PROBABILITY & STATISTICS SMALL SAMPLE TESTS

UNIT-V
SMALL SAMPLES
If the size of the sample < 30 the sample is said to be small sample or exact sample.
There are three kinds of tests for small samples.
1. Students t test using t - distribution
2. 𝜒 2 𝑡𝑒𝑠𝑡 𝑢𝑠𝑖𝑛𝑔 χ2 𝑑𝑖𝑠𝑡𝑟𝑖𝑏𝑢𝑡𝑖𝑜𝑛
3. F – test using F – distribution
DEGREES OF FREEDOM:
Degrees of freedom one number of values we can choose freely.
For example: If 𝑥1 + 𝑥2 + 𝑥3 = 50 to find 𝑥3 we have to assign any value to the two of the
variables (say 𝑥1 , 𝑥2). Similarly for four we require three values.
Degrees of freedom for 2 sample values = 2 – 1 = 1
Degrees of freedom for n sample = n – 1
SMALL SAMPLE TESTS :
(i) t-test:
Name of the test Null Level of Test statistic
Hypothesis𝑯𝟎 significance(𝜶)
1.Test for single mean 𝜇=𝜇0 5% or 1% or 10% 𝑥̅ −𝜇
t= 𝑠
√𝑛−1
2.Test for difference 𝜇 1= 𝜇 2 5% or 1% or 10% 𝑥̅ − 𝑦̅
t= 1 1
of means 𝑠 √𝑛 +𝑛
1 2
3. paired t-test 𝜇 =0 5% or 1% or 10% 𝑑̅
=𝑠
√𝑛

(ii) F-test:
Test for difference of variance
𝐻0 : 𝜎1 2 = 𝜎2 2
𝛼 : 5% or 10%.
𝑆 2 𝐺𝑟𝑒𝑎𝑡𝑒𝑟 𝑣𝑎𝑟𝑖𝑎𝑛𝑐𝑒
Test statistic F = 𝑆12 = 𝑠𝑚𝑎𝑙𝑒𝑟 𝑣𝑎𝑟𝑖𝑎𝑛𝑐𝑒
2
𝟐
(iii) Chi-square (𝝌 )test:
Test for goodness of fit and test for independence of attributes.
(𝑂−𝐸)2
Test statistic 𝜒 2 = ∑ 𝐸

T – DISTRIBUTION:
T – Distribution is used when population standard deviation is unknown.
Let 𝑥̅ be the mean of a random sample of size n, taken from a normal population having the
(𝑥𝑖 −𝑥̅ )2
mean μ and variance 𝜎 2 and sample variance 𝑆 2 = ∑𝑛𝑖=1 then test statistic is
𝑛−1
𝑥̅ −𝜇 1
𝑡= 𝑤ℎ𝑒𝑟𝑒 𝑥̅ = ∑𝑛𝑖=1 𝑥𝑖 , 𝑛 = 𝑠𝑎𝑚𝑝𝑙𝑒 𝑠𝑖𝑧𝑒 (𝑛 − 1) = 𝑑𝑒𝑔𝑟𝑒𝑒𝑠 𝑜𝑓 𝑓𝑟𝑒𝑒𝑑𝑜𝑚
𝑆⁄√𝑛 𝑛

PROPERTIES OF T – DISTRIBUTION:
1. The shape of t – distribution is bell – shaped which is similar to that of a normal distribution.

Page 1
PROBABILITY & STATISTICS SMALL SAMPLE TESTS

2. The mean of standard normal distribution and as well as t – distribution is zero. Variance of t
– distribution depends on degrees of freedom V.
3. The variance of t – distribution approaches 1 as n → ∞. In fact t – distribution with v –
degrees of freedom approaches standard normal distribution as 𝑉 = (𝑛 − 1) → ∞
APPLICATIONS OF THE T – DISTRIBUTION:
1. To test the significance of sample mean, when population variance is not given.
2. To test the significance of the mean of sample, i.e., to test if the sample mean differs
significantly from the population mean.
3. To test the significance difference between two sample means or to compare two samples.
4. To test the significance of an observed sample correlation coefficient, sample regression
coefficient.
2. CHI – SQUARED (𝝌𝟐 ) DISTRIBUTION:
Chi – squared distribution is a continuous probability distribution of continuous random
variable X with probability density function is given by
𝑣 𝑥
1
𝑉 𝑉 𝑥 2−1 . 𝑒 −2 , 𝑓𝑜𝑟 𝑥 > 0
𝑓(𝑥) = { 2 2 Γ( 2 )
0, 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒
𝑤ℎ𝑒𝑟𝑒 𝑉 𝑖𝑠 𝑡ℎ𝑒 + 𝑣𝑒 𝑖𝑛𝑡𝑒𝑔𝑒𝑟 𝑎𝑙𝑠𝑜 𝑘𝑛𝑜𝑤𝑛 𝑎𝑠 𝑑𝑒𝑔𝑟𝑒𝑒𝑠 𝑜𝑓 𝑓𝑟𝑒𝑒𝑑𝑜𝑚.
𝜒 2 distribution was used to measure of goodness of fit and to test independent attributes.
PROPERTIES OF 𝝌𝟐 DISTRIBUTION:
1. 𝜒 2 distribution curve is not symmetrical, lies entirely in the first quadrant. It is not a normal
curve since 𝜒 2 varies from 0 𝑡𝑜 ∞.

2. It depends only on the degrees of freedom v.

3. If 𝜒12 𝑎𝑛𝑑 𝜒22 are two independent distributions with 𝑉1 𝑎𝑛𝑑 𝑉2 degrees of freedom, the 𝜒12 +
𝜒22 will be 𝜒 2 - distribution with (𝑉1 + 𝑉2 ) degrees of freedom.
4. Here α denotes the area under the chi square distribution to the right of 𝜒𝛼2 .
NOTE:
1. For various values of 𝛼 𝑎𝑛𝑑 𝑣 the values of 𝜒 2 are tabulated in tables.

Page 2
PROBABILITY & STATISTICS SMALL SAMPLE TESTS
2. Test of goodness of fit are used when we want to determine whether an actual sample
distribution matches unknown theoretical distribution. This test is known as "𝜒 2 − 𝑡𝑒𝑠𝑡" of
goodness of fit.
Test statistic for 𝜒 2 – test is
(𝑂−𝐸)2
𝜒2 = ∑ ( ) 𝑤ℎ𝑒𝑟𝑒 𝑂 = 𝑂𝑏𝑠𝑒𝑟𝑣𝑒𝑑 𝑓𝑟𝑒𝑞𝑢𝑒𝑛𝑐𝑦; 𝐸 = 𝐸𝑥𝑝𝑒𝑐𝑡𝑒𝑑 𝑓𝑟𝑒𝑞𝑢𝑒𝑛𝑐𝑦
𝐸

3. i) If the data is given in series of 'n' numbers then degrees of freedom 𝑣 = 𝑛 − 1.

ii) For binomial distribution d.f. (𝑣) = 𝑛 − 1, 𝑓𝑜𝑟 𝑝𝑜𝑖𝑠𝑠𝑜𝑛 𝑑𝑖𝑠𝑡𝑟𝑖𝑏𝑢𝑡𝑖𝑜𝑛 𝑛 − 2
iii) For normal distribution d.f. (𝑣) = 𝑛 − 3.
APPLICATIONS OF 𝝌𝟐 – DISTRIBUTION:
1. To test the goodness of fit.
2. To test the independence of attributes.
3. To test the estimation of population variance.
F – DISTRIBUTION:
* F – distribution is sampling distribution of the ratio of the two sample variances.
* It is an important continuous probability distribution which plays an important role in
connection with sampling from normal populations is the F – distribution.
* Let 𝑆12 𝑎𝑛𝑑 𝑆22 are two variances of two random samples of sizes 𝑛1 𝑎𝑛𝑑 𝑛2 respectively
drawn from the normal population with variance 𝜎12 𝑎𝑛𝑑 𝜎22. To test whether the two samples
come from two populations having equal variances. Test statistic is
𝑆 2 ⁄𝜎 2
𝐹 = 𝑆12⁄𝜎12
2 2
𝜎22 𝑆12
𝐹 = 𝜎2 𝑆 2
1 2
Which follows F – distribution with 𝑉1 = 𝑛1 − 1 𝑎𝑛𝑑 𝑉2 = 𝑛2 − 1 degrees of freedom.
𝑆2
If 𝜎12 = 𝜎22 𝑡ℎ𝑒𝑛 𝐹 = 𝑆12.
2

PROPERTIES OF F – DISTRIBUTION:
i) F – distribution is free from population parameter and depends upon degrees of freedom only.
ii) F – distribution curve lies entirely in first quadrant.
iii) The F – curve depends not only on the two parameters 𝑉1 𝑎𝑛𝑑 𝑉2 but also on the order in
which they are stated.
1
iv) 𝐹1−𝛼 (𝑉1, 𝑉2 ) =
𝐹𝛼 (𝑉1 ,𝑉2 )

v) Mode of F – distribution is less than unity.

T – distribution is used when population S.D. is not known.
* For T – distribution
(𝑥 −𝑥̅ )2 ∑𝑥
𝑆 2 = ∑𝑛𝑖=1 𝑖 (𝑥̅ = )
𝑛−1 𝑛
𝑥̅ −𝜇
𝑡 = 𝑆⁄ 𝑤ℎ𝑒𝑟𝑒 𝑆 = 𝑆. 𝐷. 𝑜𝑓 𝑠𝑎𝑚𝑝𝑙𝑒
√𝑛
∑𝑛
𝑖=1 𝑥𝑖
𝑥̅ = 𝑛
𝑥̅ −𝜇
S.D. of population known - 𝑡 = 𝜎⁄
√𝑛−1

Page 3
PROBABILITY & STATISTICS SMALL SAMPLE TESTS
𝑥̅ −𝜇
1. For student t – test statistic when S.D. is t is defined as 𝑡 = 𝜎⁄
√𝑛−1

CONFIDENCE INTERVAL (LIMITS) FOR μ:

