- Faculty Taskforce
- College of Agriculture and Life Sciences
- Division of Biological Sciences
- Albert R. Mann Library
- Cornell University
To print the Journal Price Study, please download the following three files in Adobe Acrobat format:
Also, throughout this page, tables are available in mulitple formats.
Purpose and Background
- A. Methodology and the Journals
- B. Data Results and Comparisons
- 1. Core Agricultural Journals
- 2. Core Biological Journals
- 3. Comparisons
- C. Page Charges, Reprints and Copyright Waivers
- D. Implications for Scholarly Communication
- 1. Scholars
- 2. Publishers
- 3. Libraries and Academic Institutions
- A. Data Elements
- B. Guidelines for Journal Price Study Worksheet
- C. Sample Datasheet Sent to a Publisher
- D. Alphabetical List of Journals Included in Studies
- E. Data for Agricultural Journals
- F. Data for Biological Journals
This report represents the first comprehensive study of journal pricing in agriculture and biology. A total of 314 journals was surveyed. The studyís purpose was to investigate recent changes (1988 vs. 1994) in journal prices including publisher types: universities, commercial firms, governments (including the United Nations), and professional societies and associations. The study was conducted under the supervision of a group of faculty in agriculture and biology at Cornell University. The analysis was carried out by Wallace C. Olsen, Senior Research Associate, and the Collection Development staff at the Albert R. Mann Library at Cornell. Statistical analyses were performed by Professor Susan Holmes and Owen Chang of the Biometrics Unit, Cornell University, and by William H. Walters, bibliographer and data librarian at Mann Library. In addition, the Mann Library Advisory Committee provided helpful criticisms and comments. The primary method used to conduct this study involved electronic scanning and optical character recognition to determine word and character counts per page, as detailed in the attached report (p. 2). In every case, the data were sent to the relevant publisher for their examination; 80% of them responded with verifications, corrections, and comments.
The Executive Summary on the following pages reviews the main points of the study, including coverage, methodology, conclusions, and recommendations. The full study is given in the main report and the journal-by-journal data can be found in the supplementary tables. The entire report, including those tables, can also be consulted on the web at: http://adam.mannlib.cornell.edu/jps/jps.htm
The Faculty Taskforce wishes to thank the Cornell Administration for funding the staff work on this project, in particular David H. Call and Daryl B. Lund, past and present deans of the College of Agriculture and Life Sciences, and Norman R. Scott, former Vice President for Research. We also acknowledge the aid of W. Ronnie Coffman, Brian Chabot, and Robert Cooke, all professors of the College of Agriculture and Life Sciences, Cornell. In particular, we wish to acknowledge our debt to the staff of Mann Library: Jan Olsen and Janet McCue, past and present directors, and Sam Demas, and William Walters. Above all, we acknowledge Wally Olsen for his supervision of this project, his involvement in writing the final report, and for his unfailing good advice. We are greatly in his debt.
The Faculty Taskforce on Journal Price Study
- Martin Alexander (Soil, Crop and Atmospheric Science)
- Gerald F. Combs (Nutritional Sciences)
- Nelson Hairston (Ecology and Systematics)
- Harold Hintz (Animal Science)
- Kenneth Horst (Plant Pathology)
- Betty Lewis (Nutritional Sciences)
- Donald Rutz (Entomology)
- Chris Wien (Fruit and Vegetable Science)
- Kraig Adler (Neurobiology and Behavior), Chair
Journal Price Study of Core Agricultural and Biological Journals Faculty Taskforce, College of Agriculture and Life Sciences, and
Division of Biological Sciences, Cornell University November 1998
Scholarly communication patterns are being threatened by high prices of scholarly journals. Libraries are having financial difficulties with this acceleration and have resorted to cutting journal subscriptions, particularly those with extraordinary price increases. Others have chosen to cut a large number of less expensive journals, thus greatly reducing diversity. Faculties affected have been forced into the decision process which gave the impetus for an ad hoc faculty taskforce at Cornell University. In consultation with Mann Library administrators, a price study of journals over a recent period of time was undertaken.
A recent identification of the most valuable or core journals for advanced agricultural education and research done by Mann Library staff was used as the initial study group. Later, a similar process to identify the current core biological journals was undertaken by Cornell biology faculty to form a second group of core journals for investigation and comparison. The two study groups had 222 agricultural titles and 174 biological titles with an overlap of 84 titles, for a total of 312 different titles.
