North South University

Mat361
Total Marks-30 (Time – 70 min + 10min)

1) Write down the difference between binomial distribution and Bernoulli

distribution. [2]
2) Write down the difference between geometric and negative binomial
distribution. [2]
3) If you throw a dice and toss a coin how many possible outcomes can you
get? [2]

Answer any four questions from the following

1) A company produce 500 electric bulbs per day. The probability of a defective
electric bulb is 0.05.
i) Calculate the probability of getting 20 defective bulbs. [3]
ii) Calculate the probability of getting not more than 2 defective bulbs. [3]

2) A life insurance company sells on the average 2 life insurance policies per day.
i) Calculate the probability that in a given day the company can not sell any
policies. [3]
ii) Calculate the probability that in a given day the company will sell 2 or more but
less than 4 policies. [3]
3) Suppose we have NSU student’s weight data. Assume this data is normally
distributed with mean 65 kg and variance 3 kg.
i) Calculate the probability that this weight is between 62 kg to 70 kg. [3]
ii) Calculate the probability that this weight is greater than 63 kg. [3]

4) We want to test is human body’s normal temperature really lower than 99

degrees. To test this 61 samples were collected whose ages were from 20 years to
40 years. Sample mean = 98.7 degrees and sample standard deviation = 0.5
degrees. Test this at 1% level of significance. [6]

5) Do people gain or lose weight when they quit smoking? To check this 4
people’s data were collected who are quit smoking for one years. The data are
given below:
Before smoking After smoking
70 kg 67 kg
67 kg 67 kg
76 kg 70 kg
75 kg 73 kg

Test this at 5% level of significance. [6]

 Table of Common Distributions
taken from Statistical Inference by Casella and Berger

Discrete Distrbutions
distribution pmf mean variance mgf/moment

Bernoulli(p) px (1 − p)1−x ; x = 0, 1; p ∈ (0, 1) p p(1 − p) (1 − p) + pet

Γ(α+β) Γ(x+α)Γ(n−x+β) nα nαβ
Beta-binomial(n, α, β) (nx ) Γ(α)Γ(β) Γ(α+β+n) α+β (α+β)2

Notes: If X|P is binomial (n, P ) and P is beta(α, β), then X is beta-binomial(n, α, β).
Binomial(n, p) ( nx )px (1 − p)n−x ; x = 1, . . . , n np np(1 − p) [(1 − p) + pet ]n
1 N +1 (N +1)(N −1) 1
PN it
Discrete Uniform(N ) N; x = 1, . . . , N 2 12 N i=1 e

pet
Geometric(p) p(1 − p)x−1 ; p ∈ (0, 1) 1
p
1−p
p2 1−(1−p)et

Note: Y = X − 1 is negative binomial(1, p). The distribution is memoryless: P (X > s|X > t) = P (X > s − t).
N −M
(M
x )( K−x ) KM KM (N −M )(N −k)
Hypergeometric(N, M, K) N
( K)
; x = 1, . . . , K N N N (N −1) ?

M − (N − K) ≤ x ≤ M ; N, M, K > 0
³ ´r
r(1−p) r(1−p)
Negative Binomial(r, p) (r+x−1
x )pr (1 − p)x ; p ∈ (0, 1) p p2
p
1−(1−p)et

( y−1
r−1 )p (1 − p)
r y−r
; Y =X +r
e−λ λx t
−1)
Poisson(λ) x! ; λ≥0 λ λ eλ(e
Notes: If Y is gamma(α, β), X is Poisson( βx ), and α is an integer, then P (X ≥ α) = P (Y ≤ y).

1
 Continuous Distributions
distribution pdf mean variance mgf/moment
Γ(α+β) α−1 P∞ ³Qk−1 ´
tk
Beta(α, β) Γ(α)Γ(β) x (1 − x)β−1 ; x ∈ (0, 1), α, β > 0 α
α+β
αβ
(α+β)2 (α+β+1) 1 + k=1 r=0
α+r
α+β+r k!
1 1
Cauchy(θ, σ) πσ 1+( x−θ )2 ; σ>0 does not exist does not exist does not exist
σ
X
Notes: Special case of Students’s t with 1 degree of freedom. Also, if X, Y are iid N (0, 1), Y is Cauchy
p
³ ´ p2
x 2 −1 e− 2 ; x > 0, p ∈ N
1 x 1 1
χ2p p p 2p 1−2t , t< 2
Γ( p
2 )2
2

Notes: Gamma( p2 , 2).

