| Sample
Size Calculator - Means
Getting the correct sample size is critical to many projects.
Our sample size calculator (for means) makes calculating your sample
size simple and quick.
Use this calculator if you are answering a question like "How
big a sample do I need to find out how long it takes our back-office
function to answer a customer query?"
Use the
other
calculator (for proportions) if you are answering a question
like "How big a sample do I need to find out what proportion of
customer queries get answered by our back-office function within
Service Level?"
Three things affect your sample size.
- How confident you want to be in the result (The Confidence
Level, expressed as 90%, 95% or 99%)
- How accurate you need the result to be (expressed in +/- 'n'
units. For example if you are measuring in hours and want to know
how long it takes to respond to a customer to within +/- 2 hours,
you would enter 2 in the 'Acceptable Error' field). The Acceptable
Error of the Estimate is sometimes called the Confidence Interval.
- How spread out your data is, expressed as a measure of the
standard deviation of your data. If you haven't yet calculated the
standard deviation for your data, you will need to do so before
you can calculate a sample size. If you have access to Microsoft
Excel, you can do this simply by typing =STDEV(range start:range
end) and the value returned will be the value of a standard
deviation for your data.
If you need to be very confident in the results (e.g. 99%) and for
them to be highly accurate (say +/- 1 unit) and your data is very
variable (say a standard deviation is calculated as 7 hours for the
responding to customer example) then you will require a large sample
to give you this degree of confidence and accuracy.
However if you don't need to be so confident in the result (e.g.
95%) and it's not necessary to be so accurate (say +/- 3 units
(hours)) and your data is not so variable (say a standard deviation
is calculated as 1.4 hours) then you will require a much smaller
sample.
Sample size is usually a mixture of desired confidence and accuracy
balanced by the cost of obtaining the sample.
The most pragmatic approach is to put the Confidence Level and
Acceptable Error you would really want into the sample size
calculator and see what the recommended sample size is. If it is too
large, try adjusting the Confidence Level and Acceptable Error
values until you have a practical sample size.
The calculator defaults to 95% Confidence Level and 1 unit of
Acceptable Error. You can change these to suit your needs.
If you would like to learn more about Lean and 6 Sigma tools and
methodologies, click here for
open training courses,
or in-house training courses.
If you cannot see the sample size calculator,
check whether your browser is blocking active content, or whether
you have the Flash Player installed.
Note: aorist consulting ltd. provides tools and examples for
illustration only, without warranty either expressed or implied.
This includes, but is not limited to, the implied warranties of
merchantability or fitness for a particular purpose.
|