Electronic Journal of Academic and Special Librarianship

v.8 no.3 (Winter 2007)

Back to Contents

Scrutinizing Cataloging Copy

Julie Banks, Professor
Kent Library, Southeast Missouri State University
jbanks@semo.edu

Abstract

In order to determine if cataloging copy today is as acceptable as past copy, the researcher compared additions and corrections to a random sample of monographic copy that corresponded to the books that went through the cataloging department during 2006 at Southeast Missouri State University with previously reported editing work from several libraries.  The outcome was that more editing was required today, but it was not possible to identify whether this was due to the rapidly changing cataloging environment or to the higher level of editing performed.

Introduction

Changes are roiling through the cataloging community and are impacting procedures and personnel. This hot topic was spotlighted in a discussion about the level of cataloging/ material format and staffing level at the 2007 ALA annual meeting.  There is now even a blog about the future of cataloging, and library literature addresses many issues related to the current state of cataloging, including the Library of Congress (LC) no longer providing authority records for series, the cross-training of personnel, significant vs. insignificant changes to records, the acceptability of records, and the upswing in clerical staff editing catalog copy and even creating original records.  With all these changes, are cataloging records still as acceptable today as they were in the past?  In order to test this, the investigator randomly selected monographic copy not produced by LC, which came through the catalog department with the corresponding books at Southeast Missouri State University (SEMO) during 2006 to determine what changes and additions were made to the catalog record and to compare these with the changes and additions previously reported in studies dating from the 1970s.

Literature Review

As indicated above, library literature is replete with articles dealing with the state of cataloging and who does it, including cataloging paraprofessionals, acquisitions personnel, and reference staff.  Indiana University’s A White Paper on the Future of Cataloging at Indiana University contained a section on the cataloging staff and asserted that job skills and assignments would expand as the “landscape of collecting and managing information becomes more complex” (Byrd et al., 2006, p. 14). The paper posited that professional catalogers would take on more managerial and training roles, while the descriptive cataloging would move to the paraprofessionals.  Mary M. Rider agreed with this and then highlighted the paraprofessionals’ new roles when she wrote about their ability to catalog from 70 to 90% of the records with little editing (Rider, 1996, p. 29).  These staff had also been assigned more complex responsibilities such as developing original records and assigning call numbers and subject headings.   Marcum, Associate Librarian of Congress, discussed the move of much descriptive cataloging work from professionals to technicians when she delivered her address to the Ebsco Leadership Seminar (Marcum, 2004, p. 11).   Neumeister (Neumeister, 1997, p. 62) and Freeborn and Mugridge (Freeborn and Mugridge, 2002, pp. 35-43) at the State University of New York at Buffalo and Penn State University, respectively, discussed the reassignment of editing some of the cataloging copy to acquisitions personnel and the problems and benefits.  Fain, Brown, and Faix described the new role reference librarians at Coastal Carolina University’s Kimbel Library have accepted in working with copy cataloging, call numbers, 856s, and in doing other traditional tasks (Fain, Brown, and Faix, 2004, pp. 41-51). 

Besides identifying who is editing catalog copy and working on related activities in the cataloging arena, authors have considered topics that addressed the changes and/or additions to bibliographic records.   One article by Leroy and Thomas noted the problem of cutting and pasting data in a cataloging web environment (Leroy and Thomas, 2004, p. 15) and suggested that it is still useful to carefully check for errors (Leroy and Thomas, 2004, p. 15).  Beall’s and Kafadar’s article evaluated the elimination of typographical errors in shared records and found that a large number had not been corrected (Beall and Kafadar, 2004, p. 97).