The two study years were 1988 and 1994. Standard data gathered for all 312 titles were: Publisher type, availability of ads, subscription price, numbers of pages, subscription cost per page, average characters per printed page, cost per 1,000 characters, Science Citation Index (SCI) citations to each title, Institute for Scientific Information (ISI) impact factor, average figure content, page charges, and availability of reprints. Eighteen titles were scanned both electronically and manually as a trial run, and this demonstrated a discrepancy between them of less than 1%. As a result, randomly selected pages were scanned for each volume year and electronically calculated. The computer counts were converted into characters per volume and calculations made to arrive at cost per 1,000 characters and cost per printed page. The resulting data were sent to the publishers of each journal for verification and addition of information and identification of special circumstances.
The core agricultural titles had 12% published by universities, 39.5% by commercial publishers, 8% by governments including the U.N., and 40.5% by societies and associations. The core biological titles had 10.4% published by universities, 51.1% by commercial publishers, 5.2% by governments, and 33.3% by societies and associations. U.S. publishers issued 46% of the agricultural journals and 56% of the biological titles; the U.K. was second in both cases, with 27% and 25%, respectively. Individual societies and associations almost always had only one title per list. Publishers with the greatest number of titles per list are Elsevier, Blackwell, and Cambridge University Press (range of 17 to 8 titles each) in agriculture; and Academic, Blackwell, and Springer International in biology (range, 20 to 13 titles). This shows high journal concentration among a very small group of publishers.
Between 1988 and 1994, agricultural journals, on a price-per-page basis, increased 64.7% for all titles, with commercial publishers at a high of 77.8% and with societies and associations at a low of 33.3%. The lowest average price per page for agricultural journals in 1994 were those of associations and societies. The most costly titles (by cost per page and cost per character) are charted for agricultural journals of which the top 25 by character cost are all commercial publications, with one exception. The 30 most expensive biological titles by character cost are commercial publications, with four exceptions.
A positive correlation is demonstrated between the origin of the most expensive titles by per-page or per-character costs with the greatest number coming from the U.K., Netherlands, Germany and the U.S. in that order. Commercial publishers titles in both groups consistently outprice the university, government, and association and society publishers. This study addresses the influence of consumer price increases and the monetary differences due to the U.S. dollar exchange losses in France, Germany, Netherlands, and the U.K. as part of the overall journal price changes. There is a substantial increase in price above and beyond the Consumer Price Index (CPI) and monetary changes in Germany and the Netherlands. The ISI impact factor (created annually by the SCI as an indication of journal importance) is matched with cost per page on the agricultural journals, resulting in the identification of 15 commercial journals with high per page costs and low impact factors for 1994. Society, association, and government publishers consistently show the greatest
impact for the least cost per page. Costs by two or three measurements and average changes between the two years are contrasted for the agricultural and biological titles. In both cases, the countries with the highest costs and greatest numbers of high cost journals are from the U.K., the Netherlands, and Germany.
The part that page charges play is reviewed along with the influence of reprints, and the relinquishing of copyright by academic scholars. These are not measured quantitatively, but explored as factors of significance to scholars and scholarly publishing. Options available to scholars, publishers, academic and research libraries, and universities to address some of these problems are outlined.
The recommendations of the Taskforce are these:
1. Faculty groups, professional societies, university administrators, and individuals should be apprised of the data delineated in this report and stimulate discussion and actions through these avenues:
a. Publication of articles concerned with the study and its findings, as well as making the full report available.
b. Presentations to professional societies and educational associations concerning the data, their implications, and potential solutions.
c. Universities faculties and administrators should reassess their policies regarding relationships with journals.
d. Discuss the results of the study with publishers where there appear to be problems that can be solved or alleviated.
2. Academics including researchers must be aware of the implications of their publishing patterns and the costs to institutions and their readers. These include the possibility of not supporting journals (either by submission of manuscripts or provision of editorial services) which have excessive costs, journals which take copyright unto themselves for profit, and journals which restrict access to information by high use or subscription costs.
3. Academics must reassess their relationships to their servicing libraries in a period of changing scholarly publishing, particularly online texts and dissemination for profit. There must be a clearer understanding of the access to information published by faculties, societies, and commercial organizations and the financial burdens these place on academic institutions and their research libraries. Total university costs should be assessed with particular focus on the faculty investments in editorial and writing work, and the buy-back costs of that information for the same academic institutions via their libraries.
