the firm.

that exceptexample,
on
sampling
deliberateNon-probability
is Sampling
of of Sampling, depends
(e) Inquiry. isa
2. sample,calledcalled sample.
called
collected
a methods data study.oncerning
income
Introduction1.
roll the the Parameter A intormation
owWner, \f it llberate In There One
he investigator. Methods a A and In
of
he is Ashis population
is if non-probability The calculate statistical
parameter. from statistical
ofthe
8CHAPTER
highly
is (2) upon anthe of
heevident, own will he
biased statistic.
methods are is collecting the
select,wants Random collected
and members aggregate Indian.
wil
subjective. will. the
various problem. aims
choose Sampling
in g We information
of from measure A language To
towardsalthough
say tothe Sampling Since On Statistic of
investigate highly sampling shall Sampling methods ar e it the from data ofobtainstatistical
100 sample a of For
grouped statistic other such a the
highly the Let
the also study in (1)
subjective general,
groupevery objects collectingexample, such
us students available as Census
workers, or of hand
paid mothodexpenditure that called only an
Suppose and mean, member values, study
Sample
Large
Tests
itemProbability intoselecting collected
he of we
workers is for thinks in the is two from an a information
study of
Judgement about statisticalstandard Method may
he very
th e nature included
following universe of
wil that classes. th e from the
information problem a
a is
choose
to simple study
pattern are Sampling. sample the statistic called be
find an because population and
the interested
investigator the in population, measuredeviation for
at method. the (1 ) is is
out low and Sampling from large, universe (2 ) the
statistical is to
his of best sample Non-random we collected
the students the Sampling be
will. estimate obtained
calculated
population
and
paid asyrepresentatives
investigator the we toable
average to
is There on nature
population. by in analysis know
workers apply, or use sampling
some from to
studying of the
a th e predict
salary. is Purposive sampling from from Method. or the the
it othercollege
no
basis Sampling of
and suffers parameter. technique. universe. average
is
chooses
members
those data the group
if
wages of Choice the method called some
he of values In
drawback having the personal orand method, whole
is fromthe rule Sampling census characteristics oflife
towards
baised of population. Non-probability of This
information There
collection the
for 800
the of or
workers judgement the universe objects of
object obtain a group method are average
selection
students method sample
of The : two of
For of is is is
a a
 Care
(b) Applied
and =o,(X)i=Central limit and Central form.
Levy purposes. th egivenlarge. mixed to thoroughly
member paper The methods and In study
central 3. proof
random numbers sample. which
be and
variance =u theorem variance numbers numbers sufficiently
() Simple Probability
onlySamplingIn However,
"If then n"If Central Central can A numbers. An (ii) The and used
Suppose of Mathematics
and X4, Limit particular A Limit limit manner. number of th e Lottery are every probability skill
X, be are alternativeTable method 100 the in one
Var X2. is
under X2, theorem given,
the same
a commonly or Random by
oas nTheorem o Theorem limit commonly appearing ite m method.
(X) .... known ., Limit Then of namesofpopulation we drum the large anthis
= One of
form certain theorem statisticians random isthem want Method:The
n
Xh of Xn in but used and size,names
in has samplingexperiencedmethod is -
Theorem question methoda of IV
tends are as of an are two th e onbook are Sampling:Random used size an
(Lindberg-Levy (Liapounoff's
general used. all to thshape e
independently Lindeberg-Levy the independent forms, is experience th e for select has
1,2,
to n numbers
of is
drawn the
required of equal selection
above tends a The page drawing for then useful if
very can have random the 800 equal a and method all
infinity.....n conditions random it statistician,
one table
be from sample the opportunity
is
theorem to important constructed ar e method boys colour.
then infinity. Form) known legitimately
has The : the the opportunity numbermembers is likely of immediate
Theorem) random contains read.
numbers an
and theorem. shown prizes drum are of selection.
simple
S,= S,= of
lottery The to
sampling
(meaning is as The written 100 of of item reliable
identically theorem random represent (8-2)
X of variates X Liapounoff's tables is of
blindly. boys pieces
of pieces ofand being in
interest thatasked. 10,400 items the results
+ t consulted. methodlottery a on being the
X X we from therefore is
in bearing numbers. separate population selected asample results
.... + n wit th of are are population
scientific
distributed to is.... Statistical cannumbersAre random selecteddrawn are
large.) isand a folded can
+ us. + E Form relythese The tedious class prorp
X, X, (X ) these hence
popular in is required.
is The In are the and done be
is = and upon numbers all book this
pieces ofblindly. in in more
random a following variate a h; analysis. numbers 800 the written
the obtained.
normal normal and in numbers tothe for sample.
objective with If
other the four isfollow
method ofsample. Allsame this
opened name. paper, boys. accurately.
taking
Var tables this on a
yariates
variate known We really digits but certain is
form (X) are if If manner,separate If applied
Tippet's all the ensures a samplemethod Sample
LargeTp
give forarranged included at theylottery random
the of as random the The rule.