For 95% of confidence 𝑋̅ ± 𝑡0.05 𝑆⁄
√𝑛
̅
For 99% of confidence 𝑋 ± 𝑡0.01 ⁄𝑆
√𝑛
PROBLEMS ON T – DISTRIBUTION:
Find (a) 𝑡0.05 𝑤ℎ𝑒𝑛 𝑣 = 16 ⇒ 𝐴𝑛𝑠 ∶ 1.746
(b) 𝑡−0.01 𝑤ℎ𝑒𝑛 𝑣 = 10 ⇒ 𝐴𝑛𝑠 ∶ −2.764
(c) 𝑡0.995 𝑤ℎ𝑒𝑛 𝑣 = 7
𝑡0.995 = 𝑡1−0.005 = −𝑡0.005 = −3.499 (𝑣 = 7)
T – Test for single mean:
2. A random sample of size 25 from a normal population has the mean 𝑥̅ = 47.5 𝑎𝑛𝑑 𝑆. 𝐷.
𝑆 = 8.4. Does this information tend to support or refuse the claim that the mean of the
population is 𝜇 = 42.5.
Sol: 𝑛 = 25, 𝑥̅ = 47.5, 𝜇 = 42.5, 𝑆 = 8.4
1. 𝐻0 : 𝜇 = 42.5
2. 𝐻1 : 𝜇1 ≠ 42.5 (Two tail)
3. 𝛼: 0.05
𝑥̅ −𝜇 47.5−42.5 5√25
4. T𝑒𝑠𝑡 𝑠𝑡𝑎𝑡𝑖𝑠𝑡𝑖𝑐 ∶ 𝑡 = 𝑆⁄ = = = 2.98
√𝑛 8.4/√25 8.4
5. Conclusion: Value of t for 24 degrees of freedom at 0.025 is 2.064 (from t – table)
|𝑡| > 2.067, ∴ 𝑅𝑒𝑗𝑒𝑐𝑡 𝐻0.
3. A mechanic is making engine parts with axle diameter of 0.700 inch. A random sample of 10
parts shows a mean diameter of 0.742 inch with a S.D. of 0.040 inch. Compute the statistic
you would use to test whether the work is meeting the specifications.
Sol: 𝑛 = 10 < 30, 𝑥̅ = 0.742,
𝑆. 𝐷. 𝑜𝑓 𝑝𝑜𝑝𝑢𝑙𝑎𝑡𝑖𝑜𝑛 𝜎 = 0.040
𝑀𝑒𝑎𝑛 𝑜𝑓 𝑝𝑜𝑝𝑢𝑙𝑎𝑡𝑖𝑜𝑛 𝜇 = 0.700
1. 𝐻0 : 𝑃𝑟𝑜𝑑𝑢𝑐𝑡 𝑖𝑠 𝑐𝑜𝑛𝑓𝑖𝑟𝑚𝑖𝑛𝑔 𝑡𝑜 𝑠𝑝𝑒𝑐𝑖𝑓𝑖𝑐𝑎𝑡𝑖𝑜𝑛𝑠 (𝑜𝑟)𝜇 = 0.700
2. 𝐻1 : 𝜇1 ≠ 0.700 (two tailed)
3. 𝛼: 0.05
𝑥̅ −𝜇 0.742−0.700
4. T𝑒𝑠𝑡 𝑠𝑡𝑎𝑡𝑖𝑠𝑡𝑖𝑐 ∶ 𝑡 = 𝜎⁄ = 0.040/√10−1 = 3.15
√𝑛−1
5. Conclusion: |𝑡| = 3.15 > 𝑡0.025 = 2.26 𝑓𝑜𝑟 9 𝑑𝑒𝑔𝑟𝑒𝑒𝑠 𝑜𝑓 𝑓𝑟𝑒𝑒𝑑𝑜𝑚, ∴ 𝑅𝑒𝑗𝑒𝑐𝑡 𝐻0.
4. The mean life of a sample of 25 fluorescent light bulbs produced by a company is computed
to be 1570 hours. The company claims that the average life of the bulbs produced by a
company is 1600 hours using the level of significance of 0.05. Is the claim acceptable.
Sol: Given sample size 𝑛 = 25, 𝑠𝑎𝑚𝑝𝑙𝑒 𝑚𝑒𝑎𝑛 𝑥̅ = 1570,
𝑆. 𝐷. (𝜎) = 120
𝑀𝑒𝑎𝑛 𝑜𝑓 𝑝𝑜𝑝𝑢𝑙𝑎𝑡𝑖𝑜𝑛 𝜇 = 1600
Degrees of freedom = 𝑛 − 1 = 24
1. 𝐻0 : 𝑇ℎ𝑒 𝑐𝑙𝑎𝑖𝑚 𝑖𝑠 𝑎𝑐𝑐𝑒𝑝𝑡𝑎𝑏𝑙𝑒 𝜇 = 1600 ℎ𝑟𝑠
2. 𝐻1 : 𝜇1 ≠ 1600 (Two tail)
3. 𝛼: 0.05
𝑥̅ −𝜇 1570−1600
4. T𝑒𝑠𝑡 𝑠𝑡𝑎𝑡𝑖𝑠𝑡𝑖𝑐 ∶ 𝑡 = 𝜎⁄ = 120/24 = −1.22
√𝑛−1

Page 4
PROBABILITY & STATISTICS SMALL SAMPLE TESTS
5. Conclusion : |𝑡| = 1.22 < 𝑡0.025 = 2.06 𝑓𝑜𝑟 24 𝑑𝑒𝑔𝑟𝑒𝑒𝑠 𝑜𝑓 𝑓𝑟𝑒𝑒𝑑𝑜𝑚,
∴ 𝐴𝑐𝑐𝑒𝑝𝑡 𝑛𝑢𝑙𝑙 ℎ𝑦𝑝𝑜𝑡ℎ𝑒𝑠𝑖𝑠 𝐻0, i.e., the claim that the average life of the bulbs produced by
the company is 1600 hrs is acceptable.
5. The average breaking strength of steel rods is specified to be 18.5 thousand pounds to test the
sample of 14 rods were tested. The mean and S.D. obtained is 17.85 and 1.955. The result of
experiment significance?
Sol: Given sample size 𝑛 = 14, 𝑠𝑎𝑚𝑝𝑙𝑒 𝑚𝑒𝑎𝑛 𝑥̅ = 17.85,
𝑆. 𝐷. (𝑆) = 1.955
𝑀𝑒𝑎𝑛 𝑜𝑓 𝑝𝑜𝑝𝑢𝑙𝑎𝑡𝑖𝑜𝑛 𝜇 = 18.5
Degrees of freedom = 𝑛 − 1 = 13
1. Null hypothesis 𝐻0 : 𝑇ℎ𝑒 𝑟𝑒𝑠𝑢𝑙𝑡 𝑜𝑓 𝑡ℎ𝑒 𝑒𝑥𝑝𝑒𝑟𝑖𝑚𝑒𝑛𝑡 𝑖𝑠 𝑛𝑜𝑡 𝑠𝑖𝑔𝑛𝑖𝑓𝑖𝑐𝑎𝑛𝑡 𝜇 = 18.5
2. Alternative hypothesis 𝐻1 : 𝜇1 ≠ 18.5 (Two tail)
3. Level of significance 𝛼: 0.05
𝑥̅ −𝜇 17.85−18.5 −0.65
4. Test statistic 𝑡: 𝑡 = ⁄ = = = −1.199
𝑆 √𝑛−1 1.955/√13 0.542
|𝑡| = 1.199 ⇒ Calculated t = 1.199
5. Conclusion: Tabulated value at 5% level of significance for 13 degrees of freedom for two
tailed test = 2.16 , since calculated value < tabulated t we accept H0 at 5% level of
significance.
6. A random sample of six steel beams has a mean compressive strength of 58,392 (pounds per
square inch), with S.D. of 648 p.s.i. Use this information and the level of significance α = 0.05
to test whether the true average compressive strength of the steel from which this sample
came is 58,000 p.s.i.
Sol: Given sample size 𝑛 = 6, 𝑠𝑎𝑚𝑝𝑙𝑒 𝑚𝑒𝑎𝑛 𝑥̅ = 58392 𝑝𝑠𝑖,
𝑆. 𝐷. (𝑆) = 648𝑝𝑠𝑖
𝑀𝑒𝑎𝑛 𝑜𝑓 𝑝𝑜𝑝𝑢𝑙𝑎𝑡𝑖𝑜𝑛 𝜇 = 58000
Degrees of freedom = 𝑛 − 1 = 5
1. Null hypothesis 𝐻0 : 𝜇 = 58000
2. Alternative hypothesis 𝐻1 : 𝜇1 ≠ 58000 (Two tail)
3. Level of significance 𝛼: 0.05
𝑥̅ −𝜇 58392−58000
4. Test statistic 𝑡: 𝑡 = ⁄ = = 1.35
𝑆 √𝑛−1 648/√5
|𝑡| = 1.35
Calculated t = 1.35
5. Conclusion : Tabulated value = 3.365
Since calculated value < tabulated t we accept H 0.
PROBLEMS RELATED TO STUDENTS T – TEST
(WHEN S.D. OF SAMPLE IS NOT GIVEN DIRECTLY) :
1. A random sample of 10 boys has the following I.Qs 70, 120, 110, 101, 88, 83, 95, 98, 107,
100. (a) Do these data support the assumption of population mean IQ of 100.
(b) Find reasonable range in which most of the mean IQ values of samples of 10 boys lie.
Sol: (a) Here S.D. and mean of the sample is not given directly.
Given data 70, 120, 110, 101, 88, 83, 95, 98, 107, 100.
∑𝑥 972
Mean 𝑥̅ = 𝑛 = 10 = 97.2
𝒙 𝒙−𝒙 ̅ ̅)𝟐
(𝒙 − 𝒙
70 -27.2 739.84
120 22.8 519.84

Page 5
PROBABILITY & STATISTICS SMALL SAMPLE TESTS
110 12.8 163.84
101 3.8 14.44
88 -9.2 84.64
83 -14.2 201.64
95 -2.2 4.84
98 0.8 0.64
107 9.8 96.04
100 2.8 7.84
1833.60
∑(𝑥𝑖 −𝑥̅ )2 1833.60
𝑆2 = = = 203.73
𝑛−1 9
𝑆 = √203.73 = 14.27
1. Null hypothesis 𝐻0 : 𝜇 = 100
2. Alternative hypothesis 𝐻1 : 𝜇1 ≠ 100 (Two tail)
3. Level of significance 𝛼: 0.05
𝑥̅ −𝜇 97.2−100
4. Test statistic 𝑡: 𝑡 = 𝑆⁄ = 14.27/√10 = −0.62
√𝑛
|𝑡| = 0.62 ⇒ Calculated t = 0.62
5. Conclusion : Tabulated value of t for 9 degrees of freedom at 5% level of significance is
2.26 (two tailed test) . Since calculated value < tabulated t we accept H 0, IQ μ = 100.
(b) Confidence limits
𝑆
𝑥̅ ± 𝑡𝛼 . = 97.2 ± 2.26 × 4.512 = (97.2 ± 10.198) = (87, 107.4)
√𝑛

2. The height of 10 males of given locality are found to be 70, 67, 62, 68, 61, 68, 70, 64, 64, 66
inches. Is it reasonable to believe that the average height is greater than 64 inches, Test at 5%
significance level assuming that for 9 degrees of freedom.
∑𝑥 660
Sol: Mean 𝑥̅ = 𝑛 = 10 = 66

𝒙 𝒙−𝒙 ̅ (𝒙 − 𝒙̅)𝟐
70 4 16
67 1 1
62 -4 16
68 2 4
61 -5 25
68 2 4
70 4 16
64 -2 4
64 2 4
66 0 0
660 90
∑(𝑥𝑖 −𝑥̅ )2 90
𝑆2 = = = 10
𝑛−1 9
𝑆 = √10 = 3.1
1. Null hypothesis 𝐻0 : 𝐴𝑣𝑒𝑟𝑎𝑔𝑒 ℎ𝑒𝑖𝑔ℎ𝑡 𝑖𝑠 𝑛𝑜𝑡 𝑔𝑟𝑒𝑎𝑡𝑒𝑟 𝑡ℎ𝑎𝑛 64 𝑖𝑛𝑐ℎ𝑒𝑠 𝜇 = 64
2. Alternative hypothesis 𝐻1 : 𝜇 > 64 (One tail)
3. Level of significance 𝛼: 0.05
𝑥̅ −𝜇 66−64 2
4. Test statistic 𝑡: 𝑡 = 𝑆⁄ = 3.16/√9 = 1.05 = 1.9
√𝑛

