This is the report of my findings from the conspectus study, conducted from 1996 to 1999. A technical report of my methods is also available online. A hard copy of these reports, as well as of the reports for all the divisions, will be made available on reserve in the Library in the near future. The web version should be considered authoritative.
I wish to thank all the people who made this report possible with their hard, often tedious, work, as well as the positive feedback and constructive criticism they gave me.
- Paul G. Streby
Collection Development Librarian
February 18, 1999
Table of ContentsOverview of the conspectus process
--1990 accreditation report
Overview of the conspectus process
The object of the assessment was to determine the collection's strengths
and weaknesses, which will enable the library and teaching faculty to plan
its growth in a more informed way. The project involved most of the
librarians, in many cases acting in consultation with the departments for
which they serve as liaisons.
Forthe assessment, the Thompson Library used the "conspectus" approach of WLN (formerly the Western Library Network, recently merged with OCLC), published in the WLN Collection Assessment Manual, 4th ed. (1992). (A fifth edition, Using the Conspectus Method: A Collection Assessment Handbook, came out in late 1997, in the middle of the conspectus project. Unless otherwise noted, standards and citations are taken from the 1992 edition, hereinafter referred to as CAM). The conspectus assigns each item in the Library's collection to one of twenty-four subject-based divisions, which are, in turn, divided into categories and, optionally, categories into subjects. Each division is assigned three ratings: one for the current level of strength, one for the current acquisitions commitment, and one for the goal level. The ratings cover the breadth, depth, currency, and accessibility of the collection, and are measured by a mix of qualitative and quantitative methods. Their definitions are summarized at right:
Most of the divisions were assigned a goal level rating of 3b, because they support undergraduate majors. The divisions that support master's programs received the rating of 3c; divisions that do not directly or indirectly support academic programs received ratings between 0 and 3a.
The following criteria were used in assigning ratings for the current collection strength:
- The number of titles in a division. This is based upon a formula that takes into account the annual domestic publishing output for each division.
- The percentage of books in standard bibliographies.
- The percentage of subscriptions to periodicals indexed in standard sources.
- Special formats, media, and collections available at the Library.
- The language of the materials.
- The average age of the titles in the division.
- The acquisition rate.
- The readership level of books and periodicals in the division.
In addition to using these criteria, some of the librarians consulted instructors for their opinions regarding the strength of the division in question.
The conspectus method was selected for several reasons:
- It gives the Thompson Library's administrators and librarians reliable information for assessing strengths and weaknesses in the collection. This in turn allows for more systematic and informed development of the collection over the next several years.
- It provides information to respond intelligently and effectively to shifting budget priorities.
- Finally, it allows for objective comparison with the collections of other libraries that have used the WLN conspectus.
One potential objection to the conspectus approach is that it does not
take into account the allocation formula used in devising the library
budgets for the academic departments. This is actually an advantage; it
provides a fresh vantage point for assessing how well the collection is
meeting a widespread standard, and how well the allocation formula is
enhancing collection development. It may point to a need to review the
The WLN divisions are arranged around the Library of Congress call number
system (LCCN). Some academic subjects supported by the Thompson Library's
collection, e.g. public administration and women's & gender
studies, could not be studied adequately within the framework of the WLN
conspectus, because they are interdisciplinary in nature. The LCCN does
not lend itself to cross-disciplinary subjects, for such fields tend to be
scattered across the call numbers.
A separate but related problem was that the divisions are broken into categories that do not always accurately reflect the ways in which academicians subdivide their fields. For example, within the Political Science Division, there are no separate categories for political theory or comparative government, even though these are central subjects within the discipline. Arguably, these topics are implicit in the WLN categories, but they are too important and intellectually distinct to be subsumed under a more general category. To lump them under a broader category would make the results significantly harder for people, especially outsiders, to interpret or draw meaningful conclusions.
To resolve these problems, several non-WLN categories were devised
within existing divisions to better reflect the nature of those
disciplines. There were also two separate non-WLN divisions: Public
Administration and Women & Gender Studies. Future analysis of the
collection will probably necessitate creating non-WLN divisions for
African-American Studies, American Culture, and Communication.
The government documents collection was not studied as part of the
conspectus. Although it is an important source of information in the
Library, it was not included in the study for two major reasons.
- It is not arranged in a manner parallel to the rest of the collection. Materials are grouped according to the government office that produced them rather than by subject. This makes it hard to correlate holdings in government documents to the arrangement of the conspectus.
- Government documents are less accessible in practice than the rest of the Library's holdings, primarily because they are catalogued separately.
