Tiak Education Society's

Commerce,Nerul
ofScienceand Bank
Solution
S.KCollege 5
Ouestion
thas
TYBBISemSubject:RM ratng scale
questionnaire within
or
a survey
Ditferential Scale iscompany, brand or
any "entity"
i) Semantic product,
to ratea adjectives
options.gramatically on opposite likelv
asks peoplemut-point rating are unsatisfied and
rames ofa answering options satisted / between.
ii) / hate,
These survey exape. love intermediate options n moe
each cnd. For with Dièrentia l Scale is the
at to return Semantic towarde
retm / unlikely ires using
or.questionna emotional attitude
peopk's
ii) Srveys
get information on the
reiable way
to psychologist, invented
topic of interest. famvus American
iv) Charles Egerton Osgood, a combinations
semantic differential scale measred using these empbyee
variation of sobjects can be
upcoming product l:unch or
v) A wide oubok about an
lIke customers something, For
satis faction ime:sures attit.:des towards wàk
diferential scale aititude to the vord Wor"'
vi) A sernntic measure a person's
CxAmple, you coukd
the following scae: 7 nteresti
NCUSydry
Boirg
UnneC2S5a

UNIT III

Questions
Sr.
No
1 What is Data Editing
examned to detect any errors or
A Data editing s a process by wiich collected data is proceeding further.
omissions and futher these are corrected as irn:ch as possihe before
Need' Objectives ofDaia Edting
1. To check data accuracy
2. Remove Erors,
3. Inpreve Quzlity
4. Perfomance of fek st:f
5. Decisibn Making
Eliting is of tvo types:
FIELD EDITING:
This is aiypeoftdäing that reates to abbreviaied or k gble writen form
of gathered data.
 S.K Colle geTiak Educath n
TYBBI Sem ofScience
S MulSocicty's
Commme ree, Nerul
Sucih Qucstion
Subje ct: RNIBank Sohulon
day afteredtmgis
the more fective
The
nvestigaiornteview. when-üone on same day or the tery
CENTRAL EDITING:
Such type cf editing
mmist not jny to
conchus ion while domg fickl cding,
next
has been conpleted. rehtes to the tue when all
Here a sngle or data collecion
common process
Wrong
As a nleplace, entry n editor corects errors ke entry n
the
all the wrongwrong nit e.t.c. he
2 How is Data Edited answers shoukd be dronped fom the
A
Validity and completeness af data - fnal results.
respotses obtamed, based on the The validity bf the data refers tothe
Range- Ihe range sets the possibk rarge of
type of editng, the tems minimum and maximumanswers tor each varDecorrectness the
Ci
the data collecton too, on the questionna ire are expected vahues of he variahe. n ths
ibat data n a gei ild ani nchdes all the toms dividually used te recordchecked (he
or cotect thequestiona re S
are 1vihin the dula) to verity
Duplicate dafa entry- Verifying that theboudaries data of speciied îor çat feld
entersd only once, wih no each unit of the regste1 ur the
Jatabase was
Logicai cosiste ncy- Consistency dupicationis the
nterdependence between the varnbles. This presence oi hgcal relatons hini and
connections between data fiekds or variables type of editing takes ino accont the
Outliers- This type of editing folbws other checks and js used fo he
vaues, based on the distrbution of the current data and detecticn of cxtiene
Edtng for previous data series
1, The irccrect answer
2. Incompiete Answer
3. inconsistunt Arswer
4. Dort know aswer/ "No reply

3!Coding categoi ized into

A
Cassifcation of the data mpies that the colected raw data is
feature.
COMTOn group havmg conmon
characteristics arc placed n a coimon group.
Data havng commcn which convey
categorized nto various groups ot chsses,
'is
The entire deta colccted
researcber the
a meanrg to the syrbol is assigned to the answers in order to pu
amral or
In coding a particuhr or casses.
Sone defnte categories beappropriate arad
responses determined by the reScarcher shoud
resporses
The casses of responses tv categorized ito
suitabie to the study. and-efective anayss as the
Coding erables
etticiet
thequestionraire or
or desiging
meaningful casses.
considered whie devebping
are
Coding decisiinscolection tool through computer
data
any other done manually or
Codng can be
 Tilak Education Scciety's
Commerce,Neru!
nce and Soution
ge
ofrScieQuestioi1 Bank
S.KColle 5
TYBBISemSubje: RM
respondents
by
Essentials of CODING ofanswers given
types
Should cover all overap
Categories shouk notfor quick reference
Code 'sheet prepared
knowedgeable person
Done by symbols, aphabets
Use of numbers, categorized into
colected raw data
is cOmmon
Classification that the
ofthe data implies commOn group.
A Classification common fcature. placed n a classes, which
group having characteristics are
various groups or convey a
Data having Common cateyorized into
colected is
The entire data
meanng to the researcher.
Classification is done in two ways:attuibutes.
according to intervais.
I. Cassification to the class THE ATTRIBUTES:
2. Chssificatio n accordng that Culi be descriuiive
ACCORDING THE gfouunion ciai actel stCS weight
CLASSIFICATION height, nconie
> Here ihe data is
cassifici ontie basis e.f.t. er numeral
ie
marital status
horcsiy,
ike iteracy, sex, measred quanttatvely bn
e.t.c. nature and cannot be
quaitative in
Descriptivc featres are anaysis attrbutes and tie
kindy consiJered wlie naking an known as statistics of
are data is attributes.
such classified
> Amlys is used r the classification accordng to the
Cassification is known as
CLASSIFICATION UN rHE BASiSmeasured OF THE INTERVAL: analyzed with he be of
quzntitatively an
mmerical teaiure of data cen be produchon, age, weight e.LC. come uNr
Ihc ncone, tu
Surie statistical unit ike
the data relating to
statistis of vaiabks and tiie uata ib cassifed
ofdata is knoV as
hs catcgory. This tpe iNTERVAL USUAL:y
way of nterva ls. TO TiE CLASS
CLASSIFICATION ACCURDENG
FOLLOWING THREE MAIN PR0BLEMIS:
NVOLVES THHE
i. Nurber of Chsscs.
Imits.
2. How to select chss of each cass.
fcquercy
3. How to detcre the