Page 6
PROBABILITY & STATISTICS SMALL SAMPLE TESTS
|𝑡| = 1.9
Calculated t = 1.9
5. Conclusion : Tabulated value of t for 9 degrees of freedom at 5% level of significance for
single test is 1.833. Since calculated value > tabulated t we reject H0.
3. A random sample from a company's very extensive files shares that the order for certain kind
of machinery were filled respectively in 10, 12, 19, 14, 15, 18, 11 and 13 days use the level of
significance α = 0.01 to test claim that on the average such orders are filled in 10.5 days.
∑𝑥 112
Sol: n = 8, Mean 𝑥̅ = = = 14
𝑛 8
2 ∑(𝑥𝑖 −𝑥̅ )2 1
𝑆 = = 7 ((10 − 14)2 + (12 − 14)2 + (19 − 14)2 + ⋯ . +(13 − 14)2 )
𝑛−1
1 1
= 7 (16 + 4 + 25 + ⋯ . +1 + 16) = 7 (72) = 10.286
𝑆 = √10.286 = 3.207
1. Null hypothesis 𝐻0 : 𝜇 = 10.5 𝑑𝑎𝑦𝑠
2. Alternative hypothesis 𝐻1 : 𝜇 ≠ 10.5
3. Level of significance 𝛼: 0.01
𝑥̅ −𝜇 14−10.5
4. Test statistic 𝑡: 𝑡 = 𝑆⁄ = 3.207/√7 = 3.087
√𝑛
|𝑡| = 3.087
Calculated t = 3.087
5. Conclusion : Tabulated value of t for 7 degrees of freedom at 1% level of significance is
3.499. Since calculated value < tabulated t we accept H0.
STUDENTS T – TEST FOR DIFFERENCE OF MEANS:
To test the significant difference between the sample means 𝑥̅ 𝑎𝑛𝑑 𝑦̅ of two independent
samples of sizes 𝑛1 𝑎𝑛𝑑 𝑛2 with the same variance, we use statistic
𝑥̅ −𝑦̅ 𝑥̅ −𝑦̅
𝑡= 1 1
(𝑜𝑟) 1 1
√𝑆 2(𝑛 +𝑛 ) 𝑆√(𝑛 +𝑛 )
1 2 1 2
1 𝑛1 1 𝑛2
𝑥̅ = 𝑛 ∑𝑖=1 𝑥𝑖 , 𝑦̅ = 𝑛 ∑𝑖=1 𝑦𝑖
1 2
1 1
𝑆 = 𝑛 +𝑛 −2 [∑𝑖(𝑥𝑖 − 𝑥̅ )2 + ∑𝑖 (𝑦𝑖
2
− 𝑦̅)2 ] (𝑜𝑟) 𝑆 2 = 𝑛 [(𝑛1 − 1)𝑆12 + (𝑛2 − 1)𝑆22 ]
1 2 1 +𝑛2 −2
T – distribution follows (𝑛1 + 𝑛2 − 2) degrees of freedom.
PROBLEMS:
1. Samples of two types of electric light bulbs were tested for length of life and following data
were obtained.
Type I Type II
Sample size 𝑛1 = 8 𝑛2 = 7
Sample mean 𝑥̅ = 1234 ℎ𝑟𝑠 𝑦̅ = 1036 ℎ𝑜𝑢𝑟𝑠
Sample S.D. 𝑆1 = 36 ℎ𝑟𝑠 𝑆2 = 40 ℎ𝑟𝑠
Is the difference in the means sufficient to warrant that type I is superior to type II regarding
length of life.
Sol: 1. 𝐻0 : 𝜇1 = 𝜇2
2. 𝐻1 : 𝜇1 > 𝜇2 (Two tail)
3. 𝛼: 5%
4. Test statistic 𝑡:
1 1
𝑆 2 = 𝑛 +𝑛 −2 [(𝑛1 − 1)𝑆12 + (𝑛2 − 1)𝑆22 ] = 8+7−2 [7(36)2 + 6(40)2] = 1436.31
1 2

Page 7
PROBABILITY & STATISTICS SMALL SAMPLE TESTS
𝑥̅ −𝑦̅ 1234−1036
𝑡= 1 1
= 1 1
= 10.09
√𝑆 2(𝑛 +𝑛 ) √1436( + )
1 2 8 9
5. Conclusion: 𝐷. 𝐹 = 8 + 7 − 2 = 13; |𝑡| > 1.77 (0.05 𝑙𝑒𝑣𝑒𝑙 − 1.77) 𝑓𝑟𝑜𝑚 𝑡𝑎𝑏𝑙𝑒
2. Two independent samples have following values
Sample I 11 11 13 11 15 9 12 14
Sample II 9 11 10 13 9 8 10 -
Test the difference between the means.
96 70
Sol: 𝑛1 = 8, 𝑛2 = 7, 𝑥̅ = = 12, 𝑦̅ = = 10
8 7
𝒙 𝒙−𝒙 ̅ (𝒙 − 𝒙̅)𝟐 𝒚 𝒚−𝒚 ̅ (𝒚 − 𝒚̅ )𝟐
11 -1 1 9 -1 1
11 -1 1 11 1 1
13 1 1 10 0 0
11 -1 1 13 3 9
15 3 9 9 -1 1
9 -3 9 8 -2 4
12 0 0 10 0 0
14 2 4
26 16
2 1 2 2 1 42
𝑆 = 𝑛 +𝑛 −2 [∑𝑖(𝑥𝑖 − 𝑥̅ ) + ∑𝑖 (𝑦𝑖 − 𝑦̅) ] = 8+7−2 [26 + 16] = 13 = 3.23
1 2
1. 𝐻0 : 𝜇1 = 𝜇2
2. 𝐻1 : 𝜇1 ≠ 𝜇2
3. 𝛼 = 0.05
𝑥̅ −𝑦̅
4. Test statistic : 𝑡 = 1 1
= 2.15
𝑆√( + )
𝑛1 𝑛2

5. Conclusion: 𝐴𝑡 13 𝑑. 𝑓. 0.05 𝑙𝑒𝑣𝑒𝑙 𝑜𝑓 𝑠𝑖𝑔𝑛𝑖𝑓𝑖𝑐𝑎𝑛𝑐𝑒 2.15 < 2.16, therefore accept H0.
3. Two horses A and B were tested according to the time (in seconds) to run a particular track
with the following results
Horse A 28 30 32 33 33 29 34
Horse B 29 30 30 24 27 29
Test whether the two horses have to same running capacity.
Sol: 𝑛1 = 7, 𝑛2 = 6,
We first compute the sample means and S.D.
1
𝑥̅ = 𝑚𝑒𝑎𝑛 𝑜𝑓 𝑓𝑖𝑟𝑠𝑡 𝑠𝑎𝑚𝑝𝑙𝑒 = 7 (219) = 31.286,
169
𝑦̅ = 𝑚𝑒𝑎𝑛 𝑜𝑓 𝑠𝑒𝑐𝑜𝑛𝑑 𝑠𝑎𝑚𝑝𝑙𝑒 = = 28.16
6
𝒙 𝒙−𝒙 ̅ (𝒙 − 𝒙 ̅)𝟐 𝒚 𝒚−𝒚 ̅ (𝒚 − 𝒚̅ )𝟐
28 -3.286 10.8 29 0.84 0.7056
30 -1.286 1.6538 30 1.84 3.3856
32 0.714 0.51 30 1.84 3.3856
33 1.714 2.94 24 -4.16 17.3056
33 1.714 2.94 27 -1.16 1.3456
29 -2.286 5.226 29 0.84 0.7056
34 2.714 7.366
219 31.4358 169 26.8336
2 1 2 2 1
𝑆 = 𝑛 +𝑛 −2 [∑𝑖(𝑥𝑖 − 𝑥̅ ) + ∑𝑖 (𝑦𝑖 − 𝑦̅) ] = 11 [31.4358 + 26.8336]
1 2

Page 8
PROBABILITY & STATISTICS SMALL SAMPLE TESTS
1
= 11 (58.2694) = 5.23 ⇒ 𝑆 = 2.3
1. 𝐻0 : 𝜇1 = 𝜇2
2. 𝐻1 : 𝜇1 ≠ 𝜇2
3. 𝛼 = 0.05
𝑥̅ −𝑦̅ 31.286−28.16
4. Test statistic : 𝑡 = 1 1
= 1 1
= 2.443
𝑆√( + ) (2.3)√(7+6)
𝑛1 𝑛2

5. Conclusion: Tabulated value of t for 7 + 6 – 2 = 11 d.f. at 5% level of significance is 2.201.

2.443 > 2.201 hence reject H0.
4. To examine the hypothesis that the husbands are more intelligent than the wives. An
investigator took a sample of 10 couples and administered them a test which measures the IQ
the results are as follows:
Husbands 117 105 97 105 123 109 86 78 103 107
Wives 106 98 87 104 116 95 90 69 108 85
Test the hypothesis with a reasonable test at the level of significance of 0.05.
Sol: 𝑛1 = 10, 𝑛2 = 10,
1030 958
We first compute the sample means and S.D. 𝑥̅ = 10 = 103, 𝑦̅ = 10 = 95.8

