DrsuoA Or EpreIMENS

prorosoxsikd bat
40 duoue on unhroun
10At
estoWLhio
n wed 10 esolat which
roaA inart haue
PMoCALA oulpitond utot
gyeont Jnpot on
nk
inpuu chold hove
con lbe daq celuet this
ingoumation. n
boposdmengos dugo
10 eduer duuqn C¡t
seding duig
pousofus 16o4 to ocheise
By miniri zing psoCLu
imanufetaing
vasiati on and
Cost

Joad tox ipetin. saduing

îngenevor duie espuinunt epoumentat
duigo any ingonation guhening eocila
whano vaxiation pauents whhe unde he Conbol

Howeven theye Statisiy e

tox conthl the epeiment în he deign
expeximenti, epeximat u ofan inchd in e
the tk0 at menti O SOme

epsimental uniti which be peopu, Paxta o Propl,

qoups % Peopte, animals, ratsia eto.
 Expeinntol cign pMOCLu ;
Dejne Bright
objectNes Sto%A
Puoblum(s) (Hgpotheiu)

Tcoduct
întupuat Analize erpeximent
data and collet
Pxedtted dato
Resete

baic psine eapeimentol augienstojk

book the dug
hishinnovotve
Psoposeo R.A Faheyin bosic
to
Ashe the
espimenti
prio Psinciplus erperita

Ss Rardo mization
3 doro Corto tos) Blocking
Raplication an expoimenta! duign
mote than
shoulo
ssçlication. The epeimant
t300tmnt applied in many
Thus e0eh
uniti
erpeament in inexeaun
satisical accuaney 6
SOme
vaiabon intoued
 expeimental tunit Such
as indivi du au o plo plant in 0qi tulhwal
Idonticol. his epeimer
moved Jy wng a utbor of
witt we heugou
epoumuntal
pefon tha
than oNCe Thot i supeathe baic expoinarnt
Aduantagus rgliolly

Ruplicaon 3eatmant'lutt in moe alialsle eirole

than t0 possi bu oith obseswoilon.
and hea by
to incQaue pecisi on tohi ch

0 obtain mOe psuci ous etirat

heet of the uotmentouoAo oosugoc
nember eplicatios will incAQLe
The ,adeoo
ni3aolo vaxiowu thatmt erpoinent
taatt
matnial.

pibotA the pactal

3ituation we ue thar Hupli cati o in Reneal.
ay Randoni zati ont

The Serond pneiple podnanta

gandon Zation
which a sandbm prcs .augrg
43eatmatt epoimenta nit the andon
 (4)

inpie thot euy Possi ble altment theatrant taa

his andomi 2ati on
Some sobalbii t
most

mthod Ond mechanical 9andom zabon.

Randonitation' psoviuo Jogial bas, om the valch

10t sqngance.
St edueu heepovimetal o.e0r09,iosiy
i Randomization liniatu bios in
not
out
is Randomi zati on

obseweo that al latio Sowe

buing y
and
KunOved
vQNati on axa nOt, choose a
tepicotion OAewosds
mam al neplained
dagn plot undos contol.
Vauaton ( haree caue loca ontol.
rupth ue
we
Fox this ppoSe
A s0otmant sugoing
ha éxpeimentar unta.
Biocking and g'ouping loca! Conbeo
Acoxding
divids the' into seveNal, bonopeouA
blockA ano thes each Such blck
Pots we Krouwn as

is dividad intig payt eQual to thainmlor f

tha featmant ake, andomle

akigned paxto a Sock,. This caled

 blocking. Sn enena blocka aO he dwot at cohich we
hold 80 hat ve
Cas masue it Conixti on 10 to vaiabli y
kdata means oh a way aralyui "of vaiant
A0 Th concapt
in sondond zea
cdively
blockdign

lockinBlocking
g' ean hot he expoumuta! unit Bhould
be Colleohd togeth to foxm elatvely honoeneous
goup A boek a alio supcatu.

