А-22 _ | THEORETICAL DISTRIBUTIONS

garments must know thessizes of collars for which he expects maxim um

demand so that he has nó stock of unwanted sizes. Ina similar way, the
teachers in theschool, college or university should: know what they expect
from the students. It is only then that they would be in a position to
comment on good or bad performance.
Amongst theoretical or expected frequency distributions, the follow-
ing three are most popular : t
' 1. Binomial Distribution,
2. Poisson Distribution, and
3. Normal Distribution.
These. distributions from the point of view of historical interest as
well as their intrinsic importance occupy a position in the forefront of
statistical theory :
1. BINOMIAL DISTRIBUTION
The Binomial Distribution also known as ‘Bernoulli Distribution’ is
associated with the name of a Swiss Mathematician James Bernoulli (1654-
1705). Binomial distribution is a probability distribution expressing the
probability of one set of dichotomous alternatives, i.e., success or failure. -
К More precisely, the binomial distribution refers to a sequence of
events which possess the following properties :
- Т. A simple experiment is repeated a number of times where, the
outcomes are independent, ie., what happens on the first trial does not
affect the second, and so on.
s . 2. Outcomes of each trial can be classified into two —mutually ex-
clusive categories, arbitrarily called “successes” and "failures".
3. The probability of success in a single trial, denoted by p, remains
the same for all trials. If the possibility of success is not the same in
each trial we will not have a binomial distribution. For example, 5 balls
are drawn at random from an urn containing 10 white and 20 red balls.
This is a binomial experiment if each ball is replaced before another is
selected. If the balls are drawn without replacement, the probability of
drawing white ball changes each time a ball is taken from the urn and
we no longer have a binomial experiment. p t
4. In a given trial the focus is on whether or not the successful
outcomes have occurred.
5. The experiment is performed under the same conditions for a
fixed number of trials, say, п. ;
. This model is useful to answer questions such as this: If we соп-
duct.an experiment under the stated conditions п times, what is the pro-
bability of obtaining exactly п successes ? More specifically suppose 10
dice are tossed together, or one die is tossed 10 times, what is the proba-
bility of obtaining exactly two aces ?
6. How binomial distribution arises can be seen from the follow-
ing: 2
If a coin is tossed once there аге two outcomes, namely, tail ог head.
The probability of obtaining a head or p=} and the probability of obtain-
ing a tail or q=}. Thus (q+p)=1. These are terms of the binom ial
(4-ЕР). в 7
 THEORETICAL DISTRIBUTIONS s A-23
Similarly, if iwo]toins are] tossed. simultaneously there are four
possible outcomes : `
“A B

Т T
T H
H T
, H H
The probabilities corresponding to these results are :
> ЕҢ DENT. HH ге
u as 4
4 2р M
g? 24р р?
These are the terms of the binomial (44-р) because
(4+РУ#=4°--24р-Ер*
‚ In a special case where p=q=4, we have
@+@#=4+4+. "
Similarly, if three coins 4, В and C are tossed the following are the
8 possible outcomes and the probabilities corresponding to these results
afe:
і ABC АВС АВС АВС АВС АВС АВС АВС
"TIT ТТН THT HIT THH НТН HHT ННН
Aw a Nu ay EA, ee: ч.
т h ; Г ^x Y
ар Gp qp qp Kr y= GD аа DS
These are the terms of the binomial (q--p)?
(4-Ер)3=48--34°р--3ар°-Ер8 " "
where p=q=}, we have
[-HP-EHEHRH.
These probabilities can be calculated by direct count also, For
example, the chance of getting 3 tails in a single toss of 3 coins is}. The
chance of getting 2 tails (combined with one head) is 2, the chance of
. getting | tail (combined with 2 heads) is $ and the chance of getting no
tails is 5. In general in л tosses of a coin the probabilities of the various
possible events, i.e., (obtaining 0, 1, 2......... п heads) are given by the
successive terms of the binomial expansion of (g+-p)", which is
(q-I-p)^ —4" +" Cig p+" Cog” p+... "C,qnp* 4... .
p^.
These terms may be listed in the form of a probability distribution
table as follows :
PROBABILITY TABLE FOR NUMBER OF HEADS
US
E Number of heads Probability
x P
Е x 0 q
1 "Ciqh*p
ралы 23 nC,qn-3p?
-
nC,qn-3p*

