DISTRIBUTIONS

Page 1 of 36
 PLEASE REFER TO MOODLE FOR NOTES ON “DISTRIBUTIONS”.

DISCRETE DISTRIBUTIONS (P.M.F.’s).

BINOMIAL DISTRIBUTION. Parameters (N, p). B(N, p).

Necessary Criteria to be a Binomial Distribution.

1. The number of observations N is fixed.
2. Each observation is independent.
3. Each observation represents one of two outcomes ("success" or
"failure").
4. The probability of "success", p, is the same for each outcome.

Let X: Binomial random variable.

P(X = x) = p(x) = B(x; N, p) = �𝑁𝑁𝑥𝑥� 𝑝𝑝 𝑥𝑥 (1 − 𝑝𝑝)𝑁𝑁−𝑥𝑥 x = 0, 1, 2, …, N.

E[X] = Np; Var(X) = Np(1 - p).

Page 2 of 36
EXAMPLE SET 9.
Problems.
1. Experiment. Given a biased coin. P(head) = 0.8.
Coin is tossed 10 times.
A. What is the probability of obtaining 7 heads?
B. What is the probability of obtaining at most 7 heads?
C. What is the probability of obtaining more than 0 heads, but less
than 5 heads?

Page 3 of 36
EXAMPLE SET 9.
Solutions.
1. Experiment. Given a biased coin. P(head) = 0.8.
Coin is tossed 10 times.

A. What is the probability of obtaining 7 heads?

Let X: R.V. denoting total number of heads obtained in 10 tosses.

N = 10; p = 0.8.
X ~ B(N, p) = B(10, 0.8). (Notation ~ ∶ is distributed, or has
a Binomial Distribution (in this case))
P(X = 7)?

P(X = 7) = p(7) = �10

7
� 0.87 (0.2)3 = 0.2. (or use Binomial table)

B. What is the probability of obtaining at most 7 heads?

P(X ≤ 7)?

P(X ≤ 7) = ∑7𝑖𝑖=0 𝑝𝑝(𝑖𝑖) = ∑7𝑖𝑖=0 𝐵𝐵(𝑖𝑖; 𝑁𝑁, 𝑝𝑝) = ∑7𝑖𝑖=0 𝐵𝐵(𝑖𝑖; 10, 0.8)

= ∑7𝑖𝑖=0�10
𝑖𝑖
� 0.8 𝑖𝑖
(0.2)10−𝑖𝑖
= 0.3222 = 1 − ∑10
𝑖𝑖=8 𝐵𝐵 (𝑖𝑖; 10, 0.8).

(or use Binomial table)

C. What is the probability of obtaining more than 0 heads, but less

than 5 heads?

𝑃𝑃(0 < 𝑋𝑋 < 5)?

Page 4 of 36
 𝑃𝑃(0 < 𝑋𝑋 < 5) = ∑4𝑖𝑖=1 𝑝𝑝(𝑖𝑖) = ∑4𝑖𝑖=1 𝐵𝐵(𝑖𝑖; 𝑁𝑁, 𝑝𝑝) = ∑4𝑖𝑖=1 𝐵𝐵(𝑖𝑖; 10, 0.8)

= ∑4𝑖𝑖=1�10
𝑖𝑖
� 0.8𝑖𝑖 (0.2)10−𝑖𝑖 = 0.0064. (or use Binomial table)

POISSON DISTRIBUTION. Parameter 𝜆𝜆, 𝜆𝜆 > 0. p(𝜆𝜆).

Assumptions and validity.

The Poisson distribution is an appropriate model if the following

assumptions are true:

• X is the number of times an event occurs in an interval, and X can

take values 0, 1, 2, ....
• The occurrence of one event does not affect the probability that a
second event will occur. That is, events occur independently.
• The average rate at which events occur is independent of any
occurrences. For simplicity, this is usually assumed to be constant,
but may in practice vary with time.
• Two events cannot occur at exactly the same instant; instead, at
each very small sub-interval exactly one event either occurs or
does not occur.

If these conditions are true, then X is a Poisson random variable, and the
distribution of X is a Poisson distribution.

The Poisson distribution is also the limit of a Binomial Distribution, for

which the probability of success for each trial equals λ divided by the
number of trials, as the number of trials approaches infinity.
NOTA BENE: Copied from Wikipedia.