Tabulation
5
cclected has to be arranged in some knd of concise and bggcal
A The mass of data
order.
the raw daia and displays data in fom of some statistical
Tabuation surnrarizes
Tabuation is an orderly TAngement of data in ows and cohumns, l:
tables.
ivolves
Sonting and countng
Grouping, chssification and surnary
Speed and accuracy
Understand mutpk variabies
 S.KColleBBTilak
ofSciEducation
ge Sem nce andSociety's
eOuestion
5
Comme rce,Nerul
Bank
Manual TablatonTaly Marks Soution
Subject: RM
Machine Tabuation
BASICIPRINCIPLES
Tables OF
1.
should
2.
3
4.
Every table
Cohn
Units of
be

headng TABULATI
shouldclear,
be concse
di
row st:atly ON:
&adequatety
titled.
unbered for easy reference.
5. Explanatory meaSreMert
sould headngs
be of?
the table shoud be clear & brief.
6. Source of
7. The coumns &
footnotCiosn concemingspecifed
nformat of the tabk
at appropriate places.
data shoukd be shouú£indicatcd be placed at appropraie pkes.
8. Demarcation rows shoud
be clearty
9. Comparabk shovd also be madecearty separated with dark ires
10. The figures data
n be pul sidebetby
should weenside. data of one cass and ttat of arothe.
percentafigres,
11:Tie align1ment of the
spaced to
ge
shculd be approx1rated before tabuatn
enhance
12. Abbrevietions the sybosof tieetc. should be prorerty alged ard zdecuately
reatabilty
shouki be avoided. sare.
r£is
Percentugc af EA a d
BSc(?JC-15)
in e
Aunual Ex aion Passei
Boany 29
Chemisre 30
4 H2.7h

Histoy
34
Malayulen 3R
37 7.37
Mathnatics 3 20 U7.50
rhys l s 28

Zoolugy 23 76.67

6 Data Display
A Display ing data ia iesearch
comrOn in rêscarch Tab'ks i vey ckix r
Tire use oftabes O dspay data s verv vory wel and rakes the data ver;
They organe data
prescrting a arge amount of data.corton, boty, fotnoies
Ttle,
isble. A table consists of

Types ofgraphs vovcd n

-on the type ofdata and the varabes
depending
sutabk graph ype
You can use a
the data.
of.
A histogamcosst
paen
The histogram br dspkying data.rectangesbavem mE
highly sed graphTit s r two
graphthat is otheroa e varsbe 2swl f e yofect
histogra1n isadrawnpext toeach
drawnfor a snge representthe
The thatare canbe histogram vaabes
rectanges histogram
ihebarsn the contios
between them A of
Leight categoncaland
varjabkes. The forboth
than two canbedrawn
Variable, It
 re, Nerl
nc and Cemne
SKCe le ge of(ScieQuesun Ra Sot
IYBB Sen 5
Subject: RM

30

Age

Ihe Lar chart

tat t s drawa cnly òr cateonal lttveen
ogam except in spare
The har chart is mkr to a
catemrical varabks tere, i i drun wtlà
varabe Srce i sIsed for
the rectanges.

Simple Bar Chart

The piechert The pie chart is a circk, as a circk has

diferent way to dispky data circe represent the whol
The pie chart is avery percentage ard the oe pie or the
360degrees so itis taken ir divded nto sices or sections, each section represents
popubtion The pie or cick s
sub-category.
magitude of the categery or the
 icalon Socety's
eofence andCom eTCe,Nenl
1YB Sem5OSusÚbOjgnet: RMBank Sohtion

e Line Chart IMwlry Aul

Raspbety: 123

ime cart or
eies of data ine plot or line
5f chart nonts gmphisave
COmmiil mcaled
mAy 'hfekls.
of chart
acrd comedted by straightwhicineh åsplsegrcnts.
ys tioraIt's:

Agr May
ar
jan
Desktops TechCaTheNet,com

chartThescatte
ona vac
The scattergram thedata
pbt chare | heother
to
smpe way aftcctsthe
very rable ofdots.
 e ey

2001
180

100

wneooned

SAnalysis ot Data-Meaning
cclected and orgnucd so that
Datsanahss sametod n whch deia s
A

ecRt derhe hebft nfme trn kum t

arahss is fo bok at wtat fle
n oter erds, the mmn pirposn of deta
deta e tryng to feit
daia fak sÑvA
For evamgk, wiat does, the úats slass or do Wht dues the
or de?

