Illinois Interlibrary Loan Assessment Project
Phase II (1999/2000)

A Final Report Submitted to the CCMCC

By
Tina E. Chrzastowski
and
Lynn Wiley



University of Illinois at Urbana-Champaign Library
Urbana, Illinois





Introduction

The Illinois Interlibrary Loan Assessment Project, Phase II (1999-2000) was created in 1998 when the CCMCC granted University of Illinois at Urbana-Champaign librarians Tina Chrzastowski and Lynn Wiley $19,700 to repeat their initial study of interlibrary loan (ILL) in the state of Illinois1. The first study, covering 11 months of ILL requests (1995/96) provided a basic framework to measure Illinois' collections by analyzing borrowing and lending activity among the 26 largest libraries in the state.
This second study, employing the same methodology, provides comparative data and assesses the impact that increased accessibility to full-text journals could have on ILL activity in Illinois. Transaction dates for heavily requested titles were analyzed to determine the likelihood of full text availability. "Core" titles, those that are heavily requested, were confirmed by identifying the overlap between the 1995/96 and 1999/00 studies. The Phase II analysis also delves deeper into journal subject analysis by comparing the requested titles’ LC Classification frequency distribution to that of the individual transactions. Phase I only looked at the subject distribution of the transactions; the analysis by title conducted in Phase II results in a better understanding of the variety of serial titles requested.

Methodology

As in the first study, data were collected from the OCLC ILL Management Statistics Service. Interlibrary loan data from the 26 largest libraries in Illinois were downloaded monthly for twelve months beginning in July 1999. A complete list of the participating libraries is found in Appendix A. Each of the 26 libraries was contacted individually and the proper permissions were obtained to collect this data before proceeding. The data were transferred to the University of Illinois' Library Research Center for collation and analysis. To compare results to the first study, queries to the resulting database repeated those of the previous study. As in 1996, analyses were conducted for all 26 libraries together, for libraries other than the 26, but still within Illinois, and for libraries outside the state of Illinois. And although all material types were included in the original download from OCLC, only journal article photocopy requests were included in this study. Due to the widespread use of Illinois' statewide online catalog for lending and borrowing books, monograph requests through OCLC are considered secondary to the statewide online catalog and therefore provide an incomplete picture of book borrowing in the state. In addition, the focus of this research continues to be journal-use within the state of Illinois. Further analysis of the data was done to determine the frequency distribution of article requests by date of publication, and to discover the subject distribution for titles with a single request from a single journal title. These analyses were done both with the help of the Library Research Center and by a graduate assistant hired by the authors to conduct further data analysis.

Budget

The initial budget request to CCMCC for the support of this project was $19,700.00. Total funding was received from CCMCC. Expenditures included:

OCLC Management Statistics fee for data:$ 4,100.00
Library Research Center fee for data compilation and analysis:$ 9,996.00
Graduate Assistant hourly for data manipulation and additional research:$ 4,400.00
Total          $18,496.00

The authors propose to spend the remaining $1,204.00 to fund the additional analysis needed and to cover travel expenses to the 2001 ILA conference to present the data from both Phase I and Phase II studies to Illinois librarians. Our abstract has been accepted for presentation.

Hypotheses

Because this study is a follow-up to one conducted for CCMCC four years earlier, the authors' hypotheses for Phase II continue to address broad areas of interlibrary lending and borrowing to present a picture of ILL activity among the state's largest 26 libraries. In addition, comparisons are made to the previous research conducted in 1995/96. The comparative data have also suggested a predictive model to assess resource sharing in Illinois since both studies’ analyses provided quite similar results. The 1995/96 hypotheses and results and predictions on the results of the 1999/00 analysis are:

1. The largest number of photocopy requests will be filled within the state of Illinois. The percentage in 1995/96 was 58% filled within Illinois and 42% filled outside the state. For 1999/00, the authors predict a slight increase in in-state fills, approximately a 65/35 split.

2. Illinois libraries will be net lenders during the 12 months of the study. In 1995/96, Illinois libraries supplied 96,928 articles in 11 months, and borrowed 89,316 articles, supporting the net lender hypothesis. The authors predict a very similar outcome from the 1999/00 data.

3. Within Illinois, the 26 libraries will meet most of the needs of the group of 26. The 1995/96 study found this hypothesis proved, with the 26 libraries filling 57% of requests from the group of 26 libraries, with 43% being filled from other in-state libraries. In 1999/00, the authors predict a greater emphasis on partnerships between the state's 26 largest libraries, with the percentage of "26-to-26" filled photocopy requests moving from 57% to approximately 65%.

