Thomas,
many thanks to your elaborate review to my review, I will try to address
all your concerns.
>Good work, but speaking of beta versions: although the version of ATLAS.ti
>5 you used (Release Candidate 2, RC2) for the above comparison was not
>beta anymore, the final release (June) was quite another boost in
>performance, stability and ease.
>Because of the unexpected demand for our new version, we needed to press
>another stock of CDs only a few month later and took the chance to
>incorporate even more improvements originating from users' feedback.
>
>So I would like to ask you to either check the final release or to somehow
>clarify that the version used for the comparison was not the final release.
I will certainly do that ASAP (my estimate is Thursday). In fact, most of
my data stem from Build 61, but my review is a beta release itself
(particularly many data on HyperRESEARCH and N6 are still missing), I have
not yet changed the version number to be on the safe side.
A few comments on your comments:
>Let me add a few comments as regards content:
>Under "Ergonomics" you state: "The biggest advantage of computer-assisted
>content analysis is its capability to handle bulk data. " In my view
>qualitative data analysis is NOT identical to bulk "word crunching"
>content analysis, although most programs incorporate some content analysis
>(and even statistical) technology.
Actually, I disagree: Although I am not at all averse to statistical
analyses, I do not think that bulk data means quantitative analysis. If I
have only small amounts of newspaper data (and my review is certainly
slanted towards media research), I actually would not use a program at all,
for digitizing newspapers means inter alia the detcontextualization of
data. If, on the other hand, I have bulk data, I am very grateful that I
can put everything onto a CD and organize and search it neatly on the
computer. That does not mean, I need to perform word counts, but that I can
memo the interesting parts of the data and search effectively for possible
counter example of my findings on small amounts of data.
>You write "the easier intelligible a program is, the more likely it is
>that you will know all of its options, as most people do not tend to read
>through the manual. " The problem is, that reading the manual or looking
>at the online help is mandatory for the more complex or tedious tasks.
>Easy to learn and not needing a manual may be feasible for trivial or
>conceptually common software but not for the more ambitious tools. We
>(you) should encourage users of such systems that its not a toy, but
>something which must be seriously learned, just like SPSS...
Oooops, I again disgree. Alas, I learned SPSS in times of DOS, so I mainly
use its scripting language, as then I can change part of the analysis very
quickly without saving several versions of the data after a recode. But
from time to time I do use the Windows interface and it is pretty
straightforward, if you know, what kind of methodology you want to use: I
have never consulted the SPSS manual w/ respect to its Windows interface.
Sure, you should read the odd stats book to operate SPSS, but apart from
the scripting language, I think the SPSS manual is dispensable.
>The screen dump you chose for illustrating the "bad interface" of ATLAS.ti
>is really ugly and with these colors truly in support of your theses!-)
Really ugly? MY color scheme? How dare you :-) I changed it now into
Loughborough's corporate colors (as I was supposed to do anyways), but it
still does not look pretty in my view, even though I fiddled quite a bit to
achieve an agreeable result (Magenta and purple are not my favorite colors,
either). I will try to find a better solution, but usually people do not
adjust their color schemes to the application. I am though less concerned
about aesthetics, but how you should figure what the varuious symbols and
letters on the interface mean. I am not alone in my assesment here, Lewis'
article in Field Methods also singles out the query tool as not very
intuitive. Having said that, I do think that interface design is definitely
not the most crucial point for a scientific program.
>In your categorization of how Windows are handled you are using "tedious,
>non-standard, no shortcut keys" as a value. Would you mind stating the
>"standard" as you see it? We were putting much effort in assimilating the
>Windows standard in our interface however good or bad it may be.
I changed the tedious, as it is indeed pretty subjective and gives little
information; I now offer a description of the behavior
>And for the less humorous we even offer a preferences switch "Be serious"
>to get rid of (most) of such useless things like "I hate computers,"!-)
Ouf. I'm glad you did allow us to remove that key in the "be serious"
option (Build 61 still thought the button was serious).
Again thanks for the commentary, and I will let you know, when I have the
data for Build 63.
Thomas
>
--
thomas koenig, ph.d.
department of social sciences, loughborough university
http://www.lboro.ac.uk/research/mmethods/staff/thomas/index.html
|