CHAPTER FIVE 52 Fundamentals of Mathematical Statistics

4. Notalions.-If X is a-real number, the set of all <0 in S s!lch that X( <0 ) =x is
"
Random Variables - Distribution Functions denoted briefly by writing X =x. Thus
5·1. Random Variable. Intuitively by a random variable (r.v) we mean a P (X =x) = p{<o: X (<0) =x.}
real number X connected with the outcome of a random experiment E. For
example, if E consists of two tosses of a coin, we may consider the random varilble Similarly P (X ~ al= p{<o: X (<0) E l- 00, a]}
which is the number of heads ( 0, 1 or 2). . and P>[a < X ~ b) = P ( <0 : X ( (0) E (a,b] )
Outcome: HII liT Til IT Analogous meanings are given to
Value 0/ X : 2 I
,(01
1 o P ( X = a or X= b ) = P { (X = a ) u ( X = b ) } ,
Thus to each outcome 0> , there corresponds a real number X (0)). Since the P ( X =a and X = b ) = P { ( X== a ) n (X = b )}, etc.
points of the sample space S correspond to outcomes, this means that a real number , Illustrations: 1. If a coin is tossed. then
which we denote by X (0)), is defined for each 0> E S. From this standpoint, we
S = {<01, (Jh} where <01 = It, (Jh::; T
define random variable to be a real function on S as follows:
.. Let S be the sample space associated with a given random experiment. A X(<o)=
_
{I, ~f <0 = N
0, If <0 == T
real-valued/unction defined on S and taking values in R (- 00 ,00 ) is called a
olle-dimensional random variable. If the/unction values are ordered pairs o/real X (<oj is a Bernoulli random variable. Here X (<o).takes only two values. A random
numbers (i.e., vectors in two-space) the/unction is said to be a two- dimensional variable which takes only a finite number of values is called single.
random variable. More generally, an n-dimensional random variable is simply a 2. An experiment consists of rolling a die and reading the number of points
function whose domain is S and whose range is a collection 0/ n-tuples 0/ real on the upturned face. The most natural random variable X to cunsider is
numbers (vectors in n- space)." X(<o) = <0; <0 =1, 2, ... ,6
For a mathematical and rigorous definition of the random variable, let us '. If we are interested in whether the number of point,s is even or odd, we consider
consider the probability space, the triplet (S, B, P), ~here S is the sample space, a random. variable Y defined as follows:
viz., space of outcomes, B is the G-field of subsets in S, and P is a probability
function on B. y ( <0 ) = {O,1, ~f <0 ~s even
if <0 IS odd
Def. A random variable (r.v.Y is a function X (0)) with domain S and range
(__ ,00) such that for every real number a, the event [00: X (00) S; a] E B.
3. If a dart is thrown at a circular target. the sample space S is the set of all
points w <.'n the target. By imagining a coordinate system placed on the target with
Remarks: 1. The refinement above is the same as saying that the function the origin at the centre, we can assign various random variables to this experiment.
X (00) is measurable real function on (S, B). A natural one is the two dimensional random variable which ~signs to the point
2. We shall need·'to make probability statements about"a'random variable X <0, its rectangular coordinates (x,y). Another is that which assigns <0 its polar
such as P {X S; a}. For the simple example given above we sbould write coordinates (r, a ). A one dimensional random variable assigns to each <0 only one
p (X S; 1) =P {HH, liT, TH}.= 3/4. That i's, P(X S; a) is simply 'the probability of the coordlnatesxory (for cartesian system), rora (for polar system). Theevent
pfth~ set of outcomes 00 for which X (00) S; a or' E, "that the dart will land in the first qUadrant" can be described by a. random
p (X S; a) =P { 00: X (oo)S; a) variable which a<;signs to each point'W its polar coordinate a so that X (<0) = a and
Since Pis a measure on (S,B) i.e., P is defined on subsetsofB, theabovepro~bility °
then E ={<o : ~ X (<0) ~ 1t12}. '
will be defined only if [ o>:X (OO)S;~) E B, which implies thatX(oo) is a measurable 4. Ifa pair of fair dice is tossed then S= {1,2,3,4,5,6}x'{1,2,3,4,5,6} and
function on (S,B). n (S) = 36. Let X be a random variable with image set '
3. One-dimensional random variables will be denoted by capit8I leuers. XeS) =(I ,2,3,4,5,6)
X,y,z, ...etc. A typical outcome of the experiment (i.e., a typical clement of the" P(X= 1)=P{I,I} = 1/36
sample space) will be denoted by 0> or e. Thus X (00) represents the real number
which the rand<,>m variable X associates wi~ the outcome 00. The values whict P(X = 2) = P{(2,1),(2,2),(l,2)} =3/36
X, y, Z, ... etc., can assume are denoted by lower case letters viz., x, y, z, .:. etc. P(X = 3) =P{(3,1).(3,2),(3,3),(2,3).(l,3)} = 5/36
Ramdom Variables· I>istribution Functions 5·3 Fundamentals of Mathematical Statistics

P (X = 4) = P (4, I), (4,2), (4,3), (4,4), (3,4), (2,4), (1,4) J = 7/36 3. Let S:::: (e\, e2, ••. , en) be the sample space of some experiment and let
Similarly. P(X = 5) = 9/36 and P (X = 6) = 11/36 E ~ S be some event aSsociated with the experiment.
Define'l'E, the characteristic random variable of E as follows:
Some theorems ()~ Random Variables. Here we shall stale (withoUt proof)
some of the fundamental results and theorems on random variables. ' () I if eo E E-
'l'E ei ={ 0 if ei Ii!: E .
Theorem 5·1. A function X(oo) from S to R (- 00 , 00) is a random variable if
In other words, 'l'E is equal to 1 if E occurs, and 'l'E is equal to 0 if ~ does not
and only if
occur.
{oo:X(oo)<a] E B Verify the following properties of characteristic random .variables:
Theorem 5·2. If XI and Xl are-rdndom yariables and C is'a constant then (i) '1'" is identically zero , i.e., '1'" (ed = 0; i = 1,2, ... , n
CXI , XI + Xl, XIXl are also random variables. (ii) 'l's is identically one , i.e., 'l's (ei ) = 1 ; i = 1,2, ... ,n
Remark. It will follow that CIXI + C1Xl is a random vari~ble for constants e
(iii) =F ~ 'I'd ed = 'l'F (ed ; i ~ l. 2, ... ,n and conversely
CI and Cl . In particular XI - Xl is a r.v. (iv) If E ~ F then 'I'd ed s. 'l'F (ei); i = 1,2, ... ,n
. ~
Theorem 5·3. If {X. «(J), n ~ 1] arc random variabl~s then (v) 'l'E ( ei ) + 'l'E ( ei) is identically 1 : i = 1,2, ...., n
sup X. (00), in[ X. (00), lim sup X. (00) and lim in[ X. (00) are·all ran- (vi) 'l'E /"'IF ( ed '=' 'I'd ed 'l'F ( ed; i ='1~ 2, ... , n
/I /I
(ViirWEVF ( ei) ='l'E (ei) + 'l'F ( ei) - 'l'E (ei ) 'l'F (ei), for i =1, 2, ... , n.
dom va: iables, whenever-they are finite for all 00.
S.2. Distribution Function. Let X be a r.v. on (S,B"P). Then the function:
Theorem 54. If X is a random variable then
Fx(x)=P(XS.x)=P{ro:X(Cll)S. x}, - oo<x<oo
(i) ~ where ( ~ )< (0) = 00 if X (00) = 0 is called the distribution function (d,f.) of X.
(U) X + ( (0) = max [0, X ( 00 ) ] If clarity permits, we may writeF(x) instead of Fx (x). ...(5·1)
(iii) X - ( (0) = - min [0, X ( 00 ) ] ~·2·1. Properties of Distribution Function. We now proceed to derive a
Ov) IX I number of properties common to all distribution functions.
are random variables. Property 1. IfF is the df. of the r.v. X and if a < b, then
Theorem 5·5. If Xl and X2 are random variaQles then P(a<XS.b)= F(b)- F(a)
(i) max [XI. X2 1and (li) min [XI. X1 1are also random variables. Proof. The events a<Xs. b' and 'XS. a' are disj(,i~t and their union is the event
I

Theorem 5·6. If X is a r.v. andf(·) is a continuous fun~tioJl, then 'X~ b' .Hence by addition theorem of probability
[(X) is a r.v.
P(a<Xs. b)+ P(Xs. a)= P(XS. b)
Theorem 5·7. If X is. a r.v. and f(.) is an increasing function, then
[(X) is a r.v.
~ P ( a < X S. b ) = P ( X S. b ) - P ( X S. a ) =F ( b) - F(a) ...(5·2)

Corollary. If [is a function of bounded variations on every finite interval Cor.!.