withsuch with of, = all randompopulation are
Linobebelow practic items method of with
mean = rando folde
that mixedpieces samp followi isselec Weok
rand
in
 means. can Curvepopulation mean. We sample from means we mean
population
and population
Relied and O to a i.e. of 4. deviation¡ varianceoie. of Applied
tne may the
separate samples do
standard Thus, distriDution Nowsample Suppose the not as If If In then
Now, ue bewil Bypopulation wil
Sampling we X
other
Westandard trproved
Wi n constructing look get mean. population. the X
look for know are is
CorolMatlaryhematics
a
mean si(1000)
ze deviate then e the
with meanu population very Most each estimate. thwords,
deviation at standard n of that value we :
mean the mean X, (100) whatrequired Z= mean X mean
the
is the
much of and E(X) = Y- From
deviation the of these The take from is we -
whole considerably a these thus should (ii)
Distribution Althoughour X of IV
whereu although the to normally A1 of the
sizeerrorsample histogram
mean. mean like difference the get n
olWn. and getting 1000 the 1000 error
estimate t the
thing of a population js = the above
sample. th e
denoted mean?2 is normal of
sample numberof besamples a sample 2sample
standard of This samples we S.N.V.
the the such from occasionally 1000 carefully wil Var()-(no)
distributed following and u t..... theorem,
again. between be.know a of
mean above means of +
0.e., by mean: It distribution curve, a population
Means X of
can the values we items each mean. We as si z e
deviation If distribution that size
X d of sample or noted.
can (i) -, n withimportant n
is a wil inthe of need the we
is bethe We ' t hin' below by of drawn n,
the equalshown saw be
calculate a size (o mean taken get (8-3)
is chance, sample number therefore samplemean, 1
o v() in = population. called means, one close X, a
then mean to above i.e . the the 100 known) from result very
u from
that th e but mean we and
ol/n the the as important
X of that mean all we some can
isthe the standard a a
where But sampling the can Sampling wi l take population population
normallysample standard the corollary
what same
of way be
o sample the
obtain a corollary
is is close
to
sample
aDistribution deviation with of
of standard the distribution sample know having
stributed deviation mean normal a Fig8..1
parent to Central asSample
standard frequency mean Large
Tests
disizendrawn the and
population how follows.
the
X means curve. population oln Liumit
deviation of of of use
with is much and mean
the
deviation th e curve. the th e ie .
distributed Moreover,is standard Theorem
mean from sampie u
of sampleequal means sample sample and
the the The mean
u of
 deviation
one. Applied
and
that that0-27% underbetween Weis Critical
Region5.
fiducial be
corresponding and fluctuations
because Z take a
(1) 6. limits. probabilities X, (0-05). that
fact
The If
problemsbased such < Zthe
99-73% knowS.N.V. Mathematics
are from -1:96, Now, sample a
Suppose,
Similarly, means
This
th e wil beyond the we
To The The put,
Procedure
A :. :. S.N.V. curve -1-96 that is
exceed
hypothesis on set that
limits.Thus, expressed u the between highly
limits is alone. a 95 P(Z<-1-06
P(|Z|> Z=
referredprevious probability
0-05 relative Zwill ± and for
up z significant rare out of we
3.lies
lies ThisP(|zwithin regions size P(|Z|> 1-96 a
useful -
a or Similarly, event know But
exceed beyond +1-96, +3S.N.V. IV
tohypothesis between|>1-96) deviation of n probability 1-96)
experience is
a of means in0-01 from numerically the and-3. to Z=
which
percentages 100
above are which that of has the 2-58) or 99% 95% us then
statement Testing such this Z> these areas1-96, in
called it
values population =
is -1-96 we =we if of taken 0-05 In Zisa
0-05 decidecannot an zË population. =that 1-96) values are between other drawing area
a are expect
X 0-01 is 1%
statement aor and the critical event 2:58 > place of 0-05.
supposed
mayHypothesis
Aconfident
as and Z = also under standard
the be from mean wil 0-05 beyond words, statistical
1-96.bounding z 5% due i.e. -2-58
be significance
regions.to
is orbecause will z
Let P(|Z|> lie the the (8-4)
about level very u in the only
derived to that lie to -2-58 is be the is the probabilities andcurve normal
with samplingremote so , probability 2:58 5% inference.
be values of between mean the shaded
parameters. true in significant the 3)
theoretically. 95 significance
specified of < population = and area+2-58 lies variate
Z4.
probability of 00027.
till cases + an (0-01).
