470

J'undamentals of Mathematical Statistics

11

11

P(A)= r.P(AnEi)=' r.P (Ei)P (A lEi)

i=I

... (412 a)

i=I

Cor. (Bayes theorem/or future events)

The probability 0/ the materialisation 0/ another event C, given
P (C

IA n

I An E2), ,P (C I A n

E I ) P (C
11

r.

P (Ei) P (A lEi) P

(c.1

E~)

is

EinA)

P(CIA)=_i_=~I ______~ _________

11

r.

...(4Pb)

P (Ei) P (A lEi)

i=--t

Proof. Since the occurrence of event A implies the occurrence of one and only
one of the events E I E2 E~. the event C (granted thatA has occurred) can occu
in the following mutually exelusive ways:
Cn EI.C nE2 Cnf!~
C =(C n E I ) U (C n E2) U ... U (C n E~)
i.e.,
C I A = [(C n E I ) I A] u [(C n 1:.'2) I A] u ... u [(C n E~) I. A]
.. P (C I A) ~ P [(C n E I ) I A] + P [(C n E 2) I A] +...+ P [(C n E~) I A]
11

_=

r.p [(C

I A]

Ei )

i= I
11

r.

P (Ei

I A)

P [C

I (EinA)]

i= I

Substituting the value of P (Ei I A) from (*). we get

11

P (C

I A) =

r.

P (Ei) P (-1 lEi) P (C

I EinA)

l~____-----,~________

..:....i=-.....:.

11

r.

P (Ei) P (A lEi)

i= 1

Rtmark. It may happen that die materialisation of the event Ei makes C

i.ridcQendenl of A. then we have
.
.

P(C

I Ei

nA)= P(C lEi).

and the abo~e.Jormula reduces to

11

P (C I A)

l: P(Ei) P (A

i E i)

P (C

I Ei)~

=.....:l~_____________

..:...i

11.

l: P (Ei) P (A lEi)
i= I

l11e event C can be considered in r~gard to A. as Future Event.

..(412 c)

471

Example 430. In 1?89 there were three candidates for the position 0/
principal- Mr. Challerji, Mr. Ayangar and Dr. Sing/!., whose chances 0/ gelling
the appointment are in the proportion 4:2:3 respectively. The prqbability that Mr.
Challerji if selected would introduce co-education in the college is 03. The
probabilities 0/ Mr. Ayangar and Dr. Singh doing the same are respectively 0-5
and 08. What is the probability that there was co-education in the college in 199O?
(Delhi Univ. B.Sc.(Stat. Hons.), 1992; Gorakhpur Univ. B.Sc., 1992)
Solution. Let the events and probabilities be defined as follows:
A : Introduction or co-education
EI : Mr. Chauerji is selected as principal
E2 : Mr. Ayangar is selected as principal
E3 : Dr. Singh is selccted as principal.
Then
4
2
3
P (E1) =

"9'

P (A'I E1)
..

P (A)

P (E2) =

=,l
10'

=i' [(A n E

"9

P (A

a~d

P (E3) ;

I E~) = 210

"9

and P (A 'I E3)=!10

E 2), u (A n 3)]
= P(AnEd + P(AnE,.) + P(l\nE3)
P (E 1) P (A. 1 E 1) + P (E2) P (A 1 E 2) :+- P (E3) P (A
1)

u (A n

=
4 3
2 5
3 8
23
= "9 . 10 + "9 . 10 + "9 . 10 = 45

I ~3)

..:xample. 431. The contentso/urns I, II and 11/ are as/ollows:

I white, 2 black and 3 red balls,

e white, 1 black and 1 red balls, and

<I white, 5 black and 3 r(;~.balls.

One urn is chosen at random and two balls drawn. They happen Jo be white
and red. What is the probability .thatthey come /rpl!l urns I, II or 11/ ?

.[Delhi Univ. B.Sc. (Stat. Hons.), 1988)

Solution. LetE\, E l and E3 denote the events that the urn I, II and III is chosen,
respectively, and let A be the event that the two balls taken from the selected urn
are white and red. Then
P (E 1)

and

P (A

1 E 1)

P (A

= P (E = P (E3) = '31
2)

= 1 x 3 =.!
6C2

4x 3

= IT

3) = 12C2

P (A

1 2-)

= 2 x 1 = .!
'C2

3'

Fundamentals of Mathematical Statistics

472

Hence'
'p (El

I A) = ;. ~E2)

P (A

I E 2)

L P (E i) P (A lEi)
j",

Similarly
30
118

E~ample. 432. In answering a question on a multiple choice test a student

either knows the answer or he guesses. Let p be the probability that he knows the
answer an4 I-p the probability that he guesses. Assume that a student who guesses
at the answer All be correct with probability 1/5, where 5 is the number of
multiple-choice alternatives. What is the conditional probability that a student
knew the answer to a question given that he answered it correctly?
[Delhi Univ. B.Sc. (Maths Hons.), 19851
Solution. Let us define the following'events:
EI : The student knew the right answer.
E2: The student guesses the right answer.
A : The student gets the right answer.
Then we are given
P (E I ) =p, P (E,.) = 1 - p, P (A I Ez) = 115
P (A lEI):; P [student gets the right answer given that he knew the right
answer] = 1
We want P (1 I A).
Using Bayes' rule. we get:
P (E I I A) =
P (EI) :P (A lEI)
= px 1
2L
P(EI) P(A lEI) + P(Ez) P(A I Ez)
1 (1 ) 1 4p + I
. px + -p

xs

Example 433. In a boltfactorymachinesA,B and C manufacture respectively

25%.35% and 40% of the total. Of their output 5,4, 2.percent are defective bolts.
A bolt ;s drawn at random/rom the product and is/ound to be defective. What are
the probabilities that it was manufactured by machines A, B and C?

Thcory of Probability

473

Solution. Let EJ. f-z and E, denote the events that a bolt-selected at random
is manufactured by the machinesA, B aM C respectively l,Uld let E denote the event
of its being defective, Then we have
P (E 1) = 0'25, P (z) = 035, P (E3) = 040
The probability of drawing a defective .bolt manufactured by machine A is
f(E

I E1)=005.

