Normal Distribution
Normal Distribution
Normal Distribution
NORMAL DISTRIBUTION
The concept of normal distribution was initially discovered by English mathematician
Abraham De Moivre (1667-1754) in 1733. De Moivre obtained this continuous
distribution as a limiting case of binomial distribution. His work was further refined
by Pierre S. Laplace (1749-1827) in 1774. But the contribution of Laplace remained
unnoticed for long till it was given concrete shape by Karl Gauss (1777-1855) who
first made reference to it in 1809 as the distribution of errors in Astronomy. That is
why the normal distribution is sometimes called Gaussian distribution. Though,
normal distribution can be used as approximation to most of the other distributions,
here we are going to discuss (without proof) its approximation to binomial distribution
only.
Normal Distribution as a Limiting Case of Binomial Distribution
Normal distribution is a limiting case of binomial distribution under the following
conditions:
parameters ( ) and 2(>0) if it takes on any real value and its probability
density function is given by
;
which may also be written as
=
Remark
.
ii) If a random variable X follows normal distribution with mean and variance 2,
then we may write, “X is distributed to N(, 2)” and is expressed as X N(,
2).
iii) No continuous probability function and hence the normal distribution can be used
to obtain the probability of occurrence of a particular value of the random
variable. This is because such probability is very small, so instead of specifying
the probability of taking a particular value by the random variable, we specify the
probability of its lying within interval.
v) Mean of Z i.e. =
=0 [ E(X) = Mean of X = ]
[ variance of X is 2]
= 1.
by .
vii) The graph of the normal probability function with respect to is famous
‘bell-shaped’ curve. The top of the bell is directly above the mean . For large
value of , the curve tends to flatten out and for small values of , it has a sharp
peak as shown in (Fig. 13.1):
Fig. 1.1
Normal distribution has various properties and large number of applications. It can be
used as approximation to most of the other distributions and hence is most important
probability distribution in statistical analysis. Theory of estimation of population
parameters and testing of hypotheses on the basis of sample statistics have also been
developed using the concept of normal distribution as most of the sampling
distributions tend to normality for large samples. Normal distribution has become
widely and uncritically accepted on the basis of much practical work. As a result, it
holds a central position in Statistics.
Let us now take some examples of writing the probability function of normal
distribution when mean and variance are specified, and vice-versa:
Example 2: Below, in each case, there is given the p.d.f. of a normally distributed
random variable. Obtain the parameters (mean and variance) of the variable.
(i)
(ii)
Solution: (i)
Comparing it with,
we have
(ii)
Comparing it with,
we get
(i)
(ii)
E 2) Below, in each case, is given the p.d.f. of a normally distributed random
variable. Obtain the parameters (mean and variance) of the variable.
(i)
(ii)
Now, we are going to state some important properties of Normal distribution in the
next section.
CHIEF CHARACTERISTICS OF NORMAL
DISTRIBUTION
The normal probability distribution with mean and variance 2 has the following
properties:
i) The curve of the normal distribution is bell-shaped as shown in Fig. 1.1.
ii) The curve of the distribution is completely symmetrical about i.e. if we fold
the curve at both the parts of the curve are the mirror images of each other.
iii) For normal distribution, Mean = Median = Mode
iv) , being the probability, can never be negative and hence no portion
of the curve lies below x-axis.
v) Though x-axis becomes closer and closer to the normal curve as the magnitude of
and
Note: Not only but all the odd order central moments
median.
with mean .
xiv) Particularly, sum or difference of two independent normal variates is also a
normal variate. If X and Y are two independent normal variates with means 1, 2
Also, if X1, X2, …, Xn are independent variates each distributed as N( ,2), then
their mean .
xv) Area property:
= 0.6827,
= 0.9544, and
, and
Here,
i)
i.e. , and
ii)
i.e.
Solution: Here = 0, 2 = 1 σ = 1.
(ii)
Mean of
Variance of
Now,
(i) Mean of X =
= = = 80 + 180 = 260
Var
=4×25+9×36
= 100 + 324 = 424
(ii) Mean of Y =
3×40−2×(60 )
= = 120 – 120 = 0
9×25+4×36
= = 225 + 144 = 369
You can now try some exercises based on the properties of normal distribution which
you have studied in the present unit.
E3) If X1 and X2 are two independent normal variates with means 30, 40 and
variances 25, 35 respectively. Find the mean and variance of
i) X1 + X2
ii) X1 – X2
E4) If X N(50, 225), find its Quartile deviation.
E5) If X1 and X2 are independent variates with each distributed as N (50, 64),
SOLUTIONS/ANSWERS
we get
(ii)
we get
X1 + X2 N(70, 60)
E4) As 2 = 225
= 15
Now,
Put
Also, when X = , Z = 0
You need not to evaluate the integral to find the area. Table is available to find such
area for different values of z1.
Here, we have transformed the integral from
i.e. we have transformed normal variate ‘X’ to standard normal variate (S.N.V.)
.
