STATISTICS AND STATISTICAL
METHODS
Statisticsisabodyofknowledgethatdealswiththe
collection,presentation,analysisandinterpretationof
numericalandcategoricaldata.
Statisticalmethodsreferstotheproceduresusedinthe
collection,presentation,analysisandinterpretationofdata.
WHY STUDY STATISTICS?
Thefieldofstatisticsisthescienceoflearningfromdata.
Statisticalknowledgehelpsyouusethepropermethodsto
collectthedata,employthecorrectanalyses,andeffectively
presenttheresults.Statisticsisacrucialprocessbehindhowwe
makediscoveriesinscience,makedecisionsbasedondata,and
makepredictions.Statisticsallowsyoutounderstandasubject
muchmoredeeply.
TWO AREAS OF STATISTICS
DescriptiveStatisticscomprisesstatisticalmethodsconcerned
withcollectinganddescribingasetofdatasoastoyield
meaningfulinformation.
InferentialStatisticscomprisesstatisticalmethodsconcerned
withtheanalysisofasubsetofdataleadingtopredictionsor
inferencesabouttheentiresetofdata.
POPULATION AND SAMPLE
Populationconsistsofthetotalityofobservationwhichweareconcerned
about.
Sampleisasubsetofagivenpopulation.
Parameterreferstoanynumericalvaluedescribingacharacteristicofa
population.
Statisticreferstoanynumericalvaluedescribingacharacteristicofa
sample.
TYPES OF VARIABLES
QuantitativeVariableisavariablethattakesonlynumericalvalues.
Example:I.Q.,height,weight,incomeandage
QualitativeVariableisavariablethattakesonlynon-numerical
values,andnumbersareusedonlyascategories
Example:Gender,sex,religion,yearlevel,educationalattainment
andoccupation
TYPES OF VARIABLE ACCORDING TO
LEVEL OF MEASUREMENT
Nominal–numbersareusedmerelyaslabelsofthecategoriesof
thevariable
Ex.sex,religion,andoccupation
Ordinal–havethesamecharacteristicwithanominalvariableand
inaddition,thenumberscanbemeaningfullyranked
Ex.economicstatus,yearlevel,andsalarygrade
Interval–havethesamecharacteristicwithanordinalvariableand
inaddition,thecategoriesintheintervalscalearedefinedinterms
TYPES OF VARIABLE ACCORDING TO
LEVEL OF MEASUREMENT
ofa“standardunitofmeasurement”sothatequalityofdifferences
betweensuccessivecategoriesofthescaleisdefined
Ex.temperatureinDegreeCelsiusandIQscore
Ratio–havethesamecharacteristicwithanintervalvariableandin
addition,ithasatruezeropoint
Ratioandintervalvariablearewhatyoucallscalevariableswhile
ordinalandnominalvariablearecategoricalvariables.
Ex.age,height,andnumberofsiblings
FREQUENCY DISTRIBUTION
Itisatabulararrangementofdataindicatingthedifferentclasses
orcategoriesandthecorrespondingfrequencies.
Therearethreebasictypesoffrequencydistributions.Thethree
typesarecategorical,ungroupedandgroupedfrequency
distributions.
CATEGORICAL FREQUENCY
DISTRIBUTION
Thecategoricalfrequencydistributionisusedfordatathatcanbe
placedinspecificcategories,suchasnominal-orordinal-leveldata.
Forexample,datasuchaspoliticalaffiliation,religiousaffiliation,or
majorfieldofstudy.
EXAMPLE
Twenty-fivearmyinducteesweregivenabloodtesttodetermine
theirbloodtype.Thedatasetisasfollows:
RAW DATA OF BLOOD TYPE
A B B AB O
O O B AB B
B B O A O
A O O O AB
AB A O B A
EXAMPLE
Blood Type Tally Total %
A 5 25 20%
B 7 25 28%
O 9 25 36%
AB 4 25 16%
UNGROUPED FREQUENCY
DISTRIBUTION
Anungroupedfrequencydistributionisusedfornumerical
dataandwhentherange(thedifferencebetweenthehighest
andthesmallestvalues)issmall.
EXAMPLE
Considerthefollowingtable,whichliststhenumberoflaptopcomputersowned
byfamiliesineachof40homesinasubdivision.
GROUPED FREQUENCY
DISTRIBUTION
Whentherangeofthedataislarge,thedatamustbegroupedintoclassesthat
aremorethanoneunitinwidth.
Toconstructafrequencydistribution,followtheserules:
1.Thereshouldbebetween5and20classes.
2.Theclasswidthshouldbeanoddnumber.Thisensuresthatthemidpointof
eachclasshasthesameplacevalueasthedata.
3.Theclassesmustbemutuallyexclusive.Mutuallyexclusiveclasseshave
nonoverlappingclasslimitssothatdatacannotbeplacedintotwoclasses.
4.Theclassesmustbecontinuous.Thereshouldbenogapsinafrequency
distribution.
GROUPED FREQUENCY
DISTRIBUTION
Whentherangeofthedataislarge,thedatamustbegroupedintoclassesthat
aremorethanoneunitinwidth.
Toconstructafrequencydistribution,followtheserules:
5.Theclassesmustbeexhaustive.Thereshouldbeenoughclassesto
accommodateallthedata.
6.Theclassesmustbeequalinwidth.Thisavoidsadistortedviewofthedata.
GROUPED FREQUENCY
DISTRIBUTION
CONSTRUCTINGGROUPEDFREQUENCYDISTRIBUTION
1.Findtherange.
�??????���=ℎ??????�ℎ����??????���−�������??????���
2.Decideonthenumberofclassintervalsorclasses,wedenoteitby�.
➢�=�
➢5-20classes
3.Determinetheclasssizeorclasswidthoftheinterval,wedenoteitbyc.
(roundedtothenearestoddwholenumber)
4.DeterminethelowerlimitLLandtheupperlimitandtheupperlimitULofthelowest
classinterval.Thelowestclassintervalshouldcontainthelowestvalueinthedataset.
GROUPED FREQUENCY
DISTRIBUTION
CONSTRUCTINGGROUPEDFREQUENCYDISTRIBUTION
ThevalueoftheULisdeterminedusingtheequation.
????????????=????????????+(??????−1)
5.Determinetheupperclassintervalsbyconsecutivelyaddingtheclasssize??????to
thevaluesofLLandULofthelowestclassintervaluntilwegettheclassinterval
withthehighestvalueinthedataset.
6.Tallythedata,findthefrequencies.
GROUPED FREQUENCY
DISTRIBUTION
CONSTRUCTINGGROUPEDFREQUENCYDISTRIBUTION
▪Theclassboundariesareusedtoseparatetheclassessotatthereareno
gapsinthefrequencydistribution.
RuleofThumb:Classlimitsshouldhavethesamedecimalplacevalueasthe
data,buttheclassboundarieshaveoneadditionalplacevalueandendina5.
▪Theclassmidpointisfoundbyaddingtheupperandlowerboundaries(or
limits)anddividingby2.
▪Thecumulativefrequenciesareusedtodeterminethenumberofcasesfailing
below(for<cf)orabove(for>cf)aparticularvalueinadistribution.
GROUPED FREQUENCY
DISTRIBUTION
CONSTRUCTINGGROUPEDFREQUENCYDISTRIBUTION
▪Therelativefrequency(rf)ofaclassintervalistheproportionofobservations
fallingwithintheclassandmaybepresentedinpercent.Thus,
��=
�
�
×100