 Remember: important Symbols ‫رموز هامة‬

Sample Statistics Population Parameter

𝐱̅ (sample mean) Mean (or average) µ (population mean)

𝑺𝟐 (sample variance) Variance 𝛔𝟐 (population variance)
S (sample standard deviation) Standard deviation 𝛔 (population standard deviation)
‫) لما يقول‬S ‫𝐗 و‬
̅( ‫ بنستخدم رموز العينة‬:‫خلي بالك‬
selected random sample ‫تم سحب عينة‬

Remember: Summary of the test

One tail left One tail rgiht Two tail
H0: µ ≥ µ0 H0: µ ≤ µ0 H0: µ = µ0
1- Hypotheses
H1: µ < µ0 H1: µ > µ0 H1: µ ≠ µ0
∗( ∗) ̅− 𝛍𝟎
𝐗
2- Test statistic 𝐙 𝐨𝐫 𝐭 = 𝛔(𝐨𝐫 𝐒)
√𝐧

3- critical value (−)𝐙𝟏−𝛂 𝐨𝐫 (−) 𝐭𝐧−𝟏,𝛂 𝐙𝟏−𝛂 𝐨𝐫 𝐭𝐧−𝟏,𝛂 (+)𝐙𝟏−𝛂/2 𝐨𝐫 (+)−𝟏 , 𝛂/2
⚫ Reject 𝐇𝟎 if: | 𝒁∗ | 𝐨𝐫 | 𝐭 ∗ | > |𝐜𝐫𝐢𝐭𝐢𝐜𝐚𝐥 value|
4- Conclusion
⚫ Accept 𝐇𝟎 if: | 𝒁∗ | 𝐨𝐫 | 𝐭 ∗ | < |𝐜𝐫𝐢𝐭𝐢𝐜𝐚l value|
 µ𝟎 = The value to be tested ‫القيمة التي يتم اختبارها‬
Note:
 Equal sign always with null hypothesis (H0)

 Note:Types of the test depending on sign of H1 ,

H1 less than (<) ∵ left tail (tabulated value = - ve)
if : H1 greater than (>) ∵ right tail (tabulated value = + ve)
H1 not equal (≠) ∵ two tail (tabulated value = + )

Summery rules (Teat & Confidence interval)

Population Standard
Estimator Test statistic Confidence interval
parameter error
̅− 𝛍𝟎
𝐗
𝛍 ̅
𝐗 √ 𝝈𝟐
=
𝝈 𝐙 ∗ (𝐨𝐫 𝐭 ∗ ) = 𝛔(𝐨𝐫 𝐒) ̅ ± 𝐙 𝛂 (𝐨𝐫 𝐭) × 𝛔 (𝐨𝐫 𝐬)
𝐗 𝟏− 𝐧 √
𝒏 √𝒏 √𝐧 𝟐

∗ ̂ − 𝐏𝟎
𝐏
̂ 𝑷 (𝟏 − 𝑷) 𝐙 = ̂ + 𝒁 𝜶 × √𝑷̂ (𝟏−𝑷̂)
P 𝐏 √ 𝐏 (𝟏− 𝐏𝟎)
√ 𝟎
𝐏 𝟏− 𝒏
𝒏 𝐧 𝟐

Important Keyword
At least = minimum = no less than Greater than or equal (≥)
At most = maximum = not exceed = no more than less than or equal (≤)
Not different = not changes = not effected Equal (=)
= working up standard
1
 Important notes

1- Determination the sample size (n):

 From the confidence interval for (𝛍)  From the confidence interval for (P)
𝛔
E = 𝐙𝟏− 𝛂 × E = 𝐙𝟏− 𝛂 × √ n
̂
𝑷(1− 𝑷) ̂
𝟐 √𝐧
𝟐
𝐙 𝟐 × 𝛔𝟐 𝐙 𝟐 × 𝐏(𝟏−𝐏)
∵ n = 𝐄𝟐
∵ n = 𝐄𝟐

:‫𝒙( حاجتين‬
̅) ‫خلي بالك لازم نعرف عن‬
1- sampling distribution of sample mean (𝐗
̅)
 Exactly Normal  Approximately Normal
If the population is normal If the population is not normal or unknown
(Whatever the sample size n) but n ≥ 30 (central limit theorem “CLT”)

normal ‫يعني لو قال في التمرين كلمة‬ normal ‫يعني لو مقلش في التمرين كلمة‬
exactly normal ‫يبقي‬ Approximately normal ‫يبقي‬
𝛔𝟐
2- var(𝐗
̅) =
𝐧
Always is smaller than variance of the population (𝛔𝟐 )

⚫ Type one and type two error:

actuality (reality) Ho is false (x)
Ho is true (√)
Decision
Reject Ho Type I error
Accept Ho Type II error

⚫ Type I Error: occurs when we reject HO while HO is true.

Note: 𝛂 = probability of type I error (probability we reject HO while HO is true)

⚫ Type II Error: occurs when we accept HO while HO is false.

Note: 𝛃 = probability of type II error (probability we accept HO while HO is false)

⚫ (1 – 𝛂): probability we accept HO while it is true.

⚫ (1 – 𝛃): probability we reject HO while it is false.

type one ‫ يبقي‬H0 ‫⚫ لو كنت رفضت‬ :‫☺ خلي بالك لو سألك علي نوع الخطأ‬
Type two ‫ يبقي‬H0 ‫⚫ لو كنت قبلت‬

⚫ Relations between 𝛂 and 𝛃 :

 If: 𝛂⇑ , 𝛃⇓ ‫عالقة عكسية )𝛂( تقل )𝛃( تزيد والعكس‬
 If: 𝐧 ⇑, then 𝛂⇓ and 𝛃⇓ ‫ تزيد‬n ‫الأتنين يقلوا مع بعض فقط لما‬
n ≥ 30 n < 30 Distribution Normal and n < 30
σ known σ unknown σ known σ unknown
Use Z Use t

2
Ex(1): In a random sample of size 36 drawn from a normal population 𝐍(𝛍, 𝟏𝟔)
it is found that the sample mean = 24.
1-Construct 95% confidence interval for population mean (𝛍)
2-Test the hypothesis that the population mean Equals 22 at significance level 1%.

