Whenever I see statistics training using Excel it immediately make me
suspect that people who don't understand/use statistics are making the
software decision.

As you said, a modern Windows-based statistical software application
(MINITAB, JMP, SPSS, etc...) has so much more to offer the user.

------------------------
Example:
Given a single column of 1000 data values already entered into Excel, which
is easier:

1. Trying to explain to a student how to create a histogram in Excel

2. Trying to explain to a student how to copy the data and create a
histogram in MINITAB/JMP/SPSS/...
--------------------------

To a person knowledgable about statistics, modern statistical software is no
more complicated than Excel, and in most cases much easier, since it is
dedicated to specific tasks.

Anyone who says its easier to create a histogram in Excel, even with the
toolpak, needs a reality check.

Excel's toolpak requires the user to review the data, determine appropriate
bins, create the bins, run the toolpak, specify the bins and data.

Most modern statistics software simply requires the user to copy & paste the
data, run a histogram command and then specify the data. No pre-analysis of
the data is required.

And that's for the most basic analysis method (histogram). It gets even
worse for more advanced methods. Try to create a box and whisker plot in
Excel. How about multiple box plots on a single axis? Try to create a normal
probability plot in Excel. Try to a multi-factor ANOVA and output a variance
component table in Excel. Try to analyze a fractional factorial designed
experiment in Excel.

Yes, using macros and such it is surely possible to get these kinds of
analyses from Excel, but by then a lot of time, effort, and money has been
invested in turning that sow's ear into a silk purse . .  and it will still
smell like a pig.





=================================================================
Instructions for joining and leaving this list, remarks about the
problem of INAPPROPRIATE MESSAGES, and archives are available at
                  http://jse.stat.ncsu.edu/
=================================================================

Reply via email to