For those that asked for discussion... here you go! This message ended up
*much* longer than I'd intended...

First, I'd like to thank Jens for sharing her notes on the differences
between Nud*Ist and NVivo. In general, her notes closely correspond to the
impressions I had when I first saw the program demo'd at Advances in
Qualitative Research last February. She gave a great summary of the
differences between the programs that may affect which choice would be 
better under different circumstances. Thanks, Jens!! 

I am using Nud*Ist4, am learning NVivo, and have been to a workshop on
Atlas-ti and have played with demos of a couple of other programs. I have
also read widely in the literature on qualitative data analysis programs,
and talked to people who use multiple programs. (I'm in the field of
Instructional Technology, and am interested in how people use computers
for high-end creative and intellectual work.  Qualitative data analysis
programs certainly fit that description! ) 

Below is a message that I previously sent to Qualrs-l (a listserv on
qualitative research in general), in response to someone asking about
several qualitative programs. Thought the new members who wanted some
general discussion on qualitative software might find it useful. :-)

===(message on choosing among several programs)

Some general advice and observations:

All of these programs are good, and will do what you want - assuming: 
(1) you understand qualitative research well enough to set goals, 
(2) have basic computer skills going in (navigating hard disk, 
moving/copying/renaming files, etc.), and 
(3) are willing to invest some learning time and explore 
new ways of working. 
These are very, very important assumptions.

However, you will probably find that different ones may be a better "fit"
for the way you work and think. Your computer hardware will make a
difference too -- some programs only come in Mac or PC flavors. NVivo
requires a fairly high-end computer, compared to the others. (Not
surprising, since it's the newest.)

My advice is to take some time to make a selection -- you'll be glad you
did. Ask other users *how* they work with the programs, download the demos
to play with, hang out on the developers' listservs, and read about the
program. After you make your decision, I strongly recommend attending a
training session.  You'll not only learn program functions, you'll
identify useful strategies for applying them. 

Sources of information:

Go to the developer's web sites, which have bibliographies, downloadable
demos and training support materials, etc.  A central website with links
to the developers (and a lot more information) is

http://kennedy.soc.surrey.ac.uk/caqdas/

This is a *great* resource. Check out the bibliography as well -- it will
include reviews, etc.

Weitzman and Miles (1995) is the classic reference. The reviews are a bit
dated now (a new edition is on its way, I understand), but there's a great
section on choosing a program that is still *very* applicable.

Hope this is useful to you, and that you aren't disappointed at not
getting a blanket endorsement of any one product. :-) The basic answer is
that "it depends"... as much on you, your skills, and your goals as the
functions in the programs. :-) 

===end previous message.

One last topic for discussion, from my IT perspective...in evaluating
programs, consider both *functionality* and *usability*. Here's the
distinction:

Functionality is a software-related judgment; it refers to what the
program will do, and how reliable it is. Different programs have different
strengths. 

Usability is a people-and-software description. It refers to how easily
someone can accomplish goals in the software, how cumbersome or intuitive
they find the interface, etc.  

Usability varies from person to person. It can vary from novices to
experts - someone accustomed to complex interfaces may find a program
quite usable, while a beginner would be overwhelmed. Conversely, experts
may be impatient with support features that novices like. (A great example
is the interactive help in Word. I hate that paper-clip! <g>) So, though
you may have an initial judgement about ease-of-use, expect it to evolve
over time. Prior experience with the domain that the software supports
also makes a *big* difference in usability -- an accounting package will
be easier to learn if you already know accounting, because you understand
what the end goals are. 

If you look at standard business or graphics software, you can see that
very functional software can often be difficult to use, simply because
it's complex and has a lot of features. Conversely, very usable software
may have more limited functionality (that's sometimes *why* it's easier). 
So, you need to evaluate both, and think about it in terms of your own
patterns as a computer user. What functionality do you need (and not need) 
and how willing are you to spend the effort to overcome interface
annoyances?

As a rule of thumb, you can expect that any software with a lot of
functionality will have a steeper learning curve than simpler software,
just because of complexity. For equally-complex software, the usability of
the software primarily affects that learning curve, especially in the
beginning. But usability *may* also affect later working processes - after
all, it's human nature to avoid doing things that are very difficult if
you can find another way!  ;^) 

I think getting a tool that won't do the job is unwise, but getting a tool
you can't/won't learn to use is equally unwise. So, think carefully about
your computer skills and attitudes. And always plan on a learning curve
--depending on new software when you're racing against a deadline is a
recipe for frustration. You need time to play.

Just my $.02. Hope these ideas and resources are useful.

Best wishes,

Linda Gilbert
University of Georgia
[log in to unmask]




%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%