1-96 tluctuations A this area
isit event very -1-96 with
sample Reject Accept
2-5%
proved
A probabilities
out are We that is -1.96
hypothesis orare unusual standard 0-05- the
But ofthe it of
1% again and
calledalone. cannot such be Critical
Region mean
zero
false. confidence 00, 1
statistical level +1-96. is
the say event an X,
deviation very Fig.8.2 95% SampleTes
Large
can The are levels The and
sample of that be
event
called small. and
be limits, significance. levels the has due If let
of 1.96
hesis is Z is standard
hypotstahypot
tedhesis meancontioei signiticank deviation taken tovery
marked > o 2-5%Reject Accept
also sampling 1:96 and
in va place smal l we
inmay
various the The by o or
 We Statistic
statistic
Test nature Test (4)
hypothesis Confidence (3) value
IfLimits nypothesis
disproved. expressed that sampleexceed value to (2) generallyan is
thelimits ofthe random
calculate rejecting
probability lie. alternative stated.
populpopul
ationa1t0iis
on there andparameterthe wayis s
Applied
statlstic After sample or supports It the The After To tore (i)
the of
trom ne should does the
and The ) t is n A e.g.
fiducial value
probability set
setting is imits is It a in limits limits the idea setting specifies Alternative It the fcan noThe the
sample
problem rejected value only percentage up Altestatistician
rnative
can sample accepted true not is of test Mathematics
limits.within the be of of are hypothesis H¡hypothesis
: Ho called boys contradiction
rejecting significance in levels of
be Null parameter
be up meansnoted a support the lbs. is
lies hypothesis. hypothesis statistic thatfixed setting up 4=110 : a ascribed
Hypothesis not
definedmean values we the which sample the 4 in to Hypothesis.
between at it
range Hypothesis
anThe
see
generally equal -
hypothesis
use Generally,
the depending
as th e null 110 alternative a IV
does that that a th e should that is
or an truedifference up college
sample normal to specified
as the a= wil supports hypothesis Ibs. lbs. of null whethersolely between to
test hypothesis when is hypothesis the
the the not hypothesis0-01. 5% lie (Alternative (Null values
thehypothesis :
sets thequal e
the the sample hypothesis arise is The
distribution, mean a population to They up to
statistic and confidence orthe in upon 110 : the the
proportion
hypothesis. level
confidence hypothesis Hypothesis)hypothesis rather For chance. approach two given a
1% will the due we are given
after should thatvalue critical and lbs.,
example, nullbelieved
level t
is eXceed hypothesis to sethe is hypothesis value
of set
lculated fixing the 0-05 accuracy sampling it is Hypothesis) than denoted then hypothesis value
significance.limits does is up mean Such
t-distribution, etc. limits lie of tothe here up etc. (8-5)
Depending the hypothesis significance) is and
region rejected.
the one and result in
and with ensure th e if or
the rejected
not limits is it a isinstead such a the
level are 1% limits
desired.
and not
value. is byhypothesis is
from
on specified support is fluctuations null should and to
hypothesis set level called within denoted assumed
HÍ. set parameter is
the it that equal
istribution, of up is it is is hypothesis the that
way if
these upon ln In up one. of
basis significance with proved. does of a 0-05 accepted.
Generally, th e
which be
probability the
significance =the totheaddition sample the
the 5% alone. by that
rejected is
we is5% hypothesis.mean level or difference 110
present They
H¡ hypothesis,
Called
calculate
ot nature
accepted, we the one is greater than
priate or It level 0-01. lbs. to wil result
1% means of the
If If expect The are
etc. we are that it the
this, mean
be or a
level These
exxample
of
need significance
means of The exceedsbetween
limits alternative null that
and truecalled
test
theFrom the called significance
the difference one accepted.
that
if Similarly, probability the Sample
Tests
Large
that of orthe
(i) value
given the
to it of are hypothesis.
the data does
confidence hypothesis the are null the the assumption,
bability calculate significance. the the we more difference the other isNull
statistic. probability fixed th e denoted weights
sample and hypothesis hypothesis may mean
not, samplewhen limitsdoes hypothesis Hypothesis
the means (and that sample
the such false.
a make of
a is the not as: the of The that
is a or
a
 the separates
calledthe and
distributioncritlcal Decision
with Making(6)
tails denoted discussed
issamplelevel,
Selection(5of
) where, the
x-statisticas where, Where, distribution test
distribution Applied
the
is The The sample, statistic
table After x-distribution: Z-The
wewe Oi s X
t- the X
Mathematics
equal region.
critical next followingconstruct distribution
population, n isdistribution:If most
of bythe
indecide setting the is and
the regions value nthe the depending for
Za step the is
observed commonly
totest
the (or Approximately
Normal)
(Population
Normal or sample
the sample if testing
is the
next table test-statistic up the
population)(Any a the
statistic.
level l, of for to Sample size sample -
acceptance or the <30 n
sizechapter). nature test
shows nul l To he is
If: tmean, upon IV
30n mean,
of th e the a
of
The xa compute criterion. frequency test used
chosen hypothesis the hypothesis.
significance. The where ) of the sample size parentis the
the the
sample u u large
the and goodness is is test size
area conditions theofthe
and level statistic
Depending the size population
region a value its and statistics
(n
denotes and population (< population and
of
arejection, distribution sample. is >30), Different
of the isE t= the
knowns.d. o of30). small Z=
critical maysignificance. the of Population under and the fit,
z-test z-test
alternative sNn-1 X-u then can arenature (8-6)
the is itsupon expected to (n o/n
which
probability test mean mean probability
:
region lie levlecalled statistic
the < we be
on 30), considered of
independence to to calculate data.
actually
(which z-test hypothesis
nature frequency
gives one ofthe The from be then be
significane. tested, tested, distributions
side criticalvalue unknown S.d. g distribution. and
given of we
Population the at
or t-test z-test t-tests the and calculate
both value of of of s o testleast
The information
the population after an attributes is is
are the statistic
the approximately are
sides event
or
region sample used. deciding standard th e standard used
signiticanls
probability)
the test Sample
Large
Teste
of and or as
the of (t- and a etc. to
rejecud statistic distributon significance cell. we statistic calculate
size deviation deviation
compare calculate as
which of normal
as
samplining
i the af the
of

significance
are fortwo-taile for a
or P(Z<-Z) the is
criticalthe v called calle
test is When ana
5% the left)
same other In Thus,th distributior or
tests.CUrve.