Internet resourcesThe World Wide Web is an increasingly important source for information, and the Thompson Library provides access to it and has organized access to some of its resources through its homepage. In assessing the collection, it was decided, with some reservations, to leave Internet resources out of the equation, for the following reasons:
- Printed materials still enjoy primacy in academia. Because of its novelty, the Web has not gained the acceptance among students or faculty that has long been enjoyed by printed materials. Some instructors discourage or even forbid their students from using Internet bibliographic sources. Until Web resources gain a level of acceptance close to that of printed materials, it seems unwise to consider them using the same sets of standards for a traditional collection.
- There are many ambiguities and uncertainties in measuring a Web-based
collection. Because it is still so novel, formal and informal standards
for measuring the viability of a Web collection are still
- There are numerous and widely agreed-upon rules of thumb for determining the quality of a new monograph. Although some rules of thumb exist for assessing quality in a website, these rules are themselves new and relatively untested.
- Standard bibliographies, which are a major tool for rating the collection, either do not exist for the Web, or are so quickly outdated that they are reliable for only a very short time.
- The universe of materials available on the Web is extremely difficult to quantify. A "metasite" may provide links to 100 other sites, which in turn link to still more sites. There is no good answer to the question of whether this metasite should be counted as one resource, as 101, or as 1,001. Additionally, the number of websites is growing at a phenomenal rate, with many if not most sites receiving little or no public notice. There is therefore no real benchmark for determining what constitutes an adequate-sized collection.
- The reliability of materials available on the Web is extremely uneven.
Although it might be objected that with the advent of the Web, countless materials are now available to library users, this argument is essentially hollow. By such logic, the Library's collection has always been superlative, because students always had the option of going to other libraries or ordering materials from bookstores. The fallacy of this statement is fairly obvious. The conspectus is measuring what is available here, not what is hypothetically available elsewhere.
It might be objected that as part of the University of Michigan, the
Thompson Library benefits from the holdings of the Ann Arbor campus. This
is true but irrelevant. The conspectus is a measure of the collection
available locally. For many if not most students, it is
inconvenient at best to make a trip to Ann Arbor for books needed for
research. Interlibrary loan is available, but it is not any easier or
quicker to retrieve a book from the Ann Arbor campus than elsewhere.
The Thompson Library does indeed have reciprocal borrowing arrangements with the libraries of Mott Community College, Baker College of Flint, and Kettering University. However, the number of books that can be borrowed is limited, and the Thompson Library does not share an online catalog with these institutions. Furthermore, there are no agreements for cooperative collection development.
Finally, it is very arguably unethical to rely on the holdings of other
institutions to serve the needs of UM-Flint students, faculty, and staff.
Students might legitimately wonder what their tuition money pays for if
they routinely must go to another library to conduct ordinary research for
1990 accreditation report
In its report, the accreditation team in 1990 listed among its concerns:
The library facilities are a major area of concern. This concern
encompasses the collection (breadth and depth), space, and accessibility.
The growth in graduate programs requires more and better access to
relevant periodicals and references. As new programs have been added, the
library holdings have not kept pace. ... ("Report of a Visit to University
of Michigan-Flint," pp. 27-28).
UM-Flint is going to be evaluated for accreditation from the North
Central Association again in 1999. While progress has been made in many
areas, (e.g. moving to a facility of adequate size, automating
circulation and adding MIRLYN), the Library's slow and inadequate
collection growth could be a liability.
Overview of results
The Thompson Library appears to have, on the whole, a solid, albeit
slender, core collection of books, and a relatively impoverished
periodicals collection. The book collection is aging and badly in need of
both pruning and new growth in order to build on the core already in
place. More collection development tools (such as bibliographies) are
needed to help the librarians and UM-Flint teaching faculty add to the
collection in a more systematic manner. Moreover, the lack of a viable
collection development policy has probably hampered rational collection
WLN publishes the following standard for periodicals collections:
|1b||Some general periodicals + Reader's Guide and/or other major general indexes|
|2a||Some general periodicals + Reader's Guide and/or other major general indexes|
|2b||2a + wider selection of general periodicals + 30% or more of titles indexed in the appropriate Wilson subject index + access to that index|
|3a||50% or more of titles indexed in the appropriate Wilson subject index + access to that index|
|3b||75% or more of titles indexed in the appropriate Wilson subject index + access to that index + wide range of basic serials + access to non-bibliographic databases|
|3c||3b + 90% of titles indexed in the appropriate Wilson subject index + access to the major indexing & abstracting services in the field|
CAM, p. 69
The periodicals collection is inadequate. For the collection as a whole, the Library should attain a collecting level of at least 3b. Of course, individual divisions might call for a higher or lower periodicals rating. Still, once the variations are balanced out, an overall periodicals rating of 3b seems appropriate.
