You are working with the text-only light edition of "H.Lohninger: Teach/Me Data Analysis, Springer-Verlag, Berlin-New York-Tokyo, 1999. ISBN 3-540-14743-8". Click here for further information.
|
|
Variability
Whenever some measurements are taken, one has to deal with variability
(or dispersion) in the data. As a matter of fact, virtually all
natural processes vary - even those which yield "constant" results show
up fluctuations if either the observation time is long enough or the scale
of the measurement is zoomed in.
Examples:
-
The actual amount of sugar in a 1 kg package of sugar may vary between
0.95 and 1.05 kg.
-
The temperature of boiling water varies with pressure and with the amount
of dissolved substances. Thus, the boiling points of several brands of
different mineral waters differ slightly from each other.
-
A time of reaction to a certain event may vary according to your physical
and psychological condition (don't drink and drive). Let's perform an experiment
by starting the following
(reaction
time measurement before and after a beer - use this data set later on).
-
The amount of rain in July varies from place to place, and from year to
year.
Statistics helps in coping with situations in the presence of variability.
The reason for the variability may be twofold:
-
process variability - introduced by the process under investigation
-
measurement variability - introduced by the measuring procedure
The variability in the data is often referred to as the noise
that obscures the true signal.
Last Update: 2004-Jul-03