At 09/05/2003 13:07, you wrote:
>I am looking for someone experienced in computer-based content analysis of
>electronic text who could give me some advice on choosing the most
>appropriate package. I don't want to burden the list with my questions, so
>if anyone is willing to reply off-list, I would be extremely grateful.
I am not the best person to give you objective advices since I am the
author of one of those content analysis software (WordStat). But I know
that most existing software have special features that make them different
from the others. Some have built in categorization dictionaries and are
thus very good at the analysis for specific types of documents (for example
Diction for political speeches, Tabari for the analysis of international
events, PCAD for psychological or psychiatric evaluation, etc.). Their
usefulness is often greatly reduced outside those application areas. Other
software have built in dictionary that are more general in scope such as
General Inquirer or the LIWC).
There is also another category of content analysis software that don't use
any categorization dictionary, but use statistical procedures and sometimes
linguistic algorithms to summarize texts (TextSmart, TextAnalyst, CatPac,
Alceste) and perform automatic analysis.
Finally there are content analysis tools that allows you to create your own
categorization dictionary or rules or apply existing ones. WordStat is one
of those, such as Protan, TextQuest, and many others. Each of them have
unique features and you may find that one of those features is essential to
your project.
I suggest you take a look at Harald Klein's web site www.textanalysis.info
(Harald is the author of TextQuest) and look at the different software
descriptions and maybe also a their associated web sites. Maybe you will
find something there that meet your needs or you will be able to have a
clearer idea of what features you really need.
Once you eliminated some software and identified a few other ones, I
suggest you send their author a description of your study and of what you
are trying to achieve and ask them directly how their software can be
useful to you. You may also go back to public forums and ask for users
comments on those selected software.
If you do so, I suggest you subscribe to the CONTENT mailing list at:
http://bama.ua.edu/~wevans/content/CONTENT_menu.html
and post a message there.
Normand Péladeau
Provalis Research
www.simstat.com
BTW, don't forget to take a look at our own software WordStat at
www.simstat.com/wordstat.htm or at the newest version 4.0 that should be
released in a week or two ( temporary page at www.simstat.com/wordstat4.htm )
|