[a,b], hIld X is a r.v. then [(X) is a r.v. P(aS.XS.b)=P{(X= a)v. (a<X;S;b)}
(proofs o[ the above theorems are beyond the scope of this book) =P(X= a)+ f(a<XS.b)
EXERCISE 5 (a) (using additive property of P)
1. Let X be a one dimensional random variable. 0) If a< b, show that the =P ( X =a ) + [F ( b ) - F (a)] ... (5·2 a)
two events a < X ~ b and X ~ a are disjoint, (U) Determine"the union of the two Simil~ly, we get I

events in part (i), (iii) show that P ( a < X ~ b) =P( X ~ b") - P( X ~ a). P(a<X<:b) =.P(a<Xs. b)-P(X=b)
2. Let a sample space S consist of three elements 001 , roz, and ro,. Let
=F(b)- F(a)- P(X= b) ...(5·2 b)
P(OOI) = 1/4, P(roz) = l/'2.and P(0)3) = 1/4. If X is a random variable defined' on
S by X (001) = 10, X (001) = :-03, X (0)3) =15, find P ( - 2 ~ X ~ 2). P(aS. X<b)~ P(a<X<b)+ P(X=a)
Jumdom Variables· Distribution Functions 5·5 ·'undamentals of Mathematical St,tistics
= F ( b) - F (a) - P (X:::: b) + P (X:::: a) ... (5·2 c) .. O~F(-oo)~F(oo)~J
Remark:. When P (X =a) =0 and P(X =b) =0, all four events a~ X ~ b. (*)and(**)giveF(-oo)= 0 and F(oo)= 1.
a < X < b, ~ X < b and a < X ~ b have the same probability F(b) - F(a). Remarks. 1. Discontinuities of F(x) are at most countable.
Property 2. If F is the df. of one-dimensional r.v. X, then 2. F(a)- F(a- 0)= liin P(a-h~ X~ a). h> 0
(i) 0 ~ F (x) ~ 1, (it) F (x) ~ F (y) if x < y. II -+0 '
In other words, all disttibution functions are monotonically non-decreas.ing F(a)- F(a- 0)= P(X= a)
and lie between 0 and 1. and F ( a + 0) - F ( a ) =
lim P ( a ~ X ~ a+ h) = 0.' h > 0
Proof. Using the axioms of certainty and non-negativity for the probability 11-+0
function P, part (i) follows uiviality from the defiqition of F (x). => oF ( a + 0') = F ( a )
For part (ii), we have for x < y, 5·3. Discrete Random Variable. If a randorit,variable takes at most a
F(y)-F(x)=P(x<X~'y)~ 0 (Property I)
countable nomber of values, it is called a discrete random variable. In other
'words, a real valued/unction defmed on a discrete sample space is called a discrete
~ F (y) ~ 'F(x) rando~ variable.
~ F (x) ~ F(y) when x < y ...(5·3) S:)' f. Probability Mass Function (and probability distributiqn 0/ a
Property 3. IfF is df. of one-dimensional r. v. X, then discrete random variable). . -
F!-oo)= lim F(x)= 0 Suppose X is a one-dimensional discrete random variable taking at most a
.1:-+-00
countably infinite number of values Xl> X2, '" With each possible outcome Xi ,
and F(oo)= lim F(x)= 1 , we aSsociate a number Pi = P ( X = Xi ) = p ( Xi ). called the probability of Xi. The
numbers p (Xi); i:; 1,2,.,.. must satisfy the following conditions:
Proof. Let us express the whole sample space S as a countable union of
disjoint events as follows:
(i) p ( xd ~ 0 Vi, (it) 1: p ( xd = 1
00 00 . i= 1
S= r u. (- n < X ~ - n + I ) ] u [ u ' '( n'< X ~ n + I )]
This function p'is called the probability mass function of tl)e random variable
n=) n=O
00 00 X and the set (Xi, p (Xi) ) is called the probability distribution (p.d.) of the r.v. X.
~ P(S)= L P(-n<X~ -n+I)+ L 'P(n<X~ ri+l) Remarks: 1. The set ~f values which X takes is called the spectrum of the
random variable.
n-= ) n=O
2. For discrete random- variable, a knowledge of the probability mass
( '.' P is additive) iunction enables us to compute probabilities of arbitrary events. In fact, if E is a
a set of real numbers, we have
I = lim L [F ( - n + I ) - F ( - n) ] P ( X E E) = 1: p' (x), where S is the sample space.
a-+oo n=1 xe EnS
b Illustration. Toss of coin, S = {H.T}. Let X be the random variable
+ lim L [F(n+I)-F(n)] liefined by
b-+oo n=O, X (" Ji) = I, i.e., X = I, if 'Head' occurs.
= lim [F(O)-F(-a)l+ lim· [F(b+l.)- F(O)] X ( T) = 0, i.e., X = 0, if 'Tail' occurs.
a-+oo b-+oo If the coin is 'fair' the probability fUl\ction is giv~n by
= LF ( 0 ) - F ( - 00 ) ] + .[ F ( 00 ) - F ( 0 ) ] P( {H} )=P( {T} )=1
1= F(oo)- F(-oo) and we can speak. of the probabilitY,distribution of the random variable X as
Since -00<00, F.( -00) ~ F (00). Also P(X= I)=P( {H} )=1 '
F ( - 00 ) ~'O and F ( 00 ) ~ 1 ( Property 2 ) P(X=O)=P( (T) ~=1 '
Ratndom Vllriablcs· Distributiun Funl:tiuns 5-7 - 5·8 Fundamentals of Mathematical Statistics

5-3-2. Discrete Distribution Function. In this case there Ufc a Solution.·

Table 1
countable number or points Xl. X2, Xlt • .j. <!fld numbers p. ~ O. E p. = I such
1 S.No. . Elementary Random Variables
that F (X ) = E Pi. For example if Xi is just the il1teger t, F (x) is a event
(i: x. $ x) y
X Z X+Y XY
"step function" having jump Pi at i, and being constant between caeii' pair of
1 HHH 3 1 3 4 3
integers.
F()() 2 HHT 2 1 2 3 2
3 HTH 2 0 0 2 0
4 HIT 1 0 0 1 0
-5 THH 2 1 2 3 2
6 THT 1 0 0 1 0
7 ITH 1 O· 0 1 0
8 ITT 0 0 0 0 0

Theorem5·5. p(Xj)= P(X= Xj)= F(x,)- F(.t.J-I), whereFisthed/. l{ere sample space is .
ofX. =
S {HHH, HHT, HTH, HIT, THII, TilT, ITH, ITT}
Proof. Let XI < X2 < ... We have (i) Ol)vio~ly ~ is ar.v. which-can take the values 0, 1, 2, and 3
F(xj)= P(XSXj) p (3) =P (HHH) == (1f2)3 '"' 1/8
j j
p(2)=P [fIHT uHTHu THH]
= L P (X = ;t;) =. L P ( Xi')
=p (HHT ) + 'p (ilTH) + P (THII) = 1/8 + 1/8 +1/8 =3/8
l= 1 i= \
J -\ Similarly p (1) =3/8 and p (0) = 1/8.
and F(Xj_ 1)= P(X~Xj_I)= ! P(Xi) These probabilities could alsO be obtained directly from the above table i.
i= 1
.. F(Xj)- "(Xj_ 1)= p(x,) ... (5·5) Table 2
Thus, given the distribution function of discrete random variable. we can
Probability table or X
compute its probability mass function. .
Example 5·1. An-experiment consists of three independent tosses of a fair Values of X
coin. Let 0 1 2 3
(x)
X = The number of heads
Y =The number of head runs, p(x) 1/8 3/8 3/8 1/8
Z =The lenght of head runs, -
a head run being defined as consecutive occurrence ofat least two heads, its length
then being the number of heads occurring together in three tosses of the coin.
Find the probabilityfunction of (i)X. (if) Y. (iii) Z, (iv) X+Y and (v) XY and
construct probability tables and draw their probability charts.
 PeY) 5·10 Fundamentals of Mathematical Statistics
Kamdom Variables ·I)istribution Functions 5·9

Table 3
'518 Example 5·2. A 'random variable X has the follqwing probability
4(8 distributiofl :
(ii) Probability Table or Y x: 0 1 2 3 4 5 ·6 7
Values of Y, 318 p (x) : 0 k 2k 2k 3k k 2 2k2 7k 2 + k
(y)
o I 218 (i) Find k, (ii) Evaluate P (X < 6), P (X ~ 6), and P ( 0 < X < 5), (iii) If

p(y) 5/8 3/8 '/8 P (X $ c) > t,find the minimum value of c, and (iv) Determine the distribution
O..--,.f-----'y function of X. [Madurai Univ. B.Sc., Oct. 1988]
7
P(z). Probability chart of Y
This is obvious from table 1. Solution. Since L p (x) = I, .we have
518 x=o
(iii) From table 1 , we have
~ k + 2k + 2k +3k + k2 + 2k2 + 7k2 + k = l
4/8
Table 4 ~ 10k2 + 9k - 1 =0
Probability Table or 3/8 ~ ( 10k - J) (k + I) = 0 ~ k = JlIO
ValuesofZ, 218 [.: k = -I, is rejected, since probabili~y canot be negative.]
0 1 2 3 (ii) P (X < 6) =P (X =0 ) + P (X = J) + ... + P (X =5)
(z) 1/8 If
1 2 2 3 I 81
p(z) 5/8 0 2/8 1/8 1 2 3 Z = 10 + 10 + 10 + 10 + 100 = '100
Pro.bability chart of Z 19
P (X ~ 6) = I - P (X < 6) = 100
p(u)
=X + Y. P (0 < X < 5) = P (X:: 1) + P (X =2) + P (X = 3) + P (X =4) ,,; 8e= 4/5
(ivl Let U From table I, we get 5/8

418
(iii) P (X $ c) > i. By trial, we get c = 4.
Table 5 (iv) X Fx (x) = P (X$x)
.318
Probability Table or U o 0
Values of U, 2/8
1 k = 111 ()
o 1 234 1(8 2 3k =3110
(u)
O~~,~~~--~--u~ 3 5k = 5/10
p(u) 1/83/8 1/8 2/8 lIS
4 8k =4/5
Probability chan of U =X +Y 5 8k + k 2 = 81/100
p(lI')
6 8k + 3k2 =831100
5/6 7 9k + IOk 2 = 1
(v) Let V=XY 4/8
EXERCISE 5 (b)
Table 6 3/8 1. «(I) A student is to match three historical events (Mahatma Gandhi's
Probability Table or V Birthday, India's freedom, and First World War) with three years·(I.947, 1914,
Values of V; 1896). If he guesses with no knowledge of 'the correct answers, what is the
(v)
o 1 2 3 probability distribution of the number of answers he gets corre~tly ?