Similarly, we have
P (E I El ) =004, and f (E I E3) =002
Hence the probability t~at a defective bolt selected-at random is manufactured
by machine A is given by
P (El

I E) = :

(E 1) f (E lEI)

~ P (Ej ) P ( I E j )

i= I

025 x 0-05
125 25
=-=0-25 x 0-05 + 035 x 004 + 040 x 002 345 69

Similarly
p(zl'E)=

..
035 x 004.
140 28
025 x 005 + 03? x 004 + 040 x 002 =345 =69

and
P (E3 , E) = 1 - [P (El , E) + P (1 , E)]

= 1 _ 25 _ :28 :: ~

69 69 69
This example illustrates one of the chief applications of Bayes Theorem.
EXERCISE 4 (d)
1. (a) State and prove Baye's Theorem.
(b) The set of even~ Ai , (k = 1,2, ... , n) are (i) exhausti~e and (ii) pairwise
mutually exclusive. .If for all k the probabilities P (Ai) and f (E I Ai) are known,
calculate f (Ail E), where E is an arbitrary evenL Indicate where conditions OJ and
(ii) are used.
(c) 'f.he events E10 El , ... , E.. are mutually exclusive and, E =El u el U .. ,
u E... Show. that if f (A , E j ) = P (B , Ej) ; i = 1, 2, ..., n, then P(A' E) =
P(B', E). Is this conclusion true if the events,Ei are not mutually exclusive?
[Calcutta Univ. B.Sc. (Maths Hons.), 1990)
(d) What are the criticisms ~gai~t the use of Bayes theorem in probability
theory.
[Sr:i. Venketeswara Univ. B.Sc., 1991)
(e) Usillg the fundamental addition and multiplication rules Q( prob<lbility,
show that
__
P(B)P(AIB)
P (B I A) - P (8) P (A I B) + P (B) P

(.4 I B)

Where B is the event complementary to the eventB.

[Delhi Univ. M.A. (.:con.), 1981)

474

Fundamentals of Mathematical Statistics

2. (a) Two groups are competing for the positions on the Board of Directors
of a corporation. The-probabilities that the first and second groups will wiii are 06
and 004 respectively. Furthermore, if the first group wins the probability of
introducing a new product is 08 and the corresponding probability if the second
group wins is O 3. Whl\t is the probability that the new product will be introduced?
Ans. 06 x 08 + 04 x 03 =06
(b) The chances of X, Y, Zbecoming'managersol a certain company are 4:2:3.
The plOoabilities that bonus scheme will be introduced if X, Y, Z become managers.
are O 3. (f5 and 08 respectively. If the bonus scheme has been introduced. what is
the probability that X is appointed as the manager.
Ans. 051
(c) A restaurant serves two special dishes. A and B to its customers consisting
of 60% men and 40% women. 80% of men order dish A and the rest B. 70% of
women order dish B and the rest A. In what ratio of A to B should the restaurant
(Bangalore Univ. B.Sc., 1991)
prepare the two dishes?
Ans. P (A) =P [(A n M) u (,4. n W)] = 06 x 08 + 004 x 03 = ()'6
Similarly -P (B) =04. Required ratio =06 : 004 =3 : 2.

3. (a) There are three urns having the following compositions of black and
white balls.
Urn 1 : 7 white. 3 black balls
Urn 2 : 4 white. 6 black balls
Urn 3-: 2 white. 8 black balls.
One of these urns is chosen at random with probabilities 020. ()'60 and 020
respectively. From the chosen urn two balls are drawn at random without replacement Calculate the probability that both these balls are white.
(Madurai Univ. B.Sc., 1991)
Ans. 8145.
(b) Bowl I contain 3 red chips and 7 blue chips. bowl II contain 6 roo chips
and 41blue chips. A bow I is selected at random and then 1 chip 'is drawn from this
bowl. (i) Compute the probability that this chip is red. (ii) Relative to the hypothesis
that the chip is red. find the conditiona! probability that it is drawn from boWl II.
[Delhi Univ. B.Sc. (Maths 80ns.)1987]

(c) In a (actory machines A and B are producing springs of the same type. Of
this production. machines A and iJ produCe 5% and 10% defective springs.
respectively. Machines A and B produce 40% and 60% of the total output of the
factOry. One spri~g is selected at random and it is found to be defective. What is
the possibility that this defective spring was pfuduced by machine A ?
.
[Delhi Univ. M.A. (Econ.),1986]
(d) Urn A con'tains 2 white. 1 blac~ and 3 red balls. urn B contains 3 white. 2
black and 4 red balls and urn C con~ns 4 white. 3 black and 2 red balls. One urn
is chosen at random and 2 ,balls are drawn. They happen to be'red and black. What

475

Theory' of ,1'robability

is lhe probability that both balls came from urn 'B' ?

[Madras U. H.Sc. April; 1989)
(e) Urn XI. Xl. Xl. each contains 5 red and 3 white balls. Urns Yi. Yz each
contain 2 red and 4 white balls. An urn is selected at random and a billl is 'drawn.
It is found to be red. Find the probability lh~t the ball comes out of the urns of the
first type.
[Bombay 1I. B.Sc., April 1992]
if) Two shipments of parts are, recei'ved. The first shipment contains 1000, parts
with 10% defectives and the second Shipment contains 2000 parts with 5%
defectives. One shipment is selected at random. Two parts are tested and found
good. Find the,probability (a posterior) that the tested,parts were selected from the
first shipment.
[Uurdwan Univ. B.Sc. (Hons.), 1988]
(g) There are three machines producing 10,000 ; 20,000 and 30,000 bullets
per hour respectively. These machines are known to produce 5%,4% and 2%
defective bullets respectively. One bullet is taken at random from an hour's
production of the three machines. What is the probability that it is defective? If the
drawn bullet is defective, what is lhe probability lhat this was produced by the
second machine?
[Delhi. Univ. B.Sc. (Stat..uons.), 1991]
4. (0) Three urns are given each containing red and whjte chips as indicated.
Urn 1 : 6 red and 4 white.
Urn 2 :'2 red and 6 white.
Urn 3 : 1 red and 8 white.
(i) An urn is chosen at random and a ball is drawn from this urn. The ball is
red. Find the probability that the urn chosen was urn I .
(ii) An urn is chosen at random aIldtwq balls are drawn without replacement
from this urn. If both balls are red, find the probability that urn I was chosen. Under
thp.se conditions, what is the probability that urn III was chosen.
An~. 108/173, 112/12,0
[Gauhati Univ. B.Sc., 1990)
(b) There are ten urns of which each of three contains 1 white and 9 black balls,
each of other three contains 9 white and 1 black 'ball, and of the remaining four.
each contains? white and 5 black balls. One of the urns is selected at random and
a ball taken blindly from it turns out to be white. What is the probabililty that an
urn containing 1 white and 9 black balls was selected? ~Agra Univ. B.Sc., 1991)
3
3
4
Hint:
P (E 1),= 10' P (Ez) = 10 and P (E3) = 10'
Let A be the event of drawing a white blll.
3139451
P(A)= ~OxW+ iOxW+W x W=2
P (A lEI)

= l~ ~d

P (El

I A) = ;0

(c) It is known that an urn containing a1together 10 balls was' filled in, the
following manner: A coin was tossed. 10 times, and accordiJ]g a~ it showed heads
Or tails, one white or one black ball was put into the ul1.l.Balls are dfawnJrom lhis

Fundamentals ofMathematicaI Statistics

urn one at a time, to times in succession (with replacement) and everyone turns
out to he white. Find the chance that the urn contains nothing but white balls.
Ans. 00702.
.
5. (a) Frqm a vessel containing 3 white and 5 black balls, 4 balls are
transferred into an empty vessel. From this vessel a ball is drawn and is found to
be white. What is the probability that out of four balls transferred. 3 are white
and I black.
[Delhi U~i. B.Sc. (Stat. 1ions.), 1985]
Hint. Let the five mutually exclusive events for the four balls transferred be
Eo. E It E 2 , E 3 , and E 4 where E; denotes the event that i white balls are
transferred and let A be the event of drawing a white ball from the new vessel.
sC4
3C I x 5C 3
3C 2 X 5C 2
Then
' P (E 2) =
P (~o) =8C4 ' P(E I')
8C4
8C4
P (E) =

Also P(A I'Eo)

and

3C 3 X SCI

8C4

and P (E4 ) = 0

I
2
3
=0, P (A lEI) =4'
P (A I E 2) =4' (A I E3) =4'

I
P (A l'E4 ) = I. Hence P(E31 A) = =;.
(b) The contents of the urns I and 2 are as follows:

Urn 1 : 4 white and 5 black balls.

