Sampling techniques Click to edit Master subtitle style and its significance

4/28/12

 A FINITE SUB SET OF UNIVERSE IS

CALLED A SAMPLING.
A SMALL PORTION OF A UNIVERSE. USED IN DAY TO DAY LIFE.

4/28/12

TEST OF SIGNIFICANCE OF LARGE SAMPLES.

Testing of significance for single

proportion.
Testing of significance of difference of

proportion.
Test of significance of single mean. Testing of significance of difference of

mean.
4/28/12

LEVEL OF SIGNIFICANCE
1% TWO TAILED /Z/ TEST RIGHT Z TAILED LEFT TAILED Z 5% =2.58 /Z/ 10% =1.966 /Z/ =0.645

=2.33

Z =1.645 =-2.33 Z =1.645

Z =1.28 Z =-1.28

4/28/12

Testing of significance for single proportion.

Used to find the significant difference

b/w the proportion of sample and the population

4/28/12

A coin was turned 400 times and the

head turned up 216 times . Test the hypothesis that the coin is unbiased.
H0=Coin is unbiased H1 =Coin is not unbiased

n=400 ;X=No of success.=216 P=Proportion of success=X/N=216/400=0.54 Pop proportion=0.5 ;Q=1-P =0.5 Z= p-P =1-0.5

=/Z/= 0.54-0.5
4/28/12

p=Proportion of success in the

sample.
Q=1-P =Proportion of failure. P =probability of success

4/28/12

DIFFERENCE OF PROPOTIONS

A machine produced 16 defective articles in a batch of 500.After overhauling it produced 3 defectives in a batch of 100.has the machine improved? H0 H1 =Machine is not improving. = p1<p2 =right tailed. N1=500 N2=100
4/28/12 =0.032-0.03

P1=16/500=0.032 P2=3/100=0.03

/Z/=

P1-P2/

Test of significance of single mean

Test wheather the difference b/w

sample mean and population mean is significant or not.

4/28/12

Q1
A normal population has a mean of

6.8 and S.D of 1.5.A sample of 400 members gave a mean of 6.75.Is the difference significant. ?
H0 =NO SIG B/W X AND H1 = SIG B/W X AND

=6.8 S.D=1.5 x =6.75 n=400 /z/= =0.67 x - s.d /n = 6.75-6.8 1.5-400 4/28/12

Significance of difference of mean

Intelligent tests given in two groups

of boys & girls . examine the scores is significant.

Girls+ boys

H1=There is no sig b/w mean scores.

75 73 8 10 60 100

mean

S.D

size

4/28/12

T-TEST /studentsdistribution

The T distribution is used when sample size is <=30 and the population of S.D is unknown.
T

= S

X - n

s = (x-x) n-1
4/28/12

APPLICATION OF TDISTRIBUTION
To test the significance of sample

mean from a random sample.

To test the significance between two

sample mean.
To test the significance of partial and

multiple correlation of coefficients.

4/28/12

Case -1
To test the significance of sample mean from a random sample.

4/28/12

Q.1.A random sample of size 16 has 53 mean.sum of squares of deviation is 135.can this sample be taken from 56 as mean.? Use 0.10 H0=mean not cme frm pop hvng 56. T= x- =53-56 =-4

s/n 3/16 s=(x-x) =135 =3 n-1 15 / t / = 4 =16-1 =15 t = - 4<1.753 Sample mean is not come from popu hvng 56 4/28/12 mean.

Case 2
To test the significance

between two sample mean.

4/28/12

Two samples of sodium vapour bulbs

for tested for length of life,The results are:

size 8 7 TYPE 1 TYPE 2 Sample mean 1234 hrs 1036 hrs Sample S.D 36 Hrs 40 hrs

Is the difference in the means

significant to generalise that type 1 is 4/28/12

ANSWER HO= 2Types having same life time. H1 = type 1 is superior to type 2

21.04>tab1.75 ho is rejected. 4/28/12

normal population with S.D 3.5 & 5.2.mean=20.3 & 18.6.test wheather the mean of two population are the same at 5% level?