Festures of bata Analys is

In depth Evaation of orgarized data
Afor Tabation an befre derpretaticn
Process of re-arranging data
Meaningtul data
Iproes que lty of ieacarch
6 Purpose of Data Analysis
A I. Tosre that necessary niormaton s avalable o soive a probiens
2. Toconvert usc data colected ito meaingful and understandab'e orm or
aterpretalion
3. Toconfrn tat data is kgiate to see the data Fais wthin necessary parameters
4. Toverfy de relability of mezsuring istruents
5. Toconfrm that corect questions having used in a compter program
6. To deebp answers to yestiors through examinatiou arnd interpretaton of data
7. To understard resuts fom surveys
8. Tcprovide inforation on dat the apps for data quality assessmert
9. Topubish or suaries fnding
7 Types of Data Analys is
A Types of Data Analysis:
a) Descriptive Analysis
 S.KColTYBBle Tiak
ge Education Society's
ofScie nce and
Sem 5
Quantitat
Qucstect:ionRMComme
Subj Bank Sobhirce,toNn erul
For the
vetandusestheydesMmeasteridescrbesres ibhe-ng mporant- characteretes
purpose
it
of f

centprralopertetndency,
ies,
of the datasel, of
b
measres of Le. mean,
is ation rddispersion ic. rauue,
data devi
O The medan, mode
helb of suunmar ised vararce, elc. staniard
by devaton, qiatle
graphs, to and the
nmencal. grephicalresearrher, na
usetul way, with ine
represent data n an toos such as charts.
b) Infe re ntial acCLraie Way tabes, and
Analysis
Inferential Statistics is l
popuation. 1e. the about peneralofisiheingsanple
from the sampe to tle
the larger resuts of analysis
li is a popukat
convenien'
io n, from which the
way ty sampk is
cain be detuced to
aken.
poss5l to querydraw
it is not
conchusi
each and every o ns ahout the popviaion
Tle sampk
choscn is a memher of the umiverse. when
reprcsentative of theof entire pop1lation
therefore, it shoulkd coriain mportait
infereniai Statistics is used tu îcatres the popuation.
of the popuation on deternne the probability cf propertiS
the basis of the ploperties of the sannple. by
employing probability theory.
The major iferential statstics are based on
ihe
as Analysis of Varince, chi-square test, staistical cdes such
studeat's distribution.
regression analyss, etc.
c) Correlation Anaiysis
If here is absoktey no coureation presenI, the vahue giver s0.
Ne Ccri blalioa
125

10

. 7.3

3,3

Perfect linear corelation: vaheof1.

comektion is given the
perfct positive given the vabe of-1.
A correbtion is
negative
A perfect
 Tilak Education Society's
Nerul
Comme rce,
ofScience and Soh1tion
S.KCollege Bank
5 Ouestion
TYBBISemSubject: RM
Correlation
Perfect Negati ve
Perfect Posii ve Correlation

79

3.

d)
Causal Analys is
cause and eflect
"Why "ot research - consequences
tlects, resuts,
Root cause, reasons, e
popuktinn
Crcukr cause - Birth ad tcgcther
ruhted canses jorcd
Systumic cause - mult ip le
Fg: cultural change

Intc rpre tation ofDsta-Essentials ard

Essentiais of data inte:jrrtation
to provde coxhsrn
A
Interp1etation ofdta sleld be neanungfal
It is acontnuoiS proccss
the presentatinn ofdata
Data ineipretat On heps to conpkte unterpretaton
Fexibility is essental teatue of every zurpretat rn
types of data
I1 Is portant to gve equal import:aike tc al

Fyu steps i data iriepretaion

1. Evahating data
2. Dra:" trtate cochsions

3. Test tentative concision

4. Amive at fral conchsion

Impotance/Sig ni licance cf Data Interpre tution

A Evahation of colkctcd and drawing conchsions
Important. step in Marketing ReScarch
Recommexdetion only afer critical analysis
Logical conchsicrs
Creative process
10 Significance of Data Interpretation and Desciptive Analysis
 1.
emlandLogRaerprrpprerlteeasentalolonn litlon \SlNeSdiajer. and Comine
re ol
4. It
$. mlToehelrpprestalon
lo defy
test yolvalthaesite s
s

ac alr:e prces
eIce,Nerml
wterpLalon busiesevchutn
ypes
fnding and of of data

yrrieedpr ain
| Testing of hes lo ala rescarc h
A|Hypottetuihqesuesis hyarepcanGthesi - Oue
be the
Teps
fhtre Tierprel c
o io
uratryarrive
pr
al
ec
sOe1r
dchswrs fudings
ParainerK Tests.ivded rio two vanoUs 1 esed resu ts ve