Urn 2 : 3 white and 6 black balls.
One urn is chosen at random and a ball is drawn and its colour noted and
replaced back to the um. Again a ball is drawn from the same urn, colour noted
and replaced. The process is repeated 4 times and as a result one ball of white
colour and three balls of black colour are obtain, 1. What is the probability that
the urn chosen was the urn 1 ?
(Poona Univ. B.E., 1989)
Hint.
P(E I ) = P (E2) 1F 1/2,
P (~ lEI) 4/9,
I - P (A lEI) 5/9
P (A I E 2 ) = 1/3,
1 - P (A I E 2 ) = 2/3
The probability that the urn chosen was the urn I

! 1. (~)3
2 9

=-----------------------!

4
2 9

(~)3
9

1, .(~)3

+ 2 3

(c) There are five urns numbered I to 5. Each urn contains to balls. The ith
urn has i defective balls and 10 - i non-defective balls; i = 1,2, ... 5. An urn is
chosen at random and then a ball is selected at random from that urn. (i) What is
thc probability that a defective ball is selected?
(ii) If the selected ball is defective, find the probability that it came from
urn i. (i = 1,2, ... , 5).
[Delhi Univ. B.Sc. (Maths Hons.), 1987]
Hint.: Define the following events:
E;: ith urn :is selected at random.

Theory of I'robability

477

A : Defective ball is selccted.

p. (Ei) = 1/5; i = 1,2, ... , 5.
P (A lEi) = P [Defective ball from ith urn] = il1O, (i = 1,2, ... , 5)
P (Ei) . P (A lEi) ='

(i)

P (Ei

x liO= ;0 ' (i = 1 , 2 , ... , 5).

~P(Ei)P(A I Ei)= ~

P(A)=
j

(ii)

=1

I A) =

(.1..)= 1-+ 2+3+4+5

50

= 1 50

P (Ei) P (A lEi)
"LP (Ei) P (A lEi)

i/50

= 3/10 = 15 ; l =

3
10

~
1 ,2, ... ,5.

For example, the probability that the defective ball came from 5th urn
= (5/15) = 1/3.
6. (a) A bag contains six balls of different colours and a ball is drawn from it
at random. A speaks truth thrice out of 4 times and B speaks truth 7 times out of 10
times. If both A and B say that a ,ed ball was drawn, find the probability of their
joint statement being true.
[Delhi Univ. B.Sc. (Stat. Hons.),1987; Kerala Univ. B.Sc.I988]
(b) A and B are two very weak students of Statistics and their chances of
solving a problem correctly are 1/8 and 1/12 res~tively. If the probability of their
making a common mistake is 1/1001 and they obtain the same answer, find the
chance that their answ~r is correct.
[Poona Univ. B.sc., 1989]
I;S x 1112
13
R d Pr bab 'I'
A
ns. eq. 0
I Ity = 1111 x ilI2 + (l - I;S) (l ~ ilI2) . ilIool
14
7. (a) Three bOxes, practicaHy indistinguishable in appearance, have two
drawerseach. Box l.contaiils a gold 'coin in one and a sit ver coin in the other drawer,
boX'II contains a gold coin in each drawer and box III contains a silver coin in each
drawer. One box is chosen at random and one of its drawers is opened at random
and a gold coin found. What is the:probability that the other drawer contains a coin
of silver?
(Gujarat Univ. B.Se., 1992)
Ans. 113, 113.
(b) Two cannons No. I and 2 fire at the same target Carinon No. I gives.on
an average 9 shots in the time in which Cannon No.2 fires 10 projectiles. But on
an average 8 out of 10 projectiles from Cannon No. 1 and 7outofIO from Cannon
No. 2 ~trike the target. [n the course of shooting. the target is struck by one
projectile. What is the probability of a projectile which has struck the target
belonging to Cannon No.2?
(Lueknow Univ. B.Sc., ,1991)
Ans. 0493
(c) Suppose 5 mep out of 100 IlDd 75 women Ol,lt of 10.006 are colour.blind.
A colour blind person is chosen at random. What is the probability of I)is being
male? (Assume.males and females to be in equal number.)
Hint. 1 = Person is a male, E2 = PersOn is a female.

478

Fundamentals of Mathematical Statistics

A =.Person is colour blind.

Then P (E I ) = P (E1 ) = Yi, P (A lEI) = 005 , P (A I E1) = 00025.
Hence find P (EI I,A).
8. (a) Three machines X, Y, Z with capacities proportional to 2:3:4 are
producting bullets. The probabilities that the machines produce defective are 01,
02 and 01 respectively. A bullet is taken from a day's production and found to be
defective What is tflpo orohability that it came from machine X ?
[Madras Univ. B.Se., 1988]
(b) In a f~ctory 2 machines MI and M1 are used for manufacturing screws
which may be uniquely classified as good or bad. MI produces per day nl boxes of
screws, of which on the average, PI% are bad while the corresponding numbers for
M1 are n1 andpz. From the total production of both MI and M1 for a certain day, a
box is chosen ~t r~ndom, a screw taken out of it and it is found to be bad. Find the
chance that tbe selected box is manufactured (i) by M I , (if) M1
Ans. (;) nl PI/(nl PI + n1P1) (ii) n1pz/(nl PI + n2P1)
9. (a) A man is equally likley to choose anyone of three routes A, B, C from
his house to the railway station, and his choice of route is,not influenced by the
weather. If the weather is dry. the probabilities of missing the train by routes A, B,
C are respectiv~ly 1/20, 1/10, 1/5. He sets out on a.dry day and misses the train.
What is the probability that the route chosen was C ?
On a wet day, the respective probabilities of missing the train by routes A, B,
Care 1/20, 1/5, 1/2 r~pectively. On the average, one day in four is wet.lfhe misses
the train, what is the probability that the day was wet?
[Allahabad Univ. B.se., 1991]
(b) A doctor is to visit the patient and from past experience it is known that
the probabilities that he will come by train, bus or scooter arerespectively 3/10,
1/5, and 1/10, the probabililty that he will use some other means of transport being,
therefore, 2/5. If he comes by train, the probability that he will be late is, 1/4, if
by bUs 1/3 and if by scooter 1/12, if he uses some other means of transport it can
be assumed that he will not be late. When he arrives he is .late. What is the
probability that (i) he comes by train (if) he is not.late?
[Burdwan Univ. B.Se. (Hons.), 1990]
Ans. (i) 1/2, (ii) 9/34