The Poisson Dist. is the dist. of rare events. The number of earthquakes,
wars, wrong phone calls/person, or aircraft disasters in a region,
country, city, etc. per time period, are just some of the rare events.
Page 5 of 36
Let X: Poisson random variable.

𝑒𝑒 −𝜆𝜆 𝜆𝜆𝑥𝑥
P(X = x) = p(x) = p(x; λ) = x = 0, 1, 2, ….
𝑥𝑥!

λ: mean (expected) number of “successes”. Rate per unit time at which

events occur.

E[X] = λ; Var(X) = λ.

Page 6 of 36
THEOREM. Sum of two Poisson Distributed random variables.
Let 𝑋𝑋1 : Poisson Distributed random variable with parameter 𝜆𝜆1 .
𝑋𝑋2 : Poisson Distributed random variable with parameter 𝜆𝜆2 .

If X = 𝑋𝑋1 + 𝑋𝑋2 then X is Poisson Distributed with parameter 𝜆𝜆1 + 𝜆𝜆2 .

E[X] = 𝜆𝜆1 + 𝜆𝜆2 ; Var(X) = 𝜆𝜆1 + 𝜆𝜆2 .

THEOREM. Sum of N Poisson Distributed random variables.

Let 𝑋𝑋𝑖𝑖 , i = 1, 2, 3, …, N be N Poisson Distributed random variables each
with corresponding parameter 𝜆𝜆𝑖𝑖 .

If X = ∑𝑁𝑁 𝑁𝑁
𝑖𝑖=1 𝑋𝑋𝑖𝑖 then X is Poisson Distributed with parameter ∑𝑖𝑖=1 𝜆𝜆𝑖𝑖 .

E[X] =∑𝑁𝑁 𝑁𝑁
𝑖𝑖=1 𝜆𝜆𝑖𝑖 ; Var(X) =∑𝑖𝑖=1 𝜆𝜆𝑖𝑖 .

Page 7 of 36
EXAMPLE SET 10.
Problems.
1. On the average, there are 5 assembly-line failures/day. The
number of failures on different days are independent.
Let X: be the R.V. denoting the number of failures on any day.
Assume X is a Poisson R.V.

A. What is the prob. of having less than 4 failures on any given

day?
B. What proportion of days will have less than 4 failures?
C. What is the prob. of 4 failures on any given day?
D. What is the prob. that there will be 4 assembly-line failures in
exactly 6 of the next 8 days? In other words, 6 of the next 8
days must each have 4 failures/day.
E. What is the prob. that there will be 1 assembly-line failure in
exactly 2 of the next 5 days? In other words, 2 of the next 5
days must each have 1 failure/day.

2. The prob. of exactly one fault in a 1 mile stretch of highway

1
is . The prob. of two or more faults on the same one mile
1000
stretch is 0. The highway is 3000 miles long.
Let X: denote the number of faults along the entire highway.
Number of faults on nonoverlapping one mile stretches of highway
are independent. Assume X has a Poisson Dist.
What is the probability of there being 5 faults on this 3000 mile
highway?

Page 8 of 36
3. It is a fact that the number of people falling out of a certain roller
coaster/day, is Poisson Dist., with mean 5.
Number of people who fall on any given day is independent of day,
and independent on number who fall on other days. Over a two-day
span, what is the prob. that the number of fallen people does not
exceed 4?

Page 9 of 36
EXAMPLE SET 10.
Solutions.
1. On the average, there are 5 assembly-line failures/day. The
number of failures on different days are independent.
Let X: be the R.V. denoting the number of failures on any day.
Assume X is a Poisson R.V.

A. What is the prob. of having less than 4 failures on any given

day?

P(X < 4)?

λ = 5.
−5 5𝑖𝑖
3 𝑒𝑒
P(X < 4) = ∑3𝑖𝑖=0 𝑝𝑝(𝑖𝑖; 5) = ∑𝑖𝑖=0 = 0.265. (or use Poisson
𝑖𝑖!
table)

B. What proportion of days will have less than 4 failures?

Over the long run, 26.5% of days will have less than 4 failures.

C. What is the prob. of 4 failures on any given day?

P(X = 4)?