p(v) 5/8 0 2/8 1/8 o 1 2 3 t· (b)' From a lot of ·10 items containing 3 defectives, a s{lmple of 4 items is
Probability chart of V =XY drawn at random. Let the random variable X de.note the number of defective items
in the sample. Answer the following when the sample is drawn without
replacement.
Ramdom Variables· Distribution Functions Fundamentals of Mathematical Statistics

(i) Find the probability distribution of X. 4. (a) A random variable X has the following probability function:
(ii) Find P (X !'> 1), P (X < 1) and P (0 < i < 2) Valuesof X.x :
p(x) .'
....2
0·]
-1
k
0
0·2
1
2k
2
0·3
3
k
Ans. (a) x 0 1 2". 3 (i) Find the value of k. and calculate mean and variance.
. (ii) Construct the c.d.f. F(X) and draw its graph.
p(x) 1 .! 0 .! Ans. (i) 0·1,0·8 and 2·-16, (ii) F (X) = 0·1,0·2,0·4,0·6,0·9, 1·0
3 2 6
(b) Given the probability function
(b) (i) _x_+-0_l----:2,.--3_ (ii) 2/3, 5/6. 1/2
1 1 3 1 ~ 0 1 2 3
p(x) "6"2 10 30 p(~) 0·1 I 0·3 0·5 0·1
2. (a) A random variable X can take all non· negative integral values, and the Let Y = X 2 + 2X , then find (i) the probability function of Y, (ii) mean and
probability that X takes the value r is' proportional to a r ( 0 < a < 1 ). Find variance of Y.
P (X = I0). [Calcutta Univ. B.Sc.1987]. Ans. (i) y 0 3 8 15
(ii) 6·4 ,16·24
Ans. P (X = r) = A a' ; r = 0, 1, 2, .... ; A = 1 - a ; P (X = 0) = A = 1 - a p(y) 0·1 0·3 0·5 0·1
(b) ~upposc that the mndom variable X has possible values 1,2,3, ... and
5. A random variable X has the following probability distribution:
P ( X = j) = 1/2 J , j = 1,2\... (i) Compute P ( X "is even), (ii) Cq,npute
Values of X, x 0 1 2 3 4 .5 6. 7 8
P (?C ~ 5) ,and (iii) Compute P (X is divisible by 3).
p{x) a 3~ 5a 7a 9a 11a 13a 15a 17a
ADS. (i) 1/3, (ii) 1/]6, and (iii) In
(i) Determine the value of a.
3. (a) Let X be a random variable such that
(ii)FindP(X< 3),P(X~ 3),P(O< X< 5).
P(X= -2):::: P(X= -1), P(X= 2)= P(X= 1) and (iii) What is the smallest value of x for which P (X ~ x) > 0·5? and
P(X> 0):::: P(X< 0)= P(X= 0). (iv) Find out the distribution function of X ?
Oblain the probability ma~s function of X and its distribution function. =
Ans. (i) a 1/81, (ii) 9/81, 12/81,24/81, (iii) 6
ADS. X -2 -1 0 1 2 (iv) x 0 1 2 3 4 5 6 7 l8
1 1 1 1 1 F(x) a 4a 9a 160 2Sa 360 49a 64a 81a
p(x) - - - -
6 6 3 6 6 6. (a) Let p (x) be the probability function of a discrete random
F(x) -1 2
- 4- -5 variable X which assumes the values XI , X2', x, ,X. , such that 2 p (XI) =3 p (x~
6 6 6 6 =p (x,) =5 p (x.). Find probability distribution and cumulative probability dis-
'(b) A.random variable X assumes the values -3, -2, -1,0, 1,~, 3 s\lch that
tribution of X. (Sardar Patel Univ. B.Sc. !987)
P(X= -3)= r(x= -2)= P(X= -1), X x,
XI ~2 X.
P(X= 1)= P(X= 2)= P(X= 3), Ans.
P (x) 1~6 1'¥l6 3Q16 416
and P ( 'J( = 0) = P ( X > 0) = P ( X < 0), (b) The following is the distributiQll function of a discrete random
Obtain the probability mass fUnction of X and 'its distribution function, and find variable X :
further the probability mass function of Y = 2X 2 + 3X + 4. x -~ -1 0 1 2 3 5 8
[Poona Univ. B:Sc., March 1991]
f(x) 0·10 0·30 045 0·5 0·75 0·90 0·95 1'()()
Ans. -3 -2 -1 o 1 2 3
1 1 1 1 1 1 1 (i) Find the probability distribution ofX.
p(x) - - - - (ii) Find P (X is even) and P ( 1 ~ X ~ 8).
9 9 9 3 9 9 9
Y 13 6 3 4 9 18 31 (iii) Find P ( X = - 3 I X < 0) and P ( X ~ 3 I X > 0).
1 1 1 1 1 1 1 [Ans. (ii) 0·30, 0·55, (iii) 1/3, 5/11].
pry) - - - -
9 9 9 3 9 9 9
Ramdom Variables· Distribution Functions 5·13 5·14 Fundamentals of Mathematical Statistics
x
7. If p(x)= 15; x= 1,2,3,4,5 function of x so that / (x) dt represents the probability that X falls in the in-
finitesimal interval (x, x + dt). Symbolically
= 0, elsewhere P (x:5 X :5 x + dt) = /x (x) dt ... (5·5)
Find(i)P{X=lor2), and (ii)P{t< X< ~I ~> I} In the figure,f ( x ) dt represents
the area bounded by the curve r ~\'"
~
[Allahabad Univ. B.sc., April 19921 Y = /(x), x-axis and the ordinates at
the points x and ~ + dt . The func-
Hint. (i) P {X = 1 or 2 l:: P ( X = 1) + P ( X = 2) = _1 + ~ = .! tion /x (x) so defined is known as
f . 15 15. 5
probability density/unction or simply
p{(..!.<X < ~~n X> f} density function 0/ random variable
(ii) P {-2
1< X< ~21 X> I} ~ l2 2)
P(X> I)
X and is usually abbreviated as 2
p.d/. The expression,f (x) dt , usually written as dF (x), is known as the prob:
P { (X =1 or ~) n X> I; P (X =2) ~15 _.! ability differential and the curve y = / ( x) is known as the probability density
= P (X> I) --l-P(X=l) 1-(Vls)-7 curve or simply probability curve.
8. The probability mass function of a random variable X is zero Definition. p.d.f./x (x) of the r.y. X is defined as:
except at the points i = O. 1,2. At these points it has the values p (0) :: 3c3 , ( ) _ I' P (x:5 X:5 x + 5 x)
jjxX-1m 0 ... (55
·a)
p(I)=4c-IOc1 al!dp(2)=5c-1 forsomec>O. ~ Sx--. 0 x
(i) Determine the value of c.
, The probability for a variate value to lie in the interval dt is /(x) dt and hence
(ii) Compute the follow!ng probabilities, P (X < 2) and P (1 < X S 2).
the probability for a variate value to fall in the finite interval [0. , ~] is:
(iii) Describe the distribution function and draw its graph.
(iv) Find the largest x such thatF (x) < ¥2. P(o.:5X:5~)= J~ /(x)dt ... (5:5 b)
(v) Find the smallest x such thatF (x) ~~. [Poona Univ. B.Sc., 1987)
Ans. (i)1. (i.!~J,~, (iv) I, (v) 1.
which represents the area between the curve y =/ (~), x-<\xis and the ordinates at
x = 0. and x =~. Further since total probability is unity, we have Jb / (x)dx = 1,
9. (a) Suppose that the random variable}( assumes three values 0,1 and 2 a
with probabilities t, ~ and ~ respectively. Obtain_the distribution function of where [a, b ] is the range of the r~dom variableX . The range of the variab.le may
be finite or infmite.
X.. [Gujarat Univ. B.Sc., 1992]
The probability density function (p.d/.) of a random variable (r. v. ) X
(b) Given that f (x) = k (112t is a probability distribution for a random usually denoted by/x (x) or simply by / (x) has the following obvious properties
variable which can take on the values x = 0, I, 2, 3,4, 5, 6, find k and find an
expression for the corresponding cumulative probabilities F (x). (i) /(x) ~ 0, - 00 < x< 00 ... (5·5 c)
[Nagpur Univ. B.sc., 1987)
5·4. Continuous Random Variable. A random variable X is said to be (tid 00
-00
f(x) dt = 1 ... (5·5 (/)
continuous if,it can take !ill possible values between certain limits. In other wor.ds. (iii) The probability P (E) given by
a random variable is said to be continuous when its different values cannot be put
P(E)= J/(x)dt ... (5·5 e)
in 1-1 correspondence with a set o!positive integers.
E
A continuous random variable;: is a random variable that (at least concep-
is well defined for any event E.
tually) can be measured to any desired degree of accuracy. Examples of continuous
random variables are age, height, weight etc. Important Remark. In case of discrete random yariable., the probability ata
point, i.e., P (x = c) is not zero for some fixed c. However, in case of continuous
5'4·1. Probabiltty Density Function (Concept and Definition). Consider the
random variables the probability at a point is always zero, Le., P (x = c) = 0 for
small interval (x, x + dx) of length dx round the pointx. Let! (x) be any continuous
all possible values of c. This follows directly from (5·5 b) by taking 0. = ~ = c.
 . S 16 Fundamentals of Mathematical Statilotics
Ramdom Variables· Distrillution Functions 5-lS

°
This also agrees with our discussion earlier that P ( E ) = does not imply that
the event E is null or impossible event. This property of continuous r.v., viz.,
and Jl,' - 3Jli' Jll' + 2~ll'3
J.13 =
~ =~' - 4113' Il( + 6!lz' IlI'z - 311114
}
... (5·7 d)

- and hence PI and pz can be computed.