10. State and prove Bayes rule and expalin why. in spite of its easy deductibility
fromthe postulates of probability, it has been the subject of such extensive
controversy.
In th~ chest X-ray tests, it is found that the probability of detection when a
person has actually T.B. is 095 and probabiliIty ofdiagnosing incorrectly as having
T.B. is 0002. In a certain city 01 % cf the adult population is suspected to be
suffering from T.B. If an adult is selected at random and is diagnosed as having

Theory of' Probability

479

T.B. on the basis of the X-ray test, what is the probability of his.actually having a
T.B.?
(Nagpur Univ. B.E., 1991)
Ans. 097
11. A certain transistor is manufactured at three factories at Bamsley, Bradford
and BrisLOl.lt is known that the Bamsley facLOry produces twice-as many transisLOrs
as the Bradford one, which produces the same number as the BrisLOI one (during
the same period). Experience also shows that 02% of the transistors produce4 at
Bamsley and Bradford are faulty and so are 04% of those produced at BrisLOI.
A service engineer, while maintaining an electronic equipment, finds a defective transistor. What is the probability that the Bradford facLOry is to blame?
(Bangalore Univ. B.E., Oct. 1992)
12. The sample space CO!1Sists of integers from -I to 2n which are assigned
probabilities proportionalLO their logarithms. Find the probabilities and show tbat
the conditional probability of the integer 2, given that an even integer occurs, is
log 2
[n log 2 + log (n ! ) ]
(t.ucknow Univ. M.A., 1992)
(Hint. Let Ej : the event that the integer 2i is drawn, (i = 1,2, 3, ... , n ).
A : the event of drawing' an even integer.
11

=>
But

A = E.
P (Ej )

E1 U

=k log (U)

..

E~

=>

P (A)

= 1: P (E

j)

;= )

(Given)

11

11

P(A)=k 1: log (fi)=k log

;=)

p(E.1 A)=

n (2i) = k[nlog2+ log, (n!)]

;=)
10g(U)

log 2 + log (n !) ]
13. In answering a question on a multiple choice test, an examinee either
knows the answer (with probability.p), or he guesses (with probability 1 - p).
Assume that the probability of answering a question correctly is unity for an
examinee who knows the an~wer and 11m for tI)e examinee who guesses, ~here m
is the number of multiple choice alternatives. Supposing an examinee answers a
question correctly, what is the probability that he really knows the answer?
[Delhi IJniv. M.e.A., 1990; M.sc. (Stat.), 1989]
Hint. Let E. = The examinee knows the answer,
1 = The examinee guesses the answer,
and
A = The examinee answers correctly.
Then P (E.) = p, P (E1) = 1 - p, P (A I E.) = 1 and P (A IE;) = 11m
Now use Bay~ theorem LO prove
..

[n

p(E.IA)= I+(:-I)P

14. DieA has four red and two, white faces whereas dieB has two red and four
White faces. A biased coin is flipped once. If it falls heads, the game d,'ntinues by

Fundamentals of Mathematical; Statistics

480

throwing die A, if it falls taUs die B is to be used.

(i) Show that the probability of getting a red face at any throw is 1/2.
(ii) If the first two throws resulted in red faces, what is the probability of
red face at the 3rd throw?
(iii) If red face turns up at the frrst n throws, what is the probability that die
A is being used?
Ans. (U) 3/5 (iii)

2?A+1

15. A manufacturing finn produces steel pipes in three plants with daily
production volumes of 500, 1,000 and 2,000 units. respectively. According to past
experience it is known that the fraction of defective outputs produced by the three
plants are respectively 0.005, 0.008 and O.OIO.1f a pipe is selected at random from
a day's total production and found to be defective, from which plant does that pipe
come?
Ans. Third plant.
16. A piece of mechanism consists of 11 components, 5 of type A, 3 of type
B, 2 of type C and 1 of type D. The probability that any particular component will
function for a period of 24 hours from the commencement of operations without
breaking down is independent of whether or not any other component breaks down
during that period and can be obtained from the following table.
Component type:ABCD
Probability:(}60 70 30 2
(i) Calculate the probability that 2 components ch9~n at random from the 11
components will both function for a period of 24 hours from the commencement
of operations without breaking down.
(ii) If at the end of.24 hours of operations neither of the 2 components chosen
in, (i) has broken down, what is the.probability that they are both type C .COInponents.
l:ii!1t.
(i) Required probability =_1_ [ SCi x (06)1 + 3C1 (). 7)1 + lCl (03)1

"Cl
+ sC t x ~Ci x 06 x 07 + sC t x lCI x (06) x (03)
4- sC t x let x (06) x (02) + 3CI x lCt x 07 x 03
+ 3Ct x ICI x 07 x (}2 + lCI x ICI x 03 x 02]
=p (Say).
(ii) Required probability (By Bayes theorem)

lCl x (0,3)1 009

p
. p
410. Geometric probability. In renuuk 3. 431 it was pointed out that the
classical. definition of probability fails if the total number of outcomes of an
('~periment is infinite. Thus. for e~~mple. if we are interested in finding the

;l

481

Theory of Probability

probability that a point selected'at random in a given region will lie in a,specified
part of it, the classical definition of probability is modific<l and extended to what
is called geometrical probability or probability in continuum, In this case, the
general 'expression for probability 'p' is given by
_ Measure of specified part of the region
pMeasure of the whole region

where 'measure' refers to the length, area or volume of the region if we are dealing
with one, two or three dimensional space respectively.
Example 434. Two points are taken at random on the given straight line of
length a. Prove that the probability of their distance exceeding a given length

c (<: a).is equalto (I - ;

J.

[Burdwan Univ. B.Sc. (Hons.), -1992;.l)elhi Univ. M.A. (Econ.), 1987]

So~utio",. Let f and Q be any two points taken at random op the given straight
line AB of length a: . Let AP x and AQ y,

(0 ~ x ~ a, 0 ~ y ~ a).
Then we want P {I x - y I > c}.
The probability can be easily calculated geometrically. Plotting. the lines
x - y =e and y - x =e alongthe co-ordinate axes, we get the fol1owing diagram:

Since 0 ~x~ a, 0 ~y ~ a, total ~a=a. a 7' a1

Area favourable to the event I x - y I > e is given by
A LMN + h. DEF = LN. MN .f. EF .DF'

=! (a- d+! (a -d=(a _e)2

2

P(lx-yl>e)=

:1

(a 'e)2

. 2

ef

I-~)