𝑒𝑒 −5 54
P(X = 4) = p(4; 5) = = 0.1755. (or use Poisson
4!
table)

Page 10 of 36
 D. What is the prob. there will be 4 assembly-line failures in
exactly 6 of the next 8 days? In other words, 6 of the next 8
days must each have 4 failures/day.

B(i; N, p) = B(6; 8, 0.1755)?

B(i; N, p) = B(6; 8, 0.1755) = �86� (0.1755)6 (0.8245)2 = 0.0006.

E. What is the prob. there will be 1 assembly-line failure in

exactly 2 of the next 5 days? In other words, 2 of the next 5
days must each have 1 failure/day.

B(i; N, p) = B(2; 5, P(X = 1))?

𝑒𝑒 −5 51
P(X = 1) = p(1; 5) = = 0.0337.
1!
∴,
B(2; 5, 0.0337) = �52� (0.0337)2 (0.9663)3 = 0.01.

2. The prob. of exactly one fault in a 1 mile stretch of highway

1
is . The prob. of two or more faults on the same one mile
1000
stretch is 0. The highway is 3000 miles long.
Let X: denote the number of faults along the entire highway.
Number of faults on nonoverlapping one mile stretches of highway
are independent. Assume X has a Poisson Dist.
What is the probability of there being 5 faults on this 3000 mile
highway?

P(X = 5)?

λ: mean number of “successes”. (Faults are “successes”)

Page 11 of 36
 λ = (length of highway) (prob. of exactly 1 fault on 1mi. stretch)
λ = (3000) (1/1000) = 3.
∴,
𝑒𝑒 −3 35
P(X = 5) = p(5; 3) = = 0.101.
5!

3. It is a fact that the number of people falling out of a certain roller

coaster/day is Poisson Dist., with mean 5.
Number of people who fall on any given day is independent of day,
and independent on number who fall on other days. Over a two-day
span, what is the prob. that the number of fallen people does not
exceed 4?

Let 𝑋𝑋1 : denote number of fallen people on day 1.

𝑋𝑋2 : denote number of fallen people on day 2.
𝑋𝑋 = 𝑋𝑋1 + 𝑋𝑋2 .

P(X ≤ 4)?

𝑋𝑋1 , 𝑋𝑋2 are Poisson Distributed, each with mean 5.

𝑋𝑋 = 𝑋𝑋1 + 𝑋𝑋2 , is Poisson Distributed, with mean 10.

−10 10𝑖𝑖
4 𝑒𝑒
P(X ≤ 4) = ∑4𝑖𝑖=0 𝑝𝑝(𝑖𝑖; 10) = ∑𝑖𝑖=0 = 0.0293. (or use Poisson
𝑖𝑖!
table)

Page 12 of 36
EXAMPLE.

Poisson Dist. ≈ Binomial Dist. for N ≥ 20, p ≤ 0.05.

If N ≥ 100 and Np ≤ 10 then approximation is excellent.

E.G.: N = 48; p = 0.05; λ = Np = (48)(0.05) = 2.4.

x B(x; 48, 0.05) p(x; 2.4) |∆|

0 0.085 0.091 0.006
1 0.215 0.218 0.003
2 0.266 0.261 0.005
3 0.215 0.209 0.006
4 0.127 0.125 0.002
5 0.059 0.060 0.001
6 0.022 0.024 0.002
7 0.007 0.008 0.001
8 0.002 0.002 0

PLEASE REFER TO MOODLE FOR AN

“INTERESTING POISSON EXAMPLE”.

Page 13 of 36
 CONTINUOUS DISTRIBUTIONS (P.D.F.’s).
EXPONENTIAL DISTRIBUTION. Parameter λ, λ > 0. e(𝜆𝜆).

The Exponential Distribution is the probability distribution of the time

between events in a Poisson Point Process, I.E., a process in which
events occur continuously and independently at a constant average
rate.

NOTA BENE: Copied from Wikipedia.

The Exponential Dist. is the dist. of the time until the next earthquake,
or next war, or next wrong phone call, or next aircraft disaster. It is
the time from now. It could also be the waiting time between events
(between two people arriving), in a Poisson Process (same λ is used as
in the Poisson Process).

Let X: Exponential random variable.

−𝜆𝜆𝜆𝜆
P(X = x) = f(x) = � 𝜆𝜆𝑒𝑒 𝑖𝑖𝑖𝑖 𝑥𝑥 ≥ 0
0 𝑜𝑜𝑜𝑜ℎ𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒.