SaTmple"T-et
b

GeDotwhernercaasly, t-tes\tessisare uscd broad catieecghoaiueses. recom

ypothes
Te
The
ndation and hens lu

T-buttesttheS ure moTeassuis TOmaherli we have Lurge P'aSarperametK Tests andtestiNon-ng

neilous are
most vhen
ha
7-testS aerimmos mory suale r sEs
Re
Null and cogm n e Usefsed
l iionan of ie aampe
possbe nlcnces wiofnen theZtesws .lard data, se(1(n>30)30),
,
AtNalemeteHyputhypotesishe sis sanplng eers deriaion is kng
Anul
diItfference
is he
hypct hsis
exist
ta
betwecnsatsthetcal hypohesis in
ongnai or dciaut set of varables. which there is IK
HO (H-7ErD). t s Stateiment, wih ie
efect, Jien
sigif:ant
Hypoiesis: ahvays the lypothesis that is tested represeAkematented by
stat1stical
a bypohesis used in hypothesis testng,
sigificant dierence between the set of vartbeswichandstatesis otentat there 's
to as the hvnothesis other than the nul bypothesis, oteu derotedreferred
by Hl
(H-one)
It-is what tha-iesearcher seeks to prove n an ndrect way, by using the
test.
of the
 Thk Edcatio Societv's Nerut
ComLerce,
Sob1tkon
andBank
nce
cllectrenrdhg those
ofScleQuestion
S,KCollegeSem Subje
5 ct:RM
TYBBI eder eut
N o mks to
TITe than
Hypothesiy: reguhr basis likcly
IXore
meas on a
Eanpks.Null
vgetan Fennks are
who eatI II;ypothcsis: basS.
Altemative reg1lr
neas ona
Cgarun Student2
and
IIEror
(Iregn'ar
Type I&Type Student I
scoring
iess
(Regularand
marks)
Action Topper)
Incorect (Beta)
error
Correct

Corncct
Passes
ncornect (Apha)
eITOr
Fails

make a type I error.

you
Type ! Eror true aul you reject it, s the kvel
of
hypouesis is' swhich
Wien ic n l type ! error is a,
inaking a
prcbubity of hypothesis test. that you
lhe
you set or you accept 2 5% chance
Sigrificance wvilling to
irdicetes that you are
An a of0.05 nypoihesis.
when you reject the mel or a. Te l Emor
are wrong vake
risk. you mst se 2iower rejcct i, yU ake a type
To lovwer this bypothesis is fake and you fal to
When thc null
wizh depends on the power
il error.
a type IIerrmr is B,
making
The probabilty ot
of the test. type il crror by ensuring your
risk ofcommitting a
Youcan decrease your
test has encugh puwer.
mull hypothesis wBen it s ake is equal to
The probablty ofrejecting the test.
-B.This vale is the power of the
Leveiof Significance
be wiling to rsk
Maximum Level of typel eror which we woud
Can be 5 %which means on atiaverage 5% Type leor. 95% conidence
of no errOr
Can be 1% which means onan average 1% Type2 error
12 ANOVA
 S.K CollegeTiakof Lduc aton
nce Society's
1YBOI SemScle5 Ouestkn
The
andCorme
Subject: RM
rce, Neru
Bark Sokitn
any onc-way analys1s of var ance
statistically simificant (ANOVA)
ndependent (relhted) roups.dif erences
S Used lo
dctemc whcther there are
The
onc-way between the means of thrcc or More

and ANOVA copares the means between he oips Vou are nterested
detemines
from carh other. whether any of those
n

means
are statistically differe nt
sigifica ntly
Speciically, t tests the
null
hypothesis:
where p=
gop mean andk=
IE nunber of ggopS.
the however, the cne-way ANOVA
retums a s.atistically significant resut, we accept
atemaive
statistically typothesis (HA), which that tiere
that
are is are at least two goup means
sigifca ntiy differen: from cach othe1.
Exanple 1: Á
leadng beverag markctng research fum tcsts ie efiectiveness
using asanple of three new flavorings for a
peopk cach. (rop I
of 30 pcopk, diided'raxdorly into three grops of 10
3. Each t4stes
is tincn gven1avor I, goy 2tastus fzvn 2and TOip 3 tastes ayor
persor.
was. Dctermine a questionna ire which evaluates how enjoyable the beverage
favo ings. whethcr there a perceived Significant difference between the ihree
is
Or l hypothesis is that any difference betveen the three lavors is due to va

Flvcr1 la:x2ve2
13 12 7
17. 19
19 15
11 16 14
20 12
15 14 16
1 18
18 11
12 4 14
16 11 11