Purpose and Background
Scholarly communication patterns rely heavily on the use of scientific journals for the transmission of detailed information. This has been an essential and valuable method of disseminating scholarly literature for two hundred years. This pattern of distribution has recently been discussed as outmoded, undemocratic, and too expensive. Libraries have traditionally attempted to aid this communication process by collecting scholarly journals where a number of scientists will use the same titles over and over. Recent interest in the concept of publishing electronically has not reached fruition, and in the scholarly world appears not to be the panacea some had hoped.1 The result for scholars, researchers, students, librarians, and administrators in universities or research organizations is an ever-increasing cost that the organization must pay to keep abreast of scientific advancements. This is not a recent problem, and its apparent acceleration from 1980 to date has been the subject of many studies and comments.2
Various aspects of journal pricing have been studied in the past twenty years. Many subject fields have been investigated and reported.3,4,5 Some standard gross measurements of journal price increases are provided yearly by several publishing and library journals.6,7,8 Some very specific investigations in the past must have too clearly used data to point fingers which resulted in a legal suit by Gordon & Breach Publishers against the American Institute of Physics over a physics journal price survey by physicist Henry H. Barschall at the University of Wisconsin published in 1989.9 This suit has recently been resolved.10 Another suit was filed by Gordon & Breach against an update of similar data published by the American Mathematical Society.11 No Gordon & Breach titles are in this study. One extensive study of importance on serial pricing was done by the Association of Research Libraries.12 A book resulted from research and a conference sponsored by British academic libraries and the British Library although few solutions are proposed nor enacted.13 The costs of journals and the choices available to university scholars and librarians have been discussed extensively in the past ten years. In 1989 a university librarian made this observation: "The high cost of research journals has mushroomed into one of the most problematic and acrimonious issues of the decade."14 In the same year, Science editorially took up the question and some of its problems and causes.15 The questions associated with the high costs of scholarly journals are not going to go away although the price increases have slowed somewhat. The portion of library budgets going to journals continues to increase even with major cancelling of some journal subscriptions.
The implications are becoming clearer to faculties as choices must be made about which journals to cancel, how savings can be realized, and still provide acceptable local library service for scholars. One of the more thoughtful recent papers concerning scholarly publishing and library journal costs is that of Herbert White in which he covers the peer assessment system, publication worthiness, the elitism of the process, the publish or perish emphasis, journal publishing economics, and how this all works for authors and publishers.16
The questions have been pressing the Mann Library for several years and recently a faculty taskforce concerned with the problems has been working with the library staff to do some investigating and determine what actions can be taken to help the situation. Heavy additional funding annually seems not to be a solution that receives quick acceptance although this partially solves the problem. Dropping of very expensive titles has been undertaken as well as the usual method of dropping some low-priced journals. Depending on increased interlibrary borrowing or using other campus libraries for select titles is mostly shifting the costs to another checkbook. These and numerous other partial or temporary actions have been implemented, but a long-term solution has not been implemented. For this reason, the faculty task force decided to see what the situations in the agricultural and biological sciences were and to proceed to some recommendations if this appeared practical. The purposes of the study are:
- To analyze the price of agricultural and biological sciences journals over a period of time, comparing the cost per page and cost per thousand characters of journals published by university presses, commercial publishers, society publishers, and government agencies;
- To use the data resulting from this study to inform faculty initiatives at Cornell and nationally, to examine the patterns of article ownership and the pricing of published scientific research in the interest of science and of the libraries which support scientific research and instruction; and
- To hone a methodology using optical character recognition (OCR) technology which can be applied elsewhere.
A. Methodology and the Journals
Determination of appropriate journals to be studied came from a five year investigation of the current core literature of the agricultural sciences. This citation analysis and peer review of titles involved 600 scientists and counselors plus sixty-eight authors and editors who provided the results in a seven volume set entitled The Literature of the Agricultural Sciences.17 The evaluations operated on two tracks: the literature vital for instruction and research for (a) developed countries, and (b) the developing countries. The final list of core journals included 364 titles from around the world. A subset of these journals was extracted: those journals of greatest value to developed country institutions which was the focus of the journal price study. This provided 222 current journals for the journal price study, which is a large group of the most important agricultural science journals. To cover the biological sciences, faculty members at Cornell University determined which of the agricultural core journals were also vital to the biological sciences (84) to which were added titles to cover all areas of the biological sciences. These supplemental titles were identified through rankings in the Journal Reports of the Science Citation Index and from earlier faculty evaluations done for the Mann Library . This composite list of over 200 titles was ranked and reviewed by appropriate faculty of all biological subjects. The final list of 174 includes 90 added for a full coverage of the biological sciences. The biological journal study was added to determine if there were any discernible variations in pricing patterns between the two different but closely related subject areas.
