STATISTICAL INFERENCE

Sample information is used to shed some light on the population characteristics i.e. we infer
population properties based on findings on the sample. The process of using sample information
to make inferences about the population is called statistical inference.

Sample Statistics Population Parameters

Statistical inference: Population mean, µ

Sample mean,

*Estimation theory Population standard

Sample standard
deviation, s deviation, σ
**Hypothesis testing
Sample
proportion, p Population proportion,
π

Statistical inference falls into two main areas i.e. statistical estimation and hypothesis testing.
*Statistical Estimation: The characteristics of the sample (sample statistic) are used to estimate
or approximate some unknown population characteristics (parameters).
**Hypothesis testing: The population characteristics are known or assumed. The sample
characteristics are used to verify or ascertain this assumed or known population characteristic.

THEORY OF ESTIMATION
The assignment of values to a population parameter based on a sample is called estimation. The
values assigned to a population parameter based on the value of a sample statistic is called an
estimate of the population parameter. The sample statistic used to estimate a population
parameter is called an estimator.
Estimation can be undertaken in two forms namely:
 Point estimation
 Interval estimation
Point estimation: A single value is calculated from sample data to estimate the unknown
population parameter. The value of a sample statistic that is used to estimate a population
parameter is called a point estimate and the procedure is called point estimation. Thus:
a sample mean ( x ) can be used as a point estimate of the population mean 
a sample standard deviation(s) - a point estimate of the population standard deviation σ
a sample proportion (p) - a point estimate of the population proportion π
A point estimate is only as good as the representativeness of its sample. If other random samples
are taken from the population, the point estimates derived from those samples are likely to vary.

Interval estimation: a point estimate may be right or wrong. It does not indicate the degree of
error or uncertainty or precision in the estimate. In estimating business parameters an error is an
obvious expectation. Therefore, uncertainty should be accommodated in business estimations.
Interval estimation includes the degree of error or uncertainty in the estimate. In interval
estimation, an interval is constructed around a point estimate and it is stated that this interval is
likely to contain the corresponding population parameter. Each interval is constructed with
regard to a given confidence level and is called a confidence interval.

An interval estimate or confidence interval: is a range of values within which the analyst can
declare, with some confidence, the population parameter lies. Confidence intervals can be two
sided or one sided.

Estimation of Population Mean, µ

Population mean =Sample mean ± z *  X

 S
i.e. µ = X  Z n or µ = X  Z n
where:  X is the standard error of the mean,
population standard deviation known is given by   
 X
n
based on the sample standard deviation is given by S  S
 X
n

 When we do this we say that we have constructed or established the confidence interval
for the population mean at a specified/ chosen confidence level.

 The width of a confidence interval depends on the size of the maximum error which
depends on the values of z, σ and n.
 However the value of the standard error of mean  X is not within the control of the
researcher, hence the width of CI depends on: the value of z which depends on the
confidence level and the sample size n.
 The confidence level determines the value of z which in turn determines the size of the
maximum error. The most commonly used confidence levels are 95% and 99%. The z
value at these confidence levels is found in the standard normal tables and used in
computations.
 What is the z value at confidence level 95%?

At 95% confidence level z = -1.96 and + 1.96

What is the z value at confidence level 99%?

At 99% confidence level z = -2.58 and + 2.58

Exercise:
Determine the z value at 90% confidence level

Ans: At 90% confidence level z = -1.645 and + 1.645

Note: z value increases as the confidence level increases and vice versa i.e. the higher the CL, the
wider / larger the width of the confidence interval, other things remaining the same. An increase
in the sample size decreases the width of the confidence interval. Therefore, if we want to
decrease the width of the confidence interval, there are two choices:-
i. Lower the confidence level
ii. Increase the sample size
However, lowering the confidence level is not a good choice because a lower confidence may
give less reliable results. Therefore, we prefer to increase the sample size if we wish to decrease
the width of a confidence interval.

Examples
1. A study involves selecting a random sample of 256 sales representatives under the age of 35.
One item of interest is their annual income. The sample mean is Sh. 55,420 and the sample
standard deviation is Sh. 2050.
a) What is the point estimate?
b) What is the 95% confidence interval for the population mean rounded to the nearest Sh.
10?
c) What is the 99% confidence interval for the population mean rounded to the nearest Sh.
10?

2. A bank wishing to determine the average amount of time a customer must wait to be served
took a random sample of 100 customers and found that the mean waiting time was 7.2
minutes. Assuming that the population standard deviation is known to be 15 minutes, find the
90% confidence interval estimate of the mean waiting time for all the bank’s customers.
3. The sponsors of television shows targeted at the children’s market wanted to know the
amount of time children spend watching television, since the types and number of programs
and commercials greatly depend on this information. As a result, a survey in Nairobi was
conducted to estimate the average number of hours children spend watching television per
week. From past experience, it is known that the population standard deviation is 8 hours. In
a sample of 100 children, it was found out that x  27.5 hours. Find the 95% confidence
interval estimate of the average number of hours children in Nairobi spend watching
television.
Estimation of Population Proportion, π

Population proportion = Sample proportion ± z *  p

pq
i.e. π = p ± z p or π = p±z
n
where:
p pq - the standard error of the proportion
= n

q = 1- p

When we do this we say that we have constructed or established the confidence interval for the
population proportion at a specified/ chosen confidence level.

Examples
1. In a sample of 800 students, 560 were male. Estimate male population proportion at the 95%
confidence level.
2. A sample of 600 accounts was taken to test the accuracy of posting and balancing of
accounts wherein 45 mistakes were found. Construct the 99% confidence interval of the
population.
3. A market survey was conducted to estimate the proportion of homemakers who would
recognize the brand name of a cleanser based on the shape and colour of the container. Of
the 1400 homemakers, 420 were able to identify the brand name. Using the 99% degree of
confidence, calculate the confidence interval of the population proportion.
4. Suppose the Nation TV network is considering replacing one of its prime time dramas with a
new family oriented comedy. Before a final decision is made a random sample of 400 prime
time viewers is conducted. After seeing a preview of the comedy, 250 indicated that they
would watch it.
i. What is the point estimate of the proportion of viewers in the population who will
watch the new show?
ii. Develop a 95% confidence interval for the proportion of viewers who will watch
the new show.

Qualities of a good estimator

1. Unbiased: an unbiased estimator of a population parameter is an estimator whose expected
value is equal to that parameter i.e. if you were to take an infinite number of samples,
calculate the value of the estimator in each sample, and then average these values, the
average value would equal the parameter.
2. Consistency: an unbiased estimator is said to be consistent if the difference between the
estimator and the parameter grows smaller as the sample size grows large.
3. Should be efficient: an efficient estimator should have the least variance or least standard
error.
4. Should be sufficient: an estimator is said to be sufficient if it extracts from the sample such
an amount of information as no other estimator does.