Scribd
Upload
9 views

Chapter 4 Sample Size Determination

Uploaded by

Admasu
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
9 views

Chapter 4 Sample Size Determination

Uploaded by

Admasu
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 17

Estimation of sample size

o Anyone designing a survey must decide:


 what amount of sampling error in the estimates is
tolerable and,
 must balance the precision of the estimates with the
cost of the survey.

o Even though many variables may be measured, an


investigator can often focus on one or two
responses that are of primary interest in the
survey, and use these for estimating the sample size.
1
o In designing sample surveys, instead of asking about the
required precision, many people ask:
 ‘What percentage of the population should I include
in the sample?’

o This is a wrong question since precision is obtained


through the absolute size of the sample – not the
proportion of the population covered.

o Indeed the population size has little effect in large


populations.

2
o Formulas for sample size determination often
involve certain unknown properties of the
population.
o These must be estimated in order to give specific
results.
o Some methods of obtaining such estimates are:

a) A pilot sample (survey), a small sample taken to


provide information and guidance for the design
of the main survey, can be used to estimate
quantities needed for setting the sample size.

3
b) Use the results from previous sampling of the
same or a similar population.

c) Apply guesswork about the structure of the


population assisted by some mathematical
results.
• For example, if you believe the population to be
normally distributed and have an idea of the range of
the data, then estimate the standard deviation as:
– S = range/4
• as approximately 95% of the values from a normal
population are within two standard deviations of the
mean.
4
5
6
Example:
o Suppose our aim is to formulate a model that
estimates the true market value of properties
(houses/buildings) in a certain city.

o For this purpose we need to take a random sample


of properties and collect information on their market
value together with their attributes (such as floor
area, construction material, access to infrastructure,
etc.).

o It is known that the total number of properties in


this city is 8000.
7
8
9
o Note that had the population size been N =
16000, the required sample size under these
settings would be n = 840.

o That is, if the population size is doubled, the


required sample size would increase just by
42.

10
11
12
13
14
15
16
17

You might also like