4/28/12

ANSWER
Here mean 1=20.3 mean 2=18.6 n1=10 n2=14 S1=3.5 S2=5.2 H0 = The mean of 2 pop are same.

H0 = The mean of 2 pop are NOT 4/28/12

Case 3- To test the significance of partial and multiple correlation coefficients

Q1 .Memory capacity of 9 students befo and after a course if meditation for a month.state whether the cource was effective?0.05
befor 10 15 9 e after 12 17 8 3 5 7 6 12 16 17 4 11 18 20 3
4/28/12

BEFORE T 10 15 9 3 7 12 16 17 4

AFTER T 12 17 8 5 6 11 18 20 3

D=X-Y -2 -2 1 -2 1 1 -2 -3 1 d=-7

D2 4 4 1 4 1 1 4 9 1 d2=29 4/28/12

H0 =Training was not effective. H1 =2 TAILED TEST

9-1=8 =2.31 :T=1.359<2.31(Tabulated data)

4/28/12

following increases of B.P.Can it be concluded that the stimulus will in general be accompained by an increse in B.P?0.05 5,2,8,-1,3,0,-2,1,5,0,4,6

4/28/12

ANSWER

T
s=d

= S

n-1 n
4/28/12

F-TEST /SNEDECORS VARIANCE

The object is to discover

wheather two independent estimates of population varience differ significantly or wheather the two samples may be regarded as drawn from the normal population having same variance.
4/28/12

APPLICATIONS
1. Wheather two independent

samples have been drawn from the normal population with the same variance. 2.Wheather 2 independent estimates of the population are homogeneous or not.
4/28/12

Q1 in two independent samples of sizes 8 & 10 the sum of squares of deviations of the sample values from the respective sample mean were 84.4 & 102 .6.To test whether the difference of variance of the population is significant or not?

4/28/12

ANSWER
F = S1 X1-X1) S2 n1-1 S1=(

S2=(X2-X2)
4/28/12

are Test wheather the samples are drawn from the 29 13 17 same A 17 27 18 25 27 B normal population.?o.5% 16 16 20 27 26 25 21

4/28/12

X1

X1-X1

(X1-X1) X2

X2-X2

(X2-X2)

17 27 18 25 27 29 13 17

-4.625 5.735 -3.625 3.375 5.735 7.735 -8.625 -4.625

21.39 28.89 13.14 11.39 28.89 54.39 74.39 21.39

16 16 20 27 26 25 21

-2.714 -2.714 1.286 8.286 7.286 6.286 2.286

7.365 7.365 1.63 68.657 53.085 39.513 5.226

4/28/12

Ans
If 2 independent samples have been

drawn from the same pop.we have 2 test

1.Equality of mean by applying T

-TEST.
2.Equality of pop mean by applying F-

TEST.
F - TEST

4/28/12

T TEST

TAB T=2.16 ;T =0.9704 T< TAB T =0.9704<2.16 HO IS

ACCEPTED.THERE IS NO SIG B/W POP MEAN.

We conclude that 2 samples have
4/28/12

CHI SQUARE / X TEST

Its discribes the magnitude of

discripancies between theory and observation.

If x=0,the observed and expected

frequencies completely coincide.

The greater the discrepancy between

the observed and expected frequencies,the greater the value of x.

4/28/12

Q1.Following table gives no of accidents fork place in an industry.

5 acc mon 14 tue 18 wed 12 thu 11 fri 15 sat 14

4/28/12

X = (Oi-Ei) O=Observed F Ei E=Expected F OBS 14 18 12 11 15 =84/6 F(O1)

EXP F(E1) (O1E1) 14 0 14 16 14 4 14 9 14 1

14 14 0

4/28/12

No of die fre

To test whether the die is biased or not?

1 2 3 4 5 40 32 29 59 57

6 59

4/28/12

x=( O i E i ) oi 40 32 29 59 57 EI Ei 46 46 46 46 46 (oi-Ei) 36 196 289 =980/46 169 121 =21.30

59 46 169

4/28/12