Earlier journal pricing studies have determined the basic data elements which would reflect most accurately the changes in such items as journal pages, content as measured in print characteristics, and several other factors. The characters on a page are viewed as the major intellectual component of a page that can be readily quantified. It is the lowest common denominator of what you get for your money in a journal. Characters per page or some derivative of a page can be quantified and compared to the price per page or a subscription price resulting in a cost per character. Character counts equalize such difficulties as the size of a page, size of type and changes in them, and some variations in editorial or layout practices. The Cornell Faculty Committee added some data to make the information more complete and to assist in observing the reasons for variations or changes. Both studies used the same two years, 1988 and 1994, as well as all of the other data gathered. These data elements which are relatively standard for comparison of journal prices are identified in Appendix A.
Random sampling techniques were used to determine the numbers of pages as well as which pages were to be studied in the two years. The traditional method of determining the characters on a page is to have a qualified individual count them one by one. This is a very tedious process which Mann Library updated with an electronic scanner. A trial test was made with eighteen journals using scanning techniques which were compared with the manual counts. The difference was less than 1%. Manual counts, alone, have a greater degree of variation depending upon the investigator, time of day, and other human factors. Therefore, the scanning figures are considered of greater validity. As a result of this and other tests, OCR scanning was used to get the character counts for all journal pages. Raw data were run through a computer program for translation into characters per page. Guidelines were prepared for personnel gathering data and scanning (Appendix B). The results for all data elements were displayed in printouts for each title; a sample is reproduced as Appendix C.
The data elements as displayed in Appendix C were sent to publishers or editors of each journal along with an explanatory sheet. Comments and corrections were invited. Of the 200 letters sent to 125 different publishers or editors, 87% were returned with few comments or error corrections. Many letters had to be sent more then once to obtain this high a response. All suggested errors or skews were examined and corrected when appropriate. The most common suggestion was that the increases in non-U.S. journal prices were the result of variations in the value of the dollar. Similarly, data for the new biological journals were sent to publishers for their perusal where the return rate was 76% with many of the same observations. A complete list of all journals in the two studies is provided as Appendix D.
B. Data Results and Comparisons
This diagram illustrates the journals in the studies and their common titles. A total of 314 different titles were included in the two studies.
After nearly a year of scanning, reviewing data, sending letters with data to publishers, and making necessary adjustments, the data was analyzed as to where the price increases had been greatest, how the costs per characters ranked titles, the influence of figure space, along with all of the other data elements and information available. Since there are 222 core agricultural journals and 174 core biological journals, the analysis took a long time even with the assistance of many tables, figures and computer calculations. The resulting amount of data is so great that it will not be presented here in gross numbers. Instead, samples of the data elements are given and then the groups or trends are provided by tables, figures, and analysis.
Table 1 lists every tenth title of the core agricultural journals to illustrate these elements: Title, Publisher, Publisher Type, Ads paged in the scholarly text, Subscription Prices and percentage changes, Number of Pages and percentage changes, and the Subscription Cost per Page and percent changes. These same data are shown for every tenth title from the core biological journal list in Table 2. These are illustrative of the data gathered.
Table 1. Select Data Elements From a Random Sampling of Core Agricultural Journals (n=22)
Table 2. Select Data Elements from a Random Sampling of Core Biological Journals (n=18)
Tables 3 and 4 take the same journals in both groups and display the final set of data elements. The Average Characters per Page were gotten from the scanning of the random sample pages each year; these were then multiplied time the pages that year to provide the Total Character per Volume; this was then divided by the annual subscription cost per year providing a subscription cost per character which is stated as Annual Cost per 1,000 Characters. The Science Citation Index (SCI) citations to a journal and the Impact Factor of that journal are in the next two columns. The Figure Content per Page is an average from physical measurements made on all randomly sampled pages. Page Charges and Reprint information were obtained from journals when possible and verified or supplied by publishers when the data was sent to them. These titles are intended to be illustrative of the total number of journals surveyed in each core list, 222 and 174.