P(X= c)= 0, V c ... (5·5!> M Median. Median is the point which divides the entire distribution in two
leads us to the following important result : equal parts. In case of continuous distribution, median is the point which divides
p (0. 5 X ~ ~) = P (0. ~ X < ~) = P (0. < X ~ ~) = P (0. < X < ~) ... (5·5 g) the total area into two equal parts. Thuc: if M is the median, then
i.e., in case of continuous r.v., it does matter whether we include the end points of
the interval from 0. to ~. I~f(X)dx= f!f(x)dx=t '" (5·8)
However, this result is in general not true for discrete random variables. Thus solving
542. Various Measures of Central Tendency, Dispersion, Skewness, and
Kurtosis for Continuous Probability Distribution. The formulae for these J'Ma f (x) dx =2.1 or
IMb f (x) dx =2. 1 ... (5·8 a)
measures in case of discrete frequency distribqtion can be easily extended to the
for M, we get the value of median.
case of continuous probability distribution by simply replacing Pi = f;IN by
(vi) Mean Deviation. Mean deviation about the mean Ill' is given by
f (x) dx, Xi by x and the summation over' i' by integration over the spedfied range
of the variable X. M.D. = fba I x- mean I f(x) dx ... (5,9)
Letfx (x) or f(x) be the p.d! of a random variable X where X is defined
(vii) Quartiles and Deciles. QI and Q3 are given by the equations
from a to b. Then
(i) Arithmetic mean = Jb x f(x) dx
fQIa f (x) dx =1.4 and fQ3a f (x) dx =2.4 ••. (5·10)
. a ... (5·6)
(ii) Harmonic mean. Harmonic mean H is given by D;, i th decile is given by

~ = (~ J: )f (x) dx ...(5·6 a)
JV;
a
f(x) dx= ...L
10
(viii) Mode. Mode is the value of x for whichf (x) is maximum. Mode is thus
... (5·10 a)

(iii) Geometric mean. Geometric mean G is given by

the solution of
log G =.J: 199 xf(x) dx ...(5·6 b) f'(x) =0 and f"(x) < 0 ... (5·11)
(iv) ~' (about origin) = Jb x f(x) dx ... (5·7)
provided it lies in [a,b].
a Example 5·3. The diameter of an electric cable; say X, is assumed to be a
continuous random variable with p.df. f ( x ) = 6x ( 1 - x), 0 ~ x ~ 1.
~' (aboutthe point! = A) =Jba (x - A)' f(x) dx
...(5·7 a) (i) Check that above is p.d/.,
and ~ (about mean) = Jb (x - mean)' f(x).dx (ii) Determine a nwnber b such that P (X < b) P (X> b) =
a ...(5·7 b) [Aligarh Univ. B.Sc. {Hons). 1990)
In particular, from (5·7), we have Solution. Obviously, for 0 ~ x ~ 1./( x ) ~ 0
Ill; (about origin) = Mean = Jba x f(xj dx·
I Now f6 f (x) dx = 6 f6 x V- x) dx
and 1l2' == Jb ;. if (x) dx
a = 6 f1o (x - .?) dx = 611x2 x311
2
_=I
Hence 112 = Ilz' - 1l1,2 = J: x2t.~) dx - (I: xf(x) dx J. . . (5,: c~ Hencef (x) is the p.d/. of r. v. X
. 3 0

From (5·7), on putting r=3 and 4 1 respectively, we get the values of (ii) P(X<b)=P(X>b.) ... (*)
Jl{ and \.4' a~d consequently the mo~ents about mean call be obtained by using
the relations :
Ramdom Variables· Distribution Functions 5-17
518 Fundamentals of Mathematical Statistics

~ I~ f(x) dK= I! f(x) dx (i) Determine the constant a.

(ij) Compute P (X ~ 1·5). [Sardar Patel Univ. B.Sc., Nov.19881
~ 6 f~ X (1 - x) dK =6 f! x (1 - x) dx Solution. (i) Constant 'a' is
probability-is unity, i.e.,.
detenniiled from the consideration that total

I~-~I~= I~-~I~ Loo~ f(x) dx= I

~ LOoof(x)dx+ f~f(X)dx+ ftf(x)dx+ gf(X)dx+ f3°O f(x)dx= I

~ [~ -~ )=[( ~-3 }-(~ -~)]
~ 3b1 - 2b' = [ 1 - 3bz + 2b' ]
~ f~axdx+ ft a.dx+ g (-ax+3a)dx= I

~ 4b' - 6 b1 + 1 = 0
(2b - 1)(2b1 - 2b - 1) =0
~ a ~~
1 1+ a I x I~ + a 1- ~ + 3x 1~ = I
~ 2b - 1 = 0 or 2b1 - 2b - .. =0 ~ ~ + a + a [( - .% + 9 )- (- 2 + 6) ] = I
Hence b = 112 is the only real value lying between 0 and 1 and satisfying (*). a' a . I
~ -+a+-=I ~ 2a=1 =) a=-
~:xample 5·4. A continuous random variable X has a p.d/. 22' 2
f(x)= 3.?, o~.x~ 1. Find a and b such that (ii) P(X~ 1·5) =f2':, f(x) dx= LOoo f(x) dx+ f6 f(x) dx + Jli5 f(x) dx
(iJ P { X ~ a } = p { X> a}, and
(ii) P { X> b } = 0·05 . [Calicut Univ. B.Sc., Sept. 1988] n
= a JO xdx+
fl.51 a.dx
Solution. (i) Since P ( X ~ a) = f ( X > a),
each must be equal to I/2,'because total probability is always one. 1;-1 1 + a I x 1 1.5 =!! + 0·5 a

.. P(X~a)=2
1
~
fa0 f(x)dx= 1
~
= a
° 2 1 2
- a-!
- - 2 [ '.' a = ~, Part (i) 1
ExalP pie 5·6. A probability curve y = f ( x ) has a range from 0 to If
=> 3Ja x1dx-!
o -2 ~ 31 x'la =!
3 O· 2 f(x) = e-·, find the mean and variance and the third moment about mean.
00 •

=>

(iiJ
, I
a='2
p (X > b) =0·05
~

::::::>
a= cr 2. '
f!f (i) dx =0-05
Solution.
[Andhra Univ. B.sc. 1988; Delhi Univ. B.Sc. Sept. 19871

Il, (rth moment ,about origin) =fooo x' f (x) dx

=> 3 -Ix'lI =T-

3 b 20
- ::::::>
.
, I
I-b = -
20 = f; x' e- x dx= r(r-t" 1)=r!.

b' _12
b=(~)L
~ ::::::>
(Using Gamma Integral)
-~O Substituting r = 1,2 and 3 successively, we get
Example 5·5. LeeX be a conlinuous,random variate withp,d/. MeaJI = Ill' = 1 ! = I, 112 = 2 ! =1, 113' = 3 ! = 6
f(x)= ax, O~ x~ 1 Hence variance = III = 112 - III ,1 =2 - r = 1
=a, I·~x~2 and 113 = 1l3' - 3112' ttl' + 2111'3 = 6 - 3 x 2 + 2 = 2
=- ax+ 3a, 2:5 x~ 3
= 0, elsewhere
Ramdom Variables· Oisttibution ,,'uncticMls 5·19 520 Fundamentals of Mathematical StatisticS

..:xample 5·7. In a continuous distribution whose relative/reque'!cy lknsity

is given by
:: %[Ib (lx-3x2+r)dx+
"t!o
n (3r-r-lx)dx]
/ (x) = Yo . x ( 2 - x ), 0 S; x S; 2,
find mean, variance, ~I ,and ~2 and hence show that the distribution is symmetri- :: '1[lx
4
U+ X41I + \3 x _l_ lx 12]:: ~
2. -
~ 40 '34 21 8
3 2

cal. Also (i)/ind mean deviation about mean and (U) show that/or this d(stribUlion
1l24.1 = 0, (iii)find the mode, harmonic mean and median. 112&+ 1 :: fa (x.- mean ):u.+1 f(x) dx
[Delhi Univ. B.Sc.(Stat. Hons.), 1992; B.Sc., Oct. 19921
Solution. Since total probability is unity, we have = ~ Po (x - I):u.+ 1X (2 - x) dx
Po /(x) dx= 1
= 14 II-1 t :u. +1 (t + 1) (1 - t) dt (x-I=t)
yofo x(2-x)dx= 1 ~ yo=3/4

3 :: 1 II t :u. + 1 (1 - t 2) dt
/(x)= 4"x(2- x) 4 -1
Since t:u. + 1 1 is an odd function of t and (1 - t -2) is an even function of t ,
, r2
'/()dx 31'2 "1(2)dx 3·2"1 the integrand t:u.+ 1 (1 - t 2) is an odd function of t .
IJ.r=JOx x =4"Jox -x = (r+2)(r+3)
In particular Hence 112..+1 = o.