Of the 2,222 Wilson-indexed periodicals surveyed, the Library has ongoing print and/or microfilm subscriptions to 673 (30%). This, in itself, amounts to a conspectus rating of 2b for the collection as a whole. This does not include full-text access through the Web databases. Once full-text access through the Web is included, the number of subscriptions to Wilson-indexed periodicals rises to 1184, bringing the percentage up to 53%.
Although nominally this is within the 3a range, WLN does not regard electronic subscriptions as adequate substitutes for print resources unless, inter alia, they contain "graphics, charts, and other features" (Using the Conspectus Method, p.73). With the partial exception of ProQuest Direct, the full-text Web access that the Thompson Library holds is text-only. For many periodicals, this does not matter, but for some, e.g. arts periodicals, text-only access is no substitute for access in print. Moreover, quality control in the databases is uneven, and students appear, on the whole, less familiar and sometimes less comfortable using electronic full-text.
There are other problems with the Library's periodicals collection:
- There are not enough periodical subscriptions for an academic library. As of early 1998, the Thompson Library subscribed to 1,185 journals. A comparison with the libraries of peer institutions shows that this is greatly below the average of 2,254 journal subscriptions per library. (See Appendix A).
- There is inadequate coordination of points of access. This is mainly due to the novelty of electronic access, but it still needs to be addressed.
- A central journals list is needed. Patrons often have to check two or more lists to determine if a periodical is available at the Library.
- There are no MIRLYN records for electronic-access journals. Although creating them would require a substantial amount of effort both initially and over time, it is crucial to provide patrons with the information they need to most effectively use the library resources that they are paying for with their tuition.
- The Library has no control over the journals available through FirstSearch, SearchBank, ProQuest Direct, and Lexis-Nexis. Access to these periodicals can be cancelled by their owners or by the vendors. Numerous full-text periodicals have in fact been cancelled and new ones added. This creates an element of instability to the Library's access to electronic journals.
FundingAccording to WLN standards, for a rating of 3b, which is adequate to support an undergraduate collection, a library should collect 15-20% of the total number of academic books published each year. Of course, materials for some divisions should be collected at a higher or lower level, depending on what sort of academic programs they support. Nevertheless, a 3b rating seems like an appropriate goal for the collection as a whole. By this standard, the Library's budget has not grown sufficiently in many years.
According to the Bowker Annual Library and Book Trade Almanac, from 1991 to 1995 there were 307,431 academic books published or otherwise made available in North America. To attain a 3b acquisitions rate, the Library should have added between 46,115 and 61,486 titles to its collection between 1991 and 1995. During this period, however, the Library ordered 33,267 titles (11% of the total) and catalogued 31,315 (10%).
The trend seems to be worsening. In 1996, the most recent year for which statistics are available, 74,262 academic books were published or made available in North America. That year, the Library ordered 4361 monograph titles and catalogued 4636. These amount to, respectively, 5.9% and 6.2% of the total for the year. In 1997, the Library ordered 4,275 titles and catalogued 3,412. These amount to 5.8% or 4.6%, respectively, of the total for 1996. In 1998, the Library ordered 3857 titles (5.2%) and catalogued 4629 (6.2%). The actual percentage of titles acquired in these years is probably even lower, because the publishing output has risen every year since 1991.
The chart below shows the serious falling-off of the already inadequate growth rate (as noted above in the 1990 accreditation report).
|The Library's 1997-1998 budget had $117,944 in its free balance (i.e. the amount allocated to academic departments for books, and not carried over from the previous year), plus $25,000 allocated (and not carried over) in funds for reference and individual librarians, for a total of $142,944 for monograph acquisitions. For each division, both the actual free balance and a "weighted total" were factored in. The weighted total represents the free balance plus a proportional share of money from the budgets for reference and the librarians. See the table below for a more detailed summary of these data.||The figure for the 1998-99 budget is $107,758 in the free balance, plus $35,608 for librarians and reference, for a total of $143,366. This is an increase of a mere $422. Money carried over from previous budgets was not included, because presumably some money from the current budget will be carried over to the next year. Thus, the carry-over amounts balance each other out.|
Based on the 1996 publishing figure, the Library should add between 11,139 and 14,852 titles each year to collect at the 3b level. At an average cost of $49.86 per title, this would cost between $555,391 and $740,520 for book acquisitions alone. (This does not include periodicals, electronic databases or other products, or processing costs). The current budget falls far short of being able to attain this goal.
