From: Sandy Thatcher <[log in to unmask]>
Date: Sun, 5 May 2013 14:09:43 -0500

"Have you asked your university's librarians how they select?"

Well, yes, in fact I have had many such conversations. During the
entire 20 years while I was director of Penn State University Press, a
collection development librarian sat on the editorial board of the
Press. It was, indeed, that member of the board who first recommended
that detailed information about the nature and extent of revisions of
a book based on a dissertation be included as a standard part of the
materials presented to the board when this type of book was
recommended by the staff for acceptance. That got me to thinking how
impoverished, by comparison, was the information on which librarians
elsewhere were being asked to make decisions about revised
dissertations. I can understand why librarians do not want to buy
unrevised dissertations, so a vendor's providing that information
truly is helpful. It still puzzles me (and probably Alex also) what
additional information is provided by separating out books as "revised
dissertations" since no further information about the nature and
extent of revision is provided. Assumptions about how revised
dissertations compare with other books are hazardous to make at best.
Susan Okin's "Women in Western Political Thought," for example, was
among the many best-selling revised dissertations I acquired for
Princeton U.P.  It is hardly a narrowly focused monograph. The
standards that university presses use to choose revised dissertations
to publish do not differ in any way from those they use to choose
other books.

I did not rely on just the information supplied by YBP in coming to
the conclusion that a systematic bias existed against buying revised
dissertations. I did a study of all books published by Penn State
University Press over a fourteen-year period (1994-2008) in Latin
American studies to find out how sales of revised dissertations
compared with sales of other books. As reported on Liblicense on March
31, 2010, this is what I found:

> During the first part of this period, we published our books (with very few exceptions, not included in the data) simultaneously in cloth and paper, as many publishers did in that era and a few still
> do. From 1994 through 2002, a total of 29 single-authored books came out in this dual format. Of that number, 20 were revised dissertations.  The average cloth sale was 193 for all books together; 229 for books not based on dissertations; and 177 for revised dissertations [i.e., 23% less]. The average sale for the paperback editions was 1,138 for all books together; 1,284 for books not based on dissertations; and 1,071 for revised dissertations.
> During the second part of this period, as we became more aware of changes in library purchasing decisions to buy paperbacks if they were immediately available, we changed our strategy to issue the
> cloth edition first and then a paperback edition later, usually when the cloth stock had sold through completely. From 2001 through 2008, a total of 35 single-authored titles were published in cloth only to begin with. Of that number, 25 were revised dissertations. The average cloth sale was 360 for all books together; 436 for books not based on dissertations; and 330 for revised dissertations [i.e., 24% less]. I did not calculate the average sales for paperback editions because they had not been issued for all of the books in this sample yet, but it is clear from partial data that the average for lifetime sales will be way below what it was during the earlier period, more like 300 to 400 copies.
> There is a lot of consistency in these data. In both periods, about 70% of the books we published originated from dissertations. One may safely conclude that it is possible to build a very distinguished list even though it is weighted heavily with revised dissertations. (I have always believed that a good many of the best books published in this field come out of dissertations. I would not want to claim
> that this is true for other fields.)
> Despite the fact that ETDs [Electronic Theses and Dissertations, made available through the NDLTD] and ProQuest's database of theses in electronic form only became a factor in the later period, the relationship between sales of books not based on dissertations and of books originating from dissertations is remarkably similar between the two periods analyzed here. This finding suggests to me that the effect of any bias against revised dissertations based on the availability of ETDs and ProQuest's database, it it has any influence at all, is outweighed by other factors involved in library decisions about purchases.
> As some librarians have suggested, there is a general perception that books based on dissertations are more "specialized," narrower in scope, etc., and thus these books are ordered in smaller numbers.
> My guess is that this hypothesis offers a better explanation of the data I have presented here than any other. I would dispute that characterization of revised dissertations as a generalization, at least for the books we publish in this field, but that's another story.

As for "market research," I did recommend to the AAUP, which for many
years tracked sales of books in different categories (such as
translations), that it add a category to track sales specifically of
revised dissertations, so that we could gather a wider range of
longitudinal data to show whether or not Penn State's experience with
this one field was an outlier or not.  That recommendation, alas, was
never acted upon, so we simply don't have the data to confirm or
disconfirm the hypothesis in any reliable way.