, 3.22 Now f'(x):: 14 (2 - lx) = 0 => x=1

Mean = III = 3.4 = 1,

, 3·t 8 , 3·2s 16
113 = 5.6 = 5' and /l4 = 6.7 = 1"" Hence mode:: 1
,,2 6 1 1 Harmonic me~ H is given by
I"tence vanencc= 112 = 112 - III
J ·
= 5 - = 5'
1
-:: Po -1- f(x)dx
6 1. 2 0 1/ 0 x
113 = 113'-3112
"III2+' 3 8
III = 5 - 3 .5,' + =
:: ~ Po (2 - x) dx =~
, 4 ' , 6 ' ,2 3 ,4 16 4 8 1 6 6 1 3 1 3
/l4 == /l4 - 113 III + 112 III - III = 1"" - . 5' + . 5' - . == 35
H= ~
= -3~S
2
113 /l4 15 3
~I = -3 == 0 and ~2 == -2 - 2 =. -
~2 III (lis) 7 If M is the median, then
Since ~I = 0, the distributior is symmetrical.
Mean deviation about meafJ
IOM f(x)dx= i
~ rM x(2-x)dx=.!
= Po I x-I I /(x)dx 4 JO 2

= J~ I x-I I f (x) dx + ~ I x-I I / (x) dx IX2_; I~ = ~

3M2- M 3 = 2
= i [J(\ (1 - x) x (2 - x) dx + ~ (x - 1) x ('~ -~) dx ] M3_3M2+2= 0
(M-1)(M 2-2M-2)= 0
Ramdom Variables· Distribution Functions 5·21 522 Fundamentals of Mathematical Statistics

The only value of M lying in [ 0, 2 ] is M = 1. Hence median is 1. 1 9 4

= 4 (24 - 4.6.1 + 6.2.1 - 3 ) = 4 == 90'
Aliter. Since we have proved that distribution is symmetrical, b b
Mode = Median = Mean = 1
Hence ~l= ~l/~i= 40'6/0'6= 4 and ~l== ~~l= 90'4/0'4= 9
Example 5·8. The elementary probability law 01 a continuous random
variable X is Example 5·9. For the following probab.ility distribution
1 (x) = Yoe-b(x-a) , a~x<oo, b>O dF == Yo . e-Ixldx , - 00 < x < 00
where a, b and Yo are constants. show that Yo = ~, ~l' = 0, 0' = ..J2 and mean deviation about mean = 1.
Show that Yo = b = Va and a = m - cr, where m and (1 are respectively the
mean and standard deviation 01 the distribution. Show also that ~1 = 4 and Solution~ We have J..::, f(x) dx = 1
~2 = 9. [Gauhati. Univ. n.Sc.,.l99~ J
Solution. Since total probability is unity, => Yo J~oo e- Ixl dx= 1 => 2yo J; e- Ixl dx= 1,
Ja 1(x) dx = 1
00 ~ Yo Joo
a
e-b(x-a) dx= 1 (since e- I x I is an even function of x)
(sinceinOSx<oo,lxl=x)
Yo I
e-b(X-4) 100

-b Ia = 1
·
2Yo= 1, l.e.,
=> Yo= '21
~ ~=b
JA.: ( rth moment about the point 'x = a')
~.' (about origin) = J~ 00 x f(x) dx = ~ J~oo x e- I xl dx
= Ja00 (x - ar 1(x) dx = b Ja 00 (x - ar e-b(x-a) dx = 0,

=,b J; t' e-btdt [ On pUlling x - a = l]

( since the integrand x . e- I x I is an odd function of x )
~2 = J~oo Xl f(x) dx= ~ J~oo Xl e- I xl dx
_•b r (r + 1) _ r!
= !2 2 10 X- e- Ixl dx
- bH I - -b' [ Using Gamma Integral] 00

In particular
[since the integrand Xl e- I x I is an even function of x 1
IJ/ = lib, ~{= 2/b2 , ~/:±: 61b3 , J.4' = 241b4
~2 =.I.~ i" e- X
dx = r( 3 ) (on using Gamma Integral)
m= Mean =a+~{=a+(lIb)
=> ~2= 2! = 2
and ~= ~2= ~z' - ~1'2= 1/b2
Now ~= ~l= ~2 - ~l'l= 2
(1=.! and m= a+ .!!:: a+ (1
b b M.D.. about mean =J 00
-00
Ix- mean I f (x) dx
1
Hence (1 and a=m-cr
Yo=b=- = !2 Joo
-00
Ix I e- I xl dx (.: Mean = ~.' = 0)-
Also ~3= ~3'-3~2'~{+ 2~{3= ~(6-3.2+2)=~=2d
3
=~.2J; Ixl e-I"l dx
3 b b
and J4 = J.4' - 4 ~{ ~1 ' + 6 ~z' ~{ 2 - 3 ~{4 = JOoo x e-" dx = r( 2 ) = 1
Ramd~m Variables· Distribution Functions 5·23 Fundamentals of Mathematical Statistics

Example 5·10. A random variable X has the probability law:

1 2 a'- = ~,12 = ~1' - ~I' 1 = 2b1 - b2 • ~ = b2 ( 2 - ~)
dF(x)=
I
:1.e-X1lb dx, O:5x<oo

Find the distance between the quartiles and show that the ratio of this distance to
(J = b "2.- (1tI2)
Q, - QI _ "2 [ {iOg'4 - "log ( 4/3) 1
the standard devation ofX is independent of the patame.ter 'b'. Hence (J - V 2 - (1t/2) ,
Solution. If QI and Q3 are the first and third q~artiles respectively, we
'Yhich is independent of the parameter •b' .
have
Example 5·11. Prove that the geqmetric mean Gof the distribU!ion
dF;6(2-x) (x.". 1) dx, 1 ~x~ 2
isgiven by 6 log (16G) = 19. [Kanpl,lrUniv.B.sc:.,Oct.I9921
Put Solution. By definition, we have
10gG= F. logxf(x)dx= 6F. logx (2-x)(x-l)dx

= - 6 F. (.i - 3x + 2) log x dx
Integrating by parts, we get

Again we have OQ3 J f (x) dx = ~ which, on proceeding similarly, will give log G = - 6 [ H~ -3: ~ 2x ) log x I~
1-
2
e-(b Ilb
2
= 3/4 =>
2
e- lb I2b
2
= 114 _ 1'2
Jl
(Xl3 _3.i2 + 2x) .!x dx]
=> Q3 = -1 2b -1 log ( 4 )
The distance between the quartiles is given by =-4 log 2+ 6x 19 (on simplification)
36
Q3 - QI = -12b [-1 log 4 - -1 log (4/3) ]
19
Joo .. 19
logG+410g2="6 ~
~I'= 1
00 X '2 /lb2
I08G+log24 ="6
O· x f(x)dx= 0 x b 1 e- X
dx
19 19
10gG+log 16="6 IOg(16G)="6
= J; -12by'1l e-' dy
~

~ 6 log (16'G) = 19
~

Example 5·12. The time one has to wait for a bus at a downtown bus stop
is observed 10 be random phenomenon (X) with thefoJ/owing probability density
function:
fx(x)= 0, for x< 0
= ~(x+ 1). for O~ x< 1
= !(x-!)
, l'
for 1~ x< ~
Joo
-- 2b1 0 ye-'dy =!(1-x) for ~~ x< 2
'1 ·
=~(4-x). for 2~ x< 3
= 2b1 r( 2 ) = 2b1 • 1 I = ;2b1 _ I

- " for 3~ x< b

Ramdom Variables· Distribution Functions S·2S 5·26 Fundamentals or Mathematical Statistics

= O. for 6~ x. Example 5·13. The amount o[bread (inlwndredso[pounds)X that a certain

bakery is able to sell in a day is found to be a numerical valued random
Let the events A and B be defined as follows:
phenomenon, with a probability [unction specijJ.ed by the pr()bability density
A : One waits between 0 to 2 minutes inclusive:
[unctionf( x) ,given by .
B : One waits between 0 to 3 minutes inclusive.
f (x)::. A . x, for 0 ~ x < 5
(i) Draw the graph of probability densityfunction.
= A (lO-x) , for 5 ~ x< 10
(ii) Show that (a) P (B IA ~:= ~. (b) P C4" n li) = ~ = 0, otherwise
(a) Find the value of A such thatf(x) is a probability density function.
Solution. (i) The graph of p.d.f. is given below.
(b) What is the probability that the number ofpounds of bread that will be sold
f(X) tomorrow is
,.,9 (i) more than 500 pounds,
(ii) less than 500 pounds,
3r9 (iii) between 250 and 750 pounds? [Agra Univ. ~.Sc., 1989]
(c) Denoting by A, B, C the events that the pounds of bread sold are as in h
(i), b (iO and b (iii) respectively, find P (A I B), P (A Ie) . Are (i) A and B
219 independent events? (if) Are A and C independent events? ..
Solution. (a) In order thatf(x) should be a probability density function
L:, f(x) dx = I
o 1 2 3 6 X
i.e .• Jg Axdx+ J~O A(lO-x)ax=1
(ii)(a) p(A)=Fof(X)dx=J~ ~ (x+l)dx+ J~ ~(X-~)dx _ 1
=> A= - (On simplification)
25
• + J~ ~(%-x )dx (b) (i) The probability that the number of pounds of bread 'that will be sold
1 tomorrow is more than 500 pounds, i.e.,
- 2 ( on simplification)
J
IO· 1 1
P (5 ~ X ~ 10 ) == 5 25 ( 10 - x ) dx::. 25
IlOx - "2 \105
Xl

f(AnB)= P( I~X~2)=~ f(x)dx

= J~ ~(x-~)u+ ~ ~(~-x }x
i
= 2~ "( ;~ ) = = 0. 5

t t
(ii) The ·probability that the number of pounds of bread that will be sold
= ~ [ ~ - ~ n + .~ [ ~ x- ~ =
r. tomorrow is less than 500 pounds, i.e.,
\
( on simplification) . I5
P(0~X~5)= 0 25 .xdx= 25
1 I I"2 15
Xl
0=
1
2= 0·5