alanting meanhat the t%eatnint 1ghoo le

espeimunt al unita

+atmeNt Bhoul

-) hoNo

thomagenous, plos bloe sloude

inleqyual
 Analis Vaians (AND VA)
Tntodstiont
Aralyis % vaxiance
betwean
mans Qnathu ausocativel Psoteduwes Sueh as
s Vaxiati on
Lamong andiketween the
-p.A. Fisho.
The t- test bOued an he siandad e
delexentendbetoee e wed to
mean. cohe hoe aso
teut dithantes oneo ean.
han Means,
samle
to compore eoeh othet mean

t+ tut s HorwevtCondluctingmultide it text

y Jead to Suswive onhuation o
oote ANOVA trcan teatin dibfexeneesio amorg
Sewetal meas tpx signibi caneeo wihout rne seeaing
the

this teehrique
1990
4estng ths homogenitiy
want to et the tet
Populotion. Fo we
Cnd out

whethar the et theue on thems igrigicanty

the ANOUA, he tot0l vaxíatin
dioent he vaiation
due to obsewations le splited into
 t6 the e betweA Qoups (sul cauu ) o 00tmant
Vasiation d to auignatle cQu!. md vasiaton ue to
dus
to chance Caue

Auumpionw Nedthailen s
fox tha ANbVA thnigjus on e apieo only und
the touoinq aunplou. itosdo
iray The mle to de conido o ANOVA echniqu

aSuboaues oh modats tosx ohihssTaria

t0.
identib
ajection
Buen
mode i non - inuQ, it
which to ANOUA
X to
ag).
Sdepenoen
bthe.
The obseywobors ndiperdent to QQ¢h

NeKratitg Tha disti buti onu he asiouas are nokmal.

homoqerity, the vaniante Jups should

tha Coutant (eQuat vQwante).

One fatoy tox)
tosN ) Srnga facto {ANDVA
ANoUA 9echniquo uedto

mose Samlu. Suppese

 set N obseyotion
dsti betd k groups oboeMUQiou Coxupording
o sirgl acto
taciok (
( iç he
to veli
loveu Srg
ay claiked dalo.
tyatmenta ) hu oem
the clauificalin.
hue subgos
ontain eitther eqo
tawe
nuner obSeWotionu. Thene
doto wih equas nunter o ohseoti

doto oithnqas umer

caupied
coME the Sub anes.J i d o
obsevwaiOnA

that
a andom voiQle

bais
otpwesecd.
dntg
n013OSizes
SOme
gouped tHeatment
Apely
to he and
ondg
AUoVA tox claui iaten
the

obsewoli ns
tseQt
Qe in
9
celun
dixeeton

Total
..
 at l she yeld the im nit teceiving som

Mathematico! Hod
ha dcsibe, the selation
he mayhe mati col modu
-Ship bw the HIponye and 49eatment to one factoa

ohke xapkstntah obsewation on the 44e0tmont.

ect (ovesall maan fet).

the ih tyeatment, eheet ard nidi 0.

ent in he ; obsesua
teatment

Asupionu
1> AU obseswaæodi ae indupendent
Naious 9u0tmenta ano envionmint al ase

adaiive
natue.

dissi bund
to de

oi th
noknallg
mean
Qnd
indeponduny
idarticaliy
and var(ei)¥oli
Nuw and altanat ve , hy Potheu
ae:

Ho3 theei significant dufoene betweon the

d3420tmunt et.
 kae
ganat m, Ho in (atleost oe

he modelieo
Sotgwcal Analyss o

which ae to se
and o; e he paamitvu
Hee
the mihod poinepe of laaut.
Crimattd
to mini mise the
h o4 Sum

he eiidat oM

condiion,
miimQ nd
Acoxding to

-H-di) (-1):0.

mod
Noo the esimalio
 i : . +. (74i.)+ ( - )
The tota vOwauon tho

Sk

ROt
k n

T55 = 455 + SS
ESS

Singe faet ox
Soue Sum Su F-9aioF-CiHCO!.

Squaves
voiation fedon
OQtment
k-|

ESS

'Tss

Buedom
Ingoenea C9utical Vaue we acept
f- calculated valuu oiF
hypotheis ohuusise we ejeet hupothus
 P vale

1 wet sgnticant (p0.o1)

6). level of sqnificant (p4o-05) oo
Ho).
to) Not ) 3gnifi ant (accept
data
Sirglu obseug.