4. Photocopy requests filled within the state of Illinois will come from the highest-requested journal titles. Photocopy requests filled from outside Illinois will more likely be for a single request from a single (unique) title. This hypothesis predicts that the "core" titles requested will be found within Illinois, and more peripheral titles will be borrowed from outside the state. In 1995/96, this hypothesis was supported, with 62% of single-title requests filled outside Illinois and 44% filled within Illinois. For 1999/00, the authors predict this split to remain about the same.

5. A core group of serial titles (requested more than 20 times each and not found in Illinois) will be identified that will be more cost effective to own and share within Illinois and which are now being accessed from outside the state. In 1995/96, this hypothesis was proved and identified 113 titles in this category. For 1999/00, the authors predict a slight increase in core titles needed, but not found, within the state. The authors also predict that there will be a high degree of overlap in the high use titles requested between the 1995/96 study and the 1999/00 study.

6. Collection subject strengths will be identified by comparing the number of loans supplied and the number of article photocopies borrowed by subject, based on Library of Congress classifications. In 1995/96, subject strengths and high-volume subjects were identified as occurring primarily in the sciences (Q, R, S, T, and W), but also in the social sciences (H), psychology (B), and literature (P). A near identical distribution is predicted for the 1999/00 study. It is also predicted that subject analysis will show that the sciences (Q, R, S, T, and W) have the highest number of article-photocopy transactions, both in borrowing and lending, as compared to other groups such as the humanities and social sciences. This hypothesis was supported by the 1995/96 data, and is again predicted to be supported by the 1999/00 data.

7. The 1995/96 LC analysis provided data at the transaction level. But because so many transactions are single requests for a single journal title, it was necessary to look at the LC ranges for just the serial titles themselves. The authors predict that these titles will come from a wide-range of subject areas distinct from those found in the core group of requests, which centered on science call number ranges.

8. In order to address the impact that the availability of electronic full-text journals would have on ILL in Illinois, the data for the most frequently-requested titles (those requested 20 or more times), filled both in-state and out-of-state, were sorted by citation date. The authors predict that over 50% of these titles could be filled by access to electronic full-text journals.

9. The 1995/96 analysis allowed the authors to analyze patterns of resource sharing that not only identified core titles but also provided a framework to assess the use of collections statewide. The authors predicted that the patterns of use would be validated in the 1999/00 study.

Data Summary

Each of the hypotheses is addressed with data from both the 1995/96 study and the 1999/00 study.

Hypothesis 1

The largest number of photocopy requests will be filled within the state of Illinois. The percentage in 1995/96 was 58% filled within Illinois and 42% filled outside the state. The authors predict a similar 60/40 split in 1999/00.
Results Hypothesis 1

Table 1.0
Requests Made by the 26 and the Suppliers/Lenders
BorrowerNumber of Photocopies Borrowed
 1995/961999/00
Illinois Consortium Libraries (the "26")29,80541,980
      Other In-State Libraries22,37417,876
            In-State Subtotal:52,179   (58%)59,856   (57%)
      Out of State37,137   (42%)45,159   (43%)
                  Total:89,316105,015

As shown in Table 1.0, ILL activity increased between 1995/96 and 1999/00 from 89,316 copies supplied to 105,015 copies supplied. The 1995/96 data however only covered 11 months, an average of 8,000 transactions per month, which would have assumed a total of approximately 97,000 for the year, still indicating an increase of 9% in 1999/00. The in-state versus out-of-state supplier ratio remained nearly identical, with a 58/42 split in 1995/96, and a 57/43 split in 1999/00.

Hypothesis 2
Illinois libraries will be net lenders during the 12 months of the study. In 1995/96, Illinois libraries supplied 53,884 articles in 11 months, and borrowed 52,179 articles, supporting the net lender hypothesis. The authors predict a very similar outcome from the 1999/00 data.

Results Hypothesis 2

Table 2.0
Photocopies Borrowed by and Supplied by the 26, Other In-State Libraries, and Out-of State Libraries
 1995/961999/00
 BorrowedSuppliedBorrowedSupplied
Illinois Consortium Libraries (the "26")29,80533,34941,98043,510
            Other In-State Libraries22,37420,53517,87616,849
              In-State Subtotal:52,17953,88459,85660,359
     
            Out of State37,13743,04445,15943,856
            Total:89,31696,928105,015104,215

Data from Table 2.0 show that the 26 libraries slipped slightly into the net borrower category, supplying 104,215 to the 105,015 borrowed. The data suggests a heavier reliance on other libraries out of state as confirmed in Table 1 with a 1% increase in fills from out of state libraries.

Hypothesis 3

Within Illinois, the 26 libraries will meet most of the needs of the group of 26. The 1995/96 study found this hypothesis proved, with the 26 libraries filling 57% of requests from the group of 26 libraries, with 43% being filled from other in-state libraries. In 1999/00, the authors predict a greater emphasis on partnerships between the state's 26 largest libraries, with the percentage of "26-to-26" filled photocopy requests moving from 57% to over 60%.