..
P ( B IA ) = P (A n B ) = 1/3 = ~
P(A) 1/2 3 (iii) The required probability is given by
(b) It n Ii means that waiting time is more than 3 minutes. J5 1
P(2·5SX~7·5)= 2.5 25 x dx+.J5
n·5 1 3
25 (lO-x)dx= 4"
:. P(Anli):P(X>3)=J3°o f(x) dx= J36 !(x) dx-t J; f(x)dx (c) The events A, Band C are given by
J6
= 3 1:9 dx=!
9
1
x
16 =!
3 3
A:S< X~ 10; B:O~ X< 5; C:2·5< X< 7·5
Random Variab:<!s· Distribution Fundions 527 5-28 Fundamentals of Mathematical Statistics

Then from pans b (i). (ii) and (iii), we have

P (A) = 0·5, P (B) ~ 0·5, P (C) =

3
4"
(ii) P (16~X ~ 24)= 2~ g: exp ( - ;0 )dt= 1- e- zI2O I~:
= e- I6/20.- e- 24/20 = e- 415 _ e- 6/5
The events A n B and A n C are given by = ()'4493 - 0·3012 = 0·1481
An B = ~ and An C : 5 < X < 7·5
00 1 [ e- llI20 [00
.. P(AnB)= P(~)= 0 (iii) P (X ~ 30) = J 1(X) dt= 20 - 1120 30
30
and P(AnC)=
n·5
J5 f(x)dx=
1 n·5
25 J5 (lO-x)dx = e- 15 = 0.2231

1 75 3
= 25 Xg="8 EXERCISE 5 (c)
1 3 3 1. (a) A continuous random variable ~ follows the probability law
P(A).P(C)= ~ )(4"= 8= P(AnC)
I (x) = Ax2 , O~x~ 1
=> A and C are independenL Determine A and find the probability that (i) X lies between 0·2 and 0·5,
Again P(A).P(B)= t~ P(AnB)
fii)Xis less than 0·3, (iii) 1/4 <X < l/2an<1 flV)X >3/4 given X >1/2.
Ans. A = 0·3, (i) 0·117, (ii) 0·027, (iii) 15/256 and (iv) 27/56.
=> A and B are not independent. (b) If a random variable X has the density fUl\ction

P (A I B) = P (A n B) = 0 I (x) = {114, - 2 < x < 2}

P(B) 0, elsewhere. J

Obtain (i) P (X < 1), (ii) P (I XI> 1) (iii) P (2X + 3 > 5)

P (A I C) = P (A () C) = 3/8 = .! (Kerala Univ. B.Se., Sept.1992)
P (C) 3/4 2
Example 5·14. The mileage C in thousands of miles which car owners get
with a certain Idnd of tyre is a random variable having probability density function
Hint. (ii) P (IX I> 1)= P(X > 1 or X <-1)= J~ i f(x) dt+ ~/(x) dt
or pd xl> 1) = 1- P (I X 1~ 1) = 1 - P (- 1 ~X ~ 1)
f(x) == 2~ e-Jl/7JJ, for x> 0 Ans. (i) 3/4, (ii) 1/2 (iii) 1/4.
2. Are any of the following probability mass or density functions?
= 0, for xSO Prove your answer in each case.
Find the probabilities that one of these tyres will-last 1 3 1 1
(i) at most 10,000 miles, (a) {(x) = x; x= 16' 16' 4'-;
(ii) anywhere from J6,OOO to 24,000 miles.
(b) I (x) = A. e-k< ; x ~ 0; A. > 0
(iii) at least 30,000 miles. (Bombay Univ. B.sc. 1989)
2.x, 0< x< I
Solution. Let r.v. X denote the mileage (in '000 miles) with a certain kind
of tyre. Then required probability is given by:
(c) 1
f(x)= 4- lx, 1< x< 2
0, elsewhere,
(i) P·(XSI0)= JbO f(x)dx= 2~ JbO e-Jl/7JJ dx
(Calicut Univ. B. Sc., Oct. 19_
Ans. (a) and (b) 3I'ti .,.m.f./p.d.f.'s, (c) is not.
_ 1 1 10 _
I e-Jl/7JJ -11'2
3. IfII and /z are p.d.f.' s and 91 + -9z = I, check I( .
- ~O - 1/20 0 - 1 - e g (x) = 91/1 (x) + 9z/z (x) , is a p.d.f.
= 1- 0·6065 = 0·3935 Ans. g (x) is a p.d.f. if 0 ~ (Eh, Eh) ~ 1 a-' fll + Eh= 1.
Random Variables· Distribution Fundions 5·29 530 Fundamentals of Math(.'matical Statistics

4. A continuous random variable X has the probability density function: (c) Find the mean, mode and median for the distribution,
/ (x) = A + Bx, 0 ~ x ~ 1. dF (:c) = sin x dx, 0 ~ x ~ Jt/2
If the mean of the distribution is 1, find A and lJ • ADs. I, Jt/2, Jt/3

Hint: Solve I~/(x)dx= 1 and I~ x/(x)dx = 4. Find A andB. 9. If the function f(x) is defined by
f(x)= ce-'u, O~x<oo, a>O
5. For the following density function (i) Find the value of constant c.
/(x)= cr(1-x), O<x< 1, (ii) Evaluate the first four moments about mean.
fmd (i) the constant c, and (ii) mean. [Gauhati Univ. B.Se. 19871
[CalieufUniv. B.Se.(subs.), 1991J ADs. (i) c =a , (ii) 0, I/cil2/a' , 'J/a4 •
,
Ans. (i) C = 12; (ii) mean =3/5 .
10. (a) Show that for the exponential distribution
6. A continuous distribution of a variable X in the range (-3, 3) is defmed by
dP= y".e-";o dx, O~x<oo, (J>O
/ (x) = i6 (3 + X)l, - 3 ~ x ~ - 1
the mean and S.D. are both equal to (J and that the interquartile range is
=.!.(6-2xl)
16 • , -1<x<1
-- (J loge 3. Also find 1J,r' and show that ~I = 4, ~2 = 9.
= i6 (3 - x)z, 1~x ~ 3 [Agra ;):liv. B.Se., 1986 ; Madras Univ. n.Se., 1987)
(b) Define the harmonic mean (H.M.) of variable X as the reciprocal of the
(i) Verify that the area under the curve is unity. expected value of IIX, show that the H.M. of variable which ranges from 0 10
Find the mean and variance of the above distribution.
(ii)
(Madras Univ. B.Se., Oct. 1992; Gujarat Univ. B.Se., Oct. 1986) 00 WIth probability density i xl e-' is 3.

Hint: I!3/(X)dx= I~i /(x)dx+ J!l/(X)dx+ Ii /(x)dx 11. (a) Find the mean, variance ~d the co-efficients ~I , ~2 of the distribu-
tion,
Ans. Mean=O, Variance=1
dF = k x'J. e- a dx, 0 < X < 00 •
7. If the random variableX has tJ'te p.dJ.,
i
/ (x) = (x + 1). - 1 < x < 1
Ans. ,,= 112; 3,3, 4/3 and 5.
(b) Calculate PI for the distribution,
= 0,' elsewhere, dF:= k x e-adx., O<x<oo
find the coefficient of skewness and kurtosis. Ans. 2 [Delhi Univ. B.Se. (Hons. Subs.), 1988J
8. (a) A random variableX has the probability density function given by 12. A continuous random variable X has a p.d.f. given' by
/ (x) = (it (1 - x), 0 ~ x ~ 1
f (x) = k x e ).a , x ~ 0, A. > 0
Find the mean I.l , mode and S.D. cr , Compute P (IJ.- 2cr < X < I.l + 2cr).
-= 0, otherwise
Find also the mean deviation about the median.
~LCrmine the constant k ,obtain the mean dlld variance of X .
(Lueknow Univ. B.Se., 1988) (Nagpur UDiv. n.Se. 19901
(b) For the continuous distribution 0. For the probability density function,
dF = Y~ (x - r)
dx ; 0 ~ x ~ I, Yo being a constant. f ( )= ,?(b+ x) - b~ x< 0
Find (i) arithmetic mean, (ii) harmonic mean, (iii) Median, (iv) triode and (v) rth x b(a+b)'
moment about mean. Hence find ~1 and ~2 and show>,that the distribution is = 2(a-x) O~xSa
symmetrical. (Delhi Univ. B.se., 1992'; Karnatak Uni v. B.Se.,'1991) a(a+b)'
Find mean, median and variance. [Calcutta Unh'. B.Se, 1984)
Ans. Mean =Median =Mode =1Z
• >
Ans. Mean =( a - b )/3, Variance = (a + b2 of ab )/]8,
2

Median = a - " a(a+ b) /2

Random Variables· Distribution Functions 5-31 5·32 Fundamentals of Math~'I11atical Statistics