𝒙 𝒙−𝒙 ̅ (𝒙 − 𝒙 ̅)𝟐 𝒚 𝒚−𝒚 ̅ (𝒚 − 𝒚̅ )𝟐

117 14 196 106 10.2 104.04
105 2 4 98 2.2 4.84
97 -6 36 87 -8.8 77.84
105 2 4 104 8.2 67.24
123 20 400 116 20.2 408.04
109 6 36 95 -0.8 0.64
86 -17 289 90 -5.8 33.64
78 -25 625 69 -26.8 718.24
103 0 0 108 12.2 148.84
107 4 16 85 -10.8 116.64
1030 1600 958 1679.6
2 1 2 2 1
𝑆 = 𝑛 +𝑛 [∑𝑖(𝑥𝑖 − 𝑥̅ ) + ∑𝑖 (𝑦𝑖 − 𝑦̅) ] = [1606 + 1679.6]
1 2 −2 18
1
= 18 (3285.6) = 182.53 ⇒ 𝑆 = 13.51
1. 𝐻0 : 𝜇1 = 𝜇2
2. 𝐻1 : 𝜇1 > 𝜇2 (ℎ𝑢𝑠𝑏𝑎𝑛𝑑𝑠 𝑎𝑟𝑒 𝑚𝑜𝑟𝑒 𝑖𝑛𝑡𝑒𝑙𝑙𝑖𝑔𝑒𝑛𝑡 𝑡ℎ𝑎𝑛 𝑤𝑖𝑣𝑒𝑠)(𝑜𝑛𝑒 𝑡𝑎𝑖𝑙𝑒𝑑)
3. 𝛼 = 0.05
𝑥̅ −𝑦̅ 103−95.8
4. Test statistic : 𝑡 = 1 1
= 1 1
= 1.19168
𝑆√(𝑛 +𝑛 ) (13.51)√(10+10)
1 2
5. Conclusion : Tabulated value of t for 18 d.f. at 0.05 level of significance is 1.734.
1.19168 < 1.734 hence accept H0. There is no difference in IQs .
5. To compare two kinds of bumper guards 6 of each kind were mounted on a car and the car
was run into a concrete wall. The following are costs of repair.
Guard 1 107 148 123 165 102 119
Guard 2 134 115 112 151 133 129
Use 0.01 level of significance to test whether the difference between two sample means is
significant.
Page 9
PROBABILITY & STATISTICS SMALL SAMPLE TESTS
Sol: 𝑛1 = 6, 𝑛2 = 6,
764 774
𝑥̅ = 6 = 127.33, 𝑦̅ = ̅)𝟐 = 2989.34; ∑(𝒚 − 𝒚
= 129; ∑(𝒙 − 𝒙 ̅)𝟐 = 1010
6
1
𝑆2 = 𝑛 [∑𝑖(𝑥𝑖 − 𝑥̅ )2 + ∑𝑖 (𝑦𝑖 − 𝑦̅)2 ] = 399.39 ⇒ 𝑆 = 19.999
1 +𝑛2 −2
1. 𝐻0 : 𝜇1 = 𝜇2
2. 𝐻1 : 𝜇1 ≠ 𝜇2 (𝑡𝑤𝑜 𝑡𝑎𝑖𝑙)
3. 𝛼 = 5%
𝑥̅ −𝑦̅
4. Test statistic : 𝑡 = 1 1
= 0.1446
𝑆√(𝑛 +𝑛 )
1 2
5. Conclusion : Tabulated value of t for 10 d.f. at 1% level of significance (two tailed test) is
3.169
tcal < ttab hence accept H0.
6. Find the maximum difference that we can expect with probability 0.95 between the means of
samples of sizes 10 and 12 from a normal population if their S.D. are found to be 2 and 3.
Sol: 𝑛1 = 10, 𝑛2 = 12, 𝑆1 = 2, 𝑆2 = 3
1 1 1
𝑆 2 = 𝑛 +𝑛 −2 [(𝑛1 − 1)𝑆12 + (𝑛2 − 1)𝑆22 ] = 10+12−2 [9(2)2 + 11(3)2] = 20 (135) = 6.75
1 2
⇒ 𝑆 = √6.75 = 2.6
1. 𝐻0 : 𝜇1 = 𝜇2
2. 𝐻1 : 𝜇1 ≠ 𝜇2
3. 𝛼 = 0.05
𝑥̅ −𝑦̅ 1 1 1 1
4. Test statistic : 𝑡 = 1 1
⇒ 𝑥̅ − 𝑦̅ = |𝑡| × 𝑆√(𝑛 + 𝑛 ) = 2.08 × 2.6 × √2 + 3
𝑆√(𝑛 +𝑛 ) 1 2
1 2
2.08 𝑖𝑠 𝑡ℎ𝑒 𝑡𝑎𝑏𝑙𝑒 𝑣𝑎𝑙𝑢𝑒 𝑓𝑜𝑟 20 𝑑. 𝑓. 𝑎𝑡 5% 𝑙𝑒𝑣𝑒𝑙 𝑜𝑓 𝑠𝑖𝑔𝑛𝑖𝑓𝑖𝑐𝑎𝑛𝑐𝑒.
𝑥̅ − 𝑦̅ = 2.32.
Hence maximum difference between means = 2.32.
7. Measuring specimen of nylon yarn, taken from two machines, it was found that 8 specimens
from first machine has a mean denier of 9.67 with S.D. of 1.81 while 10 specimens from
second machine had a mean denier of 7.43 with S.D. of 1.48 assuming that the proportions are
normal test the hypothesis 𝐻0 : 𝜇1 − 𝜇2 = 1.5 against 𝐻1 : 𝜇1 − 𝜇2 > 1.5 𝑎𝑡 0.05 level of
significance.
Sol: 𝑛1 = 8, 𝑛2 = 10, 𝑥̅ = 9.67, 𝑦̅ = 7.43, 𝑆1 = 1.81, 𝑆2 = 1.48
1 42.64
𝑆 2 = 𝑛 +𝑛 −2 [(𝑛1 − 1)𝑆12 + (𝑛2 − 1)𝑆22 ] = 16 = 2.665
1 2
⇒ 𝑆 = √2.665 = 1.63
1. 𝐻0 : 𝜇1 − 𝜇2 = 1.5(𝛿)
2. 𝐻1 : 𝜇1 ≠ 𝜇2 > 1.5(𝛿)
3. 𝛼 = 0.05
𝑥̅ −𝑦̅−𝛿 (9.67−7.43)−(1.5) 0.74
4. Test statistic : 𝑡 = 1 1
= 1 1
= 0.7732 = 0.96
𝑆√(𝑛 +𝑛 ) 1.63√(8+10)
1 2
5. Conclusion: Table value for 16 d.f. at 5% level of significance is 1.746
𝑡𝑐𝑎𝑙 < 𝑡𝑡𝑎𝑏𝑙𝑒 , 𝐻𝑒𝑛𝑐𝑒 𝑎𝑐𝑐𝑒𝑝𝑡 𝐻0 .
PAIRED SAMPLE T – TEST:
When the two samples are not independent paired t – test is applied for n paired observation
by taking differences 𝑑1 , 𝑑2 , … . . 𝑑𝑛 of the paired data. To test whether the differences 𝑑1 from
a random sample from a population with mean μ we use statistic

Page 10
PROBABILITY & STATISTICS SMALL SAMPLE TESTS
𝑑̅ −𝜇 1
𝑡 = 𝑆⁄ 𝑤ℎ𝑒𝑟𝑒 𝑑̅ = 𝑛 ∑ 𝑑𝑖
√𝑛
1 1 (∑ 𝑑) 2
2
𝑆 = ∑𝑛𝑖=1(𝑑𝑖 − 𝑑̅ )2 𝑜𝑟 𝑆 2 = [∑ 𝑑 2 − ]
𝑛−1 𝑛−1 𝑛
Above statistic follows students t – distribution with n – 1 degrees of freedom.
NOTE:
Take mean of differences 𝜇 = 0 𝑎𝑠 𝐻0 𝑎𝑛𝑑 𝜇 > 0 𝑎𝑠 𝐻1
PROBLEMS:
1. Ten workers were given training program with a view to study their assembly time for a
certain mechanism. The result of the time and motion studies before and after training
programme are given below.
Workers 1 2 3 4 5 6 7 8 9 10
X1 15 18 20 17 16 14 21 19 13 22
Y1 14 16 21 10 15 18 17 16 14 20
𝑋1 = Time taken for assembling before training
𝑌1 = Time taken for assembling after training
Test whether there is significant difference is assembly times before and after training.
Sol: Let μ be the mean of population of differences.
1. 𝐻0 : 𝜇 = 0 (𝑡𝑟𝑎𝑖𝑛𝑖𝑛𝑔 𝑖𝑠 𝑛𝑜𝑡 𝑢𝑠𝑒𝑓𝑢𝑙)
2. 𝐻1 : 𝜇 > 0
3. 𝛼: 0.05
4. Test statistic :
Computation Differences (before and after training)
1, 2, -1, 7, 1, -4, 4, 3, -1, 2
14
𝑑̅ = 10 = 1.4
1 1
𝑆2 = ∑𝑛𝑖=1(𝑑𝑖 − 𝑑̅ )2 = (82.4) = 9.1555
𝑛−1 9
𝑆 = 3.026
𝑑̅ −𝜇 1.4−0
𝑡 = 𝑆⁄ = 3.026⁄
√𝑛 √10
⇒ |𝑡| = 1.46
5. Conclusion : Table value with 9 d.f. at 0.05 is 1.833
𝑡𝑐𝑎𝑙 < 𝑡𝑡𝑎𝑏 ⇒ 𝑎𝑐𝑐𝑒𝑝𝑡 𝐻0 i.e., training is not useful.
2. Scores obtained in a shooting competition by 10 soldiers before and after intensive training
are given below:
Before 67 24 57 63 54 56 68 33 43
After 70 38 58 58 67 68 75 42 38
Test whether the intensive training is useful at 0.05 level of significance.
Sol: 1. 𝐻0 : 𝜇 = 0 (𝑡𝑟𝑎𝑖𝑛𝑖𝑛𝑔 𝑖𝑠 𝑛𝑜𝑡 𝑢𝑠𝑒𝑓𝑢𝑙)
2. 𝐻1 : 𝜇 > 0 (𝑡𝑟𝑎𝑖𝑛𝑖𝑛𝑔 𝑢𝑠𝑒𝑓𝑢𝑙)
3. 𝛼: 0.05
4. Test statistic :
Computation Differences -3, -14, -1, -3, 7, -13, -12, -7, -9.5
1 50
𝑑̅ = ∑10 𝑑 = − = −5
10 𝑖=1 𝑖 10
1 1
𝑆2 = ∑𝑛𝑖=1(𝑑𝑖 − 𝑑̅ )2 = (482) = 53.555
𝑛−1 9
𝑆 = 7.32
𝑑̅ −𝜇 −5−0
Test statistic 𝑡 = 𝑆⁄ = 7.32⁄ = −2.16
√𝑛 √10

Page 11
PROBABILITY & STATISTICS SMALL SAMPLE TESTS
⇒ |𝑡| = 2.16
5. Conclusion: Table value with 9 d.f. at 0.05 is 1.83
𝑡𝑐𝑎𝑙 > 𝑡𝑡𝑎𝑏 ⇒ 𝑟𝑒𝑗𝑒𝑐𝑡 𝐻0 i.e., intensive training is useful.
3. Blood pressure of 5 women before and after intake of a certain drug are given below.
Before 110 120 125 132 125
After 120 118 125 136 121
Test whether there is significant change in blood pressure at 1% level of significance.
Sol: Differences -10, 2, 0, -4, 4
𝑑̅ = 2
𝑆 2 = 30 ⇒ 𝑆 = √30
1. 𝐻0 : 𝜇 = 0
2. 𝐻1 : 𝜇 > 0 (𝑜𝑛𝑒 𝑡𝑎𝑖𝑙)
3. 𝛼 ∶ 0.05
𝑑̅ −𝜇
4. Test statistic : 𝑡 = 𝑆⁄ = 0.82
√𝑛
5. Conclusion: t – table value for 4 d.f. is 4.6
𝑡𝑐𝑎𝑙 ≤ 𝑡𝑡𝑎𝑏𝑙𝑒 , accept 𝐻0 .