If If
th (7)
and at syrBy (c)Re
(b)Le (a) R (ii) O We
u
symmetric The Two
Applied N
Rig leve ()TL
Let Tw 1 c
 lor or P(Z<-z) the is test When distribution tests. ne (2) Applied
O two-tailed aleft) critical called symmetrical
micance In Thus,
at By (c) (b) called is (a) an(ii) We
and Damne other
the O) If If The Two
level ::
symmetry Relation Left Right alternative One use TWo on the
tailed
test
Lefttailed
Right test
Two
test tailed 1% the = value probability
Mathematics
i.e. P(|Z|> a P(Z2Z)= left the u> (i.e. curve. the Tailed
testthName
e of whilecritical words, test critical of Tailed Tailed Tailed areasAlternative Null Tailedrejection as
percentage significance of tailed right Ho u other we
value, at between Z or
Hypothesis
< For and
the this the value of
z)corresponding on Ho Test hand area know -
Test tailed hypothesis Test
=2a.
the a. test. Test u< bothHypothesis or example, distribution One IV
critical say, means level = = : and
If u :
of
P(Z<e) P[(Z< standard th e For th e : Ho, In: sides > Inthe lies Tailed
Z,=-2-054
=-2326| Z Z,|Zl 1-64 of (LOS) Z test. If Ho). Ho th e
of the the one rejection on
=+2-326| 1% regions significance area for for critical example,region is while two a
76||z,=t2-326|z| or Z) the parts of
a a to
Forregion test onetailed and
:u=Ho tailed Tests
2:33 for the one-tailed
is two a
+ +normal he example,
of is
sided hence, Motesting,
for for two same the P(Z> (Z>
tvalues sample of
level the
rejection of one test hypothesis
areaSIdes the
z, two-tailed onetailed
= critical 20. same Z) z)]
distribution test rejection sided. for we it lies curve
(8-7)
2% test of for the is .
1.estatistic
+ Level tailed test example, use a on on
2-054 as significance
one-tailed < lies test
value (right lies two one
of test test, is that uoon area the the the at
double isthe for on tailed rejection side two two is
Z
+1-645Z,= significance are th e z, a u lying a
le ft left > the onl tailes
y normal
=-1-645| = 5 10%critical the atest uo test. ends
t % that right on
critical in tailed side is the
1-960||zl and the and one area the are
a
side test
regions one of of distribution.
a 2% right-tailed two test. the right side liescalledistestcalled
z,1-282
=+Z region of
respectively. tailed tailed normal tailed th e oncalled is
=-1-282 =t 10% (i.e. Fi8.g3. 0 of the Sample
Tests
Large
the the
1645 the test. normal
ie. testtest: curve, test. two one the two The
probabilities) This the normal
then test. twotails
tailstailed z.cune
Let cuve,
means level the curve. of tailed the of
z, the
be test the ie
of
 Applied
2
of
as
significance Tne
discussed It Mathematics
may same
Right
test Two
be Left tailed Name
table
noted tailed tailed
C, earlier. isa
Area the of is -
IV
Left-tailed
Test =X-Zal2= again
thatthe test test the
Area area5%
a= of
area5%lowertailCritical
same Table recast
C=X-ZS.E. x (LOS) (LOS) (LOS) test
of Critical
lower
tail = rejection
value of from
rejection as
S.E. x critical
value for a=10% Criticaldifferent
10%a=
critical a=5% a=5% Z=1-64
k for
Z-1-64 H¡iu<Ho Hi
H Values
Acceptance of
value value
Acceptance of Acceptance
Hypothesis
Null of
-196 Hypothesis
area
90%
Null Region level point
Hypothesis area
95% Region of of (8-8)
area
95% Null Region of of
Fig.8.4 of % (Normal of
significance z, of 2-5%a=2:5
a= a=5%Z=1-96 Critical
Value view
k*164 1-96Z= for z,
for
two-tailed a below.
