Chapter 4: Estimation

• Central limit theorem

• If you sample is from an unknown population the sample size is
30 then the normal distribution can be approximated.
•;
• The mean number of days per year that business travellers are on
road for business is 115 with a standard deviation of 60 days. A
sample of 50 business travellers will be selected.
•
• What is the probability that the sample mean will be more than 115
days per year?
What is the probability that the sample average will be less than 125 days per year?

What is the probability that the sample mean will be within ± 5 days of the population mean?
Sampling From the normal population: - sample size doesn’t matter.

The monthly savings (in rand) of households in a certain suburb have a normal distribution with a mean of 300 and
variance of 196. If a random sample of 20 households is taken,

 What is the probability that the sample average will have monthly savings less than R295?
 Calculate the probability that the mean savings will be more than R304.

Calculate the probability that the average savings will be between R300 and R310.
Confidence Intervals
• Mean:
• Light bulbs supplied by a certain company have an average life of 750
hours with the standard deviation of 30 hours. If a random sample of
46 bulbs is taken, construct a 90% confidence interval for the
population mean of all bulbs.

•
The test marks for 55 students produced a mean of 72,8% and a variance of 275,56%. If the distribution of the marks
may be assumed to be approximately normally distributed. Construct a 95% confidence interval for the mean test
marks.

; ; ;
Mean: ; working with sample variance (s)

The marks of students’ tutorial marks are known to be normally distributed; a sample of 26 tutorial marks was taken
and gave a sample mean of 75% with a standard deviation of 13%. Construct a 95% confidence interval of the true
average marks.

In order to estimate the average time spent on the computer terminals per student at a local university, a sample of
18 students over a one-week period was taken and gave a sample mean of 9 hours with a standard deviation of 1.2
hours. Assume normality, construct a 99% confidance of the true average time spend on the computer terminals .
Mean: ; working with population variance (s)

It is known that the time it takes to drive to work is normally distributed with the population standard deviation of 20
minutes. A random sample of 28 motorists is taken and gave a sample average time of 35.8 minutes. Construct a 95%
confidence interval for the actual average time it takes all motorists in Cape Town to drive to work.
 Population variance

[ ]
2 2
( 𝑛− 1 ) 𝑠 ( 𝑛 −1 ) 𝑠
𝐶𝐼 ( 𝜎 ) =
2
2
; 2
𝜒 𝛼 𝜒 𝛼
𝑛 −1 ; 1 − 𝑛 −1 ;
2 2

A beer company uses dispensing to fill beer cans that provide a maximum variance so that cans are not overfilled
or under filled. A sample of fills for 30 cans yielded a variance of 12ml. If the amounts of beer dispensed are
normally distributed, construct a 99% confidence interval for the variance.
The test marks for 51 students produced a mean of 72,8% and a variance of 275,56%. If the distribution of the
marks may be assumed to be approximately normally distributed. Construct a 95% confidence interval for the
standard deviation of the marks.
Population proportion:

In a random sample of 500 people 160 didn’t have drivers license. Construct a 99% confidence interval for the
proportion of people without drivers’ license.

; ; ; ;
In a random sample of 100 students 35 were in favor of voting for the SRC. Construct 95% confidence interval for
the proportion of students in favor of voting for the SRC.

; ; ; ;
Construct 95% confidence interval for the proportion of students not in favor of voting for the SRC
Confidence Interval for the difference in the two population means.

s12 s22
Mean n 30  
CI 1  2 1   x1  x 2 Z
1
 
n1 n2
2

Example 1: given the following:

Construct a 99% confidence interval for the difference between the two population means

Zero is not included in the interval; we can therefore conclude that there might be difference between the two
means.
 n < 30  
CI 1   2 1   x 1  x 2 t n n  2;1  s p
1

1
s p
n1  1s12  n2  1s22
1 2 2 n1 n2 n1  n2  2

An experiment was conducted by the manufacturers of a well-known brand of cat food to examine the energy content of the
cat food they produce. The researchers monitored two groups of cats over a period of three weeks, 11 cats that were fed a diet
of canned food produced by this manufacturer and 14 cats that were fed a diet of dry cat food produced by this manufacturer.
At the end of the trial, the metabolic energy content was determined for each cat, resulting in the following

Construct a 95% confidence interval for the difference between in the mean metabolic energy content of cats fed canned
food and those fed dry food.

√ √
2 2
( 1
𝑛 −1 ) 1 ( 2
𝑠 + 𝑛 − 1 ) 2
𝑠 (10 ) ( 6.1 )2+ ( 13 ) ( 5.3 )2
= = =5.6617
𝑛1 +𝑛 2 −2 11+14 − 2
Construct a 90% confidence interval for the differences between the means.
 pˆ1 1  pˆ1  pˆ 2 1  pˆ 2  pˆ1 1  pˆ1  pˆ 2 1  pˆ 2 
Two proportions: CI p1  p2 1  pˆ1  pˆ 2 Z    standard error
1
2 n1 n2 n1 n2

A business consultant agency wants to estimate the difference between the percentages of users of two types of
toothpastes who will never switch to another toothpaste. In a sample of 100 users of Toothpaste A, 31 say that they will
never switch to another toothpaste. In another sample of 100 users of Toothpaste B, 24 say that they will never switch
to another toothpaste. Construct a 95% confidence interval for the difference between the two population proportions.
𝑛1= 100 𝑋 1 =31 𝑛2 =100 𝑋 2=24

Zero is included in the interval; we can therefore conclude that there no difference between the percentages of users of two
types of toothpastes