t-test

Seminar presented by
SAYYED HEENA RIYAZ
&
SHAIKH QAMRUNNISA ABDUL
WAHID
MSC –I MICROBIOLOGY
(QUANTITATIVE BIOLOGY)
CONTENT:
1. Biostatistic
a. Defination
b. History
c. population
d. sample
e. null hypothesis
2. t-test
a. defination
b. history
3. Types of t-test
a. t-test for two small samples
b. t-test for two large samples
c. t-test for paired samples
4. Summary
5. previous years questions.
BIOSTATISTIC:
Biostatistics is the application of statistic to a wide range of topics in biology.

History:
 Ronald Fisher developed several basic statistical methods in support of his
work studying the field experiments.

Ronald Fisher

 Sewall G. Wright developed F-statistics and methods of computing

them.

Sewall G. Wright
 POPULATION:

In statistic population is a well defined group which is being studied .

SAMPLE:

The selected part of the population is known as sample

NULL HYPOTHESIS:
There is no significance difference between population mean and sample
mean.
 t-test:
 A t-test is a statistic that checks if two means are relaibly
different from each other.
 t-test findout significantly difference exists between two
groups of data.

HISTORY:

 The t-statistic was introduced in 1908 by

William Sealy Gosset.

William Sealy
Gosset
Types of t-test:

t-test for two large samples

t-test for two small

samples

t-test for paired samples

 t-TEST FOR TWO LARGE SAMPLES:
 Sample consisting of more than 30 observations or items is
called as large samples.

 t= Difference of means of two samples

x -x
= t= 1 2
standard error of difference Sd

Where,

     
• X1 = mean of first sample = or or

     
X2 = mean of second sample= or or

• n1=number of observation in the first sample

n2= number of observation in the second sample
 Where,
 
Sd, =
• Sd =Standard error of the difference

  )     )
 
S12 = varience of first sample = or or
   

    )

  )     )
 

S22 = varience of second sample = or or

 2  2

    )

 2
 Degrees of freedom= (n1+n2-2)
 EXAMPLE-1:
Q-The following data relate to the days to flowering in two varieties of
mungbeans , G-65 &PS-16.Determine whether two means are significantly
different.
G-65 PS-16
n 30 35
mean 32 38
varience 9.62 14.23

SOLUTION:
Null hypothesis :There there is no significant difference between mean days to flowering in
both the varieties.

x1-x2  
t= Where, Sd =
Sd
 Sd =
  x1-x2
t=
Sd

 t=
Sd=  

t = 7.06
Sd=  0.32

Sd= 0.72

Sd= 0.84

Degrees of freedom= (n1+n2-2) =(30+35-2) =(65-2) =63

Conclusion: The calculated value of t(7.06) is greater than the tabulated
value of t for 63(nearest to 60)degrees of freedom (1%=2.66). It is clearly
indicated that the two means are very much different.hence the null
hypothesis stating that there is no significance difference between 2
samples is rejected at P=1% .
 EXAMPLE-2:
Q- Data recorded on the number of tomatoes per plant on two varieties of tomato. Compare
the mean of two varieties & give your conclusion.
Variety A:
6,8,10,12,12,14,11,6,8,9,12,14,13,7,8,10,12,14,15,7,8,13,16,9,10,13,14,13,14,14,9,11,
13,13,13,15,9,10,11,12,14,16,17,13,16,17,15,15,16,17.
Variety B:
8,10,12,13,15,17,19,9,8,11,13,15,17,21,14,17,16,14,14,8,9,12,15,19,12,10,13,15,18,11
,13,15,16,10,11,7,21,9,14,18,19,14,9,11,15,20,20,18,15,16.

SOLUTION:

Null hypothesis: there is no significance difference between two mean no. of tomatoes
per plant in both the varieties.
Variety A:  X =∑f1x1/∑f1 = 604/50 =
1
x1 f1 f1x1 12.08
6 2 36 72 12 =
7 2 49 98 14 =
8 4 64 256 32 =
9 4 81 324 36 = 9.03
10 4 100 400 40

11 3 121 363 33

12 5 144 720 60

13 8 169 1352 104

14 7 196 1372 98

15 4 225 900 60
Variety B:  
2
x2 f2 x2f2
= 696/50
7 1 49 49 7 = 13.92
8 3 64 192 24  
9 4 81 324 36
=
10 3 100 300 30
=
11 4 121 484 44
=13.91
12 3 144 432 36
13 4 169 676 52
14 5 196 980 70
15 7 225 1575 105
16 3 256 768 48
17 3 289 867 51
18 3 324 972 54
19 3 361 1083 57
20 2 400 800 40
21 2 441 882 42
Total 10384 696
 x1-x2
Sd =
  t=
Sd

12.08-13.92
  t=
Sd = 0.67

-1.84
t = 0.67
9.03+13.91
Sd = 50

t = 2.74
Sd = 0.45

Sd = 0.67

Degree of freedom = (n1+n2-2) = 50+50-2= 100-2 = 98

Conclusion : the calculated value of t is 2.74 greater than the
tabulated value (2.71) for 98 or 120 degree of freedom at 1% level of
significance. Hence there is no significance difference between two
 t-TEST FOR TWO SMALL SAMPLES:
 Sample consisting of upto 30 observations or items is
called as small samples.
x1-x2 or x1-x2 n 1n 2
t= t=
Sd n1+n2
Sd 1/n1+1/n2
Where,

     
• X1 = mean of first sample = or or

     
X2 = mean of second sample= or or

• n1=number of observation in the first sample

n2= number of observation in the second sample
 Where,
 
Sd, =
• Sd =Standard error of the difference

    )

S12 = varience of first sample =

 

    )

S22 = varience of second sample =

 2
  )     )
 
Pooled standard
deviation or standard +
 2
 
= error of difference:
  )   )
  +  

=
n1+n2-2

 Degrees of freedom= (n1+n2-2)

   )   )
Sd =   +  

n1+n2-2

 Degrees of freedom= (n1+n2-2)

 EXAMPLE-1:
Q . in a mutation breeding experiments, gamma irradiation effect was evaluated on 100
seed weight in grams per plant of a mungbean variety in M2 generation. The
experimeters obtained the following results .analyse the data using the t-test and give
your inference as regards the effect of gamma irradiation.

