[Blindmath] review of Minitab and JAWS documentation

[Jonathan Godfrey]

Hi all,

 

I've read the document obtained by Paul from Minitab that outlines the lack
of accessibility that is to be experienced with Minitab 15 and an
unspecified version of Jaws. It's a good start but my guess is that the
reviewer was not experienced with Minitab, or possibly statistical software
in general. The reviewer was also sighted and explains things in a way that
a sighted person at Minitab would need things explained. The flow on impact
is that the reviewer tries to act blind and work sighted. The range of
activities mentioned exceeds the ways I successfully used when I was a
Minitab user. A blind user would not necessarily use all the techniques
mentioned in the review.

 

I found one comment I disagree with. The reviewer tried to use specific Jaws
keystrokes to read out material from the session window. If a Minitab user
has enabled the command language (commands get printed after menu items are
used), and users can type in their own commands, then we should be able to
just arrow up to any point in the session window. Sure the new output (valid
or error message) isn't echoed by Jaws but we can read it.

 

One criticism of Minitab given is the poor arrangement of statistical
output. The comment pertains to tabulated output that is not in a table.
Minitab (like many stats packages) uses a monospace font with embedded
spaces to ensure output looks tabulated. Yes this is difficult to read as a
jaws user but the only solution is to use proper tables and that would
require an output window in html or similar document format. SAS has already
gone the way of formatted tables; I suspect Minitab will too in due course.
BTW: R is the same as Minitab but there is a simple work around because we
can transpose anything that looks like a table.

 

The reviewer notes that Minitab graphs are inherently inaccessible. One of
the biggest bugbears for me was the ability to re-create a graph when I
needed to fix something like the title or axis label. A sighted person just
clicks and edits objects in the graph window. A blind user must go back
through the menus and edit the appropriate spots in the dialogue box. This
is where SAS, Stata and R have it over SPSS and Minitab. This editing is not
mentioned in the review.

 

The reviewer notes that some output in Minitab could be printed in the
session window instead of a graphics window. True, and in most cases this
can be achieved via the command language. Old Minitab commands do still
work. I'm fairly sure we can still generate graphs that are made up of text
objects, printed in the session window. I'll experiment with a colleague
that loves Minitab later.

 

Finally, the reviewer notes the fact that Jaws users get some help on using
SPSS. They don't note that this optimism is unfounded as the advice for SPSS
is woefully out of date, referring to some scripts written for version 13 of
SPSS by someone in the UK over ten years ago. I've written to FS to complain
about the advice on their webpages, in 2011 I think.

 

The conclusion reflects some optimism, but lacks the experience of a
dispossessed Minitab user. There is one note that suggests different
elements of the software were written at different times and reflects
differing programming practice. This is almost inevitable with software that
has existed for over 25 years.

 

Jonathan

 

 

 

 

 

 

 

 

A.      Jonathan R. Godfrey

Lecturer in Statistics

 

Institute of Fundamental Sciences

Massey University

Palmerston North, New Zealand

 

Room: Science Tower B, 3.15

Phone (work): +64-6-356 9099 ext 7705

Phone (cell): +64-29-538 9814