SNEDECOR'S F – TEST OF SIGNIFICANCE:

When testing the significance of the differences of the means of two samples came from same
population or from populations with equal variances. If the variance of proportions are not
equal a significant difference in the means may arise. Before we apply t – test for the
significance of the difference of two means, we have to test for the equality of population
variances using F – test of significance.
If 𝑆12 𝑎𝑛𝑑 𝑆22 are the variances of two samples of sizes 𝑛1 𝑎𝑛𝑑 𝑛2 respectively then
population variances are given by
𝑛1 𝜎12 = (𝑛1 − 1)𝑆12 𝑎𝑛𝑑 𝑛2 𝜎22 = (𝑛2 − 1)𝑆22
𝑣1 = 𝑛1 − 1, 𝑣2 = 𝑛2 − 1 degrees of freedom of these estimate.
We want to test if these estimates 𝑆12 𝑎𝑛𝑑 𝑆22 are significantly different or if the samples may
be regarded as drawn from the same populations or from two populations with same variance
𝜎2.
In this case set up null hypothesis 𝐻0 : 𝜎12 = 𝜎22 , i.e., population variances are same.
𝑆2
Test statistic is 𝐹 = 𝑆12 [𝑆12 , 𝑆22 ]
2
∑(𝑥𝑖 −𝑥̅ )2 ∑(𝑦𝑖 −𝑦̅)2
𝑆12 = , 𝑆22 =
𝑛1 −1 𝑛2 −1
𝐺𝑟𝑒𝑎𝑡𝑒𝑟 𝑣𝑎𝑟𝑖𝑎𝑛𝑐𝑒
𝐹 = 𝑆𝑚𝑎𝑙𝑙𝑒𝑟 𝑣𝑎𝑟𝑖𝑎𝑛𝑐𝑒
PROBLEMS:
1. In one sample of 8 observations the sum of the squares of deviations of the sample value from
the sample mean was 84.4 and in other sample of 10 observations it was 102.6. Test whether
this difference is significant at 5% level.
Sol: Here 𝑛1 = 8, 𝑛2 = 10
∑(𝑥𝑖 − 𝑥̅ )2 = 84.4; ∑(𝑦𝑖 − 𝑦̅)2 = 102.6
∑(𝑥𝑖 −𝑥̅ )2 84.4
𝑆12 = = = 12.057
𝑛1 −1 7

Page 12
PROBABILITY & STATISTICS SMALL SAMPLE TESTS
∑(𝑦𝑖 −𝑦̅)2 102.6
𝑆22 = = = 11.4
𝑛2 −1 9
1.𝐻0 : 𝑆12 = 𝑆22
2.𝐻1 : 𝑆12 ≠ 𝑆22
3. 𝛼 ∶ 5%
𝑆2 12.057
Now 𝐹 = 𝑆12 = = 1.057
2 11.4
i.e., calculated F = 1.057
Tabulated value of F at 5% level for (7, 9) degrees of freedom is 3.29
𝐹𝑐𝑎𝑙 < 𝐹𝑡𝑎𝑏 𝑖. 𝑒. , 𝑎𝑐𝑐𝑒𝑝𝑡 𝐻0.
2. Two random samples gave the following results
Sample Size Sample mean Sum of squares of deviation from the mean
1 10 15 90
2 12 14 108
Test whether the samples came from the same normal population.
Sol: 𝑛1 = 10, 𝑛2 = 12, 𝑥̅ = 15, 𝑦̅ = 14
∑(𝑥𝑖 − 𝑥̅ )2 = 90; ∑(𝑦𝑖 − 𝑦̅)2 = 108
To test whether the samples drawn from same normal population.
i.e., 𝐻0 : 𝜇1 = 𝜇2 𝑎𝑛𝑑 𝜎12 = 𝜎22
Here we have to test
i) Test equality of variances by F - test
ii) Test equality of means by t – test
i) F – test (equality of variances)
1. 𝐻0 : 𝜎12 = 𝜎22
2. 𝐻1 : 𝜎12 ≠ 𝜎22
3. 𝛼 ∶ 5%
4. Test statistic :
Given 𝑛1 = 10, 𝑛2 = 12, ̅̅̅ 𝑋1 = 15, 𝑌̅1 = 14, ∑(𝑥𝑖 − 𝑥̅ )2 = 90, ∑(𝑦𝑖 − 𝑦̅)2 = 108
∑(𝑥𝑖 −𝑥̅ )2
𝑆12 = = 10
𝑛1 −1
∑(𝑦𝑖 −𝑦̅)2
𝑆22 = = 9.82
𝑛2 −1
𝑆12 10
𝐹= 𝑆22
= 9.82 = 1.018
i.e., calculated F = 1.018
5. Conclusion: Tabulated value of F at 5% level for (9, 11) degrees of freedom is 2.90
𝐹𝑐𝑎𝑙 < 𝐹𝑡𝑎𝑏 𝑖. 𝑒. , 𝑎𝑐𝑐𝑒𝑝𝑡 𝐻0. i.e., sample came from same normal population with same
variance.
ii) t – test : (to test equality of means)
Given 𝑛1 = 10, 𝑛2 = 12, ̅̅̅ 𝑋2 = 15, 𝑌̅2 = 14, ∑(𝑥𝑖 − 𝑥̅ )2 = 90, ∑(𝑦𝑖 − 𝑦̅)2 = 108
1 1
𝑆 2 = 𝑛 +𝑛 −2 [∑ (𝑥𝑖 − 𝑥̅ )2 + ∑ (𝑦𝑖 − 𝑦̅)2 ] = 20 [90 + 108] = 9.8 ⇒ 𝑆 = 3.15
1 2
1. 𝐻0 : 𝜇1 = 𝜇2
2. 𝐻1 : 𝜇1 ≠ 𝜇2
3. 𝛼: 0.05
𝑥̅ −𝑦̅ 15−14
4. Test statistic : 𝑡 = 1 1
= 1 1
= 0.74
𝑆√(𝑛 +𝑛 ) 315√10+12
1 2
5. Conclusion : Tabulated for 20 d.f. at 5% level (two tailed) is 2.086

Page 13
PROBABILITY & STATISTICS SMALL SAMPLE TESTS
𝑡𝑐𝑎𝑙 < 𝑡𝑡𝑎𝑏 ⇒ 𝑎𝑐𝑐𝑒𝑝𝑡 𝐻0 𝑖. 𝑒. , 𝜇1 = 𝜇2
Given samples have been drawn from the same normal population.
3. Pumpkins were grown under two experimental conditions. Two random samples of 11 and 9
pumpkins shown the sample S.D. of their weights as 0.8 and 0.5 respectively. Assuming that
the weight distributions are normal. Test the hypothesis that true variance are equal.
Sol: 𝑛1 = 11, 𝑛2 = 9, 𝑆1 = 0.8, 𝑆2 = 0.5
Sample S.D's are given
Population variances 𝜎12 𝑎𝑛𝑑 𝜎22 use the relation
𝑛 𝑆2 11×(0.8)2
𝜎12 = 𝑛 1−11 = = 0.704;
1 10
𝑛2 𝑆22 9×(0.5)2
𝜎22 = 𝑛 = = 0.281
2 −1 8
1.𝐻0 : 𝜎12
= 𝜎22
2.𝐻1 : 𝜎12
≠ 𝜎22
3. 𝛼 = 0.05
𝜎2 0.704
4. Test statistic : 𝐹 = 𝜎12 = 0.281 = 2.5 ⇒ 𝐹 = 2.5
2
5. Conclusion : Tabulated value of F for (10, 8) d.f. at 5% level of significance is 3.35
𝐹𝑐𝑎𝑙 < 𝐹𝑡𝑎𝑏 𝑖. 𝑒. , 𝑎𝑐𝑐𝑒𝑝𝑡 𝐻0.
4. Time taken by workers in performing a job by method I and method II is given below:
Method I 20 16 26 27 23 22 -
Method II 27 33 42 35 32 34 38
Do the data show that the variances of time distribution from population from which these
samples are drawn do not differ significantly.
Sol: Calculation of sample variance
𝑛1 = 6, 𝑛2 = 7,
∑𝑥 134 241
𝑥̅ = 𝑛 = 6 = 22.3, 𝑦̅ = 7 = 34.4
1
𝒙 𝒙−𝒙 ̅ ̅)𝟐
(𝒙 − 𝒙 𝒚 𝒚−𝒚 ̅ (𝒚 − 𝒚̅ )𝟐
20 -2.3 5.29 27 -7.4 54.76
16 -6.3 39.69 33 -1.4 1.96
26 3.7 13.69 42 7.6 57.76
27 4.7 22.09 35 0.6 0.36
23 0.7 0.49 32 -2.4 5.76
22 -0.3 0.09 34 -0.4 0.16
38 3.6 12.96
81.34 133.72
∑(𝑥𝑖 −𝑥̅ )2 81.34
𝑆12 = = = 16.26
𝑛1 −1 5
∑(𝑦𝑖 −𝑦̅)2 133.72
𝑆22 = = = 22.29
𝑛2 −1 6
1.𝐻0 : 𝜎12 = 𝜎22
2.𝐻1 : 𝜎12 ≠ 𝜎22
3. 𝛼 = 0.05
𝑆2 22.29
4. Test statistic : 𝐹 = 𝑆22 = 16.26 = 1.3699
1
⇒ 𝐹 = 1.3699
5. Conclusion: Tabulated value of F for (5, 6) d.f. at 5% level of significance is 4.95.
𝐹𝑐𝑎𝑙 < 𝐹𝑡𝑎𝑏 𝑖. 𝑒. , 𝑎𝑐𝑐𝑒𝑝𝑡 𝐻0.
Page 14
PROBABILITY & STATISTICS SMALL SAMPLE TESTS
5. The measurements of the output of two units have given the following results assuming that
both samples have been obtained from the normal population at 5% significant level test
whether the two populations have same variances.
Unit A 14.1 10.1 14.7 13.7 14.0
Unit B 14.0 14.5 13.7 12.7 14.1
Sol: 𝑛1 = 5, 𝑛2 = 5,
∑𝑥 66.6 69
𝑥̅ = 𝑛 = 5 = 13.32, 𝑦̅ = = 13.8
1 5
𝒙 𝒙−𝒙 ̅ ̅)𝟐
(𝒙 − 𝒙 𝒚 𝒚−𝒚 ̅ ̅ )𝟐
(𝒚 − 𝒚
14.1 0.78 0.6084 14.0 0.2 0.04
10.1 -3.22 10.3684 14.5 0.7 0.49
14.7 1.38 1.9044 13.7 -0.1 0.01
13.7 0.38 0.1444 12.7 -1.1 1.21
14.0 0.68 0.4624 14.1 0.3 0.09
13.488 1.84
∑(𝑥𝑖 −𝑥̅ )2 13.488
𝑆12 = = = 3.372
𝑛1 −1 4
∑(𝑦𝑖 −𝑦̅)2 1.84
𝑆22 = = = 0.46
𝑛2 −1 4
1.𝐻0 : 𝜎12 = 𝜎22
2.𝐻1 : 𝜎12 ≠ 𝜎22
3. 𝛼 = 0.05
𝑆2 3.372
4. Test statistic : 𝐹 = 𝑆12 = = 7.33
2 0.46
⇒ 𝐹 = 7.33
5. Conclusion: Tabulated value of F for (4, 4) d.f. at 5% level of significance is 6.39.
𝐹𝑐𝑎𝑙 > 𝐹𝑡𝑎𝑏 𝑖. 𝑒. , 𝑟𝑒𝑗𝑒𝑐𝑡 𝑛𝑢𝑙𝑙 ℎ𝑦𝑝𝑜𝑡ℎ𝑒𝑠𝑖𝑠 𝐻0 .
6. The nicotine contents in milligrams in the samples of tobacco were found to be as follows:
Sample A 24 27 26 21 25 -
Sample B 27 30 28 31 22 36
Can it be said that two samples came from same normal population.
Sol: 𝑛1 = 5, 𝑛2 = 6,
∑𝑥
𝑥̅ = 𝑛 = 24.6, 𝑦̅ = 29
1
̅)𝟐 = 21.2; ∑(𝒚𝒊 − 𝒚
∑(𝒙𝒊 − 𝒙 ̅)𝟐 = 108
∑(𝑥𝑖 −𝑥̅ )2 21.2
𝑆12 = = = 5.3
𝑛1 −1 4
∑(𝑦𝑖 −𝑦̅)2
𝑆22 = = 21.6
𝑛2 −1
F – test :
1. 𝐻0 : 𝜎12 = 𝜎22
2. 𝐻1 : 𝜎12 ≠ 𝜎22
3. 𝛼 = 5%
𝑆22 21.6
4. Test statistic : 𝐹 = =
𝑆12 5.3
𝐹𝑐𝑎𝑙 = 4.075
5. Conclusion: 𝐹𝑡𝑎𝑏 𝑎𝑡 (5. 4) d.f. is 6.2.
𝐹𝑐𝑎𝑙 ≤ 𝐹𝑡𝑎𝑏𝑙𝑒 , 𝐴𝑐𝑐𝑒𝑝𝑡 𝐻0.
T – test for difference of mean :