Fundamentals of Mathematical StatistiCs

482

Example 435. (Bertrand's Problem). If a chord is taken at random in a

circle, what is the chance that its length.l is not less than 'a' ., the ra4~us of the
circle?
Solution. Let the chord AB make an angle e with the diameter AOA ' of the
circle with centre 0 and radius OA=a. Obviously, e lies betwseen -1t/2 and 1t12.
Since all the positions of the chord AB and
consequently all the values of e are equally
likely, e may be regarded as a random variable which is unifonnly distributed c/. 81
A'
',over (~1t/2, 1t/2) with probability density
function

1
(e) = 1t

; - 1t/2 < e S

1t/2

L ABA " being the aQgle in a semicircle, is a rigt-t angle. From A ABA' we have
AB = cose

AA'
1=2acose

The required probability 'p~ is given by

p=P(1 ~ a)=P(2acose ~ a)
=p(cose ~ 1/2)=p(le I S 1t/3)
1tI3

,J

f (e) de = ~

-1tI3

1tI3

J de = ~

-1tI3

Example 436. A rod of length' a' is broken into three parts at random. What
is the probability that a triangle can beformedfrom these parts?
Solu~ion. Let the lengths of the three parts of the rod' be x, y and a - (x + y).
Obviously, we have
x>O; y>Oandli+y<a =.. y<a-x
... (*)
In order that these three parts fonn the sides of a triangle, we should have
~+y>a-~+~

and

y>--x'
2
a

x+a- (x+ y y

y<-

y + a - (x + y) > x

y<-

...(**)

2
since in a triangle, the sum of any two sides is greater than the third. Equivalently.
(..) can be written as
a

--x<y<2
2
Hence, on using (.) ~d (... ), the required probability is given by

... (.**)

Theory

or l'robability

483
aI2

al2

aI2

.::.. .[_<,;:;..,al2:;<.,.L....::.x_d---,Ydx_"
a a-x

f f
o

l [~- (~-

x )] dx

f (a -x) dx

dydx

rf I~/~

a2/8

-1-(a2-x)2'1~ = a2/2 = '4

Example 437. (Burron's Needle -Problem). A vertical Qoard is ruled with
horizontal parallel lines at constant distance 'a' apart. A needle of length IJ< a)
is thrown at random on:the table. Find the probability that it will intersect one of
the lines.
Solution. Let y denote the distance from the cenlre of the needle to the nearest
parallcl and ~ be angle fonned by the needle with this parallel. The quantities y
and ~ fully detennine the position of the needle. Obviously y ranges from 0 to
0(2 (since I < a) and, ~ from 0 to 1t
Since the needle is ,dropped randomly, all poss~ble ~alues of y and ~ may be
regarded as equally lik~ly and consequently the joint probability density function
fCy,~) of y and ~ is given by the unifonn
dislribution.( c.f. 8.1 ) QY

1
I

JCy,~)=k.; O~~S;1t,

Os; Y s; a12,
...(*)
a
where k is a constant.
The needle will jnt~lsect one of the
lines if the distance of its cenlre from me
line is less than ~ I sin~, i.e., the required
event can be represented by the inequality
0< y < I sin ~ . Hence.the required probability p is given by

('~.)12

It

J J

-0

f(Y,

~) dy df!>

0
-alZ

1= . It

J J
o

f(y,~)djiJ~

0
It

=
-

~J0

sin~d~

.
l-cos~I~=

(a/2) .1t
a1t

21
an

Fundamentals of Mathematical Statistics

484

EXERCISE 4 (e)
1. Two points are selected at random in a line AC of length 'a' so as to lie on
the opposite sides of its mid-point O. Find the probability that the distance between
them is less than a/3 .
2. (a) Two points are selected at random on a line of length a. What is the
probability that .10ne of three sections in which the line is thus divided is less than
al41

Ans. 1116.
(b) A rectilinear segment AB is divided by a point C into two parts AC=a,

CB=b.PointsXand Y are taken at random onAC and CB respectively. What is the

probability thatAX,XYand BY can form a triangle?
(c)ABG is a straight line such thatAB is 6 inches and BG is S' inches. A pOint
Y is chosen at random on the BG part of the line. If'C lies between Band G in such
a way that AC=t inches, find
(i) the probability that Y will lie' in BC.
(ii) the probability that'Y will lie in CG.
What can you say about the sum of these probabilities?
(d) The sides of a rectangle are taken at random each less than a and all lengths
are equally likely. Find the chance that the diagonal is less than a.
3. (a) Three points are taken at random on the circumference of a circle. Find
the chan~ that they lie on the same semi- circle.
(b) A chord is drawn at random in a given circle. Wliat is the probability that
iris greater than the side of an equilateral triangle inscribed in that circle?
(c) Show that the probability of choosing two points randomly from a line
segment of length 2 inches and their being at a distance of at least I inch from each
other is 1/4.
[Delhi Univ. M.A. (Econ.), 1985]
4. A point is selected at random inside a circle. Find the probability that the
point is closer to the centre of the circle than to its circumference.
S. One takes at random two points P and Q on a segment AB of length a
(i) What is the prooabiJity for the distancePQ being less than b ( <a )1
(ii) Find the chance that the distance between them is greater t~an a given
length b.
6. Two persons A and B, make an appointment to meet on a certain day at a
certain place, but without fixing the time further than that it is to be between 2 p.m.
and 3 p.m and that each is to wait not longer than ten minutes for the other.
Assuming that each is independently equally likely to arrive at any time during the
hour, find the probability that they meet.
Third person C, is to be at the same place from 210 p.m. until 240 p.m. on
the same day. Find the probabilities of C being present when A and B are there
together (i) When A and B remain after they meet, Oi) When A and B leave as soon
as they meet.

Theory of Probability

485

Hint. Denote the times of arrival of A by x and of B by. y. For the meeting to
take place it is necessary and sufficient that
Ix-yl<lO
We depict x and y as Cartesian coordinates in the plane; for the scale unit we
take one minute. All possible outcomes can be described as points of a ~qujife with
side 60. We shall finally get [cf. Example 434, with a == 60, c =lO]
P [I x - y 1< lO] = 1 - (5/6)~ = 11136
7. Tile outcome of an experiment are represented by points 49 the square
bOunded by x = 0, x = ~ and y = 2 in the ~-plane., If the probability is distributed
~niformly, determine the probability that Xl + l > 1
Hint.
~ dx dy = 1 ~ dx dy
Required probability P () =

where E is the region for whiCh

E'

Xl +

'l:> 1 and' E'

is the region for which

Xl+ l~ 1.
1 1

4P ()

=4 -

J Jdx dy =3
o

3
4

P()=-

8. A floor is paved with tiles, each tile, being a parallelogram such that the
distance between pairs of opposite sides are a and b respectively, the length of the
diagonal being I. A stick of length c falls on the floor parallel to the diagonal. Show
that the probability that il will lie entirely;on one tile is

(1-7

If a circle of diameter d is thrown on the floor, show that the probability that
it will lie on one tile is

(I-~J (I-~)
9. Circular discs of radius r are thrown at random on to a plane circular table
of radius R which is surrounded by: a border of uniform width r lying in'the same
plane as the table. {f the discs are thrown independently and at random, end N stay
on the table, show that the probability that a fixed point on the table but not ~ the
border, will be covered is