This is because, the computation of
[By Symmetry]
= 2 0.34135
= 0.6827
= 2 0.4772
= 0.9544
μ±σ
Fig. 14.4: Area within the Range
and
= 2 0.49865 = 0.9973
μ±σ
Fig. 14.5: Area within the Range
From the above discussion, we conclude that while solving numerical problems, we
need to transform the given normal variate into standard normal variate because tables
for the area under every normal curve, being infinitely many, cannot be made
available whereas the standard normal curve is one and hence table for area under this
curve can be made available and this is given in the Appendix at the end of this unit.
Example 7: If X ~ N (45, 16) and Z is the standard normal variable (S.N.V.) i.e
Solution: Here we are given that X is normally distributed with mean 80 and standard
deviation (S.D.) 5.
i.e. Mean =
(i) X = 95,
=
= 0.5 – 0.4987 [Using table area under normal curve]
= 0.0013
Fig. 14.6: Area to the Right of X= 95
(ii) X = 72,
=
= 0.5 – 0.4452 [Using table area under normal curve]
= 0.0548
(iii) X = 60.5,
X = 90,
=
=
= 0.5000+ 0.4772
= 0.9772
(iv)
=
= 0.4997 – 0.3413 [Using table area under normal curve]
= 0.1584
= 0.4993 – 0.2881
= 0.2112
(b) What is the lowest weight of the 100 heaviest male students?
(Assuming that the weights are normally distributed)
Solution: Let X be a normal variate, “The weights of the male students of the
university”. Here, we are given that µ = 60 kg, σ = 16 kg, therefore,
X ~ N(60, 256).
We know that if X ~ N(µ, σ2), then the standard normal variate is given by
.
Hence, for the given information,
Therefore,
P[X < 55] = P [Z < 0.31] = P [Z > 0.31] [See Fig. 14.11]
[ ¿ t h e table of areasunder
Normal Curve ]
= 0.3783
Number of male students having weight less than 55 kg = N P(X < 55)
= 1000 0.3783
= 378
= 1000
0.2643
= 264
For X 65,
= 0.3264 + 0.1217
= 0. 4481
Fig. 14.13: Area Representing Students having Weight between 45 kg and 65 kg
= 1000 0.4481 =
448
Now, for , .
0.4.
Example 10: In a normal distribution 10% of the items are over 125 and 35% are
under 60. Find the mean and standard deviation of the distribution.
Solution:
Fig. 14.15: Area Representing the Items under 60 and over 125
Let X ~ N(µ, σ2), where µ and σ2 are unknown and are to be obtained.
Here we are given
P[X > 125] = 0.1 and P[X < 60] = 0.35. [See Fig. 14.15]
For X 125, Z
Now
and
Putting the values of in Equations (1) and (2), we get
… (3)
… (4)
Hence
Now, you can try the following exercises:
(i) P[X < 22], (ii) P [X > 23], (iii) , and (iv)
α
E8) Suppose X ~ N (30, 16) then find in each case
(i)
(ii)
E9) Let the random variable X denote the chest measurements (in cm) of 2000
boys, where X ~ N(85, 36).
SOLUTIONS/ANSWERS
E6) We are given X ~ N(150, 9)
∴
in usual notations, we have
Now,
i)
ii)
= 0.5 + 0.3413
= 0.8413
=
= 0.4772 – 0.3413 = 0.1359
iv)
=
For
For
= 1–
= 1– [0.4987 – 0.3413]
= 1– 0.1574 = 0.8426.
Mean =
If Z is S.N.V then
i)
Now
Putting
ii) … (2)
Now
[Due to symmetry]
= 23.4
E9) We are given X ~ N(85, 36), N = 2000
i.e.
If X ~ N(µ, σ2 ) and
a) i) For X = 87,
Now P[X < 87] = P [Z < 0.33] [See Fig. 14.27]
= 0.5 + P [0 < Z < 0.33]
= 0.5 + 0.1293
= 0.6293
Therefore, number of boys having chests measurement
= = 1259
For X = 90,
= 0.2967 – 0.0675
= 0.2292
number of boys having chests measurement between 86 cm
and 90 cm
= N. P [86 ]
= 0.7967
= 2000 0.7967
= 1593
b) Let be the lowest chest measurement amongst 100 boys having the
largest chest measurements.
Now, for X , .