Table 3. Select Data Elements from a Random Sampling of Core Agricultural Journals (n=22)
Table 4. Select Data Elements from a Random Sampling of Core Biological Journals (n=18)
Both the core agricultural and the core biological journal lists were achieved through extensive citation analysis, examination of other numerical data and rankings, and evaluations by faculties of several institutions. The derivation of the agricultural list was from an extensive investigation for an unrelated purpose. It must be clear that 84 titles in the two subject lists are the same; that is, those 84 titles are evaluated as core to both the agricultural sciences and the biological sciences. This similarity of topical subject might lead to the conclusion that there are very few differences between the publishing characteristics of the two. This did not prove to be true as will be demonstrated with the data that follows. The lists are interlinked by related subject areas, by types of publishers, as well as the publishers with the greatest numbers of titles in each group. Table 5 compares the derivation of the titles.
Table 5. Numbers of Journal Titles by Publisher Type
Societies & Associations
- Countries with 10
- or More Journals
Journals (n=222) Germany
Journals (n=174) Germany
Clearly, the United States is paramount in core agricultural and biological journals having 46 to 56% of all titles. The U.K. (25-27%) and Germany (7%) are roughly the same in each list. The publishers with the major numbers of titles are these:
Societies and associations publish 33 to 41% of the titles in the two lists, but only one society has four titles in a list. Societies and associations are numerous in both lists; nearly all publish only one or two titles. The predominance of Elsevier in agriculture and Academic in biology is noteworthy. Blackwell in the last decade has taken over the publishing of a quantity of titles formerly published by societies and associations. The average cost in 1994 of all agricultural titles was $401.81, but $571.74 for the biological titles.
1. Core Agricultural Journals
Figure 1 looks at the price per page based on subscription costs of titles grouped by type of publisher. These data provide a rougher cost evaluation than character costs. The average price per page increase for all of the titles between 1988 and 1994 is 64.7%. The commercial publishers titles which constitute 40% of all titles are responsible for the average price change being so great, since all the other publisher types are between 55.5 and 33.3% price increase. The largest group of publishers titles (41%) are those of societies and associations with only a 33.3% increase. University publishers and governments as publisher are less influential with 44 journals together.
Figure 2 provides basically the same data as Figure 1 except they are plotted against the actual subscription costs of the journals within each publisher type.
The most useful measurement is the price per character that a subscriber must pay for the annual volume. This diminishes the cost of white space. Figure 3 displays the 25 specific agricultural core journals with the greatest price by characters expressed in 1,000 characters for the 1994 subscription year. Figure 3 also shows the ranking of the same titles using the costs per page based on subscription costs. The similarity of rankings of these titles in the top 25 or 35 journals proves the similarity of results whichever costing method is used.
Figure 1. Agricultural Journals; Range and Average Price Per Page by Publisher Type (n=22)
Figure 2. Agricultural Titles; Subscription Price Range and Average Price by Publisher Type
Figure 3. Twenty-Five Most Costly Agricultural Titles Ranked by Cost Per 1,000 Characters and Subscription Cost per Page (1994)
Figure 3 has several titles and ranks not displayed in the cost per page column, e.g., Rankings 4, 9, 13, etc; all are below the 25 (left column) in cost per 1,000 characters with a cost below 21.5¢. Three of these titles have low character costs, but high page costs. The cost per page for these three are 3.5 to 4.0 times their cost per 1,000 characters, whereas the more standard ratio is 2.5 or 3 to 1. These three also have ISI Impact Factors which are extraordinary: Rank No. 4: Plant, Cell and Environment is $1.00 per page with an impact factor of 2.291; No. 9, Planta, is 91.3¢ per page with an impact factor of 3.30; and No. 14, Molecular and General Genetics is 76.6¢ per page with an impact factor of 2.917. The first title is published by Blackwell and the other two by Springer International. These are highly compressed journals with influence.
- Figure 4. Agricultural Titles by Price Per Page, 1988 and 1994.
This figure plots the price per page based on institutional subscription costs for each journal within publisher types and for both years. It is clear that half of all commercial journals in 1994 were above the single, highest priced journals in any other group, university, government or society.
Figure 5. Regression Lines Showing Trends in Price of Agricultural Journals, by Journal Type (Cost per thousand characters, adjusted for U.S. inflation, 1988-1994=22.9%)
Figures 5 and 6 reaffirm that commercial agricultural journals were more expensive than nonprofit journals in both 1988 and 1994. Figure 5 shows cost per thousand characters while Figure 6 shows cost per page. There are a number of commercial journals which costs more than double those of the highestpriced university and government journals.