Page 14 of 36
EXAMPLE SET 11.
Problems.
1. Let the R.V. X denote the number of fish caught/hour at a certain
radioactive lake.
Assume X has a Poisson Distribution with mean 1.8 fish/hour.
Compute the probability a person fishing at this lake for 1 hour:

A. Will catch no fish.

B. Will catch at least 4 fish.
C. Will catch at most 2 fish.
D. What is the probability it will require less than 20 mins. between
catches?

Page 15 of 36
EXAMPLE SET 11.
Solutions.
1. Let the R.V. X denote the number of fish caught/hour at a certain
radioactive lake.
Assume X has a Poisson Distribution with mean 1.8 fish/hour.
Compute the probability a person fishing at this lake for 1 hour:

A. Will catch no fish.

P(X = 0)?

P(X = 0) = p(0; 1.8) = 0.1653.

B. Will catch at least 4 fish.

P(X ≥ 4)?

P(X ≥ 4) = 1 − P(X ≤ 3) = 1 − ∑3𝑖𝑖=0 𝑝𝑝(𝑖𝑖; 1.8) = 0.1087.

C. Will catch at most 2 fish.

P(X ≤ 2)?

P(X ≤ 2) = ∑2𝑖𝑖=0 𝑝𝑝(𝑖𝑖; 1.8) = 0.7306.

D. What is the probability it will require less than 20 mins. between

catches?

P(less than 20 mins. between catches)?

Time between events in a Poisson Process is sought ⇒ use

Exponential Distribution.
Page 16 of 36
 −1.8𝑡𝑡
f(t) = �1.8𝑒𝑒 𝑖𝑖𝑖𝑖 𝑡𝑡 ≥ 0
0 𝑜𝑜𝑜𝑜ℎ𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒.
1/3
P(less than 20 mins. between catches) = ∫0 𝑓𝑓 (𝑡𝑡) 𝑑𝑑𝑑𝑑 t is in hours

1/3
= ∫0 1.8𝑒𝑒 −1.8𝑡𝑡 𝑑𝑑𝑑𝑑

P(less than 20 mins. between catches) = 0.45.

NORMAL DISTRIBUTION. Parameters 𝜇𝜇, 𝜎𝜎 2 . N(𝜇𝜇, 𝜎𝜎 2 ).

Necessary Criteria to be a Normal Distribution.

The four characteristics of the Normal Distribution.
Normal distributions are symmetric about their mean, unimodal,
asymptotic, and the mean, median, and mode are all equal.

Page 17 of 36
Page 18 of 36
Let X: Normal random variable.

1 𝑥𝑥 − 𝜇𝜇 2
1 − � �
P(X = x) = f(x) = 𝑒𝑒 2 𝜎𝜎 − ∞ < 𝑥𝑥 < ∞.
√2𝜋𝜋 𝜎𝜎

Def.: Standard Normal Distribution;

A random variable, X, which is Normally Distributed (having a Normal
Distribution), with mean 𝜇𝜇 = 0, and variance 𝜎𝜎 2 = 1, is said to be a
Standard Normal random variable, denoted by Z.
Z is said to have a Standard Normal Distribution.

Page 19 of 36
THEOREM. Standardizing.
Let X: random variable.
E[X] = 𝜇𝜇.
Var(X) = 𝜎𝜎 2 .

The transformation of X to a Standard Normal R.V. is given by:

𝑋𝑋− 𝜇𝜇
Z= .
𝜎𝜎

NOTA BENE.
Assume X ~ N(𝜇𝜇, 𝜎𝜎 2 ).
To compute any probabilities, P(X < x), P(X ≥ x), etc., Normal Probability
tables are required.
There are no Normal Probability tables with 𝜇𝜇 ≠ 0, 𝜎𝜎 2 ≠ 1.
A Normal Probability table with 𝜇𝜇 = 0, 𝜎𝜎 2 = 1, is called a Standard
Normal Probability table. This is the only table available to compute
probabilities for a Normal R.V.
If X ~ N(𝜇𝜇, 𝜎𝜎 2 ), 𝜇𝜇 ≠ 0, 𝜎𝜎 2 ≠ 1 ⇒ X must be transformed into a
Standard Normal R.V., Z, using the standardizing transformation
Theorem above. After standardizing, the Standard Normal Probability
table may be used to compute required probabilities.

