In a message dated 08/24/2000 5:53:44 AM US Eastern Standard Time,
[log in to unmask] writes:

> To estimate the average of SIR-test results should arithmetic or geometric
>  mean used and how standard deviation should calculated?
>

There are several ways you COULD do it.  If you are working to Bellcore, or
Telcordia Technologies as they are now called, then it has a specific method
for the calculation.  In essence, it is the geometric mean for Bellcore, and
that is probably how you would do it in most cases.

What I would recommend is that you work in LogOhms, which is the base 10
logarithm of the measured resistance.  The antilog of the arithmetic average
of the LogOhms is the same as the geometric mean of the resistance.  Standard
deviations on LogOhm data are also easier to comprehend and work with.   I
prefer to work in the LogOhm system because if you are doing calculations on
a large number of high value data points, you will get an overflow error in
an Excel spreadsheet.  With LogOhms, this does not happen.

Also, make sure that you are combining together like patterns.  Don't average
the values from a B-24 pattern with those of a B-25 B/E pattern.  You can't
directly compare patterns with different geometries.

Doug Pauls
CSL

##############################################################
TechNet Mail List provided as a free service by IPC using LISTSERV 1.8c
##############################################################
To subscribe/unsubscribe, send a message to [log in to unmask] with following text in
the body:
To subscribe:   SUBSCRIBE TECHNET <your full name>
To unsubscribe:   SIGNOFF TECHNET
##############################################################
Please visit IPC web site (http://www.ipc.org/html/forum.htm) for additional
information.
If you need assistance - contact Keach Sasamori at [log in to unmask] or
847-509-9700 ext.5315
##############################################################