n C,qn-rpr

= са
 -=

; THEORETICAL DISTRIBUTIONS
А-74-
we obtain probability of
Since by expanding the binomial " (q--p)* is naturally called the
bilit y distr ibuti om
v 0, 1, 2.........7 heads the proba
the binomial distribution. The
binomial probability distribution or simply
‚ general form of the distribution is
PTS P(r)—-"C,q'-tp'
exactly r successes.
where P(r) denotes the probability of getting
and non-occur-
* Thus for an event R with probability of occurrence p number. of
of the
rénce q, if п trials are made probability distribution It is custom ary to
occurrences of A will be as set in the above table.
non-oc curren ce the
call the occurrence of an event the ‘success’ and
‘failure’.
. f we want to obtain the probable frequencies of the various out-
be used :
comes in N sets of n trials, the following expression shall
Ма+р)*
N(q- p — N(q"-" Cg p+ "Cag" p+ c EPCq7tp! ... +P")
The frequencies obtained by the above expansion are known as
expected or theoretical frequencies. On the other hand, the frequencies
experiments are called actual or observed
- .actually obtained by making the observed
frequencies. Generally, there is some difference between
and expected frequencies but the difference become s smalle r and smaller
; '
as N increases.
It should be noted that the variate in the binomial distribution is a
takes
discrete one and not continuous, i.e., the number of successes (x)
only integral values.
Obtaining Coefficients of the Binomial
obtaining coefficients and exponents for any power of the bino-
w^
> For
mial, the following rules may be remembered. To find the terms of the
expansion of (94-Р)".
yok. The first term is q".
‚2. The second term is nq™ ?p. L-
= 3. In each succeeding term the power of gis reduced by 1 and the
power of p is increased by 1.
ww. 4. The coefficient of any term is found by multiplying the coeffi-
cient of the preceding term by the power of q in that preceding term, and
dividing the product so obtained by one more than the power of p in
that preceding term. ч
» When we expand (q+7)", we get E,
` (а+р)"=4"%--"С,"-1р--"Сд%р?-+-.........р"
where 1, *C;, "C,......... are called the binomial coefficients. Thus in the
expansion of (q--p)* we will have
49-р)°--4°--54%р--104%р°--104р3--59р*+-р°
and the coefficients will be 1,5, 10, 10, 5,1.
From the above binomial expansion, the following general relation-
ships should be noted : А = SOR
hel. The number of terms in a binomial expansion is always n--1l.
+
 THEORETICAL DISTRIBUTIONS" . › А-25

2. The exponents of p and £ for’ any single term; when- added

together, always sum to 7. % ч
3. The exponents of q are т, (n— 1), (п—2).:#...... 1,0, respectively, ғ
and thë exponents of p are 0, 1, 2......... (n—1), n, respectively, (note:
р°=1„; q*—1). ;
^4. The coefficients for the n+1 terms of the distribution are always
symmetricalascending to the middle of the series and then descending,
when л is odd number, n-+1 is even and the coefficients of the two central
terms are identical. х
The coefficients of the binomial expansion can very conveniently b
obtained from the Pascal's* triangle given below : *
PASCAL'S TRIANGLE
Numberin Binominal Coefficients Sum
sample n “»
i 3
2 1 4
3 1 1 8
4 1 PA А 16
5 Ф 1:.-57230.-40- 5-4 32
.$ 1 765 153159007 155406 aL 64
? 7 1 12,01 су S| 7 1 128
8 1 8 28 56 70 56 28 1 256 =
9 1 9 36 84 126 6 8 3i 9 1 512
10 1 10 45 120 210 252 210 120 45 10 i 1.024
Ў Inspection of the above table will show that each term is derived by
adding together the two terms in the line above which lie on either side
ofit. Thus, when n=6, the fourth term 20, is found by adding together
the terms 10 and 10 in the line for п=5. The expansion yields terms
which form a symmetrical distribution, because the probability of the
events р and д are equal, that is, p—g—'5. Sd
# It should be noted that the binomial distribution is symmetricalif ,
p=0'5 and skewed if p#0'5. When p<0'5, is skewed to the right, when
p>0'5, it is skewed to the left. The interchange of p and g in any binomial
distribution yields its mirror image. The skewness of the binomial dis-
tribution irrespective of the size of p, becomes less pronounced as"
increases. * ул
Properties of the Binomial Distribution
1. The shape and location of binomial distribution changes as p
. changes for a given n or as п changes fora given p. As p increases fof
a fixed л, the binomial distribution shifts to the right. This is demons-
- trated in the diagrams (page A-2°6) where a series of line graphs have
been*prepared for п=6 and p—0:10, 0°20, 0°30, 0:50, 0°80 and 0°90. б
.32. The mode of the binomial distribution is equal to the value of x
which has the largest probability. For example, if п=6 and p—0'3, the .
mode is equal to 2. While for n=6 and p—0'9 the mode is equal to 6.,
The mean and mode are equal if zpisan integer. For example, when
n=6 and р=0:50, the mean and mode are both equal to 3. For fixed n,
both the mean and mode increase as p increases. H
3. As л increases for a fixed p, the binomial distribution moves to
the right, flattens, and spreads out. The mean of the binomial distribu-
* Blaise Pascal wasa famous French Philosopher and a Mathematician.
> p 4 à
= .
Е
+
A-26 THEORETICAL DISTRIBUTIONS
6 . =]
tion, пр, obviously increases as л increases with р held constant. E
larger п there are more possible- outcomes of а binomial experiment an
the probability associated with any particular outcome becomes smaller.
BINOMIAL PROBABILITY DISTRIBUTION FOR N=6
0-6
! 2-01 р=0:9
о | t
beat
e2- "——T