MST
F=
MSE

Where,
F= Anova Coeffcint
MST= Mean sum of squares due to treatnent
SE=Mean sum of squares due to erO.
 14
13|F-
testA
Chhsqunr SquareChi
8. 7. 6. 5. 4. 3. 2. 1.
Maki
The ng Statng
ComputSelecting
ing Chi-square than Step Step StDegree
Abypothesis.Significa ep Step variance 1here r=ol2/o22Stcvarances. ratio
statisticaily
Making relatiorships p Varances
unequalare. that Step F
generate UsesF-test
fhlhws statitcal 1. 4. 5: 4: 3. 2: the of
gereral the 1: Value F-dist1ibution. is
If ol2 the
a the kypothesis from
Goodncssa nce,Compare ofe Calcuhic variances Firstly, an naed
assurptions
the critical the LookCulkuat = tvc
decisio n five tcst: freedom is Sig ni
the research met h od we F at assined Larger
Variance
Sample
Smaller variances
ficavahue. F after
formuh sapling hetween vae statistic the the frame S.KColle
test stens an reject the the ar e nt. Fthe
of (df2)
F test The TYBBI
nd F H1:
equal . the Sample Invaue
astatistic and mferential
used deyrces to ge
Tiak
is (random fit two atthe statistic vaiue
is = statistic null order vahie morise Sem of
refers be ol2# Varlance a 5
Scie
interpretindigstrbution nul to
the greater
nul n2 HO: and toprcanovake
variables required in of arger minent Subject: Educat
be
RM nceion
Question
determie e -l ol2
(f G22.aternate compare and
hypothsampi
eses ng) how
to statistics hypo:esis. obtaned thshere
than frcedom
F = used on
table sample distributio the Bank Society
vel Here 22.
tvo to
analyFst Commerce,
and cbseorganized the in 's
tle nl
Dcgrcc detemnedistribut PLA. Soh1tion
goodnesstechnque critical
lf Step and vararce n).
hypotl12 ohihesie s. varances,
resutspeci
s fying oi akermat
aiid e Nerul
the th 2 n2 ioFin.sher.
insiguicance, cf ßZ2
a
F wth
vahue aie and The whetone hVarious
cbserved designed of statste frccdom ar e
er F-Test
f:
bivariate tlie o22 has testthe is
the atthe hypotthieesis null
the sanpic is calculate
theto
statistical is
test data we crtical
Gbtaieà (dil) smaller symbolshypothesis any
to caot requTed slales
thatthe test
statistic are 1able. lest vahie sizes nl- tests that
to tor n samnk forassnes
those signifcant level
rejec: Step obtained
he 2 of
predkted is
ru! ws .

A15
Pair5e
Ifsar The
 varables the
after be and
want
knUW the can then
to
after va and ncreases
or
quartiative corelatio n
and
DPf direction,
variable example Demand
of and aner before
goups
samples We beforevaes
weights two another different
risionethatcs mdecreaseforrectpocion ve.noven
Charactebetween
related montns. measured
same of arxd
dist:bution sets nean Prxce
Nerul
hwothe
for
between
vahues) nGmal

3 mice.
during
snil the the
been
20
and the
compare
association
Questionscorrelatiolinear
n sameis reatifotersehneitxpanple
is
Soh1tion
Commerce, fryed X
relatedof
hastreatnent
mice
IV
UNIT ass0Cztoornncrease te
s andpegative
Tilak
Education
Society's meansof ntervalsthe not treatrnent
weight mice. and denote nmove nmove
Bank
pair is 20hefre uscd
to charact
Meaning
of eristics unt egat
orpost
ive iveththeen varäab Cor
ks ekton
The.amountyarableistmenta
Subject:
RM means noi
approximates same
andOuestion
the
(le.,
distrbution
are 21 thi
sampl.
tocorrelths
ation
Copare conßdence on
receved ofynes the be afor
Analysis that
assumedecreases.
Science comjaresaplespáct can
ftwo two
vaues
increnses saple wegn of
5 an cf weightt-test
TYBBISem t
of n
of iohwodstributicn
uscd 30 cakuate shape
based
whenmicean thesets and use gWE
Correlaton f
ge
to rehted as 20 the p2ired
queston, Meanine
We
Colle iave sze tests test20
want
is tvce
is han the data,X
(-test sample for I trcatment ges situatios,
Ile
not ksst to
you quemly
Different
ttesl
he ndepender
=~0-E Sample
Test case,
sarmples
s Use AS
of thie measng
exxample to Tns
tre pared
the
les
adiic
this
We
the answer ireatrment.
trcati:.
such No.
Sr. Ans
In sample
The
paired anwhether from
samples.
Paired the As To In
OCbservel
If
5 A
 Souiety's
atan Nerul
Tikak Eda end Commerce,
ofSciene Soh1ton
S. KColle ge Ohacstn Bank
TYRBLSem S coureate
Subject: RI
ratrer pot br
of arbestng chata scaer
reatiors hin scatter gaph and
Wetancor the caed as
scaterpbt is ako
Cocti Nnt a
dugram
Types ofr Comelation
Types of Corrolation

