The history of SEMNET
SEMNET was originally established by Carl E. Ferguson, Jr., University
of Alabama, and Edward E. Rigdon, Georgia State University, in
February 1993, in an effort to facilitate the exchange of information
in the field of structural equation modeling (SEM). Ferguson
and Rigdon, two marketing professors, recognized that SEM is an
interdisciplinary field. While many people were studying or using
SEM in their research, they tended to be isolated from each other.
Departmental or disciplinary boundaries meant that even people
working at the same institution, perhaps in the same building,
might not recognize each other. The negative effects of this
isolation were intensified by (1) the rapid pace of change in
the field, (2) the complexity of the method, and (3) the poor
quality of most SEM software.
Ferguson and Rigdon had been searching for a vehicle to break
through this isolation since hosting a small interdisciplinary
SEM conference in Atlanta in early 1992, but few options looked
promising. Electronic bulletin board services required an investment
in equipment and time. Commercial services were too expensive,
given the specialized nature of the field. Several professional
associations had special interest groups (SIG's) devoted to SEM,
but organizational bureaucracies made cross-discipline exchanges
difficult. Establishing a new professional association for SEM
was too much work to contemplate, except as a last resort.
In early 1993, Ferguson learned that the Seebeck Computer Center
at the University of Alabama would be willing to play host to
an Internet discussion group. Following a meeting between Ferguson,
Rigdon, and Joyce Donnelly, director of the Seebeck Center, Darren
Evans-Young of the Seebeck Center did the necessary programming,
and Ferguson became the list owner/ postmaster.
SEMNET's subscription list grew slowly at first, driven by Internet
postings, word-of-mouth, and flyers distributed at academic meetings.
More rapid growth occurred as teachers and leaders in SEM became
subscribers, and after a description of SEMNET appeared in the
second issue of Structural Equation Modeling: A Multidisciplinary
Journal, which had begun quarterly publication in January
1994. By the end of 1994, SEMNET had over 450 subscribers around
the world. In April 1996, SEMNET has more than 950 subscribers.
Who should subscribe to SEMNET?
SEMNET is intended for everyone who studies or uses SEM methods,
whether they are a novice to the field or a leader. All disciplinary
affiliations are welcome. Even in its early days, SEMNET's subscription
list included disciplines ranging from medicine to marketing to
What are the rules for posting on SEMNET?
General "netiquette" applies here. Remember that what
you post to SEMNET will be seen by nearly one thousand very busy people
around the world, including a great many of the leading authorities
in SEM. SEMNET exists because its subscribers are willing to
read and answer SEMNET postings. This willingness comes from
the subscribers' perception that the postings are worthy of their
interest and their precious time. Junk postings erode that perception,
and threaten the vitality of the list.
Remember also that, just like an ordinary letter, once it's gone,
it's gone. SEMNET is currently unmoderated. That means that
your posting goes out automatically, to all SEMNET subscribers,
more or less. Your own e-mail system may include a sophisticated
"recall" function, but SEMNET does not have this function
intrinsically. So be careful.
Also, double-check the "TO" address, to avoid sending
personal message over the 'net. That's really embarrassing.
Finally, remember that the people who read your postings are mostly
pretty nice people, who are doing you a favor by reading your
messages. Be polite. Don't be insulting. If you feel like you've
been insulted, assume it's just a misunderstanding. Resist the
urge to charge out in defence of your name or your work. Remember
that SEMNET subscribers, by and large, are well qualified to evaluate
the merits of an argument, and know the difference between insight
and invective. They also, by and large, remember when they were
just learning about SEM, or about electronic communications, themselves--they
won't condemn you for making a mistake, because they have made
Can I use SEMNET to sell products and services?
The Internet was established as a tool for research. Many of the entities that contribute resources to maintain the Internet in general (and SEMNET in particular) do so on the understanding that they are supporting a noncommercial service. Furthermore, in general, SEMNET subscribers will prefer to see as little "advertising" on the list as possible. On the other hand, SEMNET subscribers are vitally interested in developments in SEM, including developments in SEM software and literature. Subscribers like to know about new software, or new versions of software, and about books and journals that are relevant to the field.
Consequently, SEMNET subscribers will appreciate information about new SEM software, books and journals, but not puffery. To make the best impression on this group, keep your SEMNET post brief, and either direct interested persons to your Web site or offer to provide more information by return mail or e-mail. (If you choose the latter, be sure to ask that e-mail replies be posted to your individual address and not over SEMNET.) If you follow this rule, your messages will be welcomed by the vast majority of the SEMNET audience.
What makes for a good SEMNET question?
Anything that relates to SEM is fair game. It's a good idea to
check (1) this FAQ, (2) published sources, and (3) available software
support first, but even very basic questions are welcome here.
Very advanced questions are also welcome. Given the nature of
this field, however, posters should recognize that not every question
has a known answer, and more than a few questions have multiple
answers, each one contingent on a different set of assumptions.
Understand the temptation to ignore a question whose correct
answer is very involved.
What makes for a good SEMNET answer?
(This paragraph is based on the thoughts of David Ronis, Raymond
Liedka, Joop Hox, and Kelly Wason.) For questions with brief,
concrete, and noncontingent answers (that is, answers that do
not reasonably start out with, "It depends . . ."),
start your answer by restating the question. Instead of including
a (potentially lengthy) verbatim copy of the question, restate
the question briefly. If you think that question relates closely
to a broader class of questions, describe that broader class.
A good answer is, of course, correct and complete, with assumptions
spelled out. If you are not sure about your answer, then say
so. Provide references to the literature, if you can. And if
you need additional information before you can answer a question,
encourage the original poster to provide that information details.
Most questioners will be happy to do so.
Can I cite SEMNET messages in a manuscript? How?
Published literature can be very frustrating when you want the answer to a specific question. By contrast, specific answers are what SEMNET is all about. So can a researcher cite SEMNET posts in research manuscripts?
At the least, SEMNET postings are equivalent to personal communications, which can be cited as long as they are correctly and carefully identified. More than that, because all SEMNET postings can be retrieved from SEMNET's monthly log files, these messages enjoy a greater degree of verifiability than personal communications as a whole.
On the other hand, unlike published research, SEMNET postings are not peer-reviewed, so the potential for error is probably greater. Furthermore, publication of a manuscript carries the presumption that the authors are willing for their work to be cited, while the authors of SEMNET postings may prefer that at least some of their remarks not appear in print.
So, to be polite, first get the author's permission for you to cite or quote their particular remarks. Then use the reference style for personal communications that is recommended by your style manual. As time goes by, newer editions of style manuals may include a specific style for this type of reference.
Return to the SEMNET FAQ home page.
Return to Ed Rigdon's home page.
Last updated: April 11, 1996