1- [1- (R:d J
SOME MISCELLANEOUS EXAMPLES
Example 438. A die is loaded in' such a manner thatfor n=l. 2, 3. 4. 5./6.the
probability of the face marked n. landing on top when the die is rolled is proportional to n. Find the probability that an odd number will appear on tossing the d,e.
[Madras Univ. D.SIe. (Stat. Mafn),1987]

486

Fundamentals of Mathematical Statistics

Solution. Here we.'are given

P (n) oc n or P (n) =kn, where k is the constant of proportionality.
Also P(l) + P(2) + ..'p(6) = I => .k( I + 2 + 3 + 4 + 5 + 6) = I or k = 1/21
,
.'
.,'
1+3+5
3
Requlred Probablhty =P(1) + P(3) + P(5) =
21
= '7
Example 439. In.terms ofprobability :
PI =peA) , P'1. =PCB) , P'l =peA (\ B), (PI. P'1., P3 > 0)
Express the following in tenns of PI ..p'1., P'l .
(a) peA u B). (b),PeA u Ii). (c) ,PG\ (\ B). (d) p(i\ u B). (e) peA (\ B)

if) P( A (\ Ii). (g) P (A I B), (h) P (B I A),

(i) P

[A (\ (A u

B)}

S9lution.
P( Au B) = 1- peA u B) = I - [P(A) + PCB) - P(AB)].
1- PI- P2+ P3.
(b) f( A u Ii) P (A () B)= 1- P (A (\ B) 1- Pl
(cj P( A (\ B) P (B - AB) P (B) - P (A (\ B) =,n'1. - Pl
(d) P eA u B) = P G\) + P (B) - P (A (\ B) = I - PI + P'1. - (P '1. - Pl)
::;l-PI+P'l
(e) P( A (\ Ii) = peA u B) = 1 - PI - Pz + P3.
[Part (a)]
(a)

=
=
=

if)
(g)
(h)
(i)

P( A (\ B ) P (t\ - A (\ B) = peA) - peA (\ B) PI - P3

P(AI-B)= P(A<'1B)/P(B)= p,IPz
P (B ~A1 = P( A (\B )/P (A)= <P2 - Pl)/(I- PI)
P[A(\(A~B)]~ p[(AnA),u(A(\B)

= P ( A(\ B) = P2 -

Example 440. Let peA)

tween the
(a)
(b)
(c)
(d)

= ]

[ .: A (\ A ~

P3

=P, P (A I B) = q, P (B I A) == r. Find relations be-

ill:Uhbers P. q. r for thefollowing cases:

Events A and B are mutually exclusive.
A and B are mutually exclusive and collectively exhaustive.
A is a subeyent ofB; B is a sribevent ofA.
A and B' are mutually exclusive.

tDelhi Univ. B.Sc. (Maths Hons.) 1985)

Solution. Frpm given data : P (A) =p, P (:A (\ B) =P (A) P (B.I A) =rp
P (B) = P (A (\ B)

..

P (A IB)

!I!.
q

(a) P(A(\.B)= 0
=>
rp= O.
(b) P(AilB)=O and P(A)+ P(B)= 1
=> p'(q+r)= q; rp= 0 ='> pq= q => p=1 V q=O.
(c) A~8 => A,(\B=A or'P(A(\#)='p(A) => rp=p => r=1 Vp=O.
B ~ A => ,A (\ B' B or peA (\ B) P(B)
=> rp=(rplq) or rp(q-I)=O => q=1
(d). P (:4 Ii) =. J - P (A u B) => 0 = I - [P (A) + P (B) - P (A (\ B)J

Theory of Probability

So

=1 + P (.4 n 8) =>
P (q+ r)= q (1 '+ pr).

P (A) + P (8)

4-87

p[I+(r/q):;: I+rpl

Example 441. (a) Twelve balls are distributed at random among three bOxes.
What is the probability that the first bo~ will contain 3 balls?
(b) If n biscuits be distributed among N persons, find't~ chance that a
particular person receives r ( < n ) biscuits.
[Marathwada Univ. B.sc. 1992]
Solution. (a) Since each ball can gOlo'any one ofthethree bbxes, there. are 3
ways in which a ball can go to anyone of the three boxes. Hence there are 312 ways
in which. 12 balls can be placed in the three boxes'.
N~mber of ways in whic~ 3 balis out of 12 can go. to the fIrSt box is 12C3. Now
the remaining 9 balJs are to be placed in 2 boxes and'this can be done in 2' ways.
Hence the total number of favourable cases = 11C3 x 29.
'
,
uC3 x29
:. Required probability ;::; ---:-=--311
(b) Take anyone biscuit. This can be given to any one-'of the N beggars so that
there are N ways of distributing anyone biscuit. Hence I)le total number of ways
in which n biscuit can be distributed at random among N beggars
= N . N .. , N ( n times',) = N.
1 r' biscuits can be given to an'y particular beggar in c, ways. Now we are left
with (n'- r) biscuits which are'to be distribut among the remaining (N - I)
beggars and this can be done' in (N - 1)"-' ways.
. . Number of. favourable cases = C,. (N - IX -,
Hence, required prob;,lbility

"C (N -1)"-'
N"

= '

Example 442. A car is parked among N cars in a row, not,ilt either end. On
his return the owner finds that exactly r of the N pJaces are still occupied.Whaiis.
lhe probability that both neighbouring places are empty?
Solution. Since the owner finds on return lhat exactly r of the /Ii places
(including.Qwner's car) are occupied, the exhaustlve number of cases for such an
arrangement is N-1C,_1 [since the remaining r!... 1 cars are to be parked in 'the
remainingN - I places and thiscanbedone in N-1C,_1 ways].
Let A denote the event that both the neighbouril)g places to owner's car are
empty. This requires the remaining (r - 1) cars to be parked in 'the remaining
N- 3 places and hence the num6er of cases favourable to A is N- 3C;_I. Hence
N-'3' ,
'
P(A) =
C,_~ = (N-r)(N-r-l)
N-1C,_1

(N' - I)(N - 2)

Exam pIe 443. What is the probability thai at least two out of n people have
lhe same birthday? Assume 365 days in a year and that all days are equally likely. ,

488

fundamentals of Mathematical Statistics

Solution. Since the birthday of any person can fallon any one of the 365 days,
(he exhaustive number of cases for the birthdays of n persons is 365".
If the birthdays of all the n .persons fallon different days, then the number of
favourable cases is
365 (365 -1) (365 - 2) .... [365 -(n-l)],
because in this case the birthday of the first person can fallon anyone of 365 dayS,
the birthday of -the second person can fall on anyone of the remaining 364 days
and soon.
Hence the probability (p) that birthdays of all the n persons are different is
given by:
_ 365 (365 - 1) (36~ - 2) ... [ 365 - (n - 1)]
P365"
= (I - 3!5 ) (1 -

3~5 ) (1 - ~ ) '"

