I have a table created using an Access add-in utility (FMS' Total Access
Statistics). The TAS scenario, when run, calculates correlation
coefficients (r values) and generates an output table of values in a
nonnormalized "matrix" form. Here're some sample data to illustrate
(mind the linewrap):

Datafield     Apples     Oranges     Pears     Tangerines
Apples        1.0000     0.1234      0.9876    0.2345
Oranges       0.1234     1.0000      0.5678    0.9988
Pears         0.9876     0.5678      1.0000    0.3456
Tangerines    0.2345     0.9988      0.3456    1.0000

I need to stack these data in a tall-skinny form in order to look at
each comparison and run additional analyses on the results. I am
presently using a "normalizing Union query," which looks something like
this in SQL:

SELECT DataField as Fruit1, "Apples" AS Fruit2, [Apples] AS RValue FROM
MyTable WHERE[DataField]<>"Apples"
UNION
SELECT DataField as Fruit1, "Oranges" AS Fruit2, [Oranges] AS RValue
FROM MyTable WHERE [DataField]<>"Oranges"
UNION
SELECT...yada yada yada...

This returns a recordset which resembles the following:

Fruit1     Fruit2     RValue
Apples     Oranges    0.1234
Apples     Pears      0.9876
Apples     Tangerines 0.2345
Oranges    Apples     0.1234
Oranges    Pears      0.5678

...and so on.

But this "normalizing Union query" requires me to hard-code the name of
each fruit. Is there any way to programmatically rearrange this
recordset in the slightly-more-normalized form shown above? Any help
would be very much appreciated.

TIA,

LeAnne

PS. Using Access 97, so any sample code needs to be DAO. Thanks!

I think I have code that does what you need; e-mailed it to your address. If