Solution

1) given:

Notes: normal population 𝐍 [(𝛍) Mean ‫اول رقم الـ‬, (𝛔𝟐 ) variance ‫] تاني رقم ألـ‬

normal population 𝐍(𝛍, 𝟏𝟔)

 n (sample size) = 36 (n > 30)  𝐱̅ (sample mean) = 24

 𝛔𝟐 = 16 (𝛔 [population standard deviation] = 4)

α = 5% = 0.05,
⚫ 95% C.I. for (𝛍) = 𝐗
̅ ± Z1− 𝛼 × σ
2 √𝐧 𝜶
𝟒 ∵ 1 – = 0.975,
= 𝟐𝟒 ± 𝟏. 𝟗𝟔 × 𝟐
√𝟑𝟔 ∵ 𝐙𝟏− 𝜶 = 𝐙𝟎.𝟗𝟕𝟓 = 1.96
𝟐
= 𝟐𝟒 ± 𝟏. 𝟑𝟏

L.L = 24 – 1.31 = 22.69 U.L = 24 + 1.31 = 25.31

∵ 95% (𝛍) is: (22.69 , 25.31)

2) The claim: population mean Equals 22, (𝛍 = 𝟐𝟐) µ0

3) Steps of the test:

1- null and alternative hypotheses: 𝐇𝟎: 𝛍 = 22 (the claim)

𝐇1: 𝛍 ≠ 22 (two tail)
𝐱̅ − 𝛍𝟎 𝟐𝟒 −𝟐𝟐
2- test statistic: ∗
𝐙 = 𝛔 = 𝟒 = 3  n > 30
use Z
√𝐧 √𝟏𝟔  σ known
3- critical value:
𝜶
Since 𝐇𝟏 is ≠, ∴ the test is two tail (use 𝟐)

∵ ±𝐙𝟏− 𝜶 = ±𝐙𝟎.𝟗𝟗𝟓 = ±2.576

𝜶
α = 1% = 0.01, ∵ 1 – = 0.995,
𝟐 𝟐

4- The conclusion: Since: |𝐙* (=3)| > |𝐙𝟏− 𝜶 (=2.576)|

𝟐

Then we:  reject 𝐇𝟎 &  accept 𝐇1 and

 reject the claim

3
Ex(2): The experts in an agricultural research center claim that they produced
a new brand of wheat seeds is better than the old brand whose mean
productivity is 20. A random sample of 10 similar acres was drawn and planted
with the new seeds. The sample mean and variance are 24 and 40 respectively.
Assume normality,
1- Construct 95% confidence interval for population mean (𝛍)
2- test expert’s claim at significance level 0.01

Solution

1) given:  n = 10, ̅ = 24,

 𝐗 ⚫ 𝑺𝟐 = 40,  S = √𝟒𝟎
𝐒
1- 95% C.I. for (𝛍) = 𝐗
̅ ± 𝐭
(𝐧−𝟏 ,
𝛂
) × ∵ α = 5% = 0.05,
𝟐 √𝐧
= 0.025,
𝜶 𝟎.𝟎𝟓
√𝟒𝟎 ∴ =
= 𝟐𝟒 ± 𝟐. 𝟐𝟔𝟐 × 𝟐 𝟐
√𝟏𝟎 𝐭 (𝐧−𝟏 , 𝛂
) = 𝐭 (𝟗 , 𝟎.𝟎𝟐𝟓) = 2.262
= 𝟑𝟒. 𝟓 ± 𝟒. 𝟓𝟐𝟒 𝟐

L.L = 24 – 4.524 = 19.476 U.L = 24 + 4.524 = 28.524

2- The test:
mean productivity of the old wheat seeds is 20 (𝛍𝐨𝐥𝐝 ) = 20
2) the claim: The new wheat seeds is better than the old (𝛍 > 20) µ0
(20 ‫)بذور القمح الجديدة أفضل من القديمة يعني بتدي انتاجية أكبر من القديمة يعني أكبر من‬

3) Steps of the test:

1- null and alternative hypotheses:

𝐇𝟎: 𝛍 ≤ 20
𝐇1: 𝛍 > 20 (right tail, ∵ use 𝛂) & (the claim)
2- test statistic:
̅ − 𝛍𝟎
𝐗 𝟐𝟒 − 𝟐𝟎  n < 30
𝐭∗ = = = 2 use t
𝐒 √𝟒𝟎  σ unknown
√𝐧 √𝟏𝟎

3- critical value: Since 𝐇𝟏 is greater than, ∴ the test is one tail (use α)

α = 1% = 0.01, 4-∵ 𝐭 (𝐧−𝟏 , 𝛂) = 𝐭 (𝟗 , 𝟎.𝟎𝟏) = 2.821

4- The conclusion: Since: 2 < 2.821

Then we:  accept 𝐇𝟎 & reject 𝐇1  reject the expert’s claim

4
Exam(2021): A machine is adjusted to produce pipes with mean diameter 5 cm and
standard deviation (𝝈) 0.4 cm. A sample of 25 pipes from the production of this machine
was selected at random and examined for quality, the mean diameter was found to be 4.8
cm. assume that the diameters of all units produced by this machine are normallyEC
distributed. The production supervisor claim that this machine is not working properly (not
working up to the standard)

(1) The sampling distribution of the sample average length (𝑿

̅ ) of metal bars produced by
this machine is:
a) Exactly normal with a variance than that of the population
b) Exactly normal with a smaller variance than that of the population
c) approximately normal with a larger variance than that of the population
d) approximately normal with a smaller variance than that of the population

(2) The Maximum error of estimating the population mean produced by this machine using the
sample average diameter at 95% C.L.
a) 1.1568 b) 0.1568 c) 1.1968 d) 1.6

(3) The 95% Confidence interval for the average length of all units produced by this machine is
given as?
a) (3.1568 , 5.9568) b) (4.2568 , 5.9568)
c) (3.5 , 5.568) d) (4.6432 , 4.9568)

(4) The appropriate null hypotheses and the alternative hypotheses to test the production
supervision’s claim
a) 𝐇0: μ = 5 versus 𝐇1: μ ≠ 5 b) 𝐇0: μ = 4.8 versus 𝐇1: μ ≠ 4.8
c) 𝐇0: ̅
X = 5 versus 𝐇1: ̅
X≠5 X = 4.8 versus 𝐇1: ̅
d) 𝐇0: ̅ X ≠ 4.8

(5) The calculated test statistic to test the production supervision’s claim:
a) Z* = -2.5 b) Z* = 2.801 c) t* = -2.801 d) t*=2.801