Because this figure is so much higher than the free balance, a 10% acquisition rate might be a more attainable immediate goal for the collection as a whole. Some divisions would, of course, require a higher percentage (for graduate programs), or a lower percentage (for disciplines not taught at UM-Flint). Any departments that already exceed 10% should at least maintain their existing level of funding, inasmuch as no department (except Education) currently meets WLN standards for acquisitions.
As noted above, a grand total of 74,262 titles were published or made available in 1996. An acquisitions rate of 10% would entail having added about 7426 titles at a total cost of $370,260 for monographs alone (i.e. $227,316, or 159%, more than the actual monographs allocation).
At a 10% acquisitions rate of new materials, there is no danger of running out of space. The Thompson Library building has enough shelf space to hold 350,000 volumes, and there are currently only about 225,000 bound volumes (including periodicals). The extra space, combined with a program of regular and prudent withdrawals of outdated materials (standard practice in most academic libraries), should allow for ample growth of the collection well into the 2020s.
|Monograph Acquisitions: Actual & Recommended, 1997-98|
|Division||Book allocation||Weighted allocation||Est. 1997 price per vol.||Est. no. vols. buyable||Avg. annual publishing output||Est. % of output acquired||AC||Min. allocation to buy 10% of publishing output|
|Art & Architecture||$3577||$4335||$47.90||91||2162||4%||2a||$10,346|
|Business & Economics
|Engineering & Technology||$2910||$3527||$65.63||54||5580||1%||2a||$36,521|
|Geography & Earth Science||$3261||$3952||$53.67||74||733||10%||2b/3a||$3952|
|History & Auxiliary Sciences||$7970||$9659||$29.19||331||5773||6%||2b||$16,843|
|Language, Linguistics, & Literature
includes Health Care, Medical Technology, Nursing, & Physical Therapy
|Philosophy & Religion||$6643||$8051||$31.70||254||3892||7%||2b||$12,331|
|Physical Education & Recreation||-||-||$27.81||-||146||-||3a||$779|
includes Sociology, Criminal Justice, & Social Work
* To preserve current level of acquisitions
** The Departments of Sociology and Anthropology share a budget; the
book allocation is divided based on enrollment. See the reports for Sociology and Anthropology
Summary of ratingsBelow is a summary of the conspectus ratings. It shows figures only at the division level; figures for the individual categories are contained in the division reports on the Library's homepage. (See above for an explanation of the ratings).
|Division||Current Level||Acquisitions Commitment||Goal Level|
|Art & Architecture||2b||2b||3b|
|Business & Economics||3a||2a||3b/3c|
|Engineering & Technology||2a||2a||2b/3b|
|Geography & Earth Science||2a||2b/3a||3b|
|History & Auxiliary Sciences||3a||2b||3b/3c|
|Language, Linguistics, & Literature||3a||2b||3b/3c|
|Philosophy & Religion||3a||2b||3b|
|Physical Education & Recreation||2b||3a*||2b|
|Women's & Gender Studies||3b||-||3b|
* Has no departmental allocation; based on acquisitions rate
** Based on shared budget
|Select Data Comparison with Peer Institutions|
|Drake University (Des Moines, IA)||3876||475,912||122.78||2955|
|La Sierra University (Riverside, CA)||1195||225,311||188.55||1422|
|Middle Tennessee State U.||12,091||604,017||49.96||3507|
|Pepperdine University (Malibu, CA)||5890||339,460||57.63||3186|
|University of San Francisco||6808||603,341||88.62||2381|
|University of Texas-Dallas||3312||554,547||167.44||3811|
|Northern Michigan University||7179||511,404||71.24||1932|
|University of Wisconsin-Oshkosh||10,382||400,000||38.53||1790|
|University of Southern Indiana||6280||196,419||31.28||1577|
|Moorhead State University||5610||367,334||65.48||1601|
|University of Wisconsin-Whitewater||9719||352,500||36.30||2036|
|University of Minnesota-Duluth||7497||457,000||60.96||2676|
|SUNY at New Paltz||6958||470,000||67.55||NA|
|University of Wisconsin-La Crosse||8659||356,277||41.15||1866|
* Estimated; part-time students counted as 0.5 FTE.
Sources: Internal UM-Flint data, College Blue Book, 1997, and
American Library Directory, 1997-98.