C,
single-tailed Distribution)
C=X+ZS.=E. xrejection
Area of Critical
2:33Z,=
Right-tailed upper
tail = = X a
=1% a=1%a=2%
Testrejection
Area of tail
aupperCritical
value +
= area5% Zal2 test
5%o
value (left
test
area at
S.E. * 258 Z,=
for for 05% a= a=
0-5 %
level a =1% SampletLarge
or
ight)
of
significan
at
leve a
 the table be error because tsampl
he e Error voe an
standard 8.Sampling otherstatistician llincreasingsample
Typehypothesis TypeError l
rejected reject 7. Applied
population. .aro we as decided TVpe erlror
Sampling The problems he Thus,confidence Type
value value error
Obvi o usl y, ()A it. Errors
When
deviation snall
of wishes we of or We
four ontherefore it ll is the I and (a) (v) t rue Mata hematics
Variables, now seems again and error and error never
This already situations
considering type to limits. the increased, confidence rejecting
the hypot if hesi A in
of see avoid. do arises arises the false s st ati sti cal
result lo hypothetical possibility know Testing
Variables error not
Il has that hypothetical
But -
Jnseen (ü) how falseH,
is trueH,
is arising to outhypot
comea hesis IV
is
where Sampling In
the may decide, when know then when becauselimits. when true hypothesis
is
whether
going that to certain of
i test practical
in whether
error the value false a value a (b) rejected,
of
tomeanthe is the the prove weBut true hypothesis is the isHypot the hesis
of a depending
problems of error do then is
help sample hypothesis.
Attributes. acceptedH,
isprocess to lie s exceeds
hypothesis
called accept test ed.hypothesis or
Correct
errorType Ildecision
consequences be one the hypothesis wi
ofthin not
because leads (i) tested
us hypothesis
type type type or A
in mean Fi8.g5. know the (b) tr ue there
testing of ofserious. type upon
ll to (8-9)
population
the We decision is the the is
I . l is whether
ielimits. confidence of is accepting error. hypothesis true
X shall error the this rejected are
Hence,
of
decreased is accepted of possibilities
hypothesis
about a is the or
only
normally making the nature rejecting
may true The false.
consider rejectedHo
is the a
Correct
decision errors the or error i.e. false is two
errorType I prove
andothe is can levels the
ofthat false error, i.e.hypothesis limits.
when ()
accepted, Hence results
accept
of a can when of one. and
distributed only of type The
beboth problem, to intrue there
arise
of be thereality. bethe error the The (iv) either
two described is ll (ii)
the signifícance the serious other
hypothesis
minimised i.e. then
types difference true difference can possibility Sample
Tests
Large
with types. as is of false A we
populastandard
tion or be we
possibilities. four
in to accepting are
of increased.
and by false
samples the which is minimised between (a) hypothesis
deviation mean wil between committing
decreasing in is it
mean. form have in increased,
certain type called or
ano å reality. false a
of (0) of The the bythe is we
a to of
 (v) (iv) (iü)
(iü) Sol. regarded We, we
30) Note
..E levels.
other Such
that testedbe
be
can
then deviation
(calledof GVn. 9.Applied
Decision:
null Level sample then say
. Critical
large isTest (1) : Example If therefore, be Now thatthe
hypothesis |Z|=
1-77 i. Alternative Null wethe that relative
seen numeicaly mean, IfSuppose
Hence,sample Testing
as rare a X Mathematics
The of Statistic standard we is
Hypothesis drawn use relative
value:
significance Z= take the
sample
from ovent
Since 1:A the reject
=Aisa Z the
from deviation the a
isthe The Hypothesis
Since : randomstandard deviation the has sample
accepted.
is table -
a occurred Hypothesis
standard IV
drawncomputed value H,normal hypothesis
: the PZ|> sample
a= :p=54 sampledeviation of and
of /10.24J50 /-5.46.population o standard
from 0-05 2 of X find
z, H, population
the because then deviationo
at : from
of at 1"66) mean its that
the value 5%
50population s 5% u mean normal X
population of of is
level S.D. 5-4 jtems level is the= X
with the significant from distributed
normally Ho the
2]=1·77 X-4 0-05. X. = of
of 3-2/7.07 is gives sample probabiliy variate the (8-10)
unknown mean is of If the this If
meanPopulation
significance 0.8 not significance from population
with =
is the and 1-96 population
observed with
54 in
known u
mean less mean place the of
1.77 = but at not mean
5%
|Z|=|A98
sample is
Mean
5-4.
than from sample 62 ofand ifdue with known
level o. if | mean mean zero
the the and |>Z to chosen
the we
critical the sampling X and meanu
table S.D. of
variance u) and Fu
significance? sample 196. exceeds
11-96 is
sis such slandard let
value = and the
1-96. (M.U.
1996,
2015) Similarly,
fluctuations
known size get that standar SampleTeyLarge
10-24. null
z,= small.
veryAs
is |Z1 deviaie hypothesisth
n
130, and Can is we value deviation
large oln
since it define alone of
U be (>
 cms. 4
of
4000 ()Gritical (i) (i) deviation CGii) () Sol.