Page 20 of 36
Page 21 of 36
Page 22 of 36
Page 23 of 36
E.G.
Let X ~ N(5, 4). (Read as: X is Normally Dist. with mean 5, and
variance 4).

A. P(X ≤ 6)?

There does not exist a Normal Probability table with 𝜇𝜇 = 5, 𝜎𝜎 2 = 4 !!

Must standardize (transform) X to a Standard Normal R.V., Z.

𝑋𝑋 − 𝜇𝜇 6 − 𝜇𝜇
P(X ≤ 6) = 𝑃𝑃 � ≤ �
𝜎𝜎 𝜎𝜎

𝑋𝑋 − 5 6−5
= 𝑃𝑃 � ≤ �
2 2

1
= 𝑃𝑃 �𝑍𝑍 ≤ �
2

P(X ≤ 6) = 0.6915.

B. P(X > 7)?

There does not exist a Normal Probability table with 𝜇𝜇 = 5, 𝜎𝜎 2 = 4 !!

Must standardize (transform) X to a Standard Normal R.V., Z.

𝑋𝑋 − 𝜇𝜇 7 − 𝜇𝜇
P(X > 7) = 𝑃𝑃 � > �
𝜎𝜎 𝜎𝜎

𝑋𝑋 − 5 7−5
= 𝑃𝑃 � > �
2 2

Page 24 of 36
 = 𝑃𝑃(𝑍𝑍 > 1)

= 1 – P(Z ≤ 1)

= 1 – 0.8413

P(X > 7) = 0.1587.

C. P(X ≤ 2)?

There does not exist a Normal Probability table with 𝜇𝜇 = 5, 𝜎𝜎 2 = 4 !!

Must standardize (transform) X to a Standard Normal R.V., Z.

𝑋𝑋 − 𝜇𝜇 2 − 𝜇𝜇
P(X ≤ 2) = 𝑃𝑃 � ≤ �
𝜎𝜎 𝜎𝜎

𝑋𝑋 − 5 2−5
= 𝑃𝑃 � ≤ �
2 2

−3
= 𝑃𝑃 �𝑍𝑍 ≤ �
2

= 1 – P(Z < 3/2)

= 1 – 0.9332.

P(X ≤ 2) = 0.0668.

Page 25 of 36
EXAMPLE SET 12.
Problems.
1. The amount of Marijuana which a filling machine dispenses into 6 oz.
bags is a R.V. X, having a Normal Distribution with 𝜇𝜇 = 6oz., and
𝜎𝜎 = 0.08oz.
Compute the probabilities that:

A. One of these bags will contain at least 6.10oz.?

B. One of these bags will contain less than 5.82oz.?

C. One of these bags will contain anywhere from 5.96 to 6.04oz.?

Page 26 of 36
EXAMPLE SET 12.
Solutions.
1. The amount of Marijuana which a filling machine dispenses into 6 oz.
bags is a R.V. X, having a Normal Distribution with 𝜇𝜇 = 6oz., and
𝜎𝜎 = 0.08oz.
Compute the probabilities that:

A. One of these bags will contain at least 6.10oz.?

P(X ≥ 6.10)?

𝑋𝑋 − 𝜇𝜇 6.10 − 𝜇𝜇
P(X ≥ 6.10) = 𝑃𝑃 � ≥ � (standardizing)
𝜎𝜎 𝜎𝜎

𝑋𝑋 − 6 6.10 − 6
= 𝑃𝑃 � ≥ �
0.08 0.08

= 𝑃𝑃(𝑍𝑍 ≥ 1.25)

= 1 – P(Z < 1.25)

= 1 – 0.8944

P(X ≥ 6.10 ) = 0.1056.

Page 27 of 36
B. One of these bags will contain less than 5.82oz.?

P(X < 5.82)?

𝑋𝑋 − 𝜇𝜇 5.82 − 𝜇𝜇
P(X < 5.82) = 𝑃𝑃 � < � (standardizing)
𝜎𝜎 𝜎𝜎

𝑋𝑋 − 6 5.82 − 6
= 𝑃𝑃 � < �
0.08 0.08

= 𝑃𝑃(𝑍𝑍 < −2.25)

= 1 – P(Z ≤ 2.25)

= 1 – 0.9878

P(X < 5.82) = 0.0122.

