlyat differentintervalsof a dayforitto be dependable(Geisinger, 2013).

Anotherconceptof reliabilityis thatof thecontent of theenquiry.Thisis to saythatifa studentgotsomeof thequestionson a particulartopicthenthestudentshould getotherquestionson thesamesubjecton thesamelevel. In addition,thetestshould be consistent.Thisis to saythatsince thequeryis to assessin a particularareathentheanswersshould not varyas muchtoo.Forexamplefortestin math should havemathematicalquestionthat will giveriseto numericalsolutionsandnot vocabularyanswersat a differenttime(Miller &amp Linn, 2013).Testingreliabilitycan be challengingsince there is noactualtestthat would giveaccuratelythelevel of reliabilityof a test.Someof themethodsusedto verifythereliabilityare:

1.Test-retest, thisimpliesthata testshould givecorrelating resultsifitisofferedagain.2.Alternateforms,thisshowthatevenifthesametestisprovidedin a differentformthere is to be a positivecorrelationin thetwo results.3.Internalconsistency,thismethodcomparesone-half of thetestto theotherhalfandensuresthattheyare testingon thesametopic(thetopicin assessment)(Miller &amp Linn, 2013).

Validity,on theotherhand,showstherelevanceof a trial.Itindicatesifthetestmeasureswhatitpurportsto measure.Thismeansthatan analysiscan beintendedforone kindof assessmentbutinsteadmeasurean entirelydifferentareaof an evaluation. Ifa testis to be accurate,thenitmust testaccuratelywhatitis intendedto measure(Geisinger, 2013). Validityis dependenton howbigan extentthetestclaimsto measure.Thenfora testto besaidto valid,itmust assesstheactualareaitis intendedto assess.Hence,thiswill leadto accuracyin theresultsobtainedforinterpretation.Inthistoothere are somemethodsusedto determinethevalidityof thetestthesemethodsare1. Content, thisis theextentto which thecontent of thetestmatchesobjectsin assessment.2.Criterion,thisis theextentto which theresultsof thetestagreewith an externalstandard.3.Construct, thisis thedegreeto which a testagreeswith othervariables as in accordance with a rationale.

Itis importantto havebothreliabilityandvalidityin testingandassessingsince thiswould leadto moreaccurateresults.Theaccurateresultswould be becauseof consistentresults.Asper thepartof reliability,thetestsare requiredto be stableandcoherent.Thiswill outlinethattheresultsare dependableandifthetestwasto berepeatedtheresultswould be almostsame.However,a testcan be reliablebutfailto be valid(Makkonen, 2013). Thisis becausethetestcan givea consistentflowof information.Informationthat correlatesoperfectlywith thepreviousresultsobtained.Thistestwill providereliableinformation,buthowaccurateis theinformation?Didthetestactuallymeasurewhatitwasintendedto measure?Therefore,theconceptof validityis neededto ensurethatthetestwasnot onlyconsistentandstablebutalsoassessedon whatitwasdesignedto teston.Thistwo conceptare paramountin bringingabout accurateresultsof a test.Theyensurethatthetestcoveredwhatitwasdesignedto coverandgavepreciseanddependableresults(Miller &amp Linn, 2013).

Intestingnormingis theprocessof makingnorms.Thesenormsare groupsof studentswhoare chosento representtheactualstudentpopulation.Their testresultwill describetheperformanceof a well-definedpopulation.In norm-referenced test,thestudentsare supposedto be comparedto a normorothersimilarstudents’averageperformance.Theconsistencyin thisperformanceswill enableto ruleout a learningdisorder(Cohen, Swerdlik &amp Sturman, 2013). In thenorming process,a groupof studentsis identifiedwhowill undergoa madetestitemthat will be thebasisof testingotherstudents.Thisgivesa basisof comparisonbetween students.Since his scorecan be comparedto thatof thegroupto establishhowhefaredin thetest(Makkonen, 2013).

References

Makkonen,R. (2013). Exploringthe Reliability and Validity of Pilot Teacher Ratings in a LargeCalifornia School District.Berkeley, CA.

Geisinger,K. F., &amp American Psychological Association. (2013). APAhandbook of testing and assessment in psychology: Vol. 1.Washington, D.C: American Psychological Association.

InternationalHaifa Verification Conference, In Bertacco, V., &amp In Legay, A.(2013). Hardwareand software: Verification and testing 9th International HaifaVerification Conference, HVC 2013, Haifa, Israel, November 5-7, 2013,Proceedings.

Miller,M. D., &amp Linn, R. L. (2013). Measurementand assessment in teaching.Boston: Pearson.

Cohen,R. J., Swerdlik, M. E., &amp Sturman, E. (2013). Psychologicaltesting and assessment: An introduction to tests and measurement.New York, NY: McGraw-Hill.