P[X
Fig. 14.30: Area Representing the 100 Boys having Largest Chest Measurements
If then we know
that Z ~ N (0, 1)
a) For X = 4.2,
For X = 4.5,
Fig. 14.31: Area Representing Probability of Waiting Time between 4.2 and 4.5 Minutes
b) For
Fig. 14.32: Area Representing Probability of Waiting Time Less than 5.2 Minutes
= 0.5 – 0.1915
= 0.3085
Therefore, probability that customer has to work for less than 5.2 min
= 0.3085
c) For
Fig. 14.33: Area Representing Probability of Waiting Time Greater than 6.8 Minutes
a) For X = 20,
b) For X= 26,
= 0.5 – 0.1293
= 0.3707
Therefore, probability that temperature of the city is more than
is 0.3707
c) For X= 23,
For X= 27,
between 23 and 27
P[23 < X < 27] = P[–0.17 < Z < 0.5] [See Fig. 14.36]
= P [–0.17 < Z < 0] + P [0 < Z < 0.5]
= P[0 < Z < 0.17] + P[0 < Z < 0.5]
= 0.0675 + 0.1915
= 0.2590
Therefore, probability that temperature of the city is between
23 and 27 is 0.2590
E8)
Now, Let
∴
Standard normal variate is
When X = 40,
When X = 50,
Now,
[Due to symmetry]
APPENDIX
The following table gives probability corresponding to the shaded area as shown in the
TABLE OF AREAS
↓ z→ 0 1 2 3 4 5 6 7 8 9
0.0 .0000 .0040 .0080 .0120 .0160 .0199 .0239 .0279 .0319 .0359
0.1
.0398 .0438 .0478 .0517 .0557 .0596 .0636 .0675 .0714 .0759
0.2
.0793 .0832 .0871 .0910 .0948 .0987 .1026 .1064 .1103 .1141
0.3
.1179 .1217 .1255 .1293 .1331 .1368 .1406 .1443 .1480 .1517
0.4
.1554 .1591 .1628 .1664 .1700 .1736 .1772 .1808 .1844 .1879
0.5
.1915 .1950 .1985 .2019 .2054 .2088 .2123 .2157 .2190 .2224
0.6
.2257 .2291 .2324 .2357 .2389 .2422 .2454 .2486 .2517 .2549
0.7
.2580 .2611 .2642 .2673 .2703 .2734 .2764 .2794 .2823 .2852
0.8
.2881 .2910 .2939 .2967 .2005 .3023 .3051 .3078 .3106 .3133
0.9
.3159 .3186 .3212 .3238 .3264 .3289 .3315 .3340 .3365 .3389
1.0
.3413 .3438 .3461 .3485 .3508 .3531 .3554 .3577 .3599 .3621
1.1
.3643 .3655 .3686 .3708 .3729 .3749 .3770 .3790 .3810 .3820
1.2
.3849 .3869 .3888 .3907 .3925 .3944 .3962 .3980 .3997 .4015
1.3
.4032 .4049 .4066 .4082 .4099 .4115 .4131 .4147 .4162 .4177
1.4
.4192 .4207 .4222 .4236 .4251 .4265 4279 .4292 .4306 .4319
1.5
.4332 .4345 .4357 .4370 .4382 .4394 .4406 .4418 .4429 .4441
1.6
.4452 .4463 .4474 .4484 .4495 .4505 .4515 .4525 .4535 .4545
1.7
.4554 .4564 .4573 .4582 .4591 .4599 .4608 .4616 .4625 .4633
1.8
.4641 .4649 .4656 .4664 .4671 .4678 .4686 .4693 .4699 .4706
1.9
.4713 .4719 .4726 .4732 .4738 .4744 .4750 .4756 .4761 .4767
2.0
.4772 .4778 .4783 .4788 .4793 .4798 .4803 .4808 .4812 .4817
2.1
.4821 .4826 .4830 .4834 .4838 .4842 .4846 .4850 .4854 .4857
2.2
.4861 .4864 .4868 .4871 .4875 .4678 .4881 .4884 .4887 .4890
2.3
.4893 .4896 .4898 .4901 .4904 .4906 .4909 .4911 .4913 .4916
2.4
.4918 .4920 .4922 .4925 .4927 .4929 .4931 .4932 .4934 .4936
2.5
.4938 .4940 .4941 .4943 .4945 .4946 .4948 .4959 .4951 .4952
2.6
2.7 .4953 .4955 .4956 .4957 .4959 .1960 .4961 .4962 .4963 .4964
2.8 .4965 .4966 .4967 .4968 .4969 .4970 .4971 .4972 .4973 .4974
2.9 .4974 .4975 .4976 .4977 .4977 .4978 .4979 .4879 .4980 .4981
3.0 .4981 .4982 .4982 .4983 .4984 .4984 .4985 .4985 .4986 .4986
3.1 .4987 .4987 .4987 .4988 .4988 .4989 .4989 .4989 .4990 .4990
3.2 .4990 .4991 .4991 .4991 .4992 .4992 .4992 .4992 .4993 .4993
3.3 .4993 .4493 .4994 .4994 .4994 .4994 .4994 .4995 .4995 .4995
3.4 .4995 .4995 .4995 .4996 .4996 .4996 .4996 .4996 .4996 .4997
3.5 .4997 .4997 .4997 .4997 .4997 .4997 .4997 .4997 .4997 .4998
3.6 .4998 .4998 .4998 .4998 .4998 .4998 .4998 .4998 .4998 .4998
3.7 .4998 .4998 .4999 .4999 .4999 .4999 .4999 .4999 .4999 .4999
3.9 .4999 .4999 .4999 .4999 .4999 .4999 .4999 .4999 .4999 .4999
.5000 .5000 .5000 .5000 .5000 .5000 .5000 .5000 .5000 .5000