Control 2.9 3.1 3.5 3.4 3 4 3.7 3 4 4

Treated 2.7 2.8 3 3.5 3.7 3.2 3 3.1 2.9 2.8

SOLUTION:
Null hypothesis: there is no significance difference between the mean 100 seed weigth
per plant in the control and treatment.
 Control Treated
x1 x12 x2 x 22
2.9 8.41 2.7 7.29
3.1 9.61 2.8 7.84
3.5 12.25 3 9
3.4 11.56 3.5 12.25
3 9 3.7 13.69
4 16 3.2 10.24
3.7 13.69 3 9
3 9 3.1 9.61
4 16 2.9 8.41
4 16 2.8 7.84

  34.6
X∑x =34.6 ∑x =121.52 ∑x2=30.7 ∑x22=95.17
2
1 =1 = 1 = 3.46
10

 
X2 = = 30.7 = 3.07
10
   )   )
Sd =   +  

n1+n2-2

    )
Sd =   +  

10+10-2

   
Sd =   +  

18
Sd =   -119.71 +    

18
x1-x2 n1n2
t=
Sd n1+n2

Sd
  +  
=
3.46-3.07 10*10
18 t=
0.38 10+10

 
Sd = 18 t=
0.39
0.38
100
20

t= 1.02 5
Sd =  

t= 1.02*2.23

Sd = 0.38 t=
2.27
Degrees of freedom= (n1+n2-2) = 10+10-2
=18
Conclusion:
The calculated value of t(2.27) is less than the tabulated value (2.87) for
18 degree of freedom at 1% level of significance. hence the null
hypothesis stating that there is no significance difference between the two
seed weigth of plant get accepted at P=0.01
T-TEST FOR PAIRED SAMPLES:
Tests the mean of one group twice.
Examples:
 Testing the balance before and after drinking.
 Testing IQ level before and after the training program .
d√n
t=
Sd or t= d
SEd
Where:
∑d
d = n , d = the mean of the difference between the paired values.

SEd = Standard error of the difference= Sd

√n
∑d2-(∑d)2
Sd = Standard deviation of the differnce= Sd n
= n-1
Degree of freedom,d.f.= n-1
EXAMPLE-1:
Q. Data recorded on the rainfall at two places , A and B in 10years given below.
Analyse the data and darw your inferences whether the two places have the same
mean annual rainfall.

Years Rainfall in mm at A Rainfall in mm at B

1971 177.29 69.79

1972 146.12 103.93
1973 159.89 74.29
1974 111.68 123.21
1975 96.94 91.47
1976 120.41 68.18
1977 114.95 55.50
1978 114.14 105.20
1979 137.38 101.88
1980 119.42 121.84
SOLUTION:

Null hypothesis:H0 : There is no significance difference in the rainfall of the two

places A and B.
d2
Years Rainfall in Rainfall in Differece
mm at A mm at B
x 1 -x 2 =d

1971 177.29 69.79 107.50 11556.25

1972 146.12 103.93 42.19 1779.99
1973 159.89 74.29 85.60 7327.36
1974 111.68 123.21 -11.53 132.94
1975 96.94 91.47 5.47 29.92
1976 120.41 68.18 52.23 2727.97
1977 114.95 55.50 59.45 3534.30
1978 114.14 105.20 8.94 79092
1979 137.38 101.88 35.50 1260.25
1980 119.42 121.84 -2.42 5.86
Total _ _ ∑d=382.93 ∑d 2 =28434.76
 ∑d = 382.92
d = n =38.29
10

∑d2-(∑d)2 13771.22
Sd Sd
n 9
= n-1 =

Sd 28434.76-(382.93)2/10 Sd 1530.14
= 10-1
=

Sd 39.12
Sd 28434.7-146635.38/10
=
= 9

28434.7-14663.53
Sd
9
=
 t= d√n
Sd
t= 38.29√10
39.12

38.29*3.16
t=
39.12
120.99
t=
39.12
t= 3.10
Degree of freedom= n-1 = 10-1 =9

Since the calculated t value (3.10) is less then the tabulated t value (3.25) for
90 of freedom at 1% level of significance. It is clearly indicates thet the two
means of rain fall at two places are very much similar hence the null
hypothesis stating that there is no significance difference in the rainfall of the
two places A and B get accepted at P=0.010 .
A QUICK
REVIEW
t-test type Paired large samples small samples
What it is?? Tests the mean Sample Sample consisting of upto 30
of one group consisting of observations or items is called as
twice. more than 30 small samples.
observations or
items is called as
large samples.
d√n x1-x2 x1-x2
Formula of t t= S t= t= n1n2
d Sd Sd n1+n2

Formula for ∑d2-(∑d)2

standard deviation Sd = n Sd   = Sd  =
Sd = n-1
of difference
) )
     
Formula for S2 = S2 =
varience S2 ______    
∑d  
X=  
Formula for mean d = n X=

(n1+n2-2) (n1+n2-2)
Degree of freedom n-1
d.f.
THANK YOU