A Journal of Judgment and Decision Making
Jonathan Baron
From the SJDM newsletter, December 2005
This is a proposal for an open-access SJDM journal. I have
volunteered to be the founding editor.
1 Open access
Open-access academic journals have been advocated with increasing
stridency by scholars, librarians, and policy makers in the last
20 years.
This page
provides some history, which includes contributions from
Stevan Harnad,
George Soros, and the
National Academies.
The main arguments are simple:
- Authors rightly want their work to be as widely read as
possible. When readers must pay, we have fewer readers.
University libraries in particular have their budgets stretched
thin and resist maintaining subscriptions to specialty
journals, particularly in developing countries (including
China, India, and Brazil) and non-English-speaking countries
(including most of Europe).
- The corollary is that readers would benefit too. Some of
them may even learn of our field's existence by finding a
journal article of interest.
- On-line publication is faster. The articles are timely.
Print journals in many fields have become archives only. We
should make the benefits of rapid publication available to all,
not just the well-connected insiders who get drafts.
- The most crucial labor involved in producing a journal,
reviewing, is unpaid anyway. Journals are produced mainly by
volunteers (who may get other credit for their volunteer
efforts, but that would still be true). The cost of journals
buys little.
- Policy makers are disturbed that taxpayers must pay to read
the results of research that they have already paid for.
As Eric Johnson (2001) put it, "There is something odd about a
system where neither authors nor reviewers are paid, yet they are
billed to see the result that they do not own." Given the
existence of the World Wide Web, it seems Pareto inferior for us,
as scholars, to turn over our work to publishers, who then charge
us to read it and, in order to defend their copyright, reduce our
readership. The continued existence of commercial scholarly
journals is surely a massive example of the status-quo bias.
2 Why SJDM?
SJDM is in a nearly unique position to start an open-access
journal. Most of our peer societies (Society for Medical
Decision Making, International Association for Research in
Economic Psychology) are committed to commercial publishers for
their journals. SJDM as yet has no journal, although we have a
book series. If we were to start a traditional journal, dues -
now low - would increase substantially. By taking a commercial
publisher out of the loop, we can start a journal without
increasing dues.
Based on a survey sent out to members and to the SJDM mailing
list, it seems that the best form for a journal to take would be
one that emphasizes short articles (which have almost no outlet
other than Psychological Science), practical
applications as well as contributions to basic science or theory,
readability, speed of decision making, and high quality. The
combination of the last two criteria would mean that many
articles would be rejected without review, a practice now common
in journals such as Science, New England Journal
of Medicine, and Psychological Science. Such a
practice would reduce substantially the risk of submission
(compared, for example, to the possibility of waiting months for
a rejection, which now happens even at Psychological
Science).
3 Do we need a new journal?
Worldwide interest in JDM (judgment and decision making) seems to
be increasing. JDM has become more relevant to many applied
fields. Medical policy makers are attending more to
cost-effectiveness analysis, which relies on utility measurement.
We now have "behavioral law and economics," and "behavioral
finance," and (recently) "behavioral public economics," which
draw on JDM extensively, as does the older field of experimental
economics.
Yet, the availability of journal space for basic JDM articles
does not seem to be increasing. To get some very rough data on
this question, I searched PsycInfo for "expected utility" from
1980 through 2004. Here are the numbers going backward. You can
see that there is an increase (going right to left).
62 93 73 64 37 40 23 16 18 28 30 20 35 19 34 17 33 24 14 14 13 8 5 8 7
A lot of these are book chapters, but the basic idea is that
there has not been a decline of interest in this topic.
Then I counted all of these articles in: Journal of
Experimental Psychology [all versions, including Quarterly
... and Canadian ...]; Journal of Personality and Social
Psychology; Psychological Science; and Memory & Cognition. I
combined these. The corresponding numbers are:
0 3 0 3 0 3 2 3 1 1 5 0 4 1 5 0 4 3 0 1 1 2 0 1 0
Then I took the ratio of the latter numbers to the former, year
by year. The question is, then, whether this ratio is declining.
That is, are the main experimental journals publishing less work
on a topic central to JDM.
As you can see from the graph, there has been a recent decline in
this ratio. The points are year by year. The line is the result
of applying loess, a smoother. The downward trend wasn't quite
significant over the whole 25 years. But I think it does provide
some evidence than main-line experimental journals are not
becoming more friendly to JDM articles.
We still have OBHDP and JBDM, but we no longer have such easy
access to JPSP, JEP:HPP, and JEP:Applied as we did in years past.
We are pretty much limited to JEP:General, which tends to favor
multi-experiment articles that are of interest to "a broad
readership." Psychonomic Bulletin and Review has just
announced that it will be accepting more JDM articles, but it is
a general journal and cannot accept that many.
Psychological Science went through a lull but now seems
to be publishing more JDM articles again, but it, too, is general
and cannot solve the problem. It also has a very high rejection
rate.
I conclude that there is some need for a new journal. I do not
think this is the main reason for doing it, and the increased
need is small, but the opportunity is there. In putting it this
way, I recognize that a new journal will, to some extent, compete
with existing ones. However, the shorter articles we would
encourage would mostly not be published in current journals.
4 Absent readers
In the long run, it matters how easy it is for people to read
articles in a journal. The relevant people are not just active
researchers but also students at all levels (high school to
graduate school), journalists, and bloggers. The main method of
access now is institutional (usually university) libraries.
After that, perhaps, is individual subscriptions, but very few
students and journalists have these.