However, I believe it is still true that many editors at university
presses continue to believe that libraries, for whatever reason, order
fewer revised dissertations than other books and that sales are
therefore lower, on average, than for other books.  In the absence of
hard data, perception becomes reality and the unfortunate impact on
junior scholars follows.  In my view, a world with far fewer revised
dissertations becoming books is a world with far less rich

Sandy Thatcher

> From: Reeta Sinha <[log in to unmask]>
> Date: Fri, 3 May 2013 00:51:10 -0700
> "You're right about the thread getting a bit old, but I hope you and
> others will excuse my extending it just a little bit further.  Perhaps
> I missed it, but nowhere in the discussion do I recall seeing explicit
> reasons why librarians want to know whether or not a book is a revised
> dissertation in order to make a decision on whether to add it to a
> collection."
> I think Rick covered this in his last post--but, I'm going to give it
> another shot--
> Speaking as a collection development librarian, but not for all, I
> don't believe librarians are focused on 'revised dissertations.' In
> general, we who select titles for our library collections do not
> obsess over one non-subject parameter.
> What's been lost, or obfuscated by some, from the start, is that
> 'revised dissertation' was not plucked out of thin air--it is on a
> list that a vendor uses. Assumptions made subsequently had little to
> do with the purpose of that list, or how such parameters are used by
> librarians.
> Michael Zeoli of YBP provided this a few rounds ago:
> "The designation of 'Revised Dissertation' conveys a sense that the
> treatment of the subject will likely be in depth.  This is supported -
> or not - by other profiling information such as readership level and
> 'select category'."
> Many CD librarians--selectors, remember examining the book in hand on
> a weekly basis--approval books. Now, we look at records on our screen,
> if we still select. We may have a book jacket image (pretty, but we're
> not supposed to judge a book by its cover) and the table of contents
> (ok, a little more information about the...contents). If I'm reviewing
> 300 titles with medical aspects, released each week (which I currently
> am), I need a little more than those 2 elements, and broad subject
> headings.
> Non-subject parameters such as 'anthology,' 'handbook,'
> 'revised/unrevised dissertations,' the 'audience' level, or notes
> about the book--if it includes previously published articles, whether
> some authors are affiliated with my institution, or if a book has a
> geographic focus of Australia (nothing wrong with that but with a
> limited budget, maybe that's not where I can spend my money)....etc.
> 'Revised dissertation' is just one aspect, one.
> I started in libraries 'reading' medical books...50-75 books that came
> on 'approval.' I read the front cover, the back, the jacket text, and
> through the 1st page or so of the 1st chapter. Why? To see if the book
> was based on a conference that took place 6 years earlier, or had
> content that had been published elsewhere, in part of whole, or, if
> the book was based on someone's doctoral dissertation. Yes, I'd look
> for the word 'revised' because if it didn't appear, I might be
> duplicating content my users already had access to--an unrevised
> dissertation available elsewhere. Did I 'measure' how much of the
> dissertation had been revised? Of course not--'revised dissertation,'
> 'based on', 'originated as' -- this was additional information, not a
> deal-breaker.
> I would also skim chapters, sometimes because the
> preface/TOC/foreword/jacket text didn't make it easy for me to
> determine whether the title was relevant for my library's users; other
> times I'd skim to narrow down which of the 10 molecular biology books
> profiled for my library that week I should keep.
> 27 years later, I don't have the book in hand; hardly any librarians,
> or library staff, I know of do what I did then. That's why we (I) rely
> on vendors/profilers to do the reading/skimming, so I don't have to
> guess, so I can narrow down a list of 300 medical titles every week
> down to the handful I can afford to purchase, that might be in-scope
> based on my institution's needs.
>>  Do some libraries really only collect the work of senior scholars
> in particular fields?
> With just a touch of irritation, I'm going to go out on a limb and
> say, of course not.
> This is why Rick's last post was necessary.
>>  I've never gone out of my way to conceal that a book we publish is a
>>  revised dissertation (I can't imagine publishing an unrevised one),
>>  though I confess we don't shout it out either.  Understanding why
>>  librarians care about this status would be helpful.
> I have a question for the two university press voices heard in this
> thread....I presume there's a 'university library' near the university
> press? Have you asked your university's librarians how they select?
> Perhaps it could be considered 'market research.'
>> From Rick Anderson:
>>  (And let me emphasize that I'm speaking here purely as a librarian, and
>>  not on behalf of any vendor.)
> I AM going to give a shout-out to the library 'book' vendors I've
> worked with over the decades, as a CD assistant and a CD librarian.
> They asked, understood, how we made our selections, and continue to
> ask what might make the selection process more efficient--the YBPs,
> Majors, Harrassowitz', Ballens, Blackwells, and many more. Their
> services and efforts made my collection development life easier. If
> all these years I only had publisher catalogs/blurbs and the
> incomplete information provided within to rely on...
> I think this bears repeating - the context skipped over in the original posts:
>>  Libraries set up approval plans for two reasons. They are designed, first,
>>  to include books that are obviously needed (thus freeing up librarians'
>>  scarce time so they can spend it seeking out titles that are more obscure
>>  or harder to get) and, second, to exclude books that, though they may be
>>  of high quality, are not a good fit for the library's particular needs
>>  (thus letting the library focus its scarce budget dollars on the
>>  acquisition of books that it needs more urgently). If libraries had
>>  infinite budgets and infinite staff time, approval plans would not be
>>  needed. Sadly, we have neither, so books have to be excluded as well as
>>  included.
> Reeta Sinha, MPH, MSLS
> Resource Management Librarian