(6) The critical value (tabulated) for testing the production supervision’s claim at 5%
significance level
a) ±1.645 b) ± 1.96 c) -1.711 d) 1.711

(7) At 5% significance level, do you agree with the production supervision’s claim
a) Yes, the null hypotheses (𝐇0) is rejected and the claim is accepted
b) Yes, both the claim and the null hypotheses (𝐇0) are accepted
c) No, the null hypotheses (𝐇0) is accepted and the claim is rejected
d) No, both the claim and the null hypotheses (𝐇0) are rejected

(8) The Type of error maybe committed when answering question (7)
a) Type one error b) type two error c) α d) β

(9) Assuming the population standard deviation equal 4, how many pipes should production
supervisor use to estimate the population mean within ±2 with 99% confidence
a) less than or equal 27 b) greater than or equal 27
c) less than or equal 54 d) greater than or equal 54

5
 Solution

1) given:  n = 25, ̅ = 4.8,

 𝐗  𝛔 = 0.4
1- Exactly normal (normal ‫ & )عشان قال في التمرين كلمة‬smaller variance (b)
𝛔
2- 95% C.I. for (𝛍) = 𝐗
̅ ± 𝐙 𝜶 ×
𝟏−
𝟐 √𝐧 ∵ α = 5% = 0.05,
𝟎.𝟒 𝜶
= 𝟒. 𝟖 ± 𝟏. 𝟗𝟔 × ∴ 1 – = 0.975,
√𝟐𝟓 𝟐
∴ 𝐙𝟏− 𝜶 = 𝐙𝟎.𝟗𝟕𝟓 = 1.96
= 𝟒. 𝟖 ± 𝟎. 𝟏𝟓𝟔𝟖 𝟐

L.L = 4.8 – 0.1568 = 4.6432 U.L = 4.8 + 0.1568 = 4.9568

A machine produce pipes with mean diameter 5, ∵ (𝛍) = 5

2) the claim: the machine not working up to the standards ‫لا تعمل بصورة جيدة‬
(𝛍 ≠ 5) µ0

3) Steps of the test:

1- null and alternative hypotheses:

𝐇𝟎: 𝛍 = 5
𝛂
𝐇1: 𝛍 ≠ 5 (two tail, ∵ use 𝟐) & (the claim)
2- test statistic:
̅ − 𝛍𝟎
𝐗 𝟒.𝟖 − 𝟓  n < 30
𝐙∗ = = = -2.5 use Z
 σ known
𝛔 𝟎.𝟒
√𝐧 √𝟐𝟓

3- critical value:
𝛂
Since 𝐇𝟏 is not equal, ∴ the test is two tail (use )
𝟐

∵ ±𝐙𝟏− 𝜶 = ±𝐙𝟎.𝟗𝟕𝟓 = ±1.96

𝜶
α = 5% = 0.05, ∵ 1 – = 0.975,
𝟐 𝟐

4- The conclusion: Since: 2.5 > 1.96

Then we:  reject 𝐇𝟎 & accept 𝐇1  accept the claim
(i.e: the machine not working up to the standards)

error ‫لو سأل عن الـ‬

type ‫لو السؤال عن ال‬ probability ‫لو السؤال عن الـ‬
If we reject 𝐇𝟎 Type one α
If we accept 𝐇𝟎 Type two 𝜷

No(8): type one because we rject 𝐇𝟎

No.(9): 𝝈 = 4 & E =2 & 𝐙𝟏− 𝜶 = ±𝐙𝟎.𝟗𝟗𝟓 = 2.576

𝟐
𝐙𝟐 × 𝛔𝟐 2.5762 × 42
n= 𝐄𝟐
=
22
= 26.54 ≅ 27

6
Midterm Exam(2022): it is known that the average normal height of all one-year-old
children is 73.65. The world Health Organization (WHO) claim that a particular disease
has a bad effect on the children’s normal height. To investigate, a random sample of 9
one-year-old, with this disease, was drawn giving a mean height of 70.23 centimeters
with a sample standard deviation of 3.75 centimeters. Assume that the height of all one-
year-old children are normally distributed

(1) The sampling distribution of average height of all one-year-old children (𝑿

̅ ) is :
a) Exactly normal with a larger variance than that of the population
b) Exactly normal with a smaller variance than that of the population
c) approximately normal with a larger variance than that of the population
d) approximately normal with a smaller variance than that of the population

(2) The Maximum error of estimating the average height of all one-year-old children
using the sample average at 98% confidence level:
a) 2.325 b) 3.625 c) 2.8825 d) 4.2

(3) A 98% confidence interval for the average height of all one-year-old children is
given as:
a) (66.625 , 73.875) b) (67.3675 , 73.1325)
c) (67.925 , 72.575) d) (66.05 , 74.45)

(4) The appropriate null hypotheses and the alternative hypotheses to test the WHO’s claim
a) 𝐇0: μ ≥ 73.65 versus 𝐇1: μ < 73.65 b) 𝐇0: ̅
X ≥ 70.25 versus 𝐇1: ̅X < 70.25
c) 𝐇0: X
̅ ≤ 70.25 versus 𝐇1: X
̅ > 70.25 d) 𝐇0: μ ≥ 73.65 versus 𝐇1: μ < 73.65
(5) The calculated test statistic to test the WHO’s claim :
a) t* = -2.72 b) t* = 2.72 c) Z* = 2.72 d) Z*= -2.72

(6) The critical value (tabulated) to test the WHO’s claim at 5% significant level :
a) -2.306 b) -1.86 c) 2.306 d) 1.86

(7) At 5% significance level, do you agree with the WHO’s claim?

a) Yes, the null hypotheses (𝐇0) is rejected and the claim is accepted
b) Yes, both the claim and the null hypotheses (𝐇0) are accepted
c) No, the null hypotheses (𝐇0) is accepted and the claim is rejected
d) No, both the claim and the null hypotheses (𝐇0) are rejected

(8) The Type of error maybe committed when answering question (7)
a) Type one error b) type two error c) α d) β
(9) The probability of committing type one error equal?
a) 0.005 b) 0.01 c) 0.025 d) 0.05

7
 Solution

1) given:  n = 9, ̅ = 70.25,
 𝐗  S = 3.75
(1) Exactly normal (normal ‫ & )عشان قال في التمرين كلمة‬smaller variance (b)
𝐒 𝟑.𝟕𝟓
(2) Maximum error (E) = 𝐭 (𝐧−𝟏 , 𝛂
) × = 2.9 × = 3.625 (b)
𝟐 √𝐧 √𝟗