Regression lines showing the price trends for types of journals further reveal that this trend is not limited to the few most expensive publications. (Each regression line is a "best fit" line that most effectively represents the corresponding data points. The placement of each line minimizes the sum of the vertical distances from the data points to the regression line.) The regression lines reveal that commercial journals were consistently more expensive than those published by government agencies, scholarly societies, or university presses; in both graphs, the regression line for commercial journals appears above and to the right of the lines representing the other journal types. Moreover, commercial journals had the greatest increase in cost from 1988 to 1994, demonstrated by the relatively steeper slope of the regression line for commercial journals.
Figure 6. Regression Lines Showing Trends in Price of Agricultural Journals, by Journal Type (Cost per page, adjusted for U.S. inflation, 19881994=22.9%)
Figure 7 was derived from one hundred journals with the greatest increases in subscription price between 1988 and 1994. Of the hundred, those costing 50 cents per page or over were examined. Two of the titles were from university presses and the remaining 42 were commercial publications. These 44 were sorted by country of publication and the correlation of high per page costs with European publishers is evident. Therefore, an analysis of the range and average price per page based on subscription costs was done contrasting U.S. and European prices, range and averages as represented in Figure 8. The 1994 U.S. averages do not even reach the counterpart 1988 European averages with the exception of the 1994 U.S. commercial titles. A major complication is the Europeans beginning in 1988 from a high per page cost and increasing that spread in 1994 at a greater rate than the U.S.
7. 44 Agricultural Titles with a 1994 Subscription Cost of $.50 per Page or Over
8. Range and Average Price Per Page by Publisher Type, U.S. and Europe
Figure 9. Average Cost Per Page Percent Increases, Consumer Price Index Percent Increases, and U.S. Dollar Changes, (1988+1994)
Figure 9 relates page price increases to consumer price index changes and the change of the U.S. dollar with major European currencies. Price increases for the publishers are for all titles in the study. Only the group of Association and Society publishers came out close to the U.S. price index increase of 22.9% or European Community increase of 29.5%. Publishing prices have exceeded the consumer price index increases in Europe and the U.S. for two decades. European currency value increases from 6.9% to 14% over the U.S. dollar in France, Germany, the Netherlands, and the U.K. are influential in understanding subscription costs for U.S. institutions.
Sources: The U.S. CPI and the dollar exchange rate were taken from the 116th ed. of Statistical Abstract of the United States, 1996. The CPI for the European Community and the four countries in it are from Table 2.43 in the 28th ed. (1991) and the 32nd ed. (1995) of Basic Statistics of the Community, published by EUROSTAT.
An effort also was made to determine the relationship between the cost per page and the ISI impact factor for each journal, the supposition being that most high cost journals probably should have a major scholarly impact. The only routine numerical measurement of impact is through the citation analysis of the Institute for Scientific Information. Figure 10 charts the cross of the price per page for a journal with its impact factor as registered by ISI for 1994. These are divided by type of publisher and provide a very demonstrative scatter display. There are four journals off the scale at the lower right, two being Science and Nature. The highly-priced commercial publications at 50 cents and upward on the scatter and under an impact factor of 1.0 clearly stand out. From a library and scholars point of view, the greatest benefits will probably derive from those intersecting under 40 cents a page and to the right of the 1.0 impact factor.
- Figure 10. Cost Per Page by Individual Title, Publisher Type, and Impact Factor; 1994
11. Enlarged Quadrant Showing Titles with Prices above 60¢ per Page and Impact Factors below 1.0.
B=Blackwell/Blackwell Wissen. Bu=Butterworth-Heinmann E=Elsevier, Pergamon K=Kluwer Academic S=Sauerlander's Sp=Springer International or Springer Verlag
Figure 11 enlarges the portion of the intersection of journal cost per page with the titles impact factor (ISI) where the cost is 60¢ per page or greater and the impact factor is below 1.0. Sixteen journals in this quadrant of Figure 10 offer a high price per page (and generally a high character cost also) and a relatively low impact factor. These are all commercial publishers journals from eight different publishers of which all but two have one title each in this category. Blackwell has two and Elsevier/Pergamon has half of all the titles, 8. These journals are not easily grouped by content, format or subject concentration. Four of the higher cost per page titles lean toward engineering subjects, and six toward plants and soils. Agricultural policy and social studies are also included. The only fact which stands out is that Elsevier or its subsidiary Pergamon own half the titles.
The foregoing information and data lead to several conclusions which should be of concern to universities and their libraries which must pay high fees for library copies for faculty and scholarly use. There is little doubt that libraries are bearing the brunt of these costs or the institution is by different means such as page charges common with society publications. Page charges are discussed on page 24.