(U) Show that, if terms of order ( ci - b )2/ a2 arc neglected, then Reqd. Prob. =P [ Each of a sample of 5 shots falls within a distance of 2 ft.
mean - median =(mean - mode) /4 from the centre]
14. A variable X can assume values only between 0 and 5 and the equation
of its frequency curve is
y = A sin ~ 1t x, 0 $; x$;5
= [P(O<X <2ll' = [
19. A random variable X has the p.d.f. :
Ig f(;<)dx r =(: =::: J
where A is a constant such that the area under the curve is unity. Detennine the 2x,O<X<1
value of A and obtain the median and quartiles of the distribution. f ( x ) = { 0, otherwise
Show also that the variance of the distribution is 50 { k- ~z }. Find (i) P ( X < ~ ), (ii) P ( i< X < ~ ) " (iii) P ( X > ~ I X> ~ ), and
~ IX > ~ ).
Ans. 1/10\ 2·5, 4{3, 10{3
1S. A continuous variable X is distributed over the interval [0, I] with p.d·L (iv) P ( X < (Gorakhpur Univ. B.Sc., 1988)
a XZ + b x, where a, b are constants. If the arithmetic mean of X is 0·5, find the l. (. )PO~< X< +'4)
A ( ')1/'" (")3/16 (l")P(X> 3/4) 71\6_
values of a and b. ns. I "t, II , I I P (X> In) 3/4 - 12' IV P (X> Iii)
ADS. -6,6
5·4·3. Continuous Distribution Function. If X is a continuous random
16. A man leaves his house at the same time every morning and the time
variable with the p.d.f. f (x), then the function
taken to journey to work has the following probability density function: less than
30 minutes, zero, between 30 minutes and 60 minutes, unifonn with density k ; Fx(x) = P(X~x)= toof(t)dt, -oo<x<oo. ...(5·12)
between 60 minutes and 70 minutes, uniform with density 2k ; and more than 70
minutes, zero. What is the probability that on one particular day he arrives at. work is' called the distribution function (dJ.) or sometimes the cumulative distribution
later than on the erevious day but not more than 5 minutes later. function (c.d.f.) of the random variable X.
17. The density function of. sheer strength of spot welds is given by Remarks 1. 0 ~ F {X ) ~ 1, - 00 < x < 00. Il

f(x) = A /160,000 for 0 $; X$; 400 2. From analysis ·(Riemann integral), we know mat
= (800 - x) /160,000 for 400 $; X$; 800
Find the number a such that F' (x) = ! F (x) = f(x)? 0 [.: f(x) is p.d.f.]
Prob. (X < a) =0·56 and the number b such that ==> F (x) is non-decreasing function of x •
Prob. ( X < b) =0·90. Find the mean, median and variance or X..
[Delhi Univ. B.E., 1987] 3. F.(-oo): lim F(x)= lim
.l-+-oo x __ oo
rx
-00
f(x)dx= I-co
-00
f(x)dx= 0
18. A baLCh of small calibre ammunition is accepted as satisfactory if none
of a sample of five shot falls more than 2 feet from the centre of the target at a and F(+oo)= lim FJ-x.) = lim toof(x)dx= I':--oof(x)dx::: 1
given range. If X. the distance from the centre of the target to a given impact .l-+'!" x-+oo

point, actually has the density 4. F ( x) is a continuous function of x on the right.

Z
f(x)= k.2xe-:r , 0< x< 3 5. The discontinuities of F ( x) are at the most countable.
where" is n number which J118kes i~ probability density function, what is the value 6. It may be noted that
of k and what is the probability that the baLCh will be accepted? ..
[Nagpur Univ. B.E., 1987J P(aS X~ b)= I: f(x)dx=- I!oc;f(x)dx- I~'oof(X)dx

Hint. Ii f(x) fLY::' 1 ~ k = 1I( 1- e- 9 )

Similarly
= P ( X ~ b) - P ( X ~ a) = F (b) - F (a)

P ( a < X < b) = P (a < X ~ b) = P «(l ~ X < b) == Iah f( I) d/

Ran$lom Vl!riablcs· Distribution Functions 5·33 S 34 Fundamentals of Yiath(.'matica'i Statistics

7. Since F'(x)=/(x), wQhave Exam pie 5,15. \lerify that the following is a distribution functiof.l:
d
- F (x)= / (x )
dx
~

This is known a5 probability diffe{ential of X.

dF( x ) = / ( x ) dx
F(x)=
{~;+
(
(),

J
I , -
x< - a
a~ x~ a
1" ,x> a
Remarks. l. It may be pointed oUt that the properties (2), Q) and (4) above
(Madras Univ. B.Sc., 1992)
uniquely characterise the distribution functions. This means that any function F(x)
satisfying (2) to (4) is thP, distribution function of some random variable, ari(l any Solution. ObVIOusly the properties (i), (ii), (iii) and (iv)are~atisfied.Also
function F(x) violating anyone or more of these three properties cannot be the we o~~rve that f+:t) is continuous at x = a and x = - a, as well.
distribution function of any random variable. Now
2. Often, one can -obtain a p.d.f. from a distribution function F (x) by
differentiating F (x) , provided the derivative exists. For example, consider
d' F( ),
dx"x:::;Jtl 1'1
-2' - a_« x_ a
0, for x <: 0 0, otherwise
F x( x ) = \ x, for 0 ~ x ~ = f( x), say
1, for x> 1
In order that F (x) is a distribution function, f (x) must be a p.d.f. Thus we
The graph of F (x) is given by bold liJles. Obviously we see that F (x) is
/. have to sh()w tl}at
continuous from right a<; stipulated in (4) and we also see that F (x) is.. not
continuous at x = 0 and x = 1 and hence is no(deriv~ble afX = 0 .and x = 1. Loooo f(x) dx = I

F(x)
Now f:'oo f(x) dx = f~.a f(x) dx = 2~ f~ a I. dx = I

Hence F ( x) is a d.f.
--------~---~---- Example 5·16. SW'pose'the life in hnuq of a certain kind of radio tube has
the probability density function:
100
f (x )= -2 ' when x ~ 100
x
1 )( = 0, when x < 100
Find the distrib~tionfunction of the distribution, What is the probabilty that none
(~f three such tubes in a given radio set will have to be replaced during thejirst150
hour;~ of operation? What is the probability thai all three of the original tubes will

Dirferenitating F'(x) w.r.t. x, we get have

I
been replaced during thefirst 150 hours? (Delhi Univ. B.Sc.• Oct. 1988)
Solution. Probability that a tube will last for first 150 hours is given by
!! F ( x )= { 1 , 0 < x.< 1. p ( X'~ 150) = p ( (j < X < 100) + P ( I00 ~ X ~ 150)
dx 0, otherwise
INote the strict.inequality in 0 ~ x < 1, since F (x) is not derivable at
X=' 0 andx= 11
= f!~ f(x) dx = f!~ l~. dt = ~
Let us define Hence the probabmiy that none of the three tubes will have to be replaced
l,O<X<l .dllring the first 150 hours is (1/3)3 = 1127 .
/ (x) = { 0, othef)t(ise
The probability lhal a tube will not last for the first 150 hours is 1 - i = ~.
Then / ( x) is a p.dJ. for F.
Random Variables· DistribuUon.Fundilms 5·35 536 Fundamentals or Mathematical Statistics
Hence the probability that 'allthrcc of the original tubes will have to be replaced (iii) Required'Probability = P ( I < X < 3) = P ( 1 < X ~ 3) ,
during the Iirst 150 hours is (2/3)3 = 8/27 . = F(3j- t(1)= ~- ~= ~
Exam pl~ 5·17. Suppose that the time in minutes that a person has to wait at
(c) l-etA den9te ~he eve"t that.a person has to wait for m9re than mInutes 3
a certain station for a train is found tq be a random phenomenon, a probability
and B the event that he has to wait for more than 1 minute. Then
function specified by the distributionfUilction,
F ( x ) = 0, for x ~ 0 P(A)= P(X> 3)'= ~ [c/.(b),(i)j
.t
= 2' for 0 ~ x < 1 P(B)·= P(X> 1)= I-P(X~ 1)= I-F(1)= 1- ~= ~
I
= '2' for I ~ x < 2
P(A n B) =P (X > 3 n X > 1) = P ( X> 3) = ~
= ~, for 2 ~ x < 4
OJ Required probability is
= 1, for x~ 4
(a) Is the Distribution Function continuous? If ,ro, give the formula for its P (A I B) = p.( ~ 0 B') 114 1
P(B) 112 =2
probability density/unction?
{IJ) Whgt is the probability that ,a person will.ha ...e to wait (i) more than 3 .(;;) R~iJirea prob~bil,ity = P (A I B)::; P ~~ ~: ).
miTlUtes, (uj less than 3 minutes, and (iii) between 1 and 3 minutes?
(c) W~at is .the conditional probability that the person will have to. waitfor·a No,,<p(~,nBi)=P(X~ 3(l).X>, 1)=f(,l',:: X~ 3)=F(3)'- F(1).= ~- ~= ~
trainfor OJ more than 3 minutes, given that it is more than} minute, (ii) less than
- i/4 1
3 miTlUtes given that it is more than 1 minute? (CalicutIUniv. B.Se., 1985) . :'. . .p ~ A r:B t= 112 =:'2
SoMion. (a) Since the value of the distribution fUllctioQ is t~e same ~tthe ·Example-5.'18. A petrol pump is supplied witli pet;ol once a day. /fits
points x = 0, x = I, x = 2" and x = 4· given by the Jwo fonns, o(:l ( x) ;lor daily volume 'X 'oj sales in ihousands 0/ Ii/res is distributed by
x < 0 and 0 ~ x < 1, 0 ~ x < 1 and 116 x < 2, 1 ~ x < 2 and 2~' x < 4,
/(x)=5(f-xt, O~x~,I,
2 ~ x < 4 and x ~ 4, the distribution function is continuous.
what must be the capacity..o/ its tank. in order: that the probability that its supply
Prob~bility density function =f ( ~ ) = -1; F(x ~ will be exhausted in a given day shall be c)-OJ? (Madras Univ. B.E., 1986)
Solution. Let the capacity of the tank ( in '000 of litres).be 'a' such that
.. f( x) = 0, for x < 0
=
1
'2' for 0 ~ x < L P(X?. a)= 0·01 ~ J1a i(x)dx= 0·01
= 0, for 1 ~ x < 2
~ II 5'(, l-x)4 dx= 0.01 or r~ 5 . (l-x )51 1 ~ 0·01
= ~, for 2 ~ x < 4 a (- 5) 1
= 0, for x~ 4 ~ (l-a)~=1/100 or l-a'=(1Il00)115
1/5' .'
(b) Let the random variable X represcnllhe waiting time in minutes. .. 1 - .( 1/1(0) = 1 - 0:3981 = 0·ro19
d=
Then x
Hence the capacity of the tank = 0·60 19 1000 litres = 60 1·9 litres.
(i)... R~!lired prQbability = P ( x. > a.) = .1 ~ P. ( X ~ 3) =. 1 - F (~) Example 5·19; Prove,that mean deviaiion is least when measured/rom the
=. 1- .~. 3 = ~ median. . [Delhi ·Univ. B.sc. (Maths. Hons.)~ 1989]
(ii) Required'probability =' P (X < 3) = P (X ~ tg ) ... P (x·= 3') :Solution. If / ( x) is the: probability function of a random variable X,
, = F (3) = l ' . a ~ X ~ b, -the,n mean deviation'M (A ), ~y, about the point x = A is given
4 by '-
(Since, the probabihty that a continuous variable takes a.fixed value b
7.ero) I M ( A )= . Iab I x - A I / (x) dx
 Random Variables· Distribution Functions 5·37 5·38 Vundamentals.uf Math<''l1laiical Statistics

