I can use the One Way ANOVA fine. The TWO WAY either comes up with weird
results or does not accept the number of rows I enter.

If I have a table with 10 rows and 2 cols and I select that region using the
ANOVA addin and enter 5 for rows, I get an error saying that I must have
equal number of rows.

I am using a data set that works ok in another environment.

This is with Excel 2003.

KarlsOne -

(1) If you have "a data set that works ok in another environment," why not
just use that environment?

(2) If you browse to microsoft.com and search for "excel two-factor anova
with replication" (without the quotes), you'll see Knowledge Base article

http://support.microsoft.com/kb/267281

which shows how the data needs to be arranged.

(3) If all else fails, use google.com, or refer to Ch. 13 of my book, Data
Analysis Using Microsoft Excel: Updated for Office XP.

All of the data analysis tools have been essentially unchanged in all
versions of Excel.

- Mike
http://www.mikemiddleton.com

"KarlsOne" wrote in message
...
I can use the One Way ANOVA fine. The TWO WAY either comes up with weird
results or does not accept the number of rows I enter.

If I have a table with 10 rows and 2 cols and I select that region using
the
ANOVA addin and enter 5 for rows, I get an error saying that I must have
equal number of rows.

I am using a data set that works ok in another environment.

This is with Excel 2003.

Mike,

Thanks for your prompt and very helpfull response. I (way back) did this
type of calculation by writing a basic program, but recently, was asked by a
student to answer why the excel addin didn't work. Volunteering to help, but
not being familiar with the addins, I tried one way anova and it worked fine,
but the two way didn't. Your response sent me the the MS support item (which
I had not come up with during my google or MS community search. This was
exactly what I needed as it pointed out the use of Labels, which I never
encountered in the Excel help. After putting the labels as indicated, all
worked out just fine. I will send your reference book info to them.

Thanks again and Best Regards,

KarlsOne

"Mike Middleton" wrote:

KarlsOne -

(1) If you have "a data set that works ok in another environment," why not
just use that environment?

(2) If you browse to microsoft.com and search for "excel two-factor anova
with replication" (without the quotes), you'll see Knowledge Base article

http://support.microsoft.com/kb/267281

which shows how the data needs to be arranged.

(3) If all else fails, use google.com, or refer to Ch. 13 of my book, Data
Analysis Using Microsoft Excel: Updated for Office XP.

All of the data analysis tools have been essentially unchanged in all
versions of Excel.

- Mike
http://www.mikemiddleton.com

"KarlsOne" wrote in message
...
I can use the One Way ANOVA fine. The TWO WAY either comes up with weird
results or does not accept the number of rows I enter.

If I have a table with 10 rows and 2 cols and I select that region using
the
ANOVA addin and enter 5 for rows, I get an error saying that I must have
equal number of rows.

I am using a data set that works ok in another environment.

This is with Excel 2003.