Results Hypothesis 3

Table 3.0
Photocopies Borrowed by the 26 from Illinois Libraries
 1995/961999/00
Illinois Consortium Libraries (the "26")29,805 (57%)41,980 (70%)
    Other In-State Libraries22,374 (43%)17,876 (30%)
       In-State Subtotal52,17959,856

While the 1999/00 data did indeed confirm that the 26 largest libraries in Illinois fill most of the needs of that group, the percentage of fills from the 26 increased from 57% in 1995/96 to 70% in 1999/00. One reason for this greater shift of interdependence may be that ILL service and delivery within this population has improved, and these libraries have been directing their orders to the larger libraries that are able to supply articles faster and more cost effectively using, for example, Ariel for internet delivery. The other in-state libraries supplying material are smaller academics, public and special libraries that may not have all the latest equipment available to them. It is possible that collections strengths have also shifted in the past five years. While multiple libraries are required to cover the varied research needs requested by interlibrary loan, that dependence shifted significantly.

Hypothesis 4

Photocopy requests filled within the state of Illinois will come from the highest-requested journal titles. Photocopy requests filled from outside Illinois will more likely be for a single request from a single (unique) title. This hypothesis predicts that the "core" titles requested would be found within Illinois, and more peripheral titles will be borrowed from outside the state. In 1995/96, this hypothesis was supported, with 62% of single-title requests filled outside Illinois and 44% filled within Illinois. For 1999/00, the authors predict this split to remain about the same.

Results Hypothesis 4

Table 4.0
A Comparison of Title Request Frequency By Supplier in the 2 Studies
 1995/961999/00
Journal Titles RequestedOut-Of-StateWithin IllinoisOut-Of-StateWithin Illinois
% of titles requested once62.3%43.7%61.0%44.4%
% of titles requested 2-19 times36.9%53.0%38.2%52.1%
% of titles requested over 20 times.71%3.2%1.0%3.4%

Table 4.0 indicates that Illinois libraries in the 1999/00 study supplied more of the frequently requested titles, from 43.7% in 1995/96 to 44.4% in 1999/00. The percentages for all 3 categories are very similar in both studies. Illinois libraries continue to be able to provide access to the most heavily used serial titles.

Hypothesis 5

A core group of serial titles (requested more than 20 times per title) not found in Illinois will be identified that will be more cost effective to own and share within Illinois and which are now being accessed from outside the state. In 1995/96, this hypothesis was proved and identified 113 titles in this category. For 1999/00, the authors predict a slight increase in core titles needed, but not found, within the state. The trend in collection development in the past few years has been to provide the full text access that replicates on-site print collections rather than expand and diversify the breadth of titles collected. The authors also predict that there will be significant overlap in the high use titles in both those filled by Illinois libraries and by the out of state institutions.

Results Hypothesis 5

Table 5.0
A Comparison of the Number of Titles and Their Request Frequency
 1995/961999/00
Journal Titles RequestedOut-Of-StateWithin IllinoisOut-Of-StateWithin Illinois
Number of titles requested once9,8825,41410,9086,103
Number of titles requested 2-19 times5,8656,5176,8457,163
Number of titles requested over 20 times113397188470

Table 5.1
Details on the Heavily Requested titles supplied by both Categories of Libraries in Both Studies

The authors prediction of a larger number of core titles requested and then provided by non-Illinois libraries is confirmed by the data in Table 5.0 with the out-of-state title list increasing by 66% (113 to 188 titles) and the in-state list increasing by 18% (397 to 470 titles). Table 5.1 indicates that there is a high degree of overlap between the high use lists from both studies. Of those titles supplied by Illinois libraries in 1995/96, 246 or 52% also appeared on the1999/00 list. High use titles supplied by out-of-state libraries showed an overlap rate of 28% between 1995/96 and 1999/00.

Hypothesis 6

Collection subject strengths will be identified by comparing the number of loans supplied and the number of article photocopies borrowed by subject, based on Library of Congress classifications. In 1995/96, subject strengths and high-volume subjects were identified as occurring primarily in the sciences (Q, R, S, T, and W), but also in the social sciences (H), psychology (B), and literature (P). A near identical distribution is predicted for the 1999/00 study.

Results Hypothesis 6

Table 6.0
A Comparison of LC Ranges for all Articles Requested by the 26 Libraries for Both Studies

Table 6.0 confirms that requesting was highest in the Sciences (Q,R,S,T,W) followed by a group of Social Sciences classes (B , H) then P (Literature) as predicted. The two studies showed a remarkable correlation with the exception of the R (Medicine) requests, which increased significantly both in-state and out-of-state in the 1999/00 study. However the authors are conducting further research about data reporting to determine if the W call numbers from 1995/96 have somehow been added as R call numbers in 1999/00.