Ron cat! NOUA detn

ANOVA aluo olld faco ANoUA
wo factosu
hoo a eespoe,
modab aUous tho expenimanto to teut tho
objeti ves
diet and factorigl
ANOUASo that the
ANOUA inpoues
9-perimentor the
vOiae
Sngle depndnt
Same anadgis.
measue
fos exanpo we
might nd eOmes
an boh
thsee dutgexent espuimnt
which
(Sex), dt, w the vaialle ' !
ae faetosa ngleneig
k'tseOtmenu blocxs
Suppose thesce

ven he Volues (i:, 2, -.. k, i L2,... b). his

toscns toQy cai fied. data.
ANOVA oVA fo*
o uay clauifed
oata ii asolou3.
 koygut ANOUA wo- way laui fieo data:
Thaatmentu Totat
bloct
Haan

oi.

13atmd

Hean 2

Mathemaicat Hodetr

he 'lina mathematieot modl o clauf

-cati on

d; i i tseeatmant efect ( i 0 )
paibt ton
h bloc
Bfeet due to charee caue.

indupendentvaviau teatmut
Ay ho obsewai ons
acive inatu
Qnd envionmNa!
Seting g Pypothuis >
N hypotheua > blw the
oignjcantdetfeanee
miQn
4eotment
thu
no
oggiant difonence
Hog theve
man fecta e.,
block

Aleunatie ypoheis
thee
djguenca,
HT
+9e atment mean

(atleout mean defor ).

bluw the

cihfer).
Catuat 0n, meani

Statisical Aratgss
shich
the pasHamtou
the
pinciple bat squavy
essinaho. ' Aepsudin to

minimun.
the

The lane0x mode! i

na and
No.h. minima and; marina Condition
Acokding
10.
 ton

Avo
NOW onuiden 1otal

Cloc Sumi.
Sum of SAuares
ss) (E35)
E SS = 135 -
 ANOVA Talte fo -cay cafed dota

E- 9atio
5ouhte
VOviOnon d (H$32
,55
3untmnt
(b
b-!
b
(k-)(b-i) ESS (b-1]
(b-)(-)
ota kb-I 135

Snyeene nul hypotheus other

we
hyeothuinI
Co- Va°aNe (ANCOUA)
Seryisive and poesgad
AvcovA lechnique dat
ANOVA technique . chen the
eehiqu hantthe
u avalale on a suplmento y alorg
Voiale X. ANcoVA is a
wih the data non
conbel
techniqu in whi ch u possi bu
0dciioral
stain SoweLs) Vaiations y
each ohepeimental
tang wita
obsewa on
an elpeinent that
hat
Supposey
)
(. main Vavialte
to k.
vavale which u inearty aeated
y cannOt
Corbo the erpeimant
HoreovoA
obsewed wiyh the '.
t can wed along
 t the 'aiatey i alid he coUaviati (os J Con Comitent
VQsuale.

seuponue vOxiaula os he
voviabuy. A7his a ot done then
indipendent
he HHO )adid be inuatd due t0

the 'na eleet . hu, makiig iit dhicut to

dettct he doenca uporue voialti

AneoJA psocodwe iu tha Comtinati on the

ArvouA and the deqewion anasis ohenevon, it is

Possibu o takeaddisional obsewati on On

moe the exprimentaa

idbugn along iththeeespone

vOiallo
AncouA Jy identibg
cONCOmi1ent vasialle y' Suppose we oant

8> Dsae mode Teaching

q
ieaching the coue of
who take the Sama finat ehaminaton.
he vaiati on in hai maka
may Sinpy
 t0 h dfeene meshoda %teathing Jyet ao
t0 th int1ligent QuoMentOa)tvela
the dißeMenee
Hence
Le odiuuted tho obseura
mthodu O% teaehing0 SHudent
data (x) o
the
SHudenu ponue
uaialt.