0
mus y
345

P-0-8

OF К< 23d 56 0123456

4. If N is large and if neither p nor q is too close to zero, the bino-
mialdistribution can be closely approximated by a normal distribution
with'standardized variable given by z NES * The approximation
becomes better with increasing N, and in the limiting case is exact.
Constants of the Binomial Distribution
The mean of the binomial distribution is np and standard devia-
tion 4/npd T
Y Proof. If pis the probability of success and 9 the probability of failure in one
trial then in n independent trials the probabilities of 0, 1, 2, 3,-.....n successes are given
N4 the Ist, 2nd, 3rd,...... n-lthterm of the binomial expansion (g-+p)". Thus we
Ive

E Ри к. xpi)
0 q^ 0xq^
1 "Cg" : EXT p
s RU -272
P л\п—1) n-ap?
2x1 qup

n
Н
р"
o TENE
пр"
THEORETICAL DISTRIBUTIONS A2.7

г В UNE LLEx.p(x)
The arithmetic mean by definition is — X(px)

3x p(2) 0g ongn-2p
2 0— guages... np
—ngqnp--n(n—1) 204 "m пр"
Taking пр common
=nplg" + (n— 1)gn7p4-......-pn1)
-—np(gq4-p) [since the expansion in brackets is the
expansion of the binomial (q--p)"-]
=np(1)"=np C^ 4p—1]
Thus Zx.p(x)—np ('." the sum of probabilities1)
Thus the mean of binomial distribution is np.
The standard deviation of binomial distribution is 4/npq.
Proof. с? or us—vs—v;? (where v; and уг are moments about origin, zero)
и={х°.р(х)}
уу=пр

Bx? p(x) m0 qn нта + ЛЕ) guage MOD)

2, — =

(02)
l
aie +Еп°рк

апп
1g tp З 002) gy
Ч.лар"
=nplar2+2(n—1)gr-tpt4 3070019 cap, see npn-t
Breaking second, third and following terms into parts, we get

Zx* p(x) e pl(gn714- (n—1)gn-2p + DUP

2x1
qu-sps p... par)
пер + 20202.gene sc (npet)
=nplat+p("+ (n—1)p{qr
+ (n—2)g^73p4-......... pna
=npll+(n—1)p(q+p)"]
(9+р)"—1=1 and the expression to the right is expansion of (g¢+p)""=1)
—np[14- (n—1)p. 1]
—np[1--np—p]
=пр=п?р?—пр%
Mays—v;*
(0 =np+n*p*—np*—(np)*
—np--ntp?
—np? —nip*
=np—np*
=np(1—p)
=npq Г. (1—p)=q)
409 ог Vus V/npq
: Thus the standard deviation of binomial distribution is V пра or the variance
s прд.
In a similar manner as above we can show that for a binomial
distribution
Dri ра(4— р) and
u,—-3r'p'g? J-npq(1 —6pg)
In case of theoretical distribution s (рх) is always 1.
 AS THEORETICAL DISTRIBUTIONS