𝐒
2- 98% C.I. for (𝛍) = 𝐗
̅ ± 𝐭
(𝐧−𝟏 ,
𝛂
) ×
𝟐 √𝐧
̅ ± E
=𝐗 ∵ α = 2% = 0.02,
= 0.01,
𝜶 𝟎.𝟎𝟐
∴ =
𝟐 𝟐
= 𝟕𝟎. 𝟐𝟓 ± 𝟑. 𝟔𝟐𝟓
𝐭 (𝐧−𝟏 , 𝛂
) = 𝐭 (𝟖 , 𝟎.𝟎𝟏) = 2.9
𝟐
(66.625 , 73.875) (a)

normal height of all one-year-old children is 73.65 , ∵ (𝛍) = 73.65

2) the claim: a particular disease has a bad effect on the children’s normal height
(𝛍 < 73.65) µ0

3) Steps of the test:

1- null and alternative hypotheses: (no. 4 = a)

𝐇𝟎: 𝛍 ≥ 73.65
𝐇1: 𝛍 < 73.65 (one tail, ∵ use 𝛂) & (the claim)
2- test statistic:
̅ − 𝛍𝟎
𝐗 𝟕𝟎.𝟐𝟓 − 𝟕𝟑.𝟔𝟓  n < 30
∗
𝐭 = = = -2.72 (no. 5 = a) use t
𝛔 𝟑.𝟑𝟕𝟓  σ unknown
√𝐧 √𝟗

3- critical value: Since 𝐇𝟏 is less than, ∴ the test is one tail (use 𝛂)

α = 5% = 0.05, ∵ −𝐭(𝐧−𝟏 , 𝛂) = -𝐭 (𝟖 , 𝟎.𝟎𝟓) = -1.86

4- The conclusion: Since: 2.72 > 1.86

Then we:  reject 𝐇𝟎 & accept 𝐇1  accept the claim
(i.e: a particular disease has a bad effect on the children’s normal height)

No(8): type one error, because we rject 𝐇𝟎

No(9): Probabiliy of type one error = α = 5% = 0.05 (d)

8
Q(2): A major car manufacturer wants to determine whether a new engine meets
new air pollution standards ‫ مصنع للسيارات عاوز يعرف المحرك الجديد بيحقق معايير التلوث ام لا‬.
The mean emission of all engines of this type (μ) must be less tha 20 parts per
million of carbon
20 ‫) اللي بتتطلع من المحرك اقل من‬μ( ‫معايير التلوث بتقول لازم يكون متوسط العودام‬
A random sample of 49 engines was drawn for testing purpose, and the emission level
of each determined giving a sample mean of 18 and a sample standard deviation is 5

(1) The sampling distribution of the sample mean (𝑿

̅ ) is :
a) Exactly normal with a larger variance than that of the population
b) Exactly normal with a smaller variance than that of the population
c) approximately normal with a larger variance than that of the population
d) approximately normal with a smaller variance than that of the population

(2) The Maximum error (Margin of error) of estimating the mean emission of all new
engines (𝛍) at 90% C.L
a) 1.4 b) 1.96 c) 1.175 d) 5

(3) The 95% Confidence interval for the mean emission level of all engines at that type?
a) (11.4 , 36.6) b) (12.6 , 19.4) c)(22 , 26) d) (16.6 , 19.4)

(4) The appropriate null hypotheses and the alternative hypotheses to test the new
type engines meets the pollution standard:
a) 𝐇0: μ ≥ 18 versus 𝐇1: μ < 18 b) 𝐇0: μ ≥ 20 versus 𝐇1: μ < 20
c) 𝐇0: ̅
X ≥ 20 versus 𝐇1: ̅
X < 20 d) 𝐇0: ̅
X ≥ 18 versus 𝐇1: ̅
X < 18

(5) The calculated test statistic to test the claim is:

a) Z* = -2.801 b) Z* = 2.801 c) t* = -2.801 d) t* = 2.801

(6) The critical value (tabulated) for testing the claim at 1% significance level:
a) Z = -2.326 b) Z = 2.326 c) t = -2.821 d) t =2.821

(7) At 1% significance level, do you agree with the claim that the new engine meets
the pollution standard?
a) Yes, the null hypotheses (𝐇0) is rejected and the claim is accepted
b) Yes, both the claim and the null hypotheses (𝐇0) are accepted
c) No, the null hypotheses (𝐇0) is accepted and the claim is rejected
d) No, both the claim and the null hypotheses (𝐇0) are rejected

(8) The Type of error maybe committed when answering question (7)
a) Type one error b) type two error c) α d) β
(9) What is the probability of rejecting the null hypotheses (H0) while it is true?
a) 0.01 b) 0.05 c) 0.99 d) 0.95

9
 Solution

1) given:  n = 49, ̅ = 18,

 𝐗  𝛔= 5

(1) approximately normal (normal ‫ & )عشان لم يقل في التمرين كلمة‬smaller variance (d)

∵ α = 10% = 0.10,
𝛔 𝟓
(2) Maximum error (E) = 𝐙𝟏− 𝜶 × = 1.645 × = 1.175 (b) 𝟎.𝟏𝟎
𝟐 √𝐧 √𝟒𝟗 ∴ 1 – = 0.95,
𝟐
𝛔 ∴ 𝐙𝟏− 𝜶 = 𝐙𝟎.𝟗𝟓 = 1.645
2- 95% C.I. for (𝛍) = 𝐗
̅ ± 𝐙 𝜶 ×
𝟏−
𝟐

𝟐 √𝐧
∵ α = 5% = 0.05, = 𝟏𝟖 ± 𝟏. 𝟗𝟔 × 𝟓

∴ 1 –
𝟎.𝟎𝟓
= 0.975,
√𝟒𝟗
 n > 30
= 𝟏𝟖 ± 𝟏. 𝟒 use Z
 σ known
𝟐
∴ 𝐙𝟏− 𝜶 = 𝐙𝟎.𝟗𝟕𝟓 = 1.96
𝟐
(16.6 , 19.4) (d)

2) the claim: a new engine meets new air pollution standards (𝛍 < 20)
)20 ‫المحركات الجديدة بتحقق معايير التلوث (يعني متوسط العوادم اللي بتتطلع من المحرك اقل من‬