Three articles looked at various fields and the amount of editing that was required to bring the record up to the appropriate rule standardsThough the studies were not exactly the same as the current study, they did consider the amount of editing required for various fields at the different times and places and therefore provided the base line for understanding the data from SEMO.  The earliest article was by Ryans, who reported in 1978 that copy cataloging was of high quality.  Indeed, 60% of the first 700 OCLC records that Kent State University attached its symbol to after September, 1975, “were correct and could be accepted with no changes” (Ryans, 1978, p. 128).  One hundred sixty records had just one error. Of those records that did need editing, many had only typing or spelling errors, which were deemed minor.  Major problems were found primarily in: (a) the subject fields (31% of the records) when there was the lack of a subject heading or the heading was incorrect or incomplete, (b) the physical description, especially when the size of the book was not provided (25% of the records), and (c) the main and added entry fields when an incomplete entry, incorrect entry, incorrect or omitted dates, and misspelled name(s) (Ryans, 1978, pp. 128-129) were in the records (16% each).  Fifty-five or 19% of the records contained incomplete or incorrect series.  No 08x, 09x, or 5xx fields were considered.

Hudson (Hudson, 1981, pp. 116-120) replicated Ryans’ study by examining 1017 monographic records SUNY-Albany attached its holding symbol to from November, 1979, to July, 1980.  Unlike Ryans, Hudson did not consider the treatment of series but did note their typographical errors.  She also studied changes and/or additions to the main and added entry fields, corrections and/or additions to the imprint and collation, and changes to make subject headings consistent with LC headings or to provide better access through additional subject headings and/or subdivisions.  The findings were that 43% of the records required revisions (Hudson, 1981, p. 118).  The subject heading fields needed the most frequent work, affecting 22.4% of the records, followed by the added entry fields at 11.7% of the records, the title field at 7.5%, imprint at 6.3%, and collation at 5.7% (Hudson, 1981, p. 118).  Hudson included a table that compared her findings by category with those of Ryans, which showed that the overall percentage of records needing editing was close.  Thus, Hudson concluded that the quality of member copy had not improved since 1975 though she commented that fewer records input in 1979 and 1980 needed correction, which suggested “a trend toward more careful cataloging and formatting” (Hudson, 1981, p. 119).  She ended her report though by stating that “Such a high percentage of records needing revision necessitates the review of all member records used if a library wishes to adhere to OCLC standards for cataloging” (Hudson, 1981, p. 119).

In 1991, McCue, Weiss, and Wilson took a slightly different tact from the previous two studies when they compared catalog records developed by LC with those developed by nine Research Libraries Information Network (RLIN) libraries. Using 80 records, they focused on typographical errors, enhancements by field and by libraries, tagging changes/updates/modifications, and authority-controlled fields.  Some of their findings were that:  a) 56 of the records had typographical errors, which were mostly punctuation errors with the most being found in the 2xx fields, followed by the 300 field, the 500 field, and the 400 field (McCue, Weiss, and Wilson, 1991, p. 68);  b) subject headings required more enhancements (additions, deletions, or replacements) than any of the other fields, followed by the 500 field and 260 field (McCue, Weiss, and Wilson, 1991, p. 69);  c) the 7xx fields required the most changes due to authority control (McCue, Weiss, and Wilson, 1991, p. 74); and d) more tagging changes were needed in the 245 field because of the first indicator (McCue, Weiss, and Wilson, 1991, p. 72), which was not critical for record retrieval.  However, the tagging problems found in the 4xx fields were retrieval problems.  MCCue, Weiss, and Wilson concluded by noting that there had been a shift in copy cataloging from professional librarians to support staff since the 1980s and by stating that there was no significant difference in enhancements, typographical errors, and other non-authority changes in the recordsproduced by LC and the 9 best RLIN members (McCue, Weiss, and Wilson, 1991, p. 74).  In terms of authority control, the authors called for an increase in cooperative authority work.    

Results and Discussion

In order to determine if the cataloging copy of today had more or less changes/additions than in the past when more professional librarians were developing and working with bibliographic records, the investigator randomly drew 397, non-DLC monographic records that passed through the SEMO cataloging department with the corresponding monographs in 2006 for a more than adequate sample size.  Ninety-three were supplied by UKM, NLC, or NLM and/or had pcc, ukblsr, nlmcopyc, or lcd authentication.  It was assumed that this copy would require little consideration.  However, 67 records or approximately 72% needed editing.  Of the copy that required changes or additions, 2 of the records had misspelled words that necessitated corrections, 19 lacked a Dewey Decimal Classification (DDC) number that had to be developed, and 30 included series statements that compelled authority work.  Changes or additions, including the development of data for a new field, to other authority-controlled fields were not considered in this study except subject headings when the record did not have an LC heading because the investigator did not have access to that information. 