From the various. moments the valu& of 8; and f, can also be

computed е. К> " К
ву _1р'(9—р) = (q—p*
tow) D "pq
Tf 8;—0 the distribution is symmetrical, 7.e., there is no skewness. Prof.
Fisher gave the following measure of skewness
4 uv
If y,—0 the distribution is symmetrical, if y; is more than zero the
distribution is positively skewed and ifit is less than zero the distribution
is negatively skewed. It should be noted that y, is a better measure of
skewness compared to B1 because B, will always be positive whereas ү,
сап be both positive as well as negative.
W Lm p'q*-Enpqd
—6pg) , |1—6pq
=з mpg Э mq
Prof, Fisher gave the following measure of kurtosis :
Үз=8—3
If the distribution is normal, y, would be zero and y, would also be zero.
But the converse is not true, ie. even if both v, and vy, are zero the
distribution is not necessarily normal. If y, is positive, the distribution is
leptokurtic and if y, is negative the distribution is platykurtic.
The various constants of the binomial distribution can be listed in
the following table :
Mean=np
» Standard Deviation=/npq
First Moment or pı=0
Second Moment or g, =npq
Third Moment or Us npq(q—p)
Fourth Moment or 473m p?q*--npq(À — 6pg)
S А 8,— (4@—р)?
npg
23.4 1-600
£,—-3-- uS
Importance of the Binomial Distribution
Qo. TheIhi binomial
1 probability
D distri
stri bution is a
distribution that is useful in describing an enorm dis
crete probabilit
ili
ous ea af real life
Ae Gar For example, a quality control inspec
proba ility of obtaining of bad light bulbs in a tor wants to know the
c NR ios the ү ШЕ ти defective. Не random sample of 10 bulbs
SAG
newers from can quickly obtain the
Atablesof D. the
Ue E j omia
binomi al , Probability
ili distri bution
istribu ti s, The
l. 4 The outcome or results of
ea ch trial
ial in
i the process are
chara
OE cteriНЫzed ШЫ
as one of two о ttypes of possiible outcomes . In other words
. ,. 2 The
possibility of outcome of any
independent of the results of previous Ma trial d i
PO ee deos and ie
фоль The:following g examples
i willill illustrate
i icaticions of binomial
the Applicat nori
 THEORETICAL DISTRIBUTIONS A-29

Illustration 1. A coin is t68sed six times. What is the probability of obtaining

four or more heads ?
Solution, When a coin is tossed the probability of head and tail in case ofan
unbiased coin are equal, i.e., p—q9-73.
The various possibilities for all the events are the terms of the expansion
(а+р)*.
© (q«-p)* — d'--6dp--159p *-20g*p?- 15g?p*4 Gap +p
Л The probability of obtaining 4 heads is
15p*g*—
15x(4)°(4)*=0'234
The probability of obtaining 5 heads is
6ар5=6(2) (4)°=0'094
The probability of obtaining 6 heads is
p*—(9*—6016
<. The probability of obtaining 4 or more heads is
0:234--070944-0016—0'344,
Illustration 2. Assuming that half the population is vegetarian so that the
chance of an individual being a vegetarian is $ and assuming that 100 investigators can
take sample of 10 individuals to see whether they are vegetarians, how many inves-
tigators would you expect to report that three people or less were vegetarians ?
(М.А. Econ. Delhi, 1966)

Soultion, Probability of a person being vegetarian or p =$

Ве q—-1—p-1—i-i
By expanding the binomial 100(3--3)'5, we get the number of investigators who
are expected to report 0, 1, 2.......-. 10 people who are vegetarians. The number of
кено» who are expected to report that three people or less were vegetarians is
given by
N(q)!94-(N 029p) + (№454%р°) + (N 120g? p*) "
=100 x (3)19 4-100 10 (3)* (3) 4-100 45 (3 (3)*--100 x 120(4)7(4)*
= 100, 1000, 4500 |12000
io24 1024" 1024 1024
..17600 . t
EA МЕ >
Hence 17 investigators would report that 3 or less people were vegetarian.
Illustration 3. Ifthe probability of defective bolt is 0'1, find (a) the mean and
standard deviation for the distribution of defective bolts in a total of 500, and (b) the
moment coefficients of skewness and kurtosis of the distribution. Ы

Solution, (а) p=0'l, n=500 А

Меап=лр=500х'1= 50.
Thus we can expect 50 bolts to be defective. n

о=\/ пра
n=500, p—0'1 and 4=0`9
в=\/ 500х01х0'9 —67 *
(Б) Moment coefficient of skewness, i.e., Үл

n-2Vv А
— azp _ (09-01) _ 08.
КОСУ СЮРТЕ Sale
A-210 THEORETICAL DISTRIBU TIONS
Since yı is more than zero the distribution is positively skewed. Howe ver the
skewness is yery moderate.
Moment coefficient of Kurtosis
onus "m
igp
£73 c4 npo
OPE"

zu 1—6(0 (0:9)