3) Steps of the test:

1- null and alternative hypotheses: (b)

𝐇𝟎: 𝛍 ≥ 20
‫مصنع للسيارات عمل محرك جديد‬
𝐇1: 𝛍 < 20 (left tail, ∵ use 𝛂) & (the claim)
‫وعاوز يعرف المحرك الجديد بيحقق‬
2- test statistic: .‫معايير التلوث ام ال‬
̅−𝛍
𝐗 𝟏𝟖 − 𝟐𝟎
∗
𝐙 = 𝟎
𝐒 = 𝟓 = -2.801 (a) ‫معايير التلوث بتقول الزم يكون متوسط‬
√𝐧 √𝟒𝟗 ‫) اللي بتتطلع من المحرك‬μ( ‫العودام‬
.20 ‫اقل من‬
3- critical value: Since 𝐇𝟏 is less than, (a)
49 ‫علشان كده المصنع سحب عينة من‬
∴ the test is one tail (use 𝛂) ‫محرك جديد وحسب متوسط العادم علشان‬
.‫يتأكد ان المصنع بيحقق معايير التلوث‬
α = 1% = 0.01, ∵ 1 – 𝜶 = 0.99,
‫فلو األختبار أكد أن متوسط العادم اللي‬
∵ 𝐙𝟏−𝜶 = - = −2.236 20

4- The conclusion: Since: 2.801 > 2.236

Then we:  reject 𝐇𝟎 & accept 𝐇1  accept the claim
(i.e. the new engine meets new air pollution standards)

No(8): type one error, because we rject

000

No(9): Probabiliy of of rejecting 𝐇𝟎 while 𝐇𝟎 is true

= probability of type one = α = 0.01

10
Q(2): The present time pilot takes to complete a certain corrective action is
6.9 second on average. The aviation administration is considering a new training
program for improving the performance of pilot. A random sample of 10 pilots
was tested in a flight simulator after two months of training and the time for
each to complete the target corrective action was measured in seconds to give
the follows: the sample mean = 6 and the sample standard deviation = 0.91287.
Assuming the normality:

(1) The sampling distribution of the sample mean (𝑿

̅ ) is :
e) Exactly normal with a larger variance than that of the population
f) Exactly normal with a smaller variance than that of the population
g) approximately normal with a larger variance than that of the population
h) approximately normal with a smaller variance than that of the population

(2) The appropriate null hypotheses and the alternative hypotheses to evaluate the
effectiveness of the training program
a) 𝐇0: μ ≤ 6.9 versus 𝐇1: μ > 6.9 b) 𝐇0: μ ≥ 6.9 versus 𝐇1: μ < 6.9
̅ ≥ 6.9 versus 𝐇1: X
c) 𝐇0: X ̅ < 6.9 d) 𝐇0: μ ≥ 6 versus 𝐇1: μ < 6

(3) The calculated test statistic to test the claim is:

a) Z* = -2.801 b) Z* = -3.117 c) t* = -2.801 d) t*= -3.117

(4) The critical value (tabulated) for testing the claim is:
a) Z = -2.326 b) Z = 2.326 c) t = -1.833 d) t = 2.821

(5) At 5% significance level, do you agree with the claim that the new training
program is improving the performance of pilot
e) Yes, the null hypotheses (𝐇0) is rejected and the claim is accepted
f) Yes, both the claim and the null hypotheses (𝐇0) are accepted
g) No, the null hypotheses (𝐇0) is accepted and the claim is rejected
h) No, both the claim and the null hypotheses (𝐇0) are rejected

(6) The Type of error maybe committed when answering question (8)
a) Type one error b) type two error c) α d) β

(7) The Maximum error of estimating (Margin of error) the population mean at 90% C.L.
a) 0.653 b) 1.175 c) 0.714 d) 3.6

(8) The 95% Confidence interval for the mean emission level of all engines at that type?
a) (3.4 , 8) b) (5.347 , 6.653) c) (6 , 7.714) d) (6.04 , 9.96)

11
 Solution

1) given:  n = 10, ̅ = 6,
 𝐗  S = 0.91287

(1) Exactly normal (normal ‫ & )عشان قال في التمرين كلمة‬smaller variance (b)

average time pilot takes to complete a certain corrective action (𝛍) = 6.9
2) the claim: The new training program has improved the pilots' performance,
(𝛍 < 6.9) µ0 (‫)تحسن مستوي الطيار يعني يعمل رد الفعل في زمن أقل‬

3) Steps of the test:

1- null and alternative hypotheses:

𝐇𝟎: 𝛍 ≥ 6.9
𝐇1: 𝛍 < 6.9 (left tail, ∵ use 𝛂) & (the claim)
2- test statistic:
̅ − 𝛍𝟎
𝐗 𝟔 − 𝟔.𝟗  n < 30
∗
𝐭 = = = -3.117 use t
𝐒 𝟎.𝟗𝟏𝟐𝟖𝟕
 σ unknown
√𝐧 √𝟏𝟎

3- critical value: Since 𝐇𝟏 is smaller than, ∴ the test is one tail (use α)

α = 5% = 0.05, ∵4--𝐭 (𝐧−𝟏 , 𝛂) = −𝐭 (𝟗 , 𝟎.𝟎𝟓) = - 1.833

5- The conclusion: Since: |t* (= 3.117)| > |𝐭 (𝐧−𝟏 , 𝛂) (= 1.833)|

Then we:  reject 𝐇𝟎 & accept 𝐇1
 accept the claim (i.e: the program is effective)

𝐒
2- 95% C.I. for (𝛍) = 𝐗
̅ ± 𝐭
(𝐧−𝟏 ,
𝛂
) × ∵ α = 5% = 0.05,
𝟐 √𝐧
= 0.025,
𝜶 𝟎.𝟎𝟓
𝟎.𝟗𝟏𝟐𝟖𝟕 ∴ =
= 𝟔 ± 𝟐. 𝟐𝟔𝟐 × 𝟐 𝟐
√𝟏𝟎
𝐭 (𝐧−𝟏 , 𝛂
) = 𝐭 (𝟗 , 𝟎.𝟎𝟐𝟓) = 2.262
= 𝟔 ± 𝟎. 𝟔𝟓𝟑 𝟐