To get an idea of how available our work is, I looked in the
OCLC WorldCat database,
which combines library catalogs from all over the world . The
first few items listed below estimate the number of relevant
institutional libraries in the database, probably about 2000. By
this estimate, OBHDP is in less than 40% and JBDM in less than
20%. On the assumption that institutional subscriptions
represent most readership, JDM work is not easy to access.
2206 Journal of Abnormal Psychology
2129 American Psychologist
1891 Journal of Personality and Social Psychology
1859 Psychological Bulletin
1835 American Economic Review
1639 Psychological Review
1396 Ethics
1302 Journal of Experimental Psychology: General
1049 Management Science
940 Journal of Consumer Research
845 Memory and Cognition
721 OBHDP
648 Psychological Science
409 Journal of Mathematical Psychology
309 Risk Analysis
307 Journal of Economic Psychology
264 Medical Decision Making
250 Journal of Economic Behavior and Organization
179 Journal of Risk and Uncertainty
153 JBDM
Looking at individual subscriptions, the picture is not much
different. JBDM has an estimated individual circulation of 450
(from the Journal, thanks to Lauris Olson for digging this up).
JEP:General, by contrast, has 726 individual subscriptions. Of
course, Psychological Science has about 12,000, all the members
of APS, but APS includes relatively few members from Europe or
developing countries.
Notice also that web publication makes a journal accessible to
Google and other search engines. News reporters as well as
students will find relevant articles this way. My impression is
that very few students know how to use PsycInfo compared to the
number who know how to use Google.
My conclusion is that open access, on the web, will increase the
potential readership of JDM articles by something on the order of
500%, probably even more when we count undergraduates and
journalists. Of course, if we are going to take advantage of
this boost in readership, we need to make sure that articles are
readable. Many journals appear to be write-only and do not worry
much about this.
5 How it could work
The journal would have an editorial
board drawn mostly from members of SJDM, and it would include the
publication committee. The board would have an editor, associate
(action) editors, and consulting editors. The editor would
appoint the initial board with the advice and consent of the
publication committee. The board would make major editorial
decisions, including decisions about format changes, selection of
new editors (main, associate, or consulting). Policy decisions,
such as dissolution of the journal, will require approval from
the publication committee. The editor's term would be five
years, renewable, at the discretion of the publication committee.
The review process would work much like that of
Psychological Science. The editor and action editor
would both have power to reject without review. Reviews would
include reviewers recommended by the author. The general
presumption is that articles sent out for review look like they
are publishable, but reviewers may find fatal flaws or prior art,
so probably about half of these would actually be published.
This means that probably another half of submitted articles, or
more, would be rejected without review.
One successful open-access journal is the
Journal of Vision,
produced by a society much like SJDM. Many of its features are
specific to its field, and the author charges starting at
$85/page seem excessive to me right now. (They are only a
little better than PLoS, the Public Library of Science,
which charges $1,500 per article.)
Another model is
RNews. This
journal, while shorter than what I would eventually like to see,
demonstrates the technical and financial feasibility of a
lower-cost solution. It is produced entirely by volunteers, and
no money changes hands at any stage. The contributors to
RNews all submit articles in a standard LATEXformat.
We could not do that because most non-economists use word
processors. However, if contributors follow a standard format
with minimal formatting and with EPS images, the translation to
LATEXmay be almost automatic. If this does not work, we will
charge small author fees - probably less than $100 per article
- and these will be waived for those who can submit usable copy
or those from developing countries. Of course, reviews will be
done using whatever the author submits.
If there is sufficient demand, a print version of the journal,
with a nice cover, will be produced and mailed to those who
request it. The cost of this will be much less than for a
standard journal because everything is done except the printing
and the cover.
The journal will be archived by volunteers on their computers.
Librarians now regard digital archives as more durable than
paper, and I shall seek their advice about how best to do it.
Records will be kept using free software, and backed up locally.
Volunteers will be solicited for duplicate backups, as well as
for mirrors of the entire journal site.
The journal will be indexed in PsycInfo and other standard
indices. A mailing list will announce new issues, with titles
and abstracts.
Copyright issues remain to be worked out, although the
Journal of Vision seems to have a good starting point.
My inclination is for the journal to have copyright but allow
authors to excerpt articles for other publications, to include
excerpts or entire articles from published conference proceedings
unless the publications were "generally available" (i.e., books
published by a standard publisher).
We could accept advertising, especially if we needed a little
money.
6 Conclusion
As pointed out to me by Bud Fennema, we can view this as a
decision under uncertainty, but the downside risk to SJDM is
small. This journal could succeed. If it does, it will bring
prestige to SJDM, and possibly new members, and it could
encourage more research in JDM, especially by those now largely
shut out by lack of access, including those who don't even know
we exist and will discover us through Google. It could also
maintain standards for research and provide encouragement to
novel approaches. I think that the probability of success is
greater than 50%.
It could also fail. Based on discussions with someone who edited
such a journal that failed, this would mean that the journal
would attract too few high-quality submissions, despite the
editor's beating the bushes for them. In this case, the main
loss would be the editors' time spent trying to get the journal
started. (Presumably the reviewers and editors would be reading
roughly the same papers for some other journal.) In my case, the
loss would also involve my current editing positions, with which
I am quite happy except that the journals are not open-access.
I am willing to take this risk. But the important thing, I
think, is that failure would not have much cost, if any, to SJDM.
SJDM would still have the option of starting another journal in
some other way, or giving up the journal business.
If you have gotten this far and if you are willing to help,
please let me know, at
baron@psych.upenn.edu. The early stages will require
help from someone with a better sense of design than I possess,
as well as volunteers to review or serve as consulting editors.