The other records considered in this study were input by various OCLC member libraries and had a wide variety of encoding levels.  Table 1 shows the level of encoding as defined in the 3rd edition of OCLC’s Bibliographic Formats and Standards (pp. FF-39-40) and the number of records corresponding to each level for this group, equaling 304.

Table 1
Encoding Levels and Number of Corresponding Records in the Sample

I

M

4

K

7

L

J

264

7

12

16

1

3

1

Though the majority of all the records considered were full level cataloging, the researcher found that 185, or almost one half of them, needed a Dewey number because no DDC was provided on the copy, the number was not fully developed, or it was not considered appropriate.  Indeed, more than one quarter of all the records did not have any suggested Dewey number at all, which was surprising because of the high number of full-level copy.  Thus, more records in this sample required work on the class number than any other addition/correction.  Though the McCue, Weiss, and Wilson article talked about changing 57.4% of the classification numbers (McCue, Weiss, and Wilson, 1991, p. 73), it was not clear if this included the actual development of the numbers from scratch.  Therefore, there was no way to verify if there had been any change in the frequency of work required or the depth of action.   If, as some literature asserts, this is one of the more professionally-oriented activities because it requires subject knowledge, then the large number of class numbers that needed development in this study suggests that at least some catalog copy continues to require careful editing by trained catalogers, who can correctly formulate the call number.

The next highest level of work was to the 260 (Imprint) field.  Graph 1 shows that delimiter (|) a required the most editing, followed by |c, and then |b. 

Graph 1
Additions and Changes to Data by Delimiter in the Imprint (260) Field

graph 1

As expected, the changes and additions were numerous and varied.  One example for |a was a “change of place within the same country between printings of the same edition” (OCLC, p. 57).  For |c, an example was the high frequency of the addition of the copyright symbol when required to reflect a copyright date or its deletion when included but the date of publication was the only date available.  The number of corrections for SEMO in this field at almost 37% was much higher than those reported by Ryans for Kent State and Hudson for SUNY-Albany at 4.4% and 6.3%, respectively.  For the most part, patrons would have been able to find the desired material without the additions/corrections, but when searching for books in the online catalog, the changed information provided in this field would allow researchers to verify that the book was by the correct publisher and was of a desired publication date.

The 4xx and 5xx fields were fairly equal in the number of appraisals made at 24.4% and 23.6%, respectively.  For series, Ryans and Hudson reported that their libraries edited 7.9% and 0.3% series, respectively.  Thus, SEMO once again did more evaluation, editing, addition work, and authority work than the other 2 libraries for series; but because LC no longer provides series authority records, additional work is to be expected.  The lack of LC authority records was probably not a problem when Kent State and SUNY-Albany published their research.  For the notes fields, the addition of page numbers as part of the 504 field is an example of work that was performed at SEMO.  Notes were not considered in the Ryans and Hudson studies, but McCue, Weiss, and Wilson discussed the frequency of typographical errors that occurred in the 500 fields and different enhancements made to the record, explaining that there was no way to categorize the additions.

Next in level of activity was the 300 (Physical Description) field.  Two records did not have any 300 field at all and 2 additional records had no meaningful information in this field.  Of the records that had problems, |a saw the greatest activity, followed by |b, and finally |c, as shown by Graph 2. 

Graph 2
Additions and Changes to Data by Delimiter in the Physical Description (300) Field

graph 2 

Examples for |a editing were the addition of the introductory pages because they were often not included, a change in the number of pages because they were not correctly input, and the addition of the number of plates.  An example of a problem with |b was that the record often did not indicate that the monograph had illustrations.  For |c, sometimes the copy did not show the book size.  Comparing the amount of editing with that reported by Ryans and Hudson, the researcher notes that it was not much higher at 19% than Ryans’ report at 15.9%.  Hudson indicated work on 5.7% of the records.  Though problems with this field would not prevent a researcher from locating desired material, correct information provided to patrons is the goal of cataloging.

