Statistical Software

Depends on what DOE wants to do...

Minitab is good for six sigma and value engineering - so if that's what your DOE project needs then I'd get that one.

Definitely follow the advise of what your major customer wishes to see. I prefer using a generic mathematics package like MATLAB, MATHCAD or Mathematica instead of a prepackaged statistics package. With these platforms you will have to know exactly what statistical analysis is required, for you will have to do the programming yourself. MATHCAD pioneered their program to look exactly like the mathematical equations your using, so proving that your methods are correct get to be a little easier. But all three platforms now use mathematical notation in the program space.

Depend upon what you are looking for.

Even you will find many functions is Microsoft office Excel. They are good enough for majority of statistical analysis.

As others have said - A DOE on what? What sort of DOE? Or is it just basic statistical analysis that you want to do?

The answers to these questions will guide others in helping you.

Bill,

I want to use DOE, just for basic statistical analysis.

regards,

Since you apparently don't mean Department of Energy, Department of Education, Date Of Expiration or some other easy to decipher meaning of your acronym DOE it would help if you explained what you are looking for. Maybe you mean an Odocoileus virginianus that does not have a Y chromosome?

Why don't you try R?

• www.r-project.org

It's free! And it's good.

Robert,

what is your experience with the R statistic software?

Regards,

Ramón

Hi Ramón,

My use of R is limited to when I studied statistics and we used it for student projects (actually it was in experimental design). However the lecturer was a consultant statistician (to industry and legal firms) and he used R exclusively. It's really his recommendation that I pass on rather than my own. While I did get the hang of R and found it extremely powerful, I am not an expert user by any means. (Basically we did some simple experiments and then analysed the results to see which parameters were contributing to the results and in what combination - lots of box plots / anova / regression type analysis - that sort of thing).

R does require you to learn its command line language - very simple really and lots of help available on the web - but it's not point and click like some of the expensive packages. It can be intimidating at first when you first open it and there are no buttons to press, just a cursor.

That said, it is very easy to point and click the expensive packages and not have any idea what's coming out... with R you need to be more explicit about what you are asking it to do.

Very powerful for stats and graphing - I used it in combination with excel. I would use excel to create the data file that R reads. In that way I could cross check between excel and R as I was learning how to drive R.

Hope you find the package you're after anyway.
Regards,
Robert

Hello, I have MINITAB and it is and excellent Software with all quality tool. It is a solid software with lot of tutorials and resources. The cost is over $900.00 if the money it no a issue. Is the best option have more options for any future needs. Since DoE have 45-50% time for Reduction, analysis and writing, MINITAB is the Best.

Minitab -Picture

Hi ramoncrispin,

I think you are referring to Design of Experiment (DOE), right?

I don't know of any others except Minitab. As others have said here, you can use Excel and do the design manually. That's what I would do, using the applicable statistical functions in Excel.

Mike

Don't you need to draw a line between straightforward statistical analysis and true DOE?

For example - the difference between analysing/characterising data and, perhaps, a Taguchi-style DOE?

Link

http://www.ee.iitb.ac.in/~apte/ExcelTaguchi_1.htm