Strong potcorreteion
r f t posibve corrolston

Songngave cornetatn
Aorkt neJate romelation

iom the
Pe:fect correlation
data ponts are alling on a sraight b gog up
1. Perfect rositiye: aes of x, y igh
low val1es ofx, y 2 lighpoints fom a lne sSuch tlat Lne Gocs fom a
data
2. Perfe ct Negativc: value on x axs then Varabks 2r
perfect
vaue o y-axis dwn to high
negutive
Strcng coneiaioi goirg out fom bw
patern among ponts aagan
1. Strong rositive: tie vf xy upwad tend bt they arr t iug oi
vahes ofx, y o iigh vaes
strong corekton
strong positive
strigint ine hen the variables have po|ts is agan gcng fom hgh va
patteE among
2. Strong Negative the vake un X-axs then the varables have strong
on y-axis town tohigh
negative correation
Weak correlation vaues
Positive:the pattern arong points is ag1n gong p from bw
| Weak
vaBes ofx, y mt allng on the straght ne bi e poris are
ofx, y to high to show strong rebtion thea the rektion is
neiier spread out or cose enogh
weak positive reation
Negative: the pattern aTOng pois s again gong fom ahigh vahe
2. Weak on x-axis not aiing on siraight ke buu the
on y-axis down to high vae
arerether fo spread out not too tight to show strong rbtoa
ports withou obwug auy pattern tiere s
standard
No coiTelation: the puints are jst to
no ebtion between he variables
 Tilak Education Society's
S.KColTYBBl
le ge Sem
ofScience and Comme rce, Nerul
5 Question Bank Sohition
Subje ct: RM
Formula
Kart Pearson's C
X= ov( )
6x.6y
cov (x,y) =xy
Correlation
V(x)= .- n
(3)²

SpearAn's co-clicieat of correiation

R= 1.
6(2a+2(m-m)
nn
Conreiaion -Application
Ans Meits of corelation
It is easy to co¿fficient:
t not oniy understand and easy to calculate
gves an idea about the cUraration of ie two series but aso
dicates
Demerits direction nf reationship
2 dsses a inear reiationshin between the variables even thougn t may I9
be thcre
!s tablk te be misintepreted as hich derse of coreation does not
DEcessariy mean very cse relatastip
Whei -0 ii sinply iples that there is ro 'iear rektionship beyeen two
vari|bes but it fais to give the jdea of rype of non-inear relationship

Use of correlation cceiticient relatiorship

Correaticn are isefi becase they indicate a predictive
"
precsely te degree of reationship between the variables
Itisused to derrve advertising expenditure Msaes
R demand and pre concept ofregression
Itis used in
deveoping ihe
reducng the range cfcertanty ecooric behavior among
n
It is used
feld of Economics touderstand
TH R sef in business of sales volme of sales etc.
in correation cost
ttevariables understard with Characieristics andpplication dependent
easyto meiri
e Its
Analvsis Meaning and
üs Charge n vaablks bacedon
the
Regression techniguc forestiratng more
mdependeat
varabesis k1nwn as
Satistical. or nore
3
the changen one bctween twoor
Ans yarabk to assuptors
mathematice!
regression
 OC tey S
S.KColle ge of Science and Connmerce, Nerul
TYBB|Sem 5 Question Bank Soution
Subje ct: RM
Regression Tethods are useful in developg quantitative
between variabes which can be used to predict the relaticarelationshis
Regression powerful and flexib le tool which s used to
is a
present or futre events
On thc basis of past or present forecast the past
events example rate depends
blood pressure depexis on age so length and age are independent on
Independent variables are aso known as predictor
lengh or
also known as
response
dependent varjablesvariables
Lne smplest finct io na l form is-the
Inear fnction where:

a and Bare parameters, remaining constant as x and y change. a is the

intercept and B is the skepe.
Ii tlie vaBues of aand B are known.
you çan fnd they that goes with any L.
Pitting the r into the eq1ation and soving.
iic il:1 ber o ulentil
ndepende:it IQole. iS riited aki tlhe irdel is
rcferred ic as multpl: regrcssikn
if it involves several
Someiines, a undel uses the square, squ:re-rout mdepenient varialbles
or ioie udependent varabies io or any olher power of ore
a predit üc ependet one, wlicin n:kes in
non- Inear regression
4
PegressiTn Analysis Application
Predicting the Future
The mnost common use of regression in
busin:eSs is to predct events tlhat hnve
yei to occur. Demaid araysis, for cxainpk, predicts hew nany unt's
cTrsurrers vil puchase. Viany oher key parameters olher
than demand are
dependent variables regression models