(1 -

n3~51 )

Hence the required probability that at least two persons have the same birthday
is
1 - P = 1- (1- 3!5/)

(J - 3~5') (1- 3~5 ) ... (1- n3~51 )

Example 444. A five-figure number is formed by the digits'O, 1, 2,3,4

(without repelitiofl).;Find the probabilifJI-lhtJt the number formed i~ divisible by 4.
[Delhi Univ. B.sc. (Stat. Hons.), 1990)
Solution. The total number of ways in which the five digits 0, 1, 2, 3,4 can be
arranged among dtemsel ves is 51. Out of these, the num ber of arrangements which
begin with 0 (and, therefore, will give only 4-digited numbers) is 41. Hence the
total number of five digited nwnbers that can be formed from the digits 0, 1,2,3,
4is
5! -4! = 120-24=96
The number formed will be divisible by 4 if the number formed by the two
digits on extreme right (i.e., the digits in the unit and tens places) is divisible by 4.
Such numbers are :
04 , 12 ,20 , 24 ,32 , and 40
If the numbers end in 04, the_remaining three digits, viz.,l, 2 and 3 can be
arranged among dtemselves in 3 I ways. Similarly, the number of arrangements of
the numbers ending with 20 and 40 is 3 ! ~n each case.
If the numbers end with 12, the remaining three digits 0,3 ,4 can be arranged
in 3 ! ways. Out of these we shall reject those numbers which start with 0 (i.e., have
oas the first digit). There are ( 3 - t ) ! =2 ! such cases. Hence, the number of five
digited numbers ending with 12 is
31-2!=6-2=4

Theory of Probability

Similly the number ,of 5.digited numbers ending with 24 and 32 each is 4.
Hence the total number of favourable cases is
3 x 3 ! + 3 x 4 = 18 + 12 ~ 30
. ed probabT
30 J6
5
Hencerequlf
llty= 96=
Example 445. (Huyghe,n's problem). A and B throw alternately with a pair
of ordinary dice. A wins if he throws 6 be/ore B throws?, and B wins if he throws
7 be/ore A throws 6.I/A begins, show that his cliance-o/winning is 30 161
[Dellii Univ. B.Sc. (Stat. Hons.)t 1991; Delhi Univ. B.Sc.,1987]
Solution. Let EI denote the event of A 's throwing '6' and E2 the event of B's
throwing '7 with a pair of dice. Then 1 and 2 are the complementary events.
'6' can be obtained with two dice in the following ways:
(1,5 ), (5, I ), (2, 4), (4, 2), (3, 3), i.e., in 5 distinct ways.
5
,.
5
31
..
P (E 1) = 36 and P (~I) = I - 36 = 36

'7' can be obtained with two dice. as follows:

(1,6), (6,1), (2, 5), (5, 2), (3,4), (4, 3), i.e., in 6 distinct ways.
'.
6
1
1 5
P(E2) = -,- = -

and P,(ElJ = 1-- = 36 6

'6 6
If A starts the game, he will win in the following mutually exclusive ways:
(i) EI happens (ii) 1 n 2 n EI happens
(iii) 1 n 2 n 1 n 2 n EI happens, and so on.
Hence by addition theorem of probability, the required probability of A's
winning, (say), P (A) is given by
P (A) =P (i) + P (U) + P (iil) + .. ,
= P (E 1) + P (1 n 2 n E 1) + P (1 n 2 n 1 n 2 nE1) + ....
=P(E1) + P(I) P(2) P(E1) + P(;) P(2) P(l) P(E2) P(E1) + '"
(By compound proba,bility theorem)
5 31 5 5 31 5 31 5 5
= 36 + 36
36 + 36
36
36 + .. ,

x'6 x

5/36
31
1- 36

x'6 x x'6 x

30

5 - 61

x'6

Example 446. A player tosses a coin and is'to score one point/or every head
and two points/or every tail turned up. He is to play on until his score reaches or
passes n. Ifp~ is the chance 0/ allaining exactly n score, show lha!
1
P~=2 [P~-I

and hence find the value o/p~.

P~-2],

[Delhi Univ. B.Sc. (Stat. "ons.),1992]

Fundamentals of Mathematical Statistics

Solution, The score n can be reached in the following two Jrlutually exclusive
ways:
(i)By throwing a tail when score is (n - 2), and"
(ii)By throwing a head when score is (n - I),
HenCe by addition the<,>~m of probability, we get

P~ P.(i) + P (ii)

=t ,P~-2 +! 'P~-I =t (P~-I + P~-2 ~

To find P~ explicitly, (*) may. be re-written as

P~

+ 2.1 P~-I =.P,;-I + 2.I P~-2

= Pl'+'2 P1
S~nce the score 2 can be obtained as
(i)Head in fU'St throw and head in 2nd throw,
"(ii)TaiI in tbe first throw, we have
111113
.
1
P2=- -+-=-+-=- and obvIOusly PI =2'2

Henc~,

from (u), we get

1
3 1 12 1 2 1 2
P~ + 2. P~ -I = 4 + 2.' 2. = 1 = 3" + 3" = '3 + 2. ' 3"

P~ PN - I P2 -

i =(- ~) (p. i =~ - t) (p. -

12

~)l

-:~)

t = ( -~) (PI - i)

Multiplying all the above equations. we get

P~ -i=(-~)1I-1 (PI-i>

= (_.!.)N-I (!_~)= (-I)"

::

p.=

2
(3'+
-

::

I)" I

'3'I

= .!3 [2 + ( -

2"

1)"

.1 !

2" ' '3

1.]
2"

Example 447. A coin is tossed (m+n) times, (mn). Show that the probability
. L __ J _ ' n + 2
of at Ieast m consecutive
m:uu3 IS "Z" + 1 '

- [Kurukshetra Univ. M.Sc.I990; Calcutta Univ.8.Sc:.(IIo05.),I986]

491

Theory of Probability

Solution. Since m >n, only one sequence of m consecutive heads is

possible. This sequence may start either w~th the first toss or second toss or third
toSS, and so on. the last,one will be starting with (n + l)th toss.
Let Ei denote the event that the sequence of m consecutive heads starts with
ith toss. Then the required probability is
P (EI )

+ P (E2) + ... + P (E

M:

I)

...

(*)

Now P(E I ) =P [Consecutive heads in f~rst m tosses and head or tail in the rest]
=

P (E2)

(~J

=P [Tail

in the first toss, followed by m cOJ)~utive'heads and

head or tail in the nexU

= ~ (~J = 2}+ I
In general,
P (E,)

=P [tail in the (r -

I)th trial followed by m consecutive heads

and head or tail in the next]
1"
.
= '21 ('I'2 = 2""1-'
-V r = 2, 3, ... , n + 1.