Significantly
Sol. IS 81 Sol. vears, MathApemplieadtics IV
the
veoSion Level putWe
Since Test (i) : Example Decision
null Critical Level (0) :
(): new kms. Z=
..
if Since Test
i. null Alternative Nul l afandomof hypothesis
H,:u42000
Hypothesis
Alternative Null tyres A The of X=71-8, Statistic
we 8-9 The of we Nul
change
Hypothesis bettere4:Atyre value
hypothesis Significance Hypothesis value
:Z=2400
ic ternative
has are years X=areStatistAl
hypothesis Since : Significance
8.9//100 71-8-70 3sample Since
: 4+45 -54.45,
reasonablyegardedExamplet:A
a
than'product In : given : sample Can is : given :Hypothesis
mean the the The u Z=
Hypothesis ? the The Z random
as
is = it rejected is | a
H,: t he computed value 70, standard H, be value standard
= Hypothesis
production
company rejected.
life is of not
computed 5, sample
old rejected.
:a 2.02 =
o iu=70 concluded 100
:a
HÍ sample
of = = drawn and of =2/20 O"55 o :
42000
=one.42,500 of
claims 0:05.
z, 8.9 , H,: Z0-05. =2, å from
value at deviation Indians deviation H, =5
n= #70 years. fro m the value at = of
process 5% 5% 55n=400. :u members
400 be
kms. that 100. that alternative a
oflevel has the oflevel large
Test isthe | of years the Z= of 5 (8-11)
Z|= of the an above the
believed lives
at average 55 of population
5l%
significance
2:02 sample, average
of population. is
hypothesissignificance population, is
level to tyres is greater foundto
result greater life-span
life
of have whose 5
significance is span
inbetter than 1"96.
isthan from we have
mean accepted.
of the put mean
the of an the a
71"8 critical mean
product.42,000
that
critical Fig9..6 Indian table is

kms years value = cms of

(M2006,
.U. 09) the value is cms.
4-45 Sample
Tests
Large
new test wiAth 1-96.
1-64 varianceandis
Z= (M.
2004)U. with more z,
sample S.D. = 2016)
(M.U.
196, standard than 1-96, Can
or
the it
7O
 Is
02 level
the
cms. sample a the a 150.
sample found
strength
deviation breakingturmed 165 ? of
machine machine 8. of sample of0-45 (V) (iV) (iii) (ii) Applied
7. the th e
significance 6. Can 5. that 4.
sample A A from A A out 3.sample 2.machine cm. 1. significance
the 1-96.is Level large but Test
random should
distribution this A th e of The
strength A The Decision Critical
(Hint:-u will random the 100. sample to A A .. null =1·125 |Z| .: sample Mathematics
fulfilling a sample be random machine Statistic
large be mean cable mean mean
is There of
be By justify hypothesis Z
of is sample within sample of14-5 10. : value
Significance: =
? taken breaking a of height Since
100 population
set with be breaking has new 15-6 sample is is S.D.
the i.e.,
1"96 = pounds. 50Would t he no X -
to claimed th e : Sinces : IV
purpOse plates of
0-5 from regarded
unknown of increased.
technique pounds pieces of claim? improvement. is The is
produce 900 of size strength strength random it of accepted. known
the the Test be 100 computed 42500
42000 - the
produced whose items 36 of
reasonable to value a 4000/J81
=
? population
distribution as and certain Mention produce nails population
u drawn has In in whether 0-05 and
metal O"5 has is standard of sample of
mean is 53 1850. order the gave value
found variance (M. U . cables EXERCISE
th e -I z, since
by from asmanufacturing type
thplates
e to the of at 1.125 = S.D.
is 1-96) =mean. in mean to of level 5-1 nails
the
to Can test deviation
suplied sample of suppose 100 | 5% (8-12)
order the 2007, cm. of Z is
machine of 66-2 have
1-5. we string sample
thickness population and the individuals of mean as = level unknown
cms, [Ans. support
09) significance apply.
you 1-125
a
that Use claim (M.U. their
sum by of iswas that
gave and mean length of is
the
central Ans. [ process a from
2-2 is
: of manufacturer
the a tested. the 2014) average less
an 1-5 (M.U.
2014)
Ans.
standard of probability s=sguares
having sample pounds. a (M.U. from
2-04, :|claim batch mean 5
average cms 65-3 it cm. than
limit Z|=3.54. is The a
Ans. [length.
at of population
withdeviation cms. theorem |Z|= 54 of claimed of2005)[height and the
will deviationsas 1% 50 [Ans. mean a :
thickness:|z|=54, [ is string |Z|= standard Does critical Fig.8.7
|Z=1, s] Can as be mean?
2-94, cables
level 1800 Ans. of SampleTes Large
Right
[Ans.Yestandard
it at that breaking
: the is22, the
of |Z|=3-53, having value
is be[Ans.
is with :
population TheSn 160. 1-64
|Z|=5,
:n=23 least find to Yes from
5 lait the performance deviatinn.
of CmS. (M20049) tested. No . Za=196,
deviation .U. how at significanca. test. breaking standard strength
1-52 regarded 0277%1 mean a at
iay mean
cms Yes] Itis No) No ie 5%1
 samples expression,
that en by the ifand wouldsamolemeanspairs
givbetween 4)
below. that the how two Universitmeans.