Insuraie corpanies heavily rely on regression analysis to estirate how

any polcy hoiders wil be irvolved in acridents or bc vtims of burgiaries
Trend Line Analysis
Linear regression is used n the creaticn of rend ines, which uses past data
to predict future perforance or "trerds."
Usualy, rend ines are sSeu in busness to show the ovemcnt cf fnancial
or product attributes over time.
Stock prices, ci prices, or product speciications-can al be aralyzed usng
trend Ines.
Risk Åralysis for Iryesiments
The capial asset pricing model was devebped using caa rgression
analysis, and a cOmon ocasure of the voktility ofa stock or invest1nent s
its beta--which is determined using irear
Salesor Market Forecasts
regession
 drinks anaBysis
iactor
simpiify see among cok
to coktaccording
o
ivolving
SCOre
LIse
th
easy
to
movement
vohmes, variables and
through maximu variables
ard
variables
technique
reduce
to
s
meüiod
eigen'alie and interesting
not
differentof
reiationship
characterist large
ics computations
of that condeise but
only
this forceexisting
and
training
demographics nunber of
variables
can number
compoTcit
time
writeon easier
a oher so
is
anatysis
where work all
irom uading and deternineto and
ncar-reression
market several
another, the
the
for
pans olarge
e
varables thc
the picipal
to al!
wen! inpetant
there askedSpriesalent laboriosandfaster
trend
the of reduce
varicties factor
betweca we and are Miranda,
some
facility
Nerul
sales
or
as when
1eSpcct
to
predicteeds
for
large
companes. ah matrix rethcd wher
were
factors people
pretrences e ve
irvolves ly
rel
comput a ti
eexpensi
vr
Soution variabBes)
analyss,
change the
,Comiror.
of to
relationshin
gy
out that -maybe
emorade-Coca-Col a ,
rce, to 1scd ndex used poiting
s qbecome
wed
Comme
Society's
Tiak
Education TCwth.
trend also for
nsed tics actors tominolo
corelation szalysis data iexaifpefactoränalytsiechni
atent variabks ue
observed concerning
several
Bank with prepare Characteris is scztroid
RM
Subject:
thvoforecasting
will
Resnurces plans
is
that
an
anaiysi:in rektive with quite
heptui iveal
than
variable
changes
variable uf as
andQuestion in
than for are
to hiring
companies nuber GCore
techn:gue inpotartare: actor
tactor for brden
mutivaraie
has be
nce
Scie
5
more
for
pans
accrate
one
Hunan
forces
methods
good
and common
extTcts
structure
aralysis
r
maximum
kelihood
method
aralysis
Gf actor
s
aioreand stuy the
among
preference'
underine a anal toybom
sis
dauatechnique
Sem iata
multivariate
of comprchersive
method
Mulivariate
fhavng more how work Meaningsa smaii for
further
analysis analyss
thcvariabes limitations
of echnique
relationship research
cost
heavyFactor
pe TYBBI regreession of factor
S.KColle one at This
helps
the analysis
ntu uc mto detcct factor Technique se still
thatLimitalions
is looksin future techq the
howare
method
sophisticated Regression
modificd. developnent
employees. hem fector drnks This
atmethod of
Analysisvariables
Factor
Mainiytocassifyrotation cf ard The
fiomThe are
createooks Lnear
types Thisput id In Viethos
A
This Meaing Meils
Merits 3. 4. 1. 2.
ar Fartti 1. 2.
Line
Ans
 Tikk Educ:tion.Society's
D.College ofScic nce and Commerce, Nerul
TYBBISem 5Question Bank Sohition
Subje ct: RM
3
The resut of a singk bctor analys is is considered generally ess reliable and
4. depcndable
very oftenfactor anavsis stats with set of in perfect data to overcoine
TOr
tS aTficulty has becen realized that anatysis stoull at east we done twice if
We get more or less similar resuts from all rounds of analysis Or confiderce
considering such results increases
S.
Factor anatysis is a complicated decision tool that can be used onty when
on handling this tool
Gnas thorough knowkdge and enough experiencc
6
Ans
Factor Analys is - Applicaticn
Applications of factor analysis psychological test such
ractor analysis has been mainly used in developing
as JQ test, personality test
everyhing fro1n muti
YOwadays companies can make better choices from statictical planing data is a
Channel Marketing to inventory control through
powerfül tool and actor analysis used ii to get resiks
stafhng
Factor anavsis is sefui m budgcting advertising and even and aialysis n
marketing rescarch
It is apraciical tool created for succUssful
any industry
Physical Sciences, hydrOchemistry
actor aiiatlysis has been viiely lised n logical Sciences such as Ecoiogy
astrophysits and cosmoogy as well as bio he appiratons are
MOlccuar bioog and biochcrmistry some cf
on data analysis to make vita:
Investme nf vesting in field that relies
portfolio is tu diversify stocks
choves the safest approach to creating analys is to anticpate Movenknt in
professiomls use factor
nvestme nt
cues thet woud otberwise go unncticed
variety of nd:usries t unay provide starting of a compary staff foue
Resourrcsmany ctors influence the determine if a company can be:e Git
together to
cOinbine different vaables hoise ialent testing akus proper
contraciors ard more ii
from fewer analysis
screening of empoyees using iactor
acute arcal tables ard
Statistics to create
rely on make
Ins urance companies hc Unly way isuranco copanies can
polcics studyä1g varmbles is
regarding deductibles rates and avaiable pans ascount when
decisions and target customers nlo commnily
Restaurants take deographics
competitors to ethicity breakdown of a
creating cnus from cost efective pannirg ad a successfui
for
Data Collectivn albws
restaurant