Concomi tent Vaniobe >

Concomi tent vaialble y (o%) supplmantany

Cwith the main
highyncoeaed

Vaiale he
Sub cauei ia Callednol concomient
wed methodsaiA to udue
fovouig maindg VQriatiON the
Conbio thetn

Stsatijication (ok) gMouping % heboqeneous

a matuial ito
(oM
eposimenta!
blecks homogenoui.
3> Rafinment of 8tauSiCal techniqe.
6 Concomient Vaxialee beides
yauale.
the dato on main
 t Among thue
expeimantal esO9N ohen he data on CoDCOm tent
Qvaila suides the doto ain

ANoVA oduantageu
70chniQue.a
ANcOUA techri

Pucautio in the We ConCorn tent Vosialte +

Concoi ent vOMi o.l zhoud se
uih he Nain
highuy
the conCoitent
Bhould note elect1d
Sub - caue of clauication.
weCoUt the dato on COn COmi tent

the en o he expeximent then we mut celeot

data on he etbect of sub - caues he
(.A:ar)
Uaificaion.
beoeen ANcovA ano ANOVA

ANCoVA m

and
uing ANOVA the

ugueion onalgis.
ttouneol om ANcovA u
Smalle than when
cOm paxed wih
Sum oh 3quae obtainudANCoUA.
fs0m ADVA.
8y AncoVA 1ehiQue i 3 when compaed with ANCOUA
Seiive rechinqw han ANOUA u' not setitve 1echnique
AJOVA .
 The layout the duign ANOVA he local

can. se cfined (o),thanged cortuol poineipt i ex cuted

io the at dulan staqe
eng change.

Sub clauel

|(3n) (n bin)... (hg) .. (" se (,3r)

b ( 5 ) (b b) .. ("bi.3bi)- (e (Iyb. )

obswat ons o% main

Hee i's a
vaate :
nand 's i oesnobswati onsa% main.
Mathematical Hodel;
clawi fied data
o one icay
"The dinea suq8euion
(o) CRD)

main vaxiale in

8ij obseswation, on con(omitent

h e i vasuae
 Seb-caueu.
whee
man eet.
atment ehheet

bie
ueOM 10 due 10 Chanta caLue.

Assunpionr
ty he moda! fo ANcovA 4eehinque iu a nean mode.
D> the obsewatiosxeindependent to egeh
otho.
3 du the obseswati ons i3
unknoon
hauing the 8Qna
CoNAtant vaiale
s Yhe popiatiorn :or ohich tha samply deaos in
nOknal Poplation:)
6 the ConcOmitent vaialle y iu Co9Uelatd
Vaialte. nsstoe
6 The obsewaiodlon CoACOMiHent vaiae is
rr

ot, included inguened

the caification.
Sating of hypotu rrir
bMo:'Aí hi txàatmany have the 8Qme mean fheet and

Igrcant.
AHeudlive hypohe))
Atleot on mean dßor on othe.
 baued 354349
baed on

(2.)
bk
40Qtmenty
ta ohne b blocku; k
Aent

b b

due todo

itoado

baeo

-k b

due SQuaes
ie,

b 7geatnent Semm

b be

Tseatmnt Sun of squau due to ;'3 and

B

TotaJ Sum ob sQuaKes due

(X.. S(4..)
bk
b baued 0)
SQuanu

br

Sun o souaU dua to s Y3

 hypethaus.

f-Yatio

ment mean
to
dua sijeet
Adjuut
ESS N-k
k
AdjwtmantAdimerd
cogans
N-K
otheoise
hypotheis
nu!
k-1

to
due
Qeui
Sun

simationto9iicocient
j ESSyy
ESSy Tyy

duSQONU

the

Sum
of SQuaNe
Tyy

We
ESS47
du
10
|Sum f4a
Fcau
%

degesAeder

Injexena
VOiatienTontmet
Souxce Tota!
AcoUAom 00 - toay tlauiped data tox) AaroUA o ROD
Sing oksUwati on pen celr

louput to ACOUA
eoun

hota
("3)
(3) (9)(,1i). (s Yor(13.3,.)

(33)

b (7b,tb)
(b,b)-..(oi,yb) (% ybx)
Sotal (2.1y.1) -.. (,9.) -- (1..y.+)

MOthemoUCa! Modol

Aee ij the obseswotion, i n h blok ;th

tHeatment On
Voiale
tho obsesuwati on ioth bloCk eceiving
Con Comi tet vauae

beer