Abstract:
Severalstudentssu erfromdyscalculia.Manyresearchandstudieshave
beencarriedouttotacklethisproblem.Mostoftheseresearchestriedto
identifywhetherastudentsu ersfromdyscalculiaornot,withoutanyfur-
ther detailsonthecase.Basedonresultsoftheidenti cation,researchers
proposedsolutionsandwaystotreatdyscalculicstudents.Thisresearch
tries totacklethedyscalculiaproblemamonggrade3studentsinSouth-
ern DirectorateschoolsinPalestine.Theresearchattemptstousearti cial
intelligentmethodsandtoolstoclusterstudentsaccordingtotheirdyscal-
culia case,andproposedinformationtechnology-basedtreatmentforeach
studentaccordingtohis/hercase.Ittriestogoafurtherstepintoidentify-
ing whattypeofdyscalculicstudentonsensingnumber.Theapproachused
in thisresearchisperhapsthe rsttouseAItoolstocluster;butnotclassify,
dyscalculic students,andtriestobreak-downthedyscalculiaprobleminto
three majortypesusingthisapproach.Toachievethis,anintensivelitera-
ture reviewwascarriedout,thenanexamusedbytheMinistryofEducation
and HigherEducationinPalestinewasmodi edandtestedbyexperts.The
modi edexamwasappliedtogradethreestudentsatschoolsofbothgen-
ders inHebronandYattadirectorates.Resultsoftheexamwerecodedand
input toanAItoolRtool.Thetooluseshierarchicaltechniquescluster-
ing. Weapplytwohierarchicaltechniquesalgorithms;theSinglelinkand WARDmethod.WARDalgorithmisanagglomerativehierarchicalcluster-
ing procedure,wherethecriterionforchoosingthepairofclusterstomerge
at eachstepisbasedontheoptimalvalueofanobjectivefunction.Single
link istechniquelooksinthedistancebetweentwoclusterstobeequaltothe
shortest distance.WARDalgorithmshowbetterresultthanSingleLink,af-
ter calculatedtheCopheneticCorrelalationCo cient(CPCC).SingleLink
is thelowestCPCC,whichis(0.46).ButWardhas(0.7)CPCC.Notonly
wastheuseoftheclusteringtodetermineifthestudentisdyscalculicor
not asthepreviouswork,butwealsouseWARDalgorithmclusteringmore
deeply todeterminewhatkindofdyscalculiaastudenthas.Theresultwas
clustered intosevenclusterssuchas(weaknessonthethreeskills,theab-
sence ofanyweaknessinthethreeskills,studentswhohavezeromarkinthe
exam). Resultsshowdi erencesamonggenderandamongdirectorates,and
inconsistency betweenclusteringresultsandstudentsmathachievementin
school.Basedontheresultsandtheliteraturereview,amodelhasbeenpro-
posedforthetreatmentofdyscalculicstudents,whereitconsistsalsoofthe
identi cationandclusteringstage.Themodelwasevaluatedby24experts,
and theircommentsandsuggestionwereincorporatedinthemodel.
Description:
CD, no of pages 80,30112, informatics1/2017