J
= aA Ix-Alf(x)dx+
' Jb
A Ix-Alf(x)dx
EXERCISE'S (d)
1. (a) Explain the terms (i) probability differential, (U) probability'density
= J
aA (A-x)f(x)dx+ f b'
A(x-A)f(x)ax .. , (1) function, and (iii) distribution function.
(b) Explain what is meant by a ra~do~ ~ariable. Distinguish between a
We; want to find the value of 'A' so that M (- A -) is minimum: From the djscrete and a COQtinllOUS random variaole.. Define distribution function of a
principle of maximum and minimum rn differential calculus, M (A) will be Plndom .variable and show that it is ·monotonic non-dccrcasing everywbere and
minimum for variations inA if continuous on the rigbt at every point.
aM(A) 0 d a2 M(A) 0 [Madras Uniy. B.Sc. (Stat Main), 1987)
aA = an aA2 > ... (2)
(c) Show that the distribution function F (x) of a random variable X is a
Differentiating (I) w.r.t 'A' under the integrarsign, since the runctions ~on:.decreasing fUllction of x. Determine the jump of F (x) at a point Xo of its

(A - x) f ( x) and ( x - A ) f ( x) vanish atthe pOint x = . A *, we get discontinuity in terms of the probability that the r~dom variable has the ~alue Xo ..
'[Calcutta Univ. B.Sc. (Hons.), 1984)
a~~,A) JA f(x)dx-,J b f(x)dx ... (3) 2. The length (in hours) X of a certain type of light bulb may be supposed to

J: J:
aA a A
be a continuous random ·..ariable with probability density function:
Also a~ ~A ) = f ( x ) dx ~ [ .1- - f ( x ) dx '] ,
°
[.: J: f(x)dx= I]
f(x)=
x
3 ,15.00< x< 2500'

= 0, elsewhere.
Determine the constant a, the distribution function of X, and compute the
= 2 JA
a f(x)dx-
. }.= 2F(A) ..,.. I;
probability of the event 1,700 ~ X ~ 1-.900.
where F(.) is the distribution fURction Qf X. Differentia'ii~ a~iliQ w.r.t.A, we,get t
Ans. a = 70,31,250: F (x) = 1 (22,5~,oOO - :2) and
a~2 M(A)= ~f(A) ... (4)
P( 1,700 <X < 1,900)= F(I,900)- F(I,700)= 1(28,~,OOO- 36,1~,()()())
Now a~ ~A ) 0, on using (3) gives ~
J: f(x)dx=' J1 f(x)dx
3. Define the "distribution futlction" (or cumulative distrtbution function) of
a random variable and state its essential pro~rties.
i.e., A is the median value. Show that, whatever the distribution function F (x) of a random variable
Also rrom (4), we see that X , P [ a ~ F (x) ~ b] = b - a, 0 ~ a, b ~ 1.
L
4. (a) The distribution function of a random variable X is given by
~~ ~ (~) > 0,
aA F ( x )= { 1- ( 1+ x) e - oX , for x ~ 0
assuming thatf( x) does not vanish at the median val~e. Th~s mea~ deviation is 0, for x< 0
least when taken from median. , Find'the corresponding density function of random variabJeX.
ib) Consider the distribution for X defined by
*If f ( x • a) is a continuous function of both variables x and a, possessing
for x< 0
Continuous partial derivatives a:2£9' a;2fx and a and b are differentiable 1 0,
F(x)= 1- 14" e -x 1"lor x_ >0

functions of a, then Deoonnine P ( x = 0) and P ( x > 0).

ada [J;'f(x,a)dx ]= J: ~ dx +f(b,a) :-f(a,a)::
(Allahabad Univ. B.Sc., 1992]
 I
j{andmn Var!,,',\cs. I)islrihution Functions 5 3~ 540 Fundamentals or Matht.'Illatical Statistics

5. (a) Let X be a continuous mndom variable with probability density Using F ( x ) , show that
fllnctioll given by (iii) P(3 < X < 5) = 0-1809, (iv) P( X < 4) = 0-5507, (v) P( X > 6) = 0·3012
ax, 0 ~ x.~ 1 9. A bombing plane carrying three bombs flies directly above a railroad lIack_
f( )
x = l a) 1 ~ x ~ 2
_ axt 3(l,~)~ x~ 3
If a bomb falls within 40 feet of track, the track will be sufficiently damaged to
disrupHhe traffic.Within a certain bomb site the points of impact of a bomb have
0, elsewhere the-probability-density function:
(i) Detennine the constant a. f(x)= (100+ x)/IO,OOO, when -100:s; x:s; 0
(ii) Det.cnnine F ( x ), and sketch its ,graph. = ( 100 - x) 110,000, when O:s; x:S; 100
(iii) If three independent observations are made, what is the probability that = 0, elsewhere
exactly one of these three numbers is larger than 1·5 ? where x represents the vertical deviation (in feet) from the aiming point, \Yhich is
[Rajasthal1 Univ. M.Sc., J987\ the traCk in this case_ 'find the distribution function_ If all the bombs are used, what
Ans. (i) 1/2, (iii) 3/8. is the probabjlity that track will be damaged ?
(b) For the density fx (x) = k e- ax ( 1 - e· ax) 10._ (x), find the nonnalising Hint. ProbabiJity that track will be damaged by the bomb is given by
constant k , fx (x) and evaluate P ( X > 1). ~ (I. X I < 40) = P (- 40 < X < 40)
[Delhi Univ. B.Sc. (Maths Hans.), 1989]
Ans. k= ~.;,F(x)= 1_2e-ax+e-2ai~P(X>1)= 2e- a _e- 2a = J~40 f(x) dx+ J~O f(x) dx
6. A random variable X has the density function:
f( x)= K._I_ if -oo<;:x<oo
Jo lOO+x r40 lOO-x
=, -40 10,000 dx+ JO
16
10,000 dx= 25
.. 1+x2!
= 0, otherwise . . Probabi\ity that a bomb will not damage the track = I - ~~ = ~
Det.cnnine K and the distribution function. Probability that no"e of the three bombs damages the track
Evaluate the probability P ( X 2: 0). Find also the mean and variance of X .
[Karnatak Univ. B..Sc. 1985] == (. i;J= 0-046656
Ans. K = 1 , F (x) = -i {tan- 1x + ~.}, P (x 2: 0) = 1/2" Mean =0, Required'probability that the track will be damaged = I - 0-046656 =0-953344_
Variapce does IJot exist. ... 10. The length of time (in minutes) that a certain lady speaks on the telephone
7. A- continuous random variable X has the distribution function is found to be random phenomenon,'with a probability' function specified by the
0, if x ~ 1 probability denSl~y functionf( x), as
F ( x ) = [ k ( x-I )4, if 1 < x ~ 3 f.(x)=Ae-X/5, forx~O
1, if x> 3 = 0, otherwise
Find (i) k, Oi) the probability density functionf ( x ) , and (iii) the mean and (a) Find the value of A that makes f (x) a p.d.f.
the median of X. Ans. A =1/5
(b) What is the probability that t~e number of minutes that she will talk over
Ans. (i)k= 1~' (ii)f(x)= -1- (x-:' 1')', 1 s: xS: 3 the phone is

8 G·
• lVen
f() -
x.-:-
{k0 x• ( 1- x), for 0 < x < 1
elsewhere
(i) More than 10 minutes, (ii) less tHan 5 minutes, and (iii) between.5 and 10
minutes? [Shlvaji Uitiv. BSC., 1990J
Show that Aos. ('-J 21 , (--J e - I (---J e - I
U - - , III - 2 - '
e e e
(i) k = 1/5, (ii) F (x) = 0 for x ~ 0 and F ex) = 1 - e- ZlS , for x> 0