Page 15
PROBABILITY & STATISTICS SMALL SAMPLE TESTS
1. 𝐻0 : 𝜇1 = 𝜇2
2. 𝐻1 : 𝜇1 ≠ 𝜇2
3. 𝛼 = 5%
𝑥̅ −𝑦̅ 24.6−29
4. Test statistic : 𝑡 = 1 1
= 1 1
= −1.92
𝑆√ + 3.78√5+6
𝑛1 𝑛2

5. Conclusion : t – table value at 5 + 6 – 2 = 9 d.f. is 2.26; |𝑡| < 𝑡 𝑡𝑎𝑏𝑙𝑒, 𝑎𝑐𝑐𝑒𝑝𝑡 𝐻0 .

CHI – SQUARE (𝝌𝟐 ) TEST:
𝜒 2 – test is used to test the goodness of fit.
Test statistic for 𝜒 2 – test
(𝑂𝑖 −𝐸𝑖)2
𝜒 2 = ∑𝑛𝑖=1 𝑤𝑖𝑡ℎ 𝑛 − 1 𝑑𝑒𝑔𝑟𝑒𝑒𝑠 𝑜𝑓 𝑓𝑟𝑒𝑒𝑑𝑜𝑚
𝐸𝑖
𝑂 = 𝑂𝑏𝑠𝑒𝑟𝑣𝑒𝑑 𝑓𝑟𝑒𝑞𝑢𝑒𝑛𝑐𝑦
𝐸 = 𝐸𝑥𝑝𝑒𝑐𝑡𝑒𝑑 𝑓𝑟𝑒𝑞𝑢𝑒𝑛𝑐𝑦
NOTE:
If the data is given in series of n numbers then degrees of freedom = n – 1
In case of Binomial distribution d.f. = n – 1
In case of Poisson distribution d.f. = n – 2
In case of Normal distribution d.f. = n – 3
𝝌𝟐 – TEST AS A TEST OF GOODNESS OF FIT:
1. The number of automobile accidents per week in a certain community are as follows 12, 8, 20,
2, 14, 10, 15, 6, 9, 4. Are these frequencies in agreement with the belief that accident
conditions were the same during this 10 week period.
100
Sol: Expected frequency of accidents in each week = 10 = 10
1. 𝐻0 : Accident conditions were same during the 10 week period.
2. 𝐻1 ∶ Accident conditions were not same.
3. 𝛼 ∶ 5%
Observed Expected frequency (O – E) (𝑶 − 𝑬)𝟐
frequency (O) (E) 𝑬
12 10 2 0.4
8 10 -2 0.4
20 10 10 10.0
2 10 -8 6.4
14 10 4 1.6
10 10 0 0.0
15 10 5 2.5
6 10 -4 1.6
9 10 -1 0.1
4 10 6 3.6
100 100 26.6
(𝑂𝑖 −𝐸𝑖 )2
4. Test statistic 𝜒 2 = ∑𝑛𝑖=1 = 26.6
𝐸𝑖
5. Conclusion : Table value for 9 d.f. at 0.05 level of significance is 16.9
2 2
𝜒𝑐𝑎𝑙 > 𝜒𝑡𝑎𝑏 , 𝑖. 𝑒. , 𝑟𝑒𝑗𝑒𝑐𝑡 𝐻0 .
2. A sample analysis of examination result of 500 students was made. It was found that 220
students had failed, 170 had secured a third class, 90 were placed in second class and 20 got
first class. Do these figures commensurate with the general examination result which is in the
ratio 4 : 3 : 2 : 1 for the various categories respectively.
Sol: 1. 𝐻0 : The observed results commensurate with the general examination results.
Page 16
PROBABILITY & STATISTICS SMALL SAMPLE TESTS
2. 𝐻1 : Observed results not equal to examination results.
3. 𝛼 = 5%
4. Test statistic : Expected frequencies are in the ratio 4 : 3 : 2 : 1
Total frequency = 500
If we divide the total frequency 500 in the ratio 4 : 3 : 2 : 1, we get the frequencies as 200,
150, 100, 50.
Class Observed Expected frequency (O – E) (𝑶 − 𝑬)𝟐
frequency (O) (E) 𝑬
Failed 220 200 20 2.00
Third 170 150 20 2.667
Second 90 100 -10 1.000
First 20 50 -30 18.00
500 500 23.667
(𝑂𝑖 −𝐸𝑖 )2
Calculated 𝜒 2 = ∑𝑛𝑖=1 = 23.667
𝐸𝑖
5. Conclusion : Degrees of freedom n – 1 = 4 – 1 = 3
2
For 𝑣 = 3, 𝜒0.005 = 7.81
2 2
𝜒𝑐𝑎𝑙 > 𝜒𝑡𝑎𝑏 , 𝑖. 𝑒. , 𝑟𝑒𝑗𝑒𝑐𝑡 𝐻0 .
3. The following figures show the distribution of digits in numbers chosen at random from a
telephone directory.
Digits 0 1 2 3 4 5 6 7 8 9
Frequency 1026 1107 997 966 1075 933 1107 972 964 853
Test whether the digits may be taken to occur equally frequently in the directory.
Sol: 1. 𝐻0 : The digits occur equally frequently in the directory.
2. 𝐻1 ∶The digits not occur equally frequently.
3. 𝛼 = 5%
4. Test statistic : Under this null hypothesis the expected frequencies for each of the digits 0,
10000
1, 2, 3, 4, 5, 6, 7, 8, 9 is 10 = 1000
Digits Observed Expected frequency (O – E)2 (𝑶 − 𝑬)𝟐
frequency (O) (E) 𝑬
0 1026 1000 676 0.676
1 1107 1000 11449 11.449
2 997 1000 9 0.009
3 966 1000 1156 1.156
4 1075 1000 5625 5.625
5 933 1000 4489 4.489
6 1107 1000 11449 11.449
7 972 1000 784 0.784
8 964 1000 1296 1.296
9 853 1000 21609 21.609
10000 10000 58.542
(𝑂𝑖 −𝐸𝑖 )2
Calculated 𝜒 2 = ∑𝑛𝑖=1 = 58.542
𝐸𝑖
5. Conclusion : Degrees of freedom n – 1 = 9 at 5% is 16.919
2 2
𝜒𝑐𝑎𝑙 < 𝜒𝑡𝑎𝑏 , 𝑖. 𝑒. , 𝑎𝑐𝑐𝑒𝑝𝑡 𝐻0.
4. A dice is thrown 264 times with the following results. Show that the dice is biased (Given
𝜒 2 = 11.07 𝑓𝑜𝑟 5 𝑑. 𝑓)
No. appeared on the dice 1 2 3 4 5 6
Frequency 40 32 28 58 54 52
Sol: 1. 𝐻0 : The dice is unbiased.

Page 17
PROBABILITY & STATISTICS SMALL SAMPLE TESTS
2. 𝐻1 ∶ The dice is biased
3. 𝛼 = 5%
264
4. Test statistic : The expected frequency of each of the numbers 1, 2, 3, 4, 5, 6 is = 44
6
Observed Expected frequency (O – E)2 (𝑶 − 𝑬)𝟐
frequency (O) (E) 𝑬
40 44 16 0.3636
32 44 144 3.2727
28 44 256 5.8181
58 44 196 4.4545
54 44 100 2.2727
52 44 64 1.4545
264 264 17.6362
(𝑂𝑖 −𝐸𝑖 )2
Calculated 𝜒 2 = ∑𝑛𝑖=1 = 17.6362
𝐸𝑖
2 2
5. Conclusion : Degrees of freedom n – 1 = 5 at 5% is 11.07; 𝜒𝑐𝑎𝑙 > 𝜒𝑡𝑎𝑏 , 𝑖. 𝑒. , 𝑟𝑒𝑗𝑒𝑐𝑡 𝐻0 .
5. A pair of dice are thrown 360 times and the frequency of each sum is indicated below:
Sum 2 3 4 5 6 7 8 9 10 11 12
Frequency 8 24 35 37 44 65 51 42 26 14 14
Would you say that the dice are fair on the basis of the Chi Square test at 0.05 level of
significance.
Sol: 1. 𝐻0 ∶ 𝑑𝑖𝑐𝑒 𝑎𝑟𝑒 𝑓𝑎𝑖𝑟
2. 𝐻1 : 𝑑𝑖𝑐𝑒 𝑖𝑠 𝑛𝑜𝑡 𝑓𝑎𝑖𝑟
3. 𝛼 = 0.05
4. Test statistic :
(x) 2 3 4 5 6 7 8 9 10 11 12
P(x) 1/36 2/36 3/36 4/36 5/36 6/36 5/36 4/36 3/36 2/36 1/36

Sum Observed Expected frequency (O – E)2 (𝑶 − 𝑬)𝟐

frequency (O) (E) 𝑬
2 8 10 4 0.4
3 24 20 16 0.8
4 35 30 25 0.833
5 37 40 9 0.225
6 44 50 36 0.72
7 65 60 25 0.414
8 51 50 1 0.02
9 42 40 4 0.1
10 26 30 16 0.53
11 14 20 36 1.8
12 14 10 16 1.6
7.445
(𝑂𝑖 −𝐸𝑖 )2
Calculated 𝜒 2 = ∑𝑛𝑖=1 = 7.445
𝐸𝑖
5. Conclusion : Degrees of freedom n – 1 = 11 at 5% is 18.3
2 2
𝜒𝑐𝑎𝑙 < 𝜒𝑡𝑎𝑏 , 𝑖. 𝑒. , 𝑎𝑐𝑐𝑒𝑝𝑡 𝐻0.
6. Fit a poisson distribution to the following data and test the goodness of the fit at 5% level of
significance.
(x) 0 1 2 3 4
(f) 214 92 20 3 1