Ы —34-046
=3+ _,.,
449 =3`01
123—301 —3— 4-0701
Since ү, isp ositive the distribution is platykurtic
(q--p)*—q'--
6dp 6g*p*-- 20g?p*--15g? p*--6qp*-i-p*
Illustration 4. The incidenee of occupational disease in an industry
the workmen have a 20% chance of suffering from it. What is such that
is the probability that out
of six workmen 4 or more will contact disease ?
Solution. The probability of aman suffering from disease
or
20
VEN RC
The probability of a man not suffering from disease,
ie,
q-1—l-e 4.
5
Hence the probability of 6men, 5 men, 4 men, etc., sufferi
ng from disease аге
в
terms in the binomial expansion er( +)
(4-Ер)*=4°--6°р-1-6др?-}-204°р--15д°р*-4-бар5-Ер®
The probability of 4or more, i.e., 4, 5 or 6 success
es is

NG) HD)
7 15g!p*--6gp5— p*

-15x16 | 6x4
15625 + 15625 "15625
==__265_ . 53
15625 ^ 312
Fitting a Binomial Distribution
When a binomial distribution is to be
fitted to observe data the
following procedure is adopted :
1. Determine the values of
the other'can be found out by the simple
(1—p). When p and q are equal the
relationship р=(1—4), and q =
distribution is symmetrical, for p
and q may be interchanged without altering the value
consequently terms equidistant from the two of any term, and
If p and q are unequal, ends of the series are equal.
the distribution is skew. If р is less than $, the
hen p is more than § the distri-
2. Expand the binomial (g--p)*. The power n is equal to one less
than the number:of terms in the expanded
coins are tossed (n—2) there will binomial. Thus when two
be three terms in the binomial.
Similarly when four coins are tossed
(n=4) there will be five terms, and
$0 on.
3. Multiply each term of the expanded
frequency), in order to obtain the expected frequ binomial by N (the total
ency in each category.
 THEORETICAL DISTRIBUTIONS A-211

The following example shall illustrate the procedure :

Illustration 5. Eight coins are tossed ata time, 256 times. Number ofheads
observed at each throw is recorded and the results are given below.’ Find the expected
frequencies. What are the theoretical values of mean and standard deviation
Calculate also the mean and S.D. of the observed frequencies.
"B Мо, of heads at Frequency No. of heads at Frequency
a throw a throw
0 2 5 56
1 6 6 VW. 1732
2 30 7 10
3 52 8 1
4 67
Solution, The chance of getting a head in a single throw of one coin is 4.
Hence p=}, q=}, n=8, N—256.
By expanding 256 (}+4)® we shall get the expected frequencies of 1, 2..
heads (successes).

M No. of heads (x) — N Xx^C,qn-tpf

Frequency
E o rw 5 2 256(3\5= 1
1 256 x"C (IGV = 8
2 256x ®Ca(#)*(4)°=28
3 256x*C,(1)*(1)5—56
4 256 x ®C,(4)*(4)*=70
Т 256x*C,(3)5(3)*—56
6 256x *C(3)*(3)* —28
7 256x*C,(y'(3!— 8
8 256x (4)8= 1
Total =256

The mean of the above He EURE is np

—8x1-4.

The standard deviation is /npg

-VEXEX8-/2- 1414. A.
) These are the mean and standard deviation of the expected frequencies. The
3 mean and standard deviation of the observed frequencies shall be :

x f d fd fat
и 0 2 —4 ES КУ "a2 | t
1 6 -3 —18 54
2 30 7 —60 = 120
3 52 —1 —52 52
4 67 0 0 0
5 56 1 56 56
6 32 2 64 128
7 10 3 30 90
8 1 4 4 16

N=256 Efd—16 Zfd*—548

sen eft
X=4+ 5р d
16:1
=4+ 556 =406

& SE QE
 ARE X ums THEORETICAL DISTRIBUTIONS
; aa
£ =, /38 (316 y
256 (25 À
. -4/ri4-0001-4/7137-1462, ©
Illustration 6. The following data show the number of seeds germinating out of
10 on damp filter for 80 set of seeds. Fit a binomial distribution to this data :
X: 0 1 2 3 4 5 6 7 8 9 10
Ef СКУ Жш. y ЫЙЫ иеш 0 0 0
e Ч " (B Sc., Agra, 1973)
" Solution. FITTING BINOMIAL DISTRIBUTION
x y fX
0 Wifes 226
} Э 8
= =. 3
Ter
P.
3 12 366
4 8 32
5 6 30
6 0 0
VA 0 0
8 0 0
9 0 0
10 0 0
а ECT
—80 EfX—174

X=ug MA
go 72195
i 174
= @ But теап=пр=-ыу=
Dyў p
174
300 =0°2175
_„.