2. Core Biological Journals
The 174 core biological journals being studied included 84 which were also in the list of core agricultural journals. Figure 12 profiles for the biological sciences the changes in price increases on a per page cost for the same years as those of the core agricultural journals shown in Figure 1.
2. Biological Journals: Range and Average Price Per Page by Publisher Type (n=174)
The biological titles represented in Figure 13 are comparable to the agricultural titles in Figure 2. The biological titles in all categories of publishers except commercial show a remarkable restraint in increased prices during this six year period. Government, university and society publication increases range from 25.8% through 35.5%, substantially less than the same categories of agricultural titles. The society and association publication increase of 35.5% resulted in a 1994 average cost slightly below that of university publications which went to an average $400. This commercial publisher increase is a remarkably low 36.8%.
Figure 13. Biological Titles; Subscriptions Price Range and Average Price by Publisher Type
Figure 14. Thirty Most Costly Core Biological Titles Ranked by Cost Per 1,000 Characters and Subscription Cost Per Page (1994)
Figure 15. Biological Titles by Price Per Page, 1988 and 1994
Table 6 provides a summary and comparison of gross data from these studies. The basic numbers of titles are pertinent: Agriculture=222, and Biology=174.Table 6. Summary Data for Biological and Agricultural Journals (1994).
Difference in %
Total 1994 subscription costs
Average 1994 subscription costs
Average number of pages per volume (1994)
Average 1994 cost per page on subscription price
Average 1994 cost per 1,000 characters
Average 1994 characters per page
Average characters per volume (1994)
Average figure content (1994)
Highly priced journals consistently have a near-the-top ranking whether based on price per page or cost per characters. Figures 3 and 14 display the most costly journals as organized by cost per 1,000 characters, and on these same Figures are the rankings organized by cost per page. There can be no doubt about the extraordinary cost of these journals when they rank in the top 25 or 30 by both measurement methods. Using the rankings of the cost per 1,000 characters in both cases, 15 of the 25 agricultural titles have a high correlation between both ranks, and 22 of the 30 biological titles, or 60% to 73% of all titles in each list. A close correlation was determined when the rankings were within 10 points of each other within each group. These are the most expensive 25 and 30 journals in populations of 222 and 174; a small group from a large, core journal population. Publishers and countries of publication on these 15 and 25 journals are:
Publishers with 2 or more titles in either list Elsevier & Pergamon
Country of Origin U.K.
Figure 16 shows agricultural and biological data demonstrating that the percentage increases are substantially greater with the agricultural journals. University publications, those of governments, and those of associations and societies all came off substantially lower in their overall increases. Another factor is that the average journal cost of all agricultural titles was $401 in 1994, as contrasted with $571 for the biological titles. Agricultural publishers may have been playing price catch up between 1988 and 1994, except for societies and associations.
Figure 16. Cost per Page Percentage Increases, 1988-1994, for Agricultural and Biological Titles
Figure 17. Most Costly Agricultural and Biological Titles Based on 1994 Subscription Cost Per Page
Figure 18 provides another view of the difference between the agricultural titles and the biological titles. The fact that agricultural journals are more expensive by character costs than the biological journals is counter-intuitive. With the fifty most expensive journals, the top three agricultural journals are much higher and then only slightly higher for the remainder. As shown earlier, the costs on agricultural journals during the period of this study increased at a greater rate than the biological journals. One possible explanation of this pattern is that agricultural journals are printed and sold in smaller quantities and the advantage of lower cost per item is not as great. The biological sciences have been riding a wave of influence and funding for twenty years which may make the competition among their journals more nearly that of a well-functioning open market.
Figure 18. Most Costly Titles by Cost per 1,000 Characters
The fact that we are dealing with the most valuable or core journals for academic and research work is emphasized by a comparison with the U.S. Periodical Price Index ... which has been published for 15 years in a spring issue of American Libraries.18 The computations are detailed and divided by subject using ANSI standards. An exact match by subject is possible only for the agricultural group. The average prices per journal are these for 1994.
Life Sciences data based on 41 titles (American Libraries): $372.52 , which is the closest subject to the Biological Sciences titles in this study whose 1994 average price was $569.26.
Agricultural titles based on 213 titles (American Libraries): $80.64, whereas this study of core academic agricultural titles the average is $402.81.