Page 18
PROBABILITY & STATISTICS SMALL SAMPLE TESTS
Sol: Given data:
(x) 0 1 2 3 4
(f) 214 92 20 3 1 ∑ 𝒇 =330
∑ 𝑓𝑥 0×214+1×92+2×20+3×3+4×1 145
Mean 𝜇 = ∑𝑓
= = 330 = 0.439
330
Mean 𝜇 = 𝜆 = 0.439
𝑒 −𝜆 𝜆𝑥
𝑃(𝑥) = 𝑥!
𝑒 −𝜆 𝜆0
𝑃(0) = = 𝑒 −𝜆 = 𝑒 −0.439 = 0.6446
0!
Recurrence formula for Poisson distribution
𝜆
𝑃(𝑥 + 1) = 𝑥+1 𝑃(𝑥)
𝜆
For 𝑥 = 0, 𝑃(1) = 1 𝑃(0) = 0.6446 × 0.439 = 0.28297
0.439 0.439
For 𝑥 = 1, 𝑃(2) = 𝑃(1) = × 0.2829 = 0.06211
2 2
0.439 0.439
For 𝑥 = 2, 𝑃(3) = 𝑃(2) = × 0.06211 = 0.0091
3 3
0.439 0.439
For 𝑥 = 3, 𝑃(4) = 4 𝑃(3) = × 0.0091 = 0.0009
4
N = 330
1. 𝐻0 ∶ Fitting is good
2. 𝐻1 ∶ Fitting is not good
3. 𝛼 = 5%
4. Test statistic :
x Observed Expected frequency (O – E) (𝑶 − 𝑬)𝟐
frequency (O) (E) N× P(x) 𝑬
0 214 213 1 0.0046
1 92 93 -1 0.0107
2 20 20 0 0
3 3 3 0 0
4 1 1 0 0
0.015
(𝑂𝑖 −𝐸𝑖 )2
Calculated 𝜒 2 = ∑𝑛𝑖=1 = 0.00606
𝐸𝑖
5. Conclusion : 𝜒 2 table value at 3 d.f. at 5% level is 7.815
2 2
𝜒𝑐𝑎𝑙 < 𝜒𝑡𝑎𝑏 , 𝑖. 𝑒. , 𝑎𝑐𝑐𝑒𝑝𝑡 𝐻0.
CHI – SQUARED TEST FOR INDEPENDENCE OF ATTRIBUTES:
* Attributes means quality or characteristic examples of attributes are blindness, honesty, beauty
etc.
* An attribute may be marked by its presence or absence in number of a given population
* Let us consider two attributes A and B. A is divided into two classes and B is divided into two
classes.
* The various cell frequencies can be expressed in the following table.

A a B
B c D

a B a+b
c D c+d
a+c b+d N

Page 19
PROBABILITY & STATISTICS SMALL SAMPLE TESTS
Expected frequencies are given by

(𝑎 + 𝑐)(𝑎 + 𝑏) (𝑏 + 𝑑)(𝑎 + 𝑏) 𝑎+𝑏

𝐸(𝑎) = 𝐸(𝑏) =
𝑁 𝑁
(𝑎 + 𝑐)(𝑐 + 𝑑) (𝑏 + 𝑑)(𝑐 + 𝑑) 𝑐+𝑑
𝐸(𝑐) = 𝐸(𝑑) =
𝑁 𝑁
𝑎+𝑐 𝑏+𝑑 𝑁
NOTE:
In this Chi – Square test we test if two attributes A and B under consideration are independent
or not.
Null Hypothesis 𝐻0 ∶ 𝐴𝑡𝑡𝑟𝑖𝑏𝑢𝑡𝑒𝑠 𝑎𝑟𝑒 𝑖𝑛𝑑𝑒𝑝𝑒𝑛𝑑𝑒𝑛𝑡
𝜒 2 𝑑𝑖𝑠𝑡𝑟𝑖𝑏𝑢𝑡𝑖𝑜𝑛 𝑓𝑜𝑙𝑙𝑜𝑤𝑠
Degrees of freedom (𝑟 − 1)(𝑠 − 1)
𝑟 = 𝑛𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑟𝑜𝑤𝑠, 𝑠 = 𝑛𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝑐𝑜𝑙𝑢𝑚𝑛𝑠.
PROBLEMS:
1. On the basis of information given below about the treatment of 200 patients suffering from
disease, state whether the new treatment is comparatively superior to the conventional
treatment
Favourable Not favourable Total
New 60 30 90
Conventional 40 70 110
Total 100 100 200
Sol: 1. 𝐻0 : 𝑁𝑢𝑙𝑙 ℎ𝑦𝑝𝑜𝑡ℎ𝑒𝑠𝑖𝑠: No difference between new and conventional treatment (or) new
and conventional treatment are.
2. 𝐻1 : New and conventional treatment are dependent
3. 𝛼 = 0.05
4. Computation : Expected frequencies are given in table:
90 × 100 90 × 100 90
= 45 = 45
200 200
100 × 110 100 × 110 110
= 55 = 55
200 200
100 100 200

Observed Expected frequency (O – E)2 (𝑶 − 𝑬)𝟐

frequency (O) (E) 𝑬
60 45 225 5
30 45 225 5
40 55 225 4.09
70 55 225 4.09
200 200 18.18
(𝑂𝑖 −𝐸𝑖 )2
Calculated 𝜒 2 = ∑𝑛𝑖=1 = 18.18
𝐸𝑖
Degrees of freedom = (2 – 1)(2 – 1) = 1
Tabulated 𝜒 2 for 1 d.f. at 5% level of significance is 3.841.
2 2
𝜒𝑐𝑎𝑙 > 𝜒𝑡𝑎𝑏 , 𝑖. 𝑒. , 𝑟𝑒𝑗𝑒𝑐𝑡 𝐻0 .
2. Given following table for hair colour and eye colour. Find the value of 𝜒 2 . Is there good
association between the two.
Hair Colour
Fair Brown Black Total
Page 20
PROBABILITY & STATISTICS SMALL SAMPLE TESTS
Eye Colour Blue 15 5 20 40
Grey 20 10 20 50
Brown 25 15 20 60
Total 60 30 60 150
Sol: 1. 𝐻0 : 𝑁𝑢𝑙𝑙 ℎ𝑦𝑝𝑜𝑡ℎ𝑒𝑠𝑖𝑠: Two attributes hair and eye colour are independent.
2. 𝐻1 : Alternate hypothesis : Two attributes hair and eye colour are dependent.
3. 𝛼 = 0.05
4. Computation : Expected frequencies are given in table:
60 × 40 30 × 40 60 × 40 40
= 16 =8 = 16
150 150 150
60 × 50 30 × 60 60 × 50 50
= 20 = 10 = 20
150 150 150
60 × 60 30 × 60 60 × 60 60
= 24 = 12 = 24
150 150 150
60 30 60 150
Observed Expected frequency (O – E)2 (𝑶 − 𝑬)𝟐
frequency (O) (E) 𝑬
15 16 1 0.0625
5 8 9 1.125
20 16 16 1.0
20 20 0 0
10 10 0 0
20 20 0 0
25 24 1 0.042
15 12 9 0.75
20 24 16 0.666
3.6458
(𝑂𝑖 −𝐸𝑖 )2
5. Test statistic 𝜒 2 = ∑𝑛𝑖=1 = 3.6458
𝐸𝑖
Degrees of freedom = (3 – 1)(3 – 1) = 4 d.f. at 5% level of significance is 9.488.
2 2
𝜒𝑐𝑎𝑙 < 𝜒𝑡𝑎𝑏 , 𝑖. 𝑒. , 𝑎𝑐𝑐𝑒𝑝𝑡 𝐻0.
PRACTICE PROBLEM:
3. Two researchers adopted different sampling techniques while investigating some group of
students to find the number of students falling into different intelligence level. The results are
follows:
Researchers Below Average Above Genius Total
Average average
X 86 60 44 10 200
Y 40 33 25 2 100
Total 126 93 69 12 200
Would you say that sampling techniques adopted by the two researchers are significantly
different.
Sol: 𝜒 2 𝑣𝑎𝑙𝑢𝑒 = 2.0971, 𝑑. 𝑓. (2 − 1)(4 − 1) = 3 𝑑. 𝑓. = 7.82 𝑎𝑡 5%
4. Four methods are under development for making discs of super conducting material. Fifty
discs are made by each method and they are checked for super conductivity when cooled with
liquid.
I method II method III method IV method Total
Super conductors 31 42 22 25 120
Failures 19 8 28 25 80
Total 50 50 50 50 200
Page 21
PROBABILITY & STATISTICS SMALL SAMPLE TESTS
Test significant difference between the proportions of super conductivity at 0.05 level.
Sol: 1. 𝐻0 : 𝑃1 = 𝑃2 = 𝑃3 = 𝑃4
2. 𝐻1 : 𝑃1 ≠ 𝑃2 ≠ 𝑃3 ≠ 𝑃4
3. 𝛼 = 0.05
4. Computation: Table of expected frequencies
50 × 120 50 × 120 50 × 120 50 × 120 120
= 30 = 30 = 30 = 30
200 200 200 200
50 × 80 50 × 80 50 × 80 50 × 80 80
= 20 = 20 = 20 = 20
200 200 200 200
50 50 50 50 200

Observed Expected frequency (O – E)2 (𝑶 − 𝑬)𝟐

frequency (O) (E) 𝑬
31 30 1 0.03
42 30 144 4.8
22 30 64 2.13
25 30 25 0.83
19 20 1 0.05
8 20 144 7.2
28 20 64 3.2
25 20 25 1.25
19.5
(𝑂𝑖 −𝐸𝑖 )2
5. Test statistic 𝜒 2 = ∑𝑛𝑖=1 = 19.5
𝐸𝑖
Degrees of freedom = (4 – 1)(2 – 1) = 3 d.f. at 5% level of significance is 7.815.
2 2
𝜒𝑐𝑎𝑙 > 𝜒𝑡𝑎𝑏 , 𝑖. 𝑒. , 𝑟𝑒𝑗𝑒𝑐𝑡 𝐻0 .
5. From the following data find whether there is any significant liking in the habit of taking soft
drinks among categories of employees.
Soft Drinks Employees
Clerk Teachers Officers Total
Pepsi 10 25 65 100
Thumsup 15 30 65 110
Fanta 50 60 30 140
Total 75 115 160 350
Sol: 1. 𝐻0 : There is significant liking in the habit of taking soft drinks.
2. 𝐻1 : There is no significant liking in habit of taking soft drinks.
3. 𝛼 = 0.05
4. Computation: Table of expected frequencies
75 × 100 115 × 100 160 × 100
= 21.4 = 32.9 = 45.71
350 350 350
75 × 110 115 × 110 160 × 110
= 23.6 = 36.1 = 50.3
350 350 350
75 × 140 115 × 140 160 × 140
= 30 = 46 = 64
350 350 350

Observed Expected frequency (O – E)2 (𝑶 − 𝑬)𝟐

frequency (O) (E) 𝑬
10 21.4 129.96 6.073
25 32.9 62.41 1.897
65 45.7 372.49 8.151
15 23.6 73.96 3.134

Page 22
PROBABILITY & STATISTICS SMALL SAMPLE TESTS
30 36.1 37.21 1.031
65 50.3 216.09 4.3
50 30 400.0 13.333
60 46 196 4.261
30 64 1156.0 18.062
60.2425
(𝑂𝑖 −𝐸𝑖 )2
5. Test statistic 𝜒 2 = ∑𝑛𝑖=1 = 60.2425
𝐸𝑖
Degrees of freedom = (3 – 1)(3 – 1) = 4 d.f. at 5% level of significance is 9.488.
2 2
𝜒𝑐𝑎𝑙 > 𝜒𝑡𝑎𝑏 , 𝑖. 𝑒. , 𝑟𝑒𝑗𝑒𝑐𝑡 𝐻0 .