yequal For 10. Applied
ploeios I Further, sizes thefollow of
Suppose
Distribution the two are
0f the Le. second mean for
is the samples twopopul
sampl es atiowe
of all
nsto, SometisimTeses ting
given samples samples Z= S= the of
S.E.= under Z-X- a these
means populations know equalit
Mathtehematics
by samples
populationnormal the from two
of came example,
is to we
G. n 2 pairs,
firstpopulations -
are are J(1/ o are the X the whet unlikely may IV
2 X2) curve. the from her averageDifference
large large ) drawn hypothesis
+2 - population. have have
- X2are
respectively always difference
two the we
n2 S2
2
+ n2
(41-H2) lf the that l.Q. may two
and then (1/ from S.N.V. isa 0s n equal same
normally and X
populations. have two the of
n2) subtracting
and If distinct
Between
o oS.N.V. isa the u ng. these equal two want
samplesstudents
= S.N.V. isa means?
between population
= same = Thenhaving X¡ to
o, s, 2 sample
we distributed differences
denote Let
know
populations (8-13)
= we the means. of
o, o population, means means The came Means
see can us Pune
?meanswhether
then = the sample take
s that prove u, are Suppose
procedureotherInfrom
he with means University. and
tasymptotically
He
graphed mean
difference would the
so that populations
words, we
combined mean and
that of offurther to average may
be
the
standard test identical. If
o uy-2
the wethe how we want
= distributionsampleswouldbetween we this having
standard and second can actually I.Q. to
= and
deviations draw hypothesis test
of
0, find weEvenstudentsif Sample
Large
Tests
We standard drawn
population the a testequal take whether they
deviation of
that
see pairslarge they
the samples
the the means from
o
from is of
from hypothesis are
and of
number discussed Mumbai
deviation
difference distribution from
or
equal, have
of the sample
tne o,the that
abOve and first the of
to
(i) ()
 from, s
estimator n (known)
Remarks.., with (2)
and Applied
In 2. In means
n this 1. (ii) (i) Suppose
() Procedure But
this then If then we
population
ofare If
the Then is Mathematics
case standard case
standard S.E., =
s calculate asymptotially
sufficiently samples we we
we , we
|Z calculate, calculate to E(X1-)
deviationstandard
S2 |Zl=calculate and take
= S.E.,s= test
S.E.s=
, a error are standard
X two n2+
n -
large, 2 - the unbiased IV
drawn the Xsamples n2 +
ng
S
+
of
deviations the |Z| hypothesis
the
2 (since |n standard 2
value deviation
1 from S2 and
population) 2 estimator
2 of
sample o1 of size
the error, n2+n
and o1, := 1
populations n
we
standard and G2. and n2 of
o 42 1 o. (8-14)
replace are take To
n :
not test with
deviationknown with the S)
the 2
o decision means
by common hypothesis
s (o
and s # X
is as
asymptotically o, 2) S.D. explained and
by and that
X
S, if o uË
and the i.e. from
in =
find sample if u2.
the Sample
Large
Tests
unbiasea o (i) 2
the = populations
O
above.
(3) S.E. sizes
(2) =
0 (1) .

Calculat (ü) perform

bette that Critical
Decisio () (iv)Level(iii)o Calcula(ii) inches.
2-5 68-0
Sol.inches ? C
Sol.:() TestingHy
of hypothesis u Note unknownand
assL(We of
The s. Since S :
ApplieMad th
th3.
e If
Exampl=36 hypothe (i) Examp Formula
Null CarefulI
Al Nu girl= Alte
 performna-ot
Sol. () (iv) (in) (i) inches?
Sol. 2-5 Hypothesis
68-0 Testing of
) hypothesis Note .. unknown and Applied
(We Calculation (i) :Eample hypothesis
rejected. Decision is Critical Level Since
Calculation Example1:The
(): inches, Earnulae 3.
better 36 Alternative Null Carefully If
The
assume Null girls S.E. =
S S.D. Can u the
Z= S.E,= Alternative samples of Hypothesis Mathematics
if|
than is significance:a= Value: : = fo r note s
Hypothesis 2: of the M2 standard
of 70
The Since (2"5) - the of as and
that S.E. the S Hypothesis samples given s the S.E.,S= -
Statistic:
n ss 2
with cannot The Statistic
population
n means wel l S IV
the 4 girls.average th e difference
Hypothesis standard Ho deviations
are
computed value 1000 as in
standard 2 Ho be 0-097 be for (1), the
0-5 : 1 of
a115 X, of of =0 X-
1: regarded 027% is =
Ha regarded two estimating (2),
marks z, 2000 u2 between standard2 of
=
- Hadeviation 1 the
eviations +
X, H2 value at =-5"15 |1,1 known, X, u1 samples (3) n S2
2
36 36 = 0-27% 0"097 = = and
scored as as deviations two
:: 72-70 1 of 67"5- M2 drawn the the
/3= W2 6. drawn |Z|=515 populations (8-15)
|2 Test level of interval (4)
denominators
o by sizes are
and 1-15 = =2 68-0 from
at 32from of :. to of
oz boys significance
1% is =-0"5 the 1000 ofbe the
the greater |Z|
of level same and taken samples ÞË
the is
same = and
5-15. of
two 72 the
of than population2000 into o2
significance with
population. from formulae ofare
lations (M.U. the
consideration sizes
standard respectively
the equal
critical
2004, table of n
(3) Sample
too, Tests
Large
whether standard and
equal.)
notare deviation value and
09, is
3. are fo r n,say,
10, (M.U. (4). then
za 67-50 testing
the deviation and
15, 8 =3,
boyswhile 2015) are
16) the and the (4)
Applied Mathematics -
IV
(8-16)
Large Sampley Applied M
: a= 1%
(ii) Level of significance is .eo (iv) Criti
Critical Value: The value of Z, at 1% levelof significance from the table
(iv)
(v) Decision: Since the computed value of|Z=1·15 is less than the critical value Z,=256 Deci
(v)
T
hypothesis isaccepted. population
girls.
