Chrissy,

You can do it easily enough. Put your lists of sixteen variables in colums
A to P starting in row 17, with one set per row. Then copy those sets and
transpose them to Q1. (If you have different sets, enter them in columns
starting in cell Q1, to Q16, then R, S, etc.)

Then in cell Q17, enter the formula

=CORREL($A17:$P17,Q$1:Q$16)

and then copy that formula into the block that matches your filled rows and
columns. That will create your matrix of values, that you can then plot.

HTH,
Bernie
MS Excel MVP

"Chrissy Wissy" wrote in message
...
Hi,
I should have made myself clearer. The correlations are based on 16
measurements for each variable. I want to plot every correlation
coefficient
(for all variable combinations) as an intensity contour. It is quite hard
to
find an application that can plot a correlation matrix in its normal form.
The only way with Excel in its bare form as I see it is to convert the
matrix
into a very long list of x (variable), y (variable), z (correlation
coefficient) values!

"Bernie Deitrick" wrote:

Chrissy,

You cannot have a meaningful correlation value for less than three sets
of data points. Are there
more points that go into it?

HTH,
Bernie
MS Excel MVP


"Chrissy Wissy" wrote in message
...
Hi
this is a point I have also raised in the graphs section. I was
wondering if
there was an excel addin that allows a coorelation matrix to be ploteed
as a
contour plot. What I want to essentially view is a contour plot showing
the
"intensity" (r
= 0 - 1) of the correlations between each variable for all
combinations. This
is primarily because I have complex data and such a realisation would
greatly
enhance my interpretation. For example, the "diagonal" in the contour
plot
would show a maximum intensity of 1 for each combination of x1/x1,
x2/x2,
etc. It would be quite straightforward to place the mirror image of the
matrix data above the triangular matrix should that be required by the
application.