Table 6.1
A Comparison of the Article Requests made by the 26 by Broad Disciplines

Figure 6.2

A Comparison of the Filled Requests by Discipline

The pattern of borrowing by disciplines is again confirmed by data in Table 6.1 and figure 6.2. Requesting in the three areas is the approximately the same in both studies. The sciences see the heaviest volume followed by social sciences and then humanities.

Hypothesis 7

The 1995/96 LC analysis provided data at the transaction level. But because so many transactions are single requests for a single journal title, it was necessary to look at the LC ranges for just the serial titles themselves. The authors predict that these titles will come from a wide-range of subject areas distinct from those found in the core group of requests, which centered on science call number ranges

Results Hypothesis 7

Table 7.0
A Comparison of the LC Classes for Serial Titles Vs. Transactions

Figure 7.1
A Comparison of Transactions to Titles by LC Class

Once again the data (in table 7.0 and Figure 7.1) show ILL transactions in Illinois being driven by science disciplines. The sciences also have the highest average transaction per title rate, making their journals higher-use and therefore more cost-effective. As would be predicted, the humanities have the lowest transaction per title rate, demonstrating that discipline’s reputation for a long "comet tail" - those single requests from a single journal. It is gratifying to see that in each discipline, the in-state transaction per title is higher than out-of state, although marginally in the humanities.

Hypothesis 8

In order to address the impact that the availability of electronic full-text journals would have on ILL in Illinois, the data for the most frequently-requested titles (those requested 20 or more times), filled both in-state and out-of-state, were sorted by citation date. The authors predict that over 50% of these titles could be filled by access to electronic full-text journals.
Results Hypothesis 8

Table 8.0
Transactions by Citation Date for Articles Received by the 26 in 1999/00

The data available in Table 8.0 confirm that the majority of transactions for the most highly requested titles were from the more current years, volumes more likely to be available in full text format. The titles are being researched now to see which are offered in a full text format by the publisher. This information will be appended to the study at a later date.

Conclusions

The 1995/96 analysis allowed the authors to identify patterns of resource sharing that not only identified core titles but also provided a framework to assess the use of collections statewide. The authors predicted that the patterns of use would be validated in the 1999/00 study. The comparative data suggest that for Illinois Libraries, these issues are constant:
Summary

Resource sharing follows very predictable patterns in both studies. The number of requests made, the material supplied in-state versus out-of-state, and the subject matter for the material requested are all strikingly similar. Illinois libraries continue to see most of their needs met by other Illinois libraries. More requests per-title are being filled, indicating that the statewide collection is serving researchers well.

There are however, two core groups of titles that were examined to see if researchers are best served with a change in collection management. The first group is those high use titles that are supplied to Illinois researchers by libraries out-of-state. Fifty-two of these titles appear on both the 1995/96 study list and the1999/00 list. The high overlap suggests that these titles should be purchased for in-state library collections. An analysis of holdings and subscription costs is being completed now to determine costs and to advise CCMCC on whether these were titles once owned but now cancelled by Illinois libraries.

The second group includes those high use titles supplied by Illinois libraries to the 26 libraries. The overlap was 62% with 246 titles appearing on both the 1995/96 and 1999/00 lists. Interlibrary Loan offices throughout Illinois had to supply almost 20,000 photocopies to answer requests for these high demand titles. Since this is a core group of material, it is recommended that CCMCC look at the possibility of purchasing full text access to these titles to best serve the needs of researchers in Illinois and to avoid the indirect ILL costs absorbed by all the Illinois libraries supplying copies across the state.

One purpose of the study was to correlate any changes in ILL use with the greater availability of electronic full text titles. The data indicate that ILL use has increased 9% since the 1995/96 study. Full text availability, now found in many Illinois libraries, is not decreasing the need for more resource sharing. Many full text titles are simply supplementing resources already available by providing print material in another format. The fact that these materials are easier to access and generally more available at the academic campuses may help to explain the decrease in use of multi-type libraries from 1995/96 to 1999/00. The investigators are still researching the full text availability of the in-state core list of titles and how much they would cost. This information will help inform future decisions on how researchers in Illinois may access this group of highly requested serials.

The investigators were impressed overall not only with the large amount of material being shared in Illinois but also by how well the state-wide collection serves all libraries and their users. Collection managers and Interlibrary loan office staff throughout the state should be proud of their accomplishments.

Recommendations

  • Investigate the purchase of the high use titles supplied by out of state libraries and which appeared on both the 1995/96 list and the 1999/00 list. (Appendix B)
  • Continue to research the high use titles supplied within the state to see if a consortia purchase of those titles in full text is possible. (Appendix C)

    Appendix A

    List of the 26 Illinois Libraries
    OCLC lending/borrowing data available from 1999/00 (and for 11 months 1995/96) Notes

    1. http://libws66.lib.niu.edu/ccm/illstudy.html