request for comment

["Andreas J. Guelzow" <aguelzow math concordia ab ca>]

Hi,

I am in the process of migrating the data analysis tools to use the 
gnumeric expre-entry widget (I really don't like to type in the ranges 
when I could select them instead). The descriptive stats tool dialog is 
already in CVS, I have also completed the correlation, covariance and 
ranking tools.

WIth respect to the 3 t-tests and the z-test I have th efollwing question:

Currently they appear as 4 differnt tools (with four different dialogs). 
Conceptually these tests are closely related, I would therefore prefer 
considering them a single tool. This yields 3 possible ways to handle them:

1) (my preference) a single tool with a single dialog that changes 
slightly based on selections (ie. equal or unequal variances, etc.) 
similar to the random number dialog.

2) 4 tools that lead to the same dialog preloaded with the approriate 
selections. (This may be the way to go if it is considered important to 
have those 4 different tools, for example in deference to XL)

3) 4 tools with distinct dialogs as currently implemented (but of course 
using the expression entry widget).

Any opinions?

Andreas

--
Prof. Dr. Andreas J. Guelzow                    
Assoc. Prof of Mathematics
Concordia University College of Alberta
http://www.math.concordia.ab.ca/aguelzow