Charncteistics
Analysis Meaning and cassifying variabks in two chuster
7 Clus te r consists vf methods of
Chster analys is each group are having sirilar
Ans members of
for gropssuch that
characteristics
 Tiak Education Society's
S.KColle ge of
TYBBI Ssience and
SemS Ouestion Comme
Bank rce, Nerul
Subject: RM Sohuton

The basic
and
exhaust obj
ive ectv e of
entitics realy existgroupsin theclareisterchustanalvsis is to
eIs based ondetcrmine howof many maually
Cluste ring popuht i on smiarities proiles among6
eHierayrueschicaltechichhtneirqinuesg techstchnieqriuesng techniques
teHichni and
can be chssified into

tinat rarchical
Non-hierarchical chustering
custering techriques as the rame suggest in is an aigorithm
Thisbulds hierarcy oî chsters a
aleorithm stats with
OWn then tu
nearcst
all the
data ponts 33SI0ied to a chister of the
algorihn chusters are most
terminates vten there is ontyntosinge
saine chster in the end this
Two mporant chuster le
the algorihm hasthings that you should know about hierarchical chstering are
been
a'so possible to folow im pkment ed above
top down approach sng bottom up approach it is
staring wik. al Gata
assigned n be sams chster and
po s assigred a separate recusively perfoming spis ilpoms cacii at4
The deCiSion of mergng two chaster
chsters chusters is taken on the basis of causes ot
these
In noniierarchical chstering
technique and
permited to jon the other chster this meareobject
of one custer wil be
ateral
one chuster to another hieiarchicai chsrering levhnneshifting
Sumeotcommonlv
bject tYomused
hierarchical chstering tecihniques are sipe nkage chustering method
conylete irag chstering method average hkage chstering mehod
centroid method
A dendrogran is a type of tea tree diagram stowing hierarchica! studying
reaticnslips between silar sets of data n other words a dendrog ann isa
graphical deice ior dispaying pis two resuts
Cluster Anaiysis -Ånplication
As Applications of chstcr anaiysis along with, other analysis suchas discrininant
Chuster anaysis ofen used
analysis the chisier aralysis based on their
interpret to
One ust beable are actually meaningful
understarding ofdata
resuls 2nalysis produced byaralysis folloys
anses are as
Todeternine f th cluster
the.quesEons that Dagnostic chsters
Someof the
Medcine:whatare
1.
 nmatix,
Applmderyi
any
disimikrdersj
Ynng orpersuasiareovarn, iables)
icatorismiarities
ty ors.
sezments marketing
niedium, store
cfcharacteri_tics objects. more
heavy,
politikcal(e.g, simiarity
and meanngful
of
he probkms between
a observed obj e ctsor
even
be
of
does biomedca! of markei
perceptions tics vest ig at e
coordid a te s,
cansensitive
characternstics
who
direct
i or
custoner
andto
sove
their Distingujsn ris detectexplain
Charactc between of ortrce,scaling
or psychology
independent
varible howhabits knd two,
any matrices.nsional,
een stmulisiikrities
weight non-price
attrbMes physical
and between nct atix.
respond to toresearcter
the
of a
thea
notresearch-identify
marketing-identify
the to betwmap
faces, corelanalatyiozenone,
Hunanities terms
+
bk
Xk and
understard
lifestylcs do favorabie consumptionwill and s
enalysis
whoMeaning
of and
rce,Nerul

or
fekù
psychological
coeficient sensitive
he

oyaty
the
n
exampes
to store
Distinguish
and
ir
-research
product
who
thosethethe cobrs, the can
conceptuá!
(distances) cor rofeconsi
latiosnt :idime
mul
n Artstaxonoy and those
exhibit will
of allow
Sohiton used
+ fs price possessing
artlysis ,
Applications
their
Tilak Comme b3X3
Educaion
Socety's be anthropological
in
Perception/imagewho -Scaling
goal tte you
media and thatdissimilaorrities MDS,
Bank anaysis
discriminant theof habits be miy
factorinexpressedtoaddimaptonof
can Sciences research
-
Mul thetiddimenspns
iniensio nal
Subject:
RM
a
whoand
Application
Question
and
+
52X2
predictorbetween
terms
it
custonkr
dscriminantbusimess,
and of
Advertising
company
Consumers
Cainpaign
general
can real
Objects
nce
Sce
in
r0ycty users
phcemeni
Social
consumption
CUstomers of With lne
of
5
Sem
-
ie groceries
differentia
a store
in
light
in Direct
differ kind In
geTYBBI +
Analysis of
characteristics
of appied
n Product
S.KColle blXI where selcction In
of
nekd
have Sciences
attrihutes
to buyers
and or
criminant the is
Areas
used nutIt
+
Db0 In "
where
b'sX's Dis
11 Ans
10 Ans