CHI – SQUARE TEST FOR POPULATION VARIANCE:

Suppose we want to test if a random sample Xi has been drawn from normal population with a
specified variance 𝜎 2 . Under the null hypothesis H0 that the population variance is 𝜎 2 test
statistic is
(𝑥𝑖 −𝑥̅ )2
𝜒 2 = ∑𝑛𝑖=1 𝜎2
𝑛𝑆 2
𝜒2 = (𝑛 − 1)𝑑. 𝑓.
𝜎2

EXAMPLE:
A firm manufacturing rivets wants to limit variations in their length as much as possible. The
lengths in (cms) of 10 rivets manufactured by new process are
2.15 1.99 2.05 2.12 2.17
2.01 1.98 2.03 2.25 1.93
Examine whether the new process can be considered superior to the old if the old population
has S.D. of 0.145 cm.
∑𝑥 20.68
Sol: We have 𝑛 = 10, 𝑥̅ = 𝑁 𝑖 = 10 = 2.068
∑(𝑥 −𝑥̅ )2 0.09096
𝑆 2 = 𝑛−1 𝑖
= 9 = 0.01010
𝜎 = 0.145
1. 𝐻0 ∶ 𝜎 2 = 𝜎02 (𝑜𝑟)𝑁𝑒𝑤 𝑝𝑟𝑜𝑐𝑒𝑠𝑠 𝑠𝑢𝑝𝑒𝑟𝑖𝑜𝑟 𝑡𝑜 𝑜𝑙𝑑
2. 𝐻1 : 𝜎 2 ≠ 𝜎02
3. 𝛼: 0.05
𝑛𝑆 2 10(0.01010)
𝜒2 = = = 4.8
𝜎2 (0.145)2
𝜒 2 table value of 9 d.f. is 16.916.
2 2
𝜒𝑐𝑎𝑙 < 𝜒𝑡𝑎𝑏 , 𝑖. 𝑒. , 𝑎𝑐𝑐𝑒𝑝𝑡 𝐻0.
Key Points:
SMALL SAMPLE TESTS :
(i) t-test:
Name of the test Null Level of Test statistic
Hypothesis𝑯𝟎 significance(𝜶)
1.Test for single mean 𝜇=𝜇0 5% or 1% or 10% 𝑥̅ −𝜇
t= 𝑠
√𝑛−1
2.Test for difference 𝜇 1= 𝜇 2 5% or 1% or 10% 𝑥̅ − 𝑦̅
t= 1 1
of means 𝑠 √𝑛 +𝑛
1 2

Page 23
PROBABILITY & STATISTICS SMALL SAMPLE TESTS
3. paired t-test 𝜇 =0 5% or 1% or 10% 𝑑̅
=𝑠
√𝑛

(ii) F-test:
Test for difference of variance
𝐻0 : 𝜎1 2 = 𝜎2 2
𝛼 : 5% or 10%.
𝑆1 2 𝐺𝑟𝑒𝑎𝑡𝑒𝑟 𝑣𝑎𝑟𝑖𝑎𝑛𝑐𝑒
Test statistic F = =
𝑆2 2 𝑠𝑚𝑎𝑙𝑒𝑟 𝑣𝑎𝑟𝑖𝑎𝑛𝑐𝑒

(iii) Chi-square (𝝌𝟐 )test:

Test for goodness of fit and test for independence of attributes.
(𝑂−𝐸)2
Test statistic 𝜒 2 = ∑ 𝐸
PRACTICE PROBLEMS
t-TEST
1.Test for single mean
1. A mechanist id making engine parts with axle diameters of 0.700 inch. A random sample of 10 parts shows a
mean diameter of 0.742 inch with a S.D of 0.040 inch. Compute the statistic you would use to test whether the
work is meeting the specifications.
2. A Sample of 26bulbs give mean life of 990 hours with a S.D of 20 hours. The manufacturer claims that the
mean life of bulbs is 1000 hour. Is the sample not up to the standard?
3. The average breaking strength of the steel rods is specified to be 18.5 thousand pounds. To test this sample
of 14 rods were tested. The mean and S.D obtained were 17.85 and 1.955 respectively. Is the result of
experiment significant?
4. A Random sample of six steel beams has a mean compressive strength of 58,392 p.s.i (pounds per square
inch) with a standard deviation of 648p.s.i Use this information and the level of significance 0.05 to test the
steel from which this sample came is 58,000p.s.i.
Problems related to student’s t-test (When S.D of the sample is not given derectly)
1. A random sample of 10 boys had the following IQ ‘s : 70,120, 110 , 101 ; 88 ; 83 ,95 , 107 , and 100.
(a) do these data support the assumption of a population mean IQ of 100 ?
(b) Find a reasonable range in which most of them mean I Q values of samples of 10 boys ?
2. The heights of 10 males of a given locality are found to be 70, 67 , 62, 68 , 61, 68, 70, 64, 64, 66 inches. Is it
reasonable to believe that the average height is greater than 64 inches ? . Test at 5% level of siginificance that
for 9 degrees of freedom (t = 1.833 at α = 0.05) .
3.Eight students were given a test in STASTICS and after one month coaching they were given another test of
the similar nature . The following table gives the increase I their marks in second test over the first
Student no 1 2 3 4 5 6 7 8
Increase of 4 _2 6 _8 12 5 _7 2
marks
Do the marks indicate that the students have gained from the coaching .
4 . A random sample of 10 bags of pesticide are taken whose weights are 50, 49, 52, 44, 45, 48, 46, 45, 49, 45
(in kgs) .Test whether the average packing can be taken to be 50 kgs .
Student t_ test for difference of means :
1. A group of 5 patients treated with medicine A weigh 42, 39, 48, 60 and 41 kgs . second group of 7
patients from the same hospital treated with medicine B weigh 38, 42, 56, 64, 68 , 69 and 62 kgs. Do
you agree with the claim that medicine B increases the weigh siginificantly.

Page 24
PROBABILITY & STATISTICS SMALL SAMPLE TESTS
2. To examine the hypothesis that the husbands are more intelligent than the wives , an investigator
took a sample of 10 couples and administered them a test which measures the I.Q. .the results are as
follows .
Husband 117 105 97 105 123 109 86 78 103 107
Wives 106 98 87 104 116 95 90 69 108 85
Test the hypothesis with a reasonable test at the level of siginificance of 0.05 .
3 . two independent samples of 8 & 7 items respectively had the following values.
Sample I 11 11 13 11 15 9 12 14
Sample II 9 11 10 13 9 8 10
Is the difference between the means of samples siginificant ?
4 . Two compare two kinds of bumper guards , 6 of each kind were mounted on a car and then the car was
run into a concrete wall . The following are the costs of repairs
Guard 1 107 148 123 165 102 119
Guard 2 134 115 112 151 133 129
Use the 0.01 level of significance to test whether the difference between two sample means is significant.
5 . the table gives the biological values of protein from 6 cow’s milk and 6 buffalo’s milk. Examine the
difference are significant.
cow’s milk 1.8 2.0 1.9 1.6 1.8 1.5
buffalo’s milk 2 1.8 1.8 2.0 2.1 1.9

Paired _ sample t _ test :

1. The blood pressure of 5 women before and after intake of a certain drug are given below :
before 110 120 125 132 125
after 120 118 125 136 121
2. Memory capacity of 10 students were tested before and after training . state whether the training was
effective or not from the following scores
Before 12 14 11 8 7 10 3 0 5 6
after 15 16 10 7 5 12 10 2 3 8
3. The average losses of workers, before and after certain program are given below. Use 0.05 level of
significance to test whether the program is effective (paired sample t-test). 40 and 35,70and 65,45, 42, 120
and 116, 35 and 55 and 50,77 and 73.
F-TESTTest for equality of two population variances:
1. In one sample of 8 observations from a normal population, the sum of the squares of deviations of the
sample values from the sample mean is 84.4 and in another sample of 10 observations it was 102.6. Test at
5%level whether the populations have the same variance.
2. Pumpkins were grown under two experimental conditions. Two random samples of 11 and 9 pupkins, show
the sample standard deviation of their weights as 0.8 and 0.5 respectively. Assuming that the weight
distributions are normal, test hypothesis that the true variances are equal.
3. In one sample of 10 observations, the sum of the deviations of the sample values from sample mean was
120 and in the other sample of 12 observations it was 314. Test whether the difference is significant at 5%
level.
4. Two random samples reveal the following results
Sample Size Sample mean Sum of squares of deviations from
the mean
1 10 15 90
2 12 14 108
Test whether the sample came from the same normal population.
5. the time taken by workers in performing a job by method I and method II is given below:
Method I 20 16 26 27 23 22 ---
Page 25
PROBABILITY & STATISTICS SMALL SAMPLE TESTS
Method II 27 33 42 35 32 34 38
Do the data show that the variances of time distribution from population from which these samples are drawn
do not differ significantly?
CHI-SQUARE TEST FOR GOODNESS OF FIT:
1.. The number of automobile accidents per week in a certain community are as follows: 12, 8, 20, 2, 14, 10,
15, 6, 9, 4. Are these frequencies in agreement with the belief that accident conditions were the same during
this 10 period week.
2.. A die is thrown 264 times with the following results. Show that the die is biased.(given chi-square 0.05
=11.07 for 5 d.f).
3. A pair of dice is thrown 360 tunes and the frequency of each sum is indicated below:
Sum 2 3 4 5 6 7 8 9 10 11 12
Frequency 8 24 35 37 44 65 51 42 26 14 14
Would you say that the dice are fair on the basis of the chi-square test at 0.05 level of significance.
4. A survey of 320 families with 5 children each revealed the following distribution.
No. of Boys 5 4 3 2 1 0
No. of Girls 0 1 2 3 4 5
No. of Families 14 56 110 88 40 12

CHI-SQUARE TEST FOR INDEPENDENCE OF ATTRIBUTES

1.The following table gives the classification of 100 workers according to gender and nature of work. Test
whether the nature of work is independent of the gender of the worker.
Stable Unstable Total
Male 40 20 60
Female 10 30 40
Total 50 50 100
2. From the following data, find whether there is any significant liking in the habit of taking soft drinks among
the categories of employees.
Soft drinks Clerks Teachers officers
Pepsi 10 25 65
Thumsup 15 30 65
Fanta 50 60 30
3. In an investigation on the machine performance, the following results are obtained.
No.of units inspected No.of defective
Machine1 375 17
Machine2 450 22
4. A survey of 240 families with 4 children each revealed the following distribution
Male Births 4 3 2 1 0
Female Births 10 55 105 58 12
Test whether the male and female births are equally popular.

********

Page 26
PROBABILITY & STATISTICS SMALL SAMPLE TESTS

Page 27