🚀 go-pugleaf

I started the current comparison by saying,

So many people say "Minitab is easy to start using, Glim is difficult".

Russ Lenth makes a point I don't fully agree with.

>I'm not sure that I agree that GLIM and Minitab are equally easy to learn,
>primarily because GLIM lacks an online help facility.  (That will change with
>GLIM-4, however.)

On-line helps are not in my experience a good way of learning any package;
they are an excellent aide-memoire if you know roughly what you are doing
but have forgotten the exact syntax. What they still lack is a good
hypertext feel - you're looking up a command JUST BECAUSE you've forgotten
its name!

>I think that an important distinction between GLIM and just about any other
>package is the fact that on has to specifically ask for most results, such as
>regression coefficients, etc.  That makes it seem less "user-friendly", but I'd

Agreed. It's partly historical, in that GLIM was developed on interactive
machines with typewriter-like terminals, but also part of the philosophy
that you should know what you are doing.


Murray Jorgensen (glad to have you on the list - how's autumn?) seems to
make my point by a contradiction.

>I have been interested in the Minitab v Glim/Genstat discussion because I
>am familiar with all three packages.
>
>I have used Genstat off and on since the late 70s when I encountered it
>while working as a government-employed statistical consultant [NZ Ministry
>of Agriculture and Fisheries]. It's great strength is in the analysis of
>designed experiments which is original reason we first purchased it.

Agreed. That's exactly the split I make. Genstat is the first recommendation
for analysing DESIGNED experiments, especially if the design was carried out.

>When I became an academic a few years ago I adopted Minitab
>because of its simplicity, ...
>
>My current research involves generalizations of GLMs [ G^{2}LMs ? ] and
>I find myself writing collections of Minitab macros to fit these models
>using IRLS. I do this because the Minitab environment is like being
>inside a friendly debugger all the time. You can get a lot of insight
>into how an algorithm is progressing by doing things like plotting

But do most people analysing data want to write their own algorithms?
(That's a rhetorical question. Anyone before answering should read the
archives on the Great Spreadsheet Debate. )

>One final remark : I am getting tired of all the dodges that I have to get
>up to to compensate for Minitab's lack of even rudimentary language
>control structures like IF ... THEN ... ELSE . I am thinking of shifting
>to XLISP-STAT or S+ for my research computing.

So what happened to simplicity? My view on Minitab is that it may be great
for getting people started very quickly, but then blinkers them by providing
a limited range of "cookbook" analyses. It's the same as the argument
against teaching Basic as a main language for people who want to write
serious and efficent and readable computer programs. Let's put it to
students (wide sense) from the outset that analysing real data is a non-trivial
task that will require them to (gasp) think, so the earlier they start
thinking the better. Cookbooks are great if you do know what you are choosing
from them, but what would you think of a host who served steak and kidney pie
with custard because he looked up "deserts" and cross-referenced on "pie"?

... adjourn for lunch.