LR q= 1—p=0'7825,
.Hence the binomial distribution to be fitted to the data is
. = А 80(0778254-02175)1* -
, The theoretical frequencies are th
30078250215)"andaretabulated below: — si
(6705 іп the expansion 0 f
i i
UF
z х Lv Theoretical frequencies
COR EN Tal a Aree So
T ud А 80x(7825)— 69
1% 80x 10(7825)*(21 75)1—
191
a 80% 45(/7825)*(2175)2— 240
3 80x 120(°7825)7(-2175)3=17'8 »
“4 80x 21007825)*(2175)4—
PX.
7
80x 252(7825)5(2175)5— 8'629
80Х210(°7825)4(°2175)в—= 07

: s, 80x 120(7825)*(2175)— 01
80% 45(°7825)2(2175)в—
G
Э» 80x 10(7825)(21759— 00 00
[Q0 ^ BOX(2075)- 00
‘ *» Total
otal =%
80 9
ў Illustration 7
Twelve 7.
appearing was considered to be dice were thrown 4096 ti E Ea h 4, 5, or6 spot
a success,
late the theoretical frequencies for 0, 1, 2 tswhixo Mr: em wa: a ilu Calcu-
zi мое. i
Notes.Solution
apes , ThereThere areare 4,4,096 trials.
i Since
ince either
еї 4, 5 ог 6isi conside red’ a success
ideré *
 THEORETICAL DISTRIBUTIONS - T ПА
Р The terms of the binomial (4-Ер)% will give the probabilities of 0, 1, 2...
successes. — Ф
Hee ^C n=12, q—À, and p-i. "
By expanding 4096 (3-1)
we get frequencies corresponding to 0, 1, 2... -.12 successes.
1., 12 , 66 , 20 , 495 792 , 924 , 792 1 495
4096( то + 4996 ^ 4096 ^ 4096 1 4096 4096 4096 7 4095 + 4096
20. 66 , 12 , +d
+ 2096 + 4096 7 4096 +0)
fo will denote observed frequencies and f, expected frequencies. The observed
frequencies cannot be in fraction but thewhole expected frequencies may be in fractions.
number.
However, they may be approximated to the
The results can be tabulated as follows :
Number of Theoretical Number of Theoretical
# successes frequencies successes frequencies

0 1 7 792
12 8 495
1
2 66 9 220
, 10 *66
3 220
495 1 12
4 "
* 5 792 12
924

Tiüustration 8. The following is the frequency distribution of 128 throws of `

heads :
seven coins, according to the number of
1 2 3 w*4 s 6 7 Total
No. of heads 0
6 19 35 30 23 7 lè 128
Throws 7
Fit a binomial distribution under the hypothesis that the 'coins are unbiased
*
distribution ?
«What is the mean and the standard deviation of the fitted M. Com. Delhi, 1971).
On the hypothesis that the coins are unbiased, p=q=.
Solution,
N=128 and n—7.
given by expansion (24-5). *
THE probability of 0, 1, 2----..-+ 7 heads will be
(BER)? "ei (9) + "es (3* QD)" 7*6 (0* Q)* "e (D* „
-F'cg(3) (B *4-*es*(3)40*3-*ez(. p
"Ж -
=(4)[1+7+214+35+35+21+7+4 1}
have to multiply . each term by N,
In order to obtain the frequencies we will A
i.e., 128.
(3+) —128x -p (1+T+214+35+35+2
т
1) Ф
+741
are :
Thus the expected frequencies
1 2 3 4 5 6 >
x 0
1 7 21 35 35 21 iTe i
fe
distribution
Mean and standard deviation of this
deviation vnpg
Mean of the binomial ditsribution is пр and standard
Here n=T, p=}, I=? s "
А
Д2 mean=7X}=3'5 and
-AHxixi-VYl75-132 *
Standard deviation
П. POISSON DISTRI BUT ION “ys
x
probability distribution and is very
Poisson distribution is a discrete ch mathe-
widely used in statistical works It was originated by ang,Frenthe Poisson
in 1837 . Stri ctly spe aki
matician, Simeon Denis Poisson,