1 − |x−µ| eµt
Double Exponential(µ, σ) 2σ e
σ ; σ>0 µ 2σ 2 1−(σt)2
1 −x θ ; x ≥
1 1
Exponential(θ) θe 0, θ > 0 θ θ2 1−θt , t <
1
q θ
Notes: Gamma(1, θ). Memoryless. Y = X γ is Weibull. Y = 2X β is Rayleigh. Y = α − γ log β is Gumbel.
X

ν +ν
Γ( 1 2 2 )
³ ´ ν21 ν1 −2 ν1 +2n ν −2n ³ ´n
ν1 ν2 +ν2 −2 Γ( )Γ( 2 2 )
Fν1 ,ν2 ν1 ν2
Γ( 2 )Γ( 2 ) ν2
x 2
¡ ν1 ¢ ν1 +ν 2
; x>0 ν2 −2 , ν2 > 2 2( ν2ν−2
2
)2 νν11 (ν2 −4)
, ν2 > 4 EX n = 2
ν ν
Γ( 21 )Γ( 22 )
ν2
ν1 , n< ν2
2
2
1+( ν )x
2
χ2ν1 /ν1
Notes: Fν1 ,ν2 = χ2ν2 /ν2 , where the χ s are independent. F1,ν = t2ν .
2

³ ´α
α−1 − β
x
1 1 1
Gamma(α, β) Γ(α)β α x e ; x > 0, α, β > 0 αβ 2αβ 1−βt , t< β
q
Notes: Some special cases are exponential (α = 1) and χ2 (α = p2 , β = 2). If α = 23 , Y = X β is Maxwell. Y =
1
X is inverted gamma.
x−µ
−
π2 β 2
Logistic(µ, β) 1
β
h e β
x−µ
i2 ; β > 0 µ 3 eµt Γ(1 + βt), |t| < 1
β
−
1+e β

1
Notes: The cdf is F (x|µ, β) = −
x−µ .
1+e β

(log x−µ)2 σ2 n2 σ 2
1 − 2 2
Lognormal(µ, σ 2 ) √1
2πσ x
e 2σ 2 ; x > 0, σ > 0 eµ+ 2 e2(µ+σ ) − e2µ+σ EX n = enµ+ 2

(x−µ)2 σ 2 t2
Normal(µ, σ 2 ) √ 1 e− 2σ 2 ; σ>0 µ σ2 eµt+ 2
2πσ
βαβ βα βα2
Pareto(α, β) xβ+1
; x > α, α, β > 0 β−1 , β>1 (β−1)2 (β−2) , β>2 does not exist
Γ( ν+1
2 ) Γ( ν+1
2 )Γ(ν− 2 )
n n
tν √1 1
0, ν > 1 ν
EX n = √
Γ( ν2 ) νπ 2 ν+1 ν−2 , ν >2 πΓ( ν2 )
ν2, n even
(1+ xν ) 2
Notes: t2ν = F1,ν .
(b−a)2 ebt −eat
b−a , a ≤ x ≤ b
1 b+a
Uniform(a, b) 2 12 (b−a)t
Notes: If a = 0, b = 1, this is special case of beta (α = β = 1).
xγ
h i
γ γ−1 − β 1 2 n
Weibull(γ, β) βx e ; x > 0, γ, β > 0 β γ Γ(1 + γ1 ) β γ Γ(1 + γ2 ) − Γ2 (1 + γ1 ) EX n = β γ Γ(1 + nγ )
Notes: The mgf only exists for γ ≥ 1.

2
• Single mean test
Case 1: X has a normal distribution with known population variance
𝑥̅ −µ0
Test statistic is 𝜎 ~ 𝑁(0,1)
√𝑛

Case 2: X has a normal distribution with unknown population variance

𝑥̅ −µ0 ∑(𝑥𝑖 −𝑥̅ )2
Test statistic is ~ 𝑡(𝑛−1) , Where 𝑠 2 =
2 𝑛−1
√𝑠
𝑛

Case 3: X has a general distribution, but we have a large sample size (n≥30).
When population variance is known i.e.
𝑥̅ −µ0
The test statistics is 2
~ 𝑁(0,1)
√𝜎
𝑛

When population variance is unknown,

𝑥̅ −µ0
The test statistic is, 2
~ 𝑁(0,1)
√𝑠
𝑛

̅
𝐷
• Matched Pairs t test: Test statistic = 2
~ 𝑡𝑛−1
√𝑆 𝐷
𝑛

2 ∑10 ̅ 2
𝑖=1(𝐷𝑖−𝐷 )
Where 𝑠 𝐷 =
𝑛−1

• Independent sample t test

𝑋̅−𝑌̅
Test statistic = 1 1
~ 𝑡𝑛1 +𝑛2 −2
√𝑠 2 𝑝 (𝑛 +𝑛 )
1 2

𝑛1 (𝑥 −𝑥̅ )2 +∑ 2 (𝑦 −𝑦 𝑛 2
∑𝑖=1 𝑖 𝑗=1 𝑗 ̅)
2
Where 𝑠 𝑝 =
𝑛1 +𝑛2 −2