C. One of these bags will contain anywhere from 5.96 to 6.04oz.?

P(5.96 < X < 6.04)?

5.96 − 𝜇𝜇 𝑋𝑋 − 𝜇𝜇 6.04 − 𝜇𝜇
P(5.96 < X < 6.04) = 𝑃𝑃 � < < �
𝜎𝜎 𝜎𝜎 𝜎𝜎

5.96 − 6 𝑋𝑋 − 6 6.94 − 6
= 𝑃𝑃 � < < �
0.08 0.08 0.08

= P(-0.5 < Z < 0.5)

= P(Z < 0.5) – [1 – P(Z < 0.5)]

Page 28 of 36
 = 2 P(Z < 0.5) – 1

= 2 (0.6915) – 1

P(5.96 < X < 6.04) = 0.3830.

THEOREM.
Let R.V.'s Xi , i = 1, 2, 3, …, n: be Normally Distributed, each with
mean μi , and variance σi2.
𝐼𝐼𝐼𝐼 𝑋𝑋 = ∑𝑛𝑛𝑖𝑖=1 𝑋𝑋𝑖𝑖 , then X is Normally Distributed, with mean
∑𝑛𝑛𝑖𝑖=1 𝜇𝜇𝑖𝑖 , and variance 𝜎𝜎 2 = ∑𝑛𝑛𝑖𝑖=1 𝜎𝜎𝑖𝑖 2 .
Xi and Xj are independent ∀𝑖𝑖𝑖𝑖 , 𝑖𝑖 ≠ 𝑗𝑗.

Page 29 of 36
EXAMPLE.

CARGO TRANSPORTATION PROBLEM.

A small single engine aircraft with a maximum cargo weight capacity of

350 lbs. is to be employed to transport boxes of seed to a very remote
community in the northern U.S.
The weights of the boxes are Normally Distributed. The average
weight of each box is 10 lbs., and each box has a variance of 4 lbs. Any
shipment whose total weight exceeds 350 lbs. will be rejected.
What is the minimum number of boxes, n, which could be loaded onto
the aircraft so as to have a probability of at least 0.90 that the total
weight of the n loaded boxes is greater than or equal to 300 lbs.?

Page 30 of 36
EXAMPLE.

CARGO TRANSPORTATION PROBLEM SOLUTION.

A small single engine aircraft with a maximum cargo weight capacity of

350 lbs. is to be employed to transport boxes of seed to a very remote
community in the northern U.S.
The weights of the boxes are Normally Distributed. The average
weight of each box is 10 lbs., and each box has a variance of 4 lbs. Any
shipment whose total weight exceeds 350 lbs. will be rejected.
What is the minimum number of boxes, n, which could be loaded onto
the aircraft so as to have a probability of at least 0.90 that the total
weight of the n loaded boxes is greater than or equal to 300 lbs.?

SOLUTION. (Problem with an unexpected solution).

Let 𝑋𝑋𝑖𝑖 : denote the weight of the i th. loaded box.

n: denote the minimum number of boxes to be loaded.

𝑆𝑆𝑛𝑛 = ∑𝑛𝑛𝑖𝑖=1 𝑋𝑋𝑖𝑖 ; 𝐸𝐸[𝑋𝑋𝑖𝑖 ] = 10; 𝑉𝑉𝑉𝑉𝑉𝑉(𝑋𝑋𝑖𝑖 ) = 4.

𝑆𝑆𝑛𝑛 ~ N(10n, 4n).

Must compute min. n ∋ 𝑃𝑃(𝑆𝑆𝑛𝑛 ≥ 300) ≥ 0.90?

𝑃𝑃(𝑆𝑆𝑛𝑛 ≥ 300) ≥ 0.90

𝑆𝑆𝑛𝑛 − 10𝑛𝑛 300 − 10𝑛𝑛

𝑃𝑃 � ≥ � ≥ 0.90
√4𝑛𝑛 √4𝑛𝑛

Page 31 of 36
 300 −10𝑛𝑛
𝑃𝑃 �𝑍𝑍 ≥ � ≥ 0.90 [Nota Bene: P(Z ≥ −1.29) = 0.90]
√4𝑛𝑛

∴,
300 −10𝑛𝑛
≤ −1.29 Constraint 1.
√4𝑛𝑛

Determine min. n from Constraint 1 ⇒ min. n = 32.