J-"

Substituting in (*),
R
'ed probabT
1
n 2+n
eqUlC
llty:: 2'" + 2"'+ 1= 2"'+ I
Examp~e 448. Cards are dealt one by one from a well-shuffled pgck until an
ace appears. Show that the probability that exactly n cards are dealt before the
first ace appears is
4(51 - n) (50 - n) (49 - n)
52.51.50.49
[Delhi Univ. B.Sc. 1992]
Solution. Let Ei denote the event diat an ace appears when the ith card is
dealt. Then the required probability 'p' is given by
p:;= P '[Exactly n cards are dealt before the first ace appears]
= P [The first ace appears at the (n + l)th dealing]
= P (E I fi E2 fi E3 fi .,. fi EM-I fi EM fi Eu l ) ,

= P (EI )

P (E2 1E I ) P (E3 lEI fi EJ ...

x P (EM I EI
, fi E2 fi . '.' fi EM-I) X P (E. + I I EI

fi

E2 fi

... fi

E.)

(*)

Now
P

~~

(EI ) =
-

P (E2

47

I EI ) = 51

492

Fun(lamentals of Mathematical Statistics

4
P(E~_IIElnE2n ... nE~_J= 52-(n-2)

5O-n

P(E~_IIEI nE2n ... nE._J'~ 52-(n-2)

p ( E~ I EI n E2 n ... n E~ -I) = 52 ~ (n - l)

P (E~ I EI

49- n
n E2 n ... n E._I) = 52 _ (n _ 1)

P ( Eu d EI

4
n E2 n ... n E~) = 52 _ n

, :- Hence. from (*) we get

~

-[48 47 46 45 44 43
_ 52 - n
p= 52x5tx50x49x48x47x ... x 52 - (n - 4)
.x
51 "
- n x 50 - n x 49 - n x 4-]
52- (n- 3) 52- (n- 2) 52- (n- I) 5+- n
_ (51 - n)(50 - n)(49 - n) 4
52x51 x50x49

Example 4-49. If/our squares are cliosen at random on d chess-board,find

the chance 'that they slwuJd be in a diagonal line.
[Delhi Univ. B.Sc. (Stat. Hons.), 1988]
Solution. In a chess-board there are 8 x 8 = 64 squares as shown in the
followim~ diagram.
Let us consider the number of ways in
A
which the 4 squares selected at random are
Al ~~-+--~~~-+-+~
in a diagonailine parallel to AB. Consider
the II ABC. Number of ways in which 4
AI I~~~__~~~-+-+~
A3
selected squares are along the lines ~ B
~ ~~~~~--~~~

.A,B,. A2B2. AIBI

and AB are C4 5C

"C '7 C4 and sC. respectively.

Similarly. in llABD there are an

equal number of ways of selecting 4
squares in a diagonal line parallel to AB.
Hence, total number of ways in which
the 4 selected squares are in a diagonal
line
parallel
to
AB
an
2 (C. + 'C4 + 6C. + 7C.) + C

Theory of Probability

493

Since there is an equal number of ways in which 4 selected squares are in a

diagonal line parallel to CD, the required number of favourable cases is given by
2 [ 2( 4C4 + sC4 + 6C4 + 7C4) + 8C4]

Since4 squares can be selected outof 64 in 64C4 ways, the required probability
is

= 2 [2( 4C4 + sC~ + 6C~ + 7C4) + 8C4]

64C4
_ [4 ( 1+ 5 + 15 + 35,) + 140] x 4 !, _
91
94 x 63-x62 x.61
-.158844
Example 450. An urn contains four tickets "f(lrked with numbers 112, 121,
211,222 and one ticket is drawn 'at random. Let Ai, (i=1, 2, 3) b.e the event that ith
digit of the number of the ticket drawn is 1. DiscuSs the independence of the events
A"Al and A3. [Qelhi Univ.I}.Sc.(Stat. Hons.),1987; Poona Univ. B.Sc.,1986]

Solution. We have
P(A I) = ~ =

t = P(A

= P(A3)

1)

AI n A2 is the event that the fIrst two digits in the number which the selected ticket
bears are each equal to unity and the only favourable case is ticket with number
112.
1

P(A I nA1) = '4 = 22

= P(A I) P(A 1 )

Similarly,
1

P(Al n A3) = '4 = P(Al) P(A3)

and

P(A3 n AI) = ~ = P(A3) P(A I)

Thus we conclude that the events Ai> A2 and A3 are pairwise independeni.
Now P(AI n:A3 nA3) = P {all the three digits in the number are 1's}
= P(cI
= 0 :F P(A I) p{.('h) P(A3)
Hence Ai> A2 and A3 though pairwise independent are not mutually independent.
Example 451. Two fair dice are thrown independently. three events A, B
and C are defined asfollows:
A: Oddface withfirst dice
B : Oddface with second dice
C .: Sum of points on two dice is odd.
Are the events A, B and C mutually independent?
[DelJti Univ. BoSe. (Stat. Hons.) 1983; M.S. Baroda Univ. B.Sc.1987)

Fundamentals o( Mathematical Statistics

494

Solution. Since each of the two 4if:e can show anyone of the six faces 1,2,3,
4,5, 6, we get:
P(A) = 3 x 6 =

.!

P(B) = 3 x 6 =

.!

36

36

[.: A= (1,3,5) x (1,2,3,4,5,6)]

[ .: B

= (1,2,3,4,5,6) x

,.

(1,3,5) ]

The sum Qf points on two dice will be 0<14 if one shows odd number and the
other shows even number. Hence favourable cases for C are :
(1,2), (1,4), (1,6);
(2, I), (2,3), (2,5);
(3,2), (3,4), (3.6);
i.e., 18 cases in all.
18
I
Hence P(C) = 36 = 2"

(4, I), (4,3), (4,5)

(5,2), (5,4), (5,6)
(6,
(6,3)" (6,5)

n,

Cases favourable to the events An B, A (') C, B (') C and A (') B (') C are
given below:

Event
AnB

Fav,ql,Uable cases
(1,1), (i l 3)~ (1, 5), (3,. '1), (3, 3), (3, 5), (5, 1) (5, 3)

(5,5), i.e., 9 in all.

A(')C

(1,2), (1,4). (1.6), (3.2), (3, 4), (3, 6), (5, 2), (5, 4)

.(5,6). i.e., 9 in all.

B(')C
AnB(')C

(2, 1), (4,1), (6, 1) (2, 3), (4, 3), (6, 3), (2, 5), (4. 5).

(6. 5), i.e., 9 in alf

Nil, because AnB implies that sum of points on two dice is
even and hence (AnB )(')C ell
.
9
1

P(A (') B)

= -36 = -4 = P(A).P(B)
9

P(A (') C) = -

36

P(B.(') C)

and

= 4

= P(A) P(C)

= 369 = 41 =P(B) P( C)

P(A (') B (') C) = P(eII) = 0." P(A) P(B) P(C)

_
Hence the events A, B and C are pairwise independent but not mutually

inde~ndent

Example 452. Let A.,Az, .... A. be independent events and P (At) =Pl.
Further, let P be the probabiJiJy thoi fWne of the events occurs; then sho'(fl thol
p ~ e - tPI

[Agra Univ. M.Sc., 1987]