L.L = 6 – 0.653 = 5.347 U.L = 6 + 0.653 = 6.653

12
midterm(2022): The management of some industry is considering a new method
of assembling its products ‫طريقة جديدة لتجميع المنتج‬. Using the present method,
technicians assemble 40 units a day on average with the standard deviation (𝝈)
of 5 units a day. the average number of daily assembled units for a random
sample of 64 units using new method is 42 unit. Ali claim that the new method
is better than the present one

(1) The sampling distribution of the sample average number of daily assembled units (𝑿
̅ ):
a) approximately normal with a variance of 0.625
b) Exactly normal with a variance of 0.390625
c) approximately normal with a variance of 0.390625
d) Exactly normal with a variance of 0.625

(2) What are appropriate null and alternative hypotheses to test the claim
a) 𝐇0: X
̅ ≥ 42 versus 𝐇1: X
̅ < 42 b) 𝐇0: X
̅ ≤ 42 versus 𝐇1: X ̅ > 42
c) 𝐇0: μ ≥ 40 versus 𝐇1: μ < 40 d) 𝐇0: μ ≤ 40 versus 𝐇1: μ > 40

(3) The calculated test statistic to test the claim is:

a) 1.3 b) 3.2 c) 1.69 d) 1.25

(4) The critical value (tabulated) for testing the claim at 0.5% significant level:
a) 1.645 b) 1.96 c) 2.326 d) 2.576

(5) At 0.5% significance level, do you agree with the claim

a) Yes, the null hypotheses (𝐇0) is rejected and the claim is accepted
b) Yes, both the claim and the null hypotheses (𝐇0) are accepted
c) No, the null hypotheses (𝐇0) is accepted and the claim is rejected
d) No, both the claim and the null hypotheses (𝐇0) are rejected

(6) The Type of error maybe committed when answering question (8)
a) Type one error b) type two error c) α d) β

(7) At which of the following significant levels, there will be no effect on your conclusion in (5)
a) 0.1% b) 0.6% c) 0.2% d) 0.4%

(8) What is the standard error of the sample average number of daily assembled units (𝑿
̅ ):
a) 1.3 b) 1.4 c) 0.625 d) 1.25

13
 Solution

1) given:  n = 64, ̅ = 42,

 𝐗  𝛔= 5

(1) approximately normal (normal ‫ & )عشان لم يقل في التمرين كلمة‬smaller variance

̅) 𝛔𝟐 𝟓𝟐
Var(𝐗 = = = 0.390625 (c)  n > 30
𝐧 𝟔𝟒 use Z
 σ known

assemble 40 units a day on average (𝛍) = 40

2) the claim: new method is better (𝛍 > 40) ‫بتجمع عدد وحدات أكتر‬

3) Steps of the test:

1- null and alternative hypotheses:

𝐇𝟎: 𝛍 ≤ 40
𝐇1: 𝛍 > 40 (right tail, ∵ use 𝛂) & (the claim)
2- test statistic:
̅−𝛍
𝐗 𝟒𝟐 − 𝟒𝟎
𝐙 = ∗ 𝟎
𝐒 = 𝟓 = 3.2
√𝐧 √𝟔𝟒

3- critical value: Since 𝐇𝟏 is greater than,

∴ the test is one tail (use 𝛂)
α = 0.5% = 0.005, ∵ 1 – 𝜶 = 0.995,
∵ 𝐙𝟏−𝜶 = ∵ 𝐙𝟗𝟗𝟓 = 𝟐. 𝟓𝟕𝟔

4- The conclusion: Since: 3.2 > 2.576

Then we:  reject 𝐇𝟎 & accept 𝐇1  accept the claim
(i.e. new method is better)

No(6): type one error, because we rject

000

No(7):
(0.5%) ‫( عند أي )𝛂( أكبر من‬not effect) ‫القرار مش حيتغير‬
0.6% = (𝛂) ‫اذا الأجابة عند‬

14
 Inference about population proportion

Q(1): A company makes auto batteries. The company claims that 80% of its batteries are
good. A consumer agency wanted to check if this claim is true. The agency took a random
sample of 40 such batteries and found that 75% of them were good.
a) Using the 1% significance level, can you conclude that the company's claim is false?
b) Construct a 95% C.I for the proportion of good batteries

Solution

1) given:  n = 40, ̂ = 75% = 0.75

𝐏
2) claim: 80% of its batteries are good (P = 0.80)

Steps of the test

1- null and alternative hypotheses:
𝐇𝟎: P = 0.80 (the claim good batteries)
𝜶
𝐇1: P ≠ 0.80 (two tail, ∵ use )
𝟐

̂ −𝟎.𝟖𝟎 𝟎.𝟕𝟓 −𝟎.𝟖𝟎

= - 0.79
𝐩
2- test statistic: 𝐙 ∗ = =
𝐩 (𝟏− 𝐩𝟎 ) 𝟎.𝟖𝟎(𝟏− 𝟎.𝟖𝟎)
√ 𝟎 √
𝐧 𝟒𝟎

3- critical value: 𝐙𝟏− 𝜶 = 𝐙𝟎.𝟗𝟗𝟓 = ±2.576

𝟐

4- Conclusion: Since: |0.79|<|2.576|, we Accept 𝐇𝟎 and reject 𝐇1 & accept the claim

̂ ̂
a) 95% C.I for Population proportion (P): ̂ ± Z 𝛼 × √𝑷(1− 𝑷)
𝐏 1− n
2

𝟎.𝟕𝟓(𝟏− 𝟎.𝟕𝟓)
= 𝟎. 𝟕𝟓 ± 𝟏. 𝟗𝟔 × √ 𝟒𝟎
= 0.75 ± 0.134
L.L = 0.75 – 0.134 = 0.616 U.L = 0.754 + 0.134 = 0.884

Q(1): two candidates (A & B) compete for a single seat in local elections. A random
sample of 225 voters was selected and the number of persons those favoring candidate
(A) is found to be 126. An observer claims that candidate (A) is more likely to win than
candidate (B). test the claim at significance level 1%

Solution
‫ من األصوات‬%50 ‫الحظ المرشحان بيتعادلو لو كل واحد اخد‬
)the claim ‫ من األصوات‬%50 ‫ يكسب يبقي الزم ياخد اكثر من‬A ‫ (يعني عشان‬%50 ‫المرشح الفائز اللي حياخد أكثر من‬
𝟏𝟐𝟔
 given:  n = 225, X (A ‫ = )عدد اصوات‬126 𝑷
̂ = = 0.56
𝟐𝟐𝟓

