Bob is correct. Paul's suggestion isn't what I was looking for. I tried it myself Tuesday night also. Bob's first comment about putting all of the data in C1 is what I needed. I had about 5700 pieces of data, and I forgot that the student version of Minitab will only take 3500, so I had to scale it down proportionally. Bob's suggestion gave me what I was looking for. From his explanation, I understand now why Minitab is programmed that way, but it is frustrating when my little TI-83 gave me what I was looking for with a lot less effort. I'll try Bob's macro this weekend.
The main reason I wanted to use Minitab for this data is because I was trying to make a normal quantile plot, or a normal probability plot as Minitab calls it. I didn't see a quick way to do it on the 83, so I turned to Minitab. That's when I got stuck.
Thanks to all of you for your help.
Bill Roloff Lake Park HS Roselle, IL
In a message dated 96-08-29 20:50:13 EDT, firstname.lastname@example.org (Bob Hayden) writes:
<< I tried what Paul suggested and I don't think it does what Bill wants, but I could be misinterpreting what he wants. My interpretation is that he has data like this
1 24 2 14 3 12 4 0 5 1
representing 50 numbers (all integers between 1 and 5 inclusive) and he wants ONE mean for all 50 numbers. I do not get ONE mean from the TABLE command. I can do this with a macro I wrote. See end of message. I'd be curious if this is what Bill wanted and if the results below are what Paul is getting by his method for this data.
----- Forwarded message from Paul James -----
I think what you are looking for is in the table command under stats. 1. go to - tables 2. cross tabulation 3. in Classification variables enter c1 4. trun on Frequencies are in box and enter c2 5 go to summaries, pick what you want (very limited and not all seem to work) 6. in associated variables enter c1 and you are off to the races >>