In terms of subject headings, 6 were either developed or verified so that the records would have at least one DLC subject heading.  This was much less than that reported by Ryans and Hudson at 12.6% and 22.4%, respectively.  The development of subject headings, like the development of class numbers, though is a professional activity because appropriate subject headings provide access to desired material and because subject knowledge is needed for the development of appropriate headings.

The other fields required less editing, see Table 2, but it is important to note that there were 55 problems with punctuation, 23 indicator changes, 7 spelling corrections, and 6 delimiter revisions.  Though most of these would not have prevented a library patron from locating material, it was important to correct them.

Table 2
Fields and Number of Additions/Corrections to Records

010

020

100

245

246

250

600

700

12

9

2

48

16

9

6

17

Conclusion

In comparing the findings of this study with those reported earlier in the literature, it appears that more editing was done at SEMO than at the other libraries when each field was compared to the same field.  This may be due in part to the decision of the managers about what should be edited.  For example, in Local Cataloging Policies in ARL Libraries, some of the libraries identified the fields that were to be carefully examined and those that would pass without change.  More editing at SEMO may also have been due to the fact that the copy was not as good as that previously reported for whatever reason, keeping in mind that some ALA-accredited library schools do not require a cataloging course for graduation and some of the cataloging responsibilities at some libraries have been moved to clerical staff.  Thus, there does not appear to be a way to verify whether the copy was better or worse because of all the changes in cataloging procedure and verification that have occurred since the previous studies were completed or because of each library’s rules of operation.   Though many of the additions and corrections noted in this study were not crucial for locating material, the fact that so many changes were made suggests the need for careful editing.  When one considers that almost one half of the records required work on the call number, it is clear that there is a true need for catalogers with their training, experience, and subject knowledge.

References

Beall, Jeffrey, and Karen Kafadar.  “The Effectiveness of Copy Cataloging at Eliminating Typographical Errors in Shared Bibliographic Records.”  Library Resources & Technical Services 48, no. 2 (2004): 92-101

Byrd, Jackie, et al., A White Paper on the Future of Cataloging at Indiana University  (January 15, 2006). <http://www.iub.edu/~libtserv/pub/Future_of_Cataloging_White_Paper.pdf>

Fain, Margaret, Micheline Brown, and Allison Faix.  “Cross-Training Reference Librarians to Catalog.”  Technical Services  Quarterly 22, no. 1 (2004): 41-51.

Freeborn, Robert B., and Rebecca L. Mugridge.  “The Reorganization of Monographic Cataloging Processes at Penn State University Libraries.” Library Collections, Acquisitions, & Technical Services 26 (2002): 35-45.

Hudson, Judith. “Revisions to Contributed Cataloging in a Cooperative Cataloging Database.” Journal of Library  Automation 14 (1981): 116-20.

Leroy, Sarah Yoder, and Suzanne Leffard Thomas.  “Impact of Web Access on Cataloging.” Cataloging & Classification  Quarterly 38, no 2 (2004):  7-16.

Local Cataloging Policies in ARL Libraries.  SPEC Kit #110 (Washington, D.C.: Office of Management Studies, Assn. of Research Libraries, 1985).

Marcum, Deanna B. Address. Ebsco Leadership Seminar. Boston: 16 Jan. 2005. <http://www.loc.gov/library/reports/CatalogingSpeech.pdf>.

McCue, Janet, Paul J. Weiss, and Marijo Wilson.  “An Analysis of Cataloging Copy:  Library of Congress vs. Selected RLIN Members.” Library Resources & Technical Services 35 (1991): 65-75.

Neumeister, Sue.  “New Innovation in Cataloging:  the Impact of Technology on Copy Cataloging.”  Library Acquisitions:  Practice & Theory 21, no. 1 (1997):  62-3.

OCLC.  Bibliographic Formats and Standards. 3rd ed.  Dublin, Ohio: OCLC, 2002--.

Rider, Mary M.  “Developing New Roles for Paraprofessionals in Cataloging.”  The Journal of Academic Librarianship 22 (1996): 26-32.

Ryans, Cynthia C. “A Study of Errors Found in Non-MARC Cataloging in a Machine-Assisted System.” Journal of Library  Automation 11 (1978): 125-32.

Back to Contents