The U.S. Periodical Price Index is averaged on popular as well as some scholarly journals and is limited to U.S. imprints. The core list is based entirely on citation analysis, evaluation by a university faculty, and has 40% non-U.S. imprints. The agricultural price ratio is 5:1 reflecting the American Librariesí survey of popular farm journals and the less expensive U.S. imprints.
Among the important facts are these which should be noted:
- The journal costs charged by commercial publishers as institutional costs are extraordinary compared to costs charged by other types of journal publishers. Regardless of any mitigating circumstances of rent, cost of editors, distribution, page charges or other factors, it is clear that the institutional subscriptions paid by libraries are carrying the major portion of expenses of scholarly journals published by commercial publishers.
- Not all commercially published journals have extraordinary subscription costs. Those with very high costs can be identified with a few commercial publishers.
- The core journal list with which we are working was carefully established on the basis of peer citing of the journals, careful citation analysis of current literature patterns, and comparisons of the most valuable journals with other citation analysis and peer evaluations. This is unmistakedly a list of the most valuable current journals for research and education in all broad areas of the agricultural sciences; we are dealing with core journals valuable to scholars. Within this group of 224 titles, the commercially published
C. Page Charges, Reprints and Copyright Waivers
- journals do not show greater quality or greater impact as demonstrated in Figure 8. In fact, slightly more than half are below the 1.0 impact factor and above the highest cost of the journals of other types of publishers.
- The data demonstrate that the heavy commercial charges for library subscriptions come largely from European publishers with the greatest influence coming from those in Germany, the Netherlands and the United Kingdom. It is also clear that a portion of the costs result from cost of living increases and the low value of the U.S. dollar in these countries. Commercial agricultural titles show an increase of 77.8% (Figure 7); only the United Kingdom accounts for over half of this increase by CPI (+34.8%) and the pound increase over the dollar (+14.0%). The German increase on CPI and dollar value is 29.7%; the Netherlands 25.0% of the total increase of 77.8%. Germany and the Netherlands have an additional near 50% increase not attributed to CPI or dollar value changes.
The foregoing tables and data make it unmistakably clear that libraries are paying high institutional prices for the journals of Elsevier, Pergamon, Springer and Academic publishers which would appear with agricultural journals to be out of proportion to their value as measured by ISI Impact Factors (see Figure 8). Of particular interest in the unlabeled table on page 19 is the fact that two of the nine journals published in the U.S. are owned by Springer and Academic. These probably have mix production between Europe and U.S. which is influencing the pricing of these U.S. journals.
The thrust of this study was to determine costs of buying journals for the library use of faculty and students in an academic institution. A multitude of details influence overall institutional costs such as space, heat, processing, salaries and page charges. These and numerous other costs were considered outside the purpose of this study and extraneous to determining the out of pocket costs of journal subscriptions. Although this fact was understood, the Faculty Taskforce decided to gather information about other influences which they felt were important in their dealings with journals and were an additional hidden cost or benefit for faculty. Therefore, page charges and reprints were examined although not translated into a quantifiable cost or benefit.
Many academics feel that having reprints of articles provides a distinct advantage and are sometimes influential in determining where an article is published. Policies were asked of publishers when the datasheets went to them for verification. The primary practice among these 314 titles is that a small number of reprints are provided free of charge. A few journals charge authors, most at reasonable prices. Some journals do not offer reprints, free or otherwise. Other than journals which offer a small number of reprints free, the pattern is very idiosyncratic. With the widespread availability of photocopy machines, this question may be of limited importance.
Page charges are direct costs authors must pay some journals to get their articles published and although this does not come as part of a library journal subscription, it probably has some influence on the subscription price. No commercial publisher in the survey levies page charges; these are entirely an activity of professional societies or associations. From one source or another, universities acknowledge the investments they make through page charges. Pressure brought on this practice and its perceived unfairness in the past ten years has altered this charge. In responses from society and association officers, it was clear that journal page charges are being dropped in a few cases and being lowered in others. The stipulation is now widespread that lack of money to pay page charges does not mean that an article cannot be accepted. Policy variations are becoming more common. An informal survey was made of the 18 organizational units of agriculture and biology units at Cornell University by the chairman of this Taskforce inquiring about the numbers and amount of money paid to journals as page charges in a year. Four of the 18 covered page charges at an average cost per unit of $982. The university does not pay page charges from any other administrative unit than these 18, although some faculty members pay from their individual grant or other special funds. The Taskforce felt that the charges paid at this time by universities