300 −10(31)
n = 31 violates Constraint 1: = −0.898 > −1.29.
�4(31)

N 300 − 10𝑛𝑛 Prob.

𝑃𝑃 �𝑍𝑍 ≥ �
√4𝑛𝑛
31 P(Z ≥ −0.898) 0.8159
32 P(Z ≥ −1.77) 0.9616

∴,
n ≥ 32. But 𝑃𝑃(𝑆𝑆32 ≥ 300) = 0.9616 ≫ 0.90.

NOTA BENE.
If the problem was rephrased as:

Determine max. n ∋ 𝑃𝑃(𝑆𝑆𝑛𝑛 ≤ 350) ≥ 0.90 ⇒ n ≤ 33.

Page 32 of 36
CHI-SQUARE DISTRIBUTION. Parameter n. ℵ2𝑛𝑛

Necessary Criteria to be a Chi-Square Distribution.

In Probability Theory and Statistics, the Chi-Square distribution (also
Chi-Squared or ℵ2𝑛𝑛 -distribution) with n degrees of freedom is the
distribution of a sum of the squares of n independent Standard Normal
random variables.

Page 33 of 36
Let X: Chi-Square random variable.

𝑒𝑒−𝑥𝑥/2 𝑥𝑥 (𝑛𝑛 − 2)/2

�2 �
P(X = x) = f(x) = 2
𝑛𝑛 x > 0.
𝛤𝛤� �
2

𝛤𝛤 (𝛽𝛽 ): Gamma Function.

If 𝛽𝛽 ∉ ℕ, 𝛽𝛽 ∈ ℝ ∶
∞
𝛤𝛤 (𝛽𝛽 ) = � 𝑒𝑒 −𝑦𝑦 𝑦𝑦 (𝛽𝛽 − 1) 𝑑𝑑𝑑𝑑.
0

If 𝛽𝛽 ∈ ℕ, 𝛽𝛽 > 0 ∶

𝛤𝛤 (𝛽𝛽 ) = (𝛽𝛽 − 1)! 𝛤𝛤 (1).

Page 34 of 36
𝛤𝛤 (𝛽𝛽 ) = (𝛽𝛽 − 1)! 𝛤𝛤 (1) = 1.

EXAMPLE.
Let 𝑋𝑋𝑖𝑖 ∶ Normally Distributed R.V.’s. 𝐸𝐸[𝑋𝑋𝑖𝑖 ] = 0; 𝑉𝑉𝑉𝑉𝑉𝑉(𝑋𝑋𝑖𝑖 ) = 9.

𝑃𝑃�∑24 2
𝑖𝑖=1 𝑋𝑋𝑖𝑖 > 90�?

Solution.
Standardize.

𝑋𝑋𝑖𝑖 − 𝜇𝜇 𝑋𝑋𝑖𝑖 − 0 𝑋𝑋𝑖𝑖

𝑍𝑍𝑖𝑖 = = = i = 1, 2, 3, …, 24.
𝜎𝜎 3 3

𝑍𝑍𝑖𝑖 are Standard Normal R.V.’s.

Let 𝑍𝑍 = ∑24 2
𝑖𝑖=1 𝑍𝑍𝑖𝑖 .

𝑍𝑍 ~ ℵ2𝑛𝑛 = ℵ224 (Z has a Chi-Square Distribution with 24

degrees-of-freedom).
∴,
𝑃𝑃�∑24 2 24 2
𝑖𝑖=1 𝑋𝑋𝑖𝑖 > 90� = 𝑃𝑃�∑𝑖𝑖=1 9𝑍𝑍𝑖𝑖 > 90�

= 𝑃𝑃�9 ∑24 2
𝑖𝑖=1 𝑍𝑍𝑖𝑖 > 90�

= 𝑃𝑃�∑24 2
𝑖𝑖=1 𝑍𝑍𝑖𝑖 > 10�

= 𝑃𝑃(𝑍𝑍 > 10)

𝑃𝑃�∑24 2 2
𝑖𝑖=1 𝑋𝑋𝑖𝑖 > 90� = 𝑃𝑃 (ℵ24 > 10).

From Chi-Square Probability table:

Page 35 of 36
0.975 < 𝑃𝑃(ℵ224 > 10) < 0.995.

Page 36 of 36