.:. Boys do not perform better than the
Exan
Example3: Test the significance of the difference between the means of twO normal nee that of the
with the sáme standard deviation from the following data. difference
Size Mean S.D. deviation.

64 6
Sol. : We
Sample I 100
Sample ll 200 67 8
Sol. : () Null Hypothesis Ho :y= M2
Alternative Hypothesis Ha: u4 H2
(0) Calculation of Statistic : X- X, = 67 -64 = 3
Since the standard deviations of the two populations are equal but We v
unknown
2 Divid
36 64
S.Es=
n V200 100
=J0.18 + 0.64 =J0-82 = 0.91
Z=-2 =3.3 .:. |Z=3.3
S 0.8
(ii) Level of
Significance: a =5%
(iv) Critical value: z, at 5% LOS is
For S
1-96. for uy = u2
(V) Decision : Since the
computed value of | Z|= 3:3 is greater than the critical
Zu=1-96, the null hypothesis is rejected.
. The samples do not
support the hypothesis that the two
although they may have the same 1. A
standard
deviation. populations have the saiie
weight of 2
Example 4: Two samples drawn from two the same k
Size
diferent populations gave the following between th
Mean S.D.
Sample 125 2.. A
340
Sample I| 150
25 stock for te
Test the
380 30 hours,
of theseandres
Sol. : (i) Nullhypothesis 5% LOS that the difference of the means of the two
at
populations is45
Hypothesis Ho: -, = 45 3. Th
380 grams
(U1) Alternatlve Hypothesis Ha:H * H2 600 working
Calculation of Statistic: X- X= 40, H-H2 we justified
S.Es =
=35 5% level of
|252, 30²
V125 150 = 3.32 4. Tw
. Z=(X1- X2)- (41 -H2) 40- 35 66-1 inches
(i) Level of = 1.5 3.32 |2]= 15
Signlflcance :a =5%
 Large Sample Tests
Applied Mathematics-IV (8-17)

LOS is 1-96.
value : Z, at 5% than the critical value z, =196.
(iv) Critical
the computed value of | Z| = 1-5 is less means of the two
Decision: Since hypothesis that the difference between the
(v) supports the
. The data 45.
populations may be but the standard deviation of one
is twice
populations have the same mean random conditions the
Example 5 : Two samples, each of size 500, drawn under simple smaller standard
that in where o is the
that ofthe
other. Show probability, will not exceed 0-3o. (M.U. 2007)
difference of the
means, in all
500.
deviation.
u;¡ = 0, O2 = 20,n = 500, n, =
u, M2 =
Sol.: We have My =
2 2 462
O1+ O2 +
S.E. S= 500 500

5g2
10
V500
We want|X-X>l< 0-3a
Dividing both sides by S.E.,
0.3o [: o is positive ]
o/10 o/10

| X- X |<3
S.E.
almost impossible. Hence,
For S.N.V.Z=3, 99-73 % area lies under the curve. The event is
for u = 2;|X-X| in all probability willbe <0.3o.
EXERCISE - II

1. A
sample of 200 fish of a particular kind taken at random from one end of a lake had mean
weight of 20 Ibs. and standard deviation of 2 lbs. At the otherend of the lake, a sample of 80 fish of
the same kind had mean weight of 20-5 lbs. and standard deviation of 2 Ibs. also. Is the difference
between the mean weights significant ?? [Ans. : |Z|=1·89, No )
2. A man buys 100 electric bulbs of each of two well-known makes taken at random from
Stock for testing purpose. He finds that 'make A' has a mean life of 1300 hours with a S.D. of 82
nours, and 'make B' has a mean life of 1248 hours with S.D, of 93 hours. Discuss the
of these results. significance
[Ans. :|Z]= 419, Significant ]
3. Ihe mean consumption of food grains among 400
sampled middle class consumers is
Ograms per day per person with a standard deviation of 120 grams. A similar sample survey of
600 working class
we justified in consumers
gave a mean of 410 grams with standard deviation of 80 grams. Are
5% level of saying that the the difference between the averages orf the two classes is 40 ? Use
4. significance.
Two groups consisting [Ans. :|z = 5-86, No ]
66-1 inches and of 400 and 500 persons have mean heights 68-5 inches and
variance 6"4 and 6-0 respectively. Examine whether the difference is significant.
[Ans. :|z| =1454, Yes ]