𝐇𝟎: P ≤ 0.5
𝐇1: P > 0.5 (the claim A ‫)فوز المرشح‬
̂ − 𝟎.𝟎𝟒
𝐩 𝟎.𝟓𝟔 −𝟎.𝟓
test statistic: 𝐙∗ = = = 1.8
𝐩 (𝟏− 𝐩𝟎 ) 𝟎.𝟓(𝟏− 𝟎.𝟓)
√ 𝟎 √
𝐧 𝟐𝟐𝟓

critical value: 𝐙𝟏− 𝜶 =𝐙𝟏− 𝟎.𝟎𝟏 𝐙𝟎.𝟗𝟗 = 𝟐. 𝟐𝟑𝟔

Conclusion: Since: 1.8 < 2.236, we Accept 𝐇𝟎 and reject 𝐇1 & reject the claim

15
Q(1): When working properly, a machine that is used to make chips for calculators does
not produce more than 4% defective chips. To check if the machine is working properly,
the quality control department at the company took a sample of size 200 chips recently
produced by this machine. The sample contained 14 defective chips
a) At 5% significance level, does this information provide enough evidence that the
machine needs adjustment?
b) Construct a 95% confidence interval for the proportion of defective units

Solution

1) given:  n = 200,
𝟏𝟒
defectiv chips(X) = 14 𝑷
̂ = = 0.07
𝟐𝟎𝟎

2) claim: machine needs adjustment (P > 0.04)

Note: When working properly (no needs to adjustment): p ≤ 0.04

Steps of the test

1- null and alternative hypotheses:
𝐇𝟎: P ≤ 0.04
𝐇1: P > 0.04 (needs to adjustment)
̂ − 𝟎.𝟎𝟒 𝟎.𝟎𝟕 −𝟎.𝟎𝟒
= 2.165
𝐩
2- test statistic: 𝐙 ∗ = =
𝐩 (𝟏− 𝐩𝟎 ) 𝟎.𝟎𝟒(𝟏− 𝟎.𝟎𝟒)
√ 𝟎 √
𝐧 𝟐𝟎𝟎

3- critical value: 𝐙𝟏− 𝜶 =𝐙𝟏− 𝟎.𝟎𝟓 𝐙𝟎.𝟗𝟓 = 𝟏. 𝟔𝟒𝟓

4- Conclusion: Since: |2.165| > |1.645|,

then we reject 𝐇𝟎 and accept 𝐇1
∵ the machine needs adjustment

̂ ̂
2) 95% C.I for Population proportion (P): ̂ ± Z 𝛼 × √𝑷(1− 𝑷)
𝐏 1− n 2

𝟎.𝟎𝟕(𝟏− 𝟎.𝟎𝟕)
= 𝟎. 𝟎𝟕 ± 𝟏. 𝟗𝟔 × √
𝟐𝟎𝟎

= 0.07 ± 0.035

L.L = 0.07 – 0.035 = 0.035 U.L = 0.07 + 0.035 = 0.105

̂
𝑷(1− 𝑷) ̂
Note: the maximum of error (E) = Z1− 𝛼 ×√
2 n

𝟎.𝟎𝟕(𝟏− 𝟎.𝟎𝟕)
= 1 . 𝟗𝟔 × √ = 0.035
𝟐𝟎𝟎

16
  Determination the sample size (n)

 From the confidence interval for (𝛍)  From the confidence interval for (P)
𝛔
E = 𝐙𝟏− 𝛂 × 𝑷(1− 𝑷) ̂ ̂
𝟐 √𝐧 E = 𝐙𝟏− 𝛂 × √
𝟐 n
𝐙 𝟐 × 𝛔𝟐 𝐙 𝟐 × 𝐏(𝟏−𝐏)
∵ n = ∵ n =
𝐄𝟐 𝐄𝟐

Ex(1): what is the smallest sample size required to be selected from a normal
distribution N(µ ,16) in order to get the error of estimating by sample mean (𝑿
̅ ) at
most one unit with probability 95%?

Solution

⚫ Normal distribution N(µ ,16) then: 𝛔𝟐 = 16 ,

⚫ Error of estimating at most one unit then: ∵ Margin of error E = 1
⚫ 95% C.I. ∴ α = 0.05: 𝐙𝟏− 𝜶 = 𝐙𝟎.𝟗𝟕𝟓 = 1.96
𝟐

𝐙 𝟐 × 𝛔𝟐 𝟏.𝟗𝟔𝟐 × 𝟏𝟔
∴ sample size required: n = = = 61.47 ≈ 62 units ∵ n ≥ 62
𝐄𝟐 𝟏𝟐

Ex(2): what is the sample size needed to estimate the population mean within ±5
with 95% confidence if we know that the population standard deviation = 50?

Solution

⚫ Within ±5: E(margin of error) = 5 (E = ± ‫)خلي بالك اللي بعد‬

⚫ population standard deviation (𝛔 ): 𝛔 = 50
⚫ 95% C.I. ∴ α = 0.05: 𝐙𝟎.𝟗𝟕𝟓 = 1.96
𝐙 𝟐 × 𝛔𝟐 𝟏.𝟗𝟔𝟐 × 𝟓𝟎𝟐
∴ sample size required: n = = = 384.2 ≈ 385 units ∵ n ≥ 385
𝐄𝟐 𝟓𝟐

Q(5): How large a sample would be necessary to estimate the true proportion defective in
a large population within ±3%, with 90% confidence? (Assume a pilot sample yields p=
0.12).

Solution

⚫ Margin of error E = 3% = 0.03 P = 0.12

⚫ 90% C.I. ∴ α = 0.05: 𝐙𝟎.𝟗𝟓 = 1.645

𝐙 𝟐 ×𝐏 ×(𝟏−𝐏) 𝟏.𝟔𝟒𝟓 𝟐 ×𝟎.𝟏𝟐 × (𝟏− 𝟎.𝟏𝟐)
∴ sample size required: n = = = 317.5 ≅ 318
𝐄𝟐 𝟎.𝟎𝟑𝟐
∵ n ≥ 318
0.5 = ‫) غير معطي بالتمرين نفترضها من عندنا‬P( ‫ لو‬:‫هام جدا‬

17
 Imp. Notes

For any confidence interval: ̅ ± E

𝐗

L.L = 𝐗
̅ - E & U.L= 𝐗
̅ + E
L.L < µ < U.L

 We can find: ̅ = L.L + E

𝐗 OR ̅ = U.L − E
𝐗

 With of the confidence interval = U.L. – L.L.

𝐔.𝐋. − 𝐋.𝐋
 We can find:  Margin of error E = 𝟐
𝐔.𝐋. + 𝐋.𝐋
 Point estimator (𝐗
̅) =
𝟐

 For any confidence interval:

If n (or σ [or 𝛔𝟐 ] )

Standard Margin of width precision Confidence

error (SE) error (E) (1 – 𝛼 ) 𝛼

Imp Note:

GENRAL PROBLEM

(1) the width of the confidence interval equal

a) upper limit of the confidence  Lower limit of the confidence
b) upper limit of the confidence - Lower limit of the confidence
c) Maximum error of estimate  2 d) All the above

(2) Increasing the sample size leads to:

a) Increasing the width if the confidence interval
b) b) Increasing the error of estimation
c) Increasing the standard error Decreasing the margin of error (E)

(3) Increasing the sample size leads to:

a) Increasing the width if the confidence interval
b) Increasing the error of estimation
c) Decreasing the precision of estimation d) None of the above

18
(4) The width of a confidence interval can be decreased by:
a) Decreasing confidence level or increasing the sample size
b) Increasing confidence level or increasing the sample size
c) Increasing confidence level or decreasing the sample size
d) Decreasing confidence level or decreasing the sample size

(5) If (𝐗
̅) the sample mean, then the variance of sampling distribution of the
sample means (𝐗
̅) equal:
a) Variance of the population  sample size
b) Variance of the population
c) Standard deviation of the population  sample size
d) The mean  sample size

(6) The lower limit of the confidence interval of the population mean (μ) equal
a) The sample mean (𝐗
̅) + margin of error (E)
b) The sample mean (𝐗
̅) - margin of error (E)
c) The sample mean (𝐗
̅) + margin of error (E)
d) The sample mean (𝐗
̅) + variance

(7) The probability (𝜷) is defined as

a) Probability of committing type one error
b) Probability of committing type two error
c) Probability of accepting null hypotheses while it is true
d) Probability of rejecting null hypotheses while it is false

(8) The probability (1 - 𝜷) is defined as

a) Probability of committing type one error
b) Probability of committing type two error
c) Probability of accepting null hypotheses while it is true
d) Probability of rejecting null hypotheses while it is false

(9) The probability (1 - 𝛂) is defined as

a) Probability of committing type one error
b) Probability of committing type two error
c) Probability of accepting null hypotheses while it is true
d) Probability of rejecting null hypotheses while it is false

(10) Increasing the sample size leads to:

a) Increasing the width if the confidence interval
b) Increasing the error of estimation
c) Decreasing the precision of estimation
d) Decreasing error of type (I) and type (II) in testing statistical hypotheses

19
Q(1): two candidates (A & B) compete for a single seat in local elections. A random
sample of 225 voters was selected and a 100(1 – 𝜶)% confidence interval for the
population proportion of those favoring candidate (A) is found to be (0.536 , 0.664)
1- Find point estimate for proportion of those favoring candidate (A)
2- Find maximum error of estimating the proportion

Solution

𝐮𝐩𝐩𝐞𝐫 𝐥𝐢𝐦𝐢𝐭+𝐥𝐨𝐰𝐞 𝐥𝐢𝐦𝐢𝐭 𝟎.𝟔𝟔𝟒 + 𝟎.𝟓𝟑𝟔

1- point estimate for proportion (𝐏
̂) = = = 0.6
𝟐 𝟐
𝐮𝐩𝐩𝐞𝐫 𝐥𝐢𝐦𝐢𝐭− 𝐥𝐨𝐰𝐞𝐫 𝐥𝐢𝐦𝐢𝐭 𝟎.𝟔𝟔𝟒 − 𝟎.𝟓𝟑𝟔
2- maximum error = = = 0.064
𝟐 𝟐

Q(1): if a 90% confidence interval for population mean is (11 , 15),

1- find the sample mean (or point estimate for population mean)
3- find the maximum error of estimate the population mean

Solution
𝐮𝐩𝐩𝐞𝐫 𝐥𝐢𝐦𝐢𝐭+𝐥𝐨𝐰𝐞 𝐥𝐢𝐦𝐢𝐭 𝟏𝟓+𝟏𝟑
1- sample mean (or point estimate for population mean) = = = 13
𝟐 𝟐
𝐮𝐩𝐩𝐞𝐫 𝐥𝐢𝐦𝐢𝐭− 𝐥𝐨𝐰𝐞𝐫 𝐥𝐢𝐦𝐢𝐭 𝟏𝟓 − 𝟏𝟑
2- maximum error of estimate = = = 1
𝟐 𝟐

Q(3): each of the following is a confidence interval for the population mean based on the
same random sample: (12 , k) and (5 , 25) at two difference confidence level. The upper
limit of the first confidence interval (K) is:

Solution

⚫ same random sample = same (𝑿

̅)
𝐮𝐩𝐩𝐞𝐫 𝐥𝐢𝐦𝐢𝐭+𝐥𝐨𝐰𝐞 𝐥𝐢𝐦𝐢𝐭 𝟓+𝟐𝟓
 From the second confidence interval: (𝑿
̅) = = = 15
𝟐 𝟐

 From the first confidence interval: E(Maximum error) = 𝑿

̅ – lower limit
= 15 – 12 = 3
∵ the upper limit of the first confidence interval = 𝑿
̅ + E = 15 + 3 = 18

Q(Very imp.): given the null and alternative hypotheses “𝐇𝟎 the caused person is innocent
and 𝐇1 the accused person is guilty”, in which of the following cases a type II is being
committed”
a) an innocent person is set free (‫)قرار صحيح‬ b) an innocent person goes to jell
c) a guilty person is set free d) a guilty person goes to jell (‫)قرار صحيح‬

Q(Very imp.): given the null and alternative hypotheses “𝐇𝟎 the caused person is innocent and 𝐇1
the accused person is guilty”, in which of the following cases a type I is being committed”
a) an innocent person is set free (‫)قرار صحيح‬ b) an innocent person goes to jell
c) a guilty person is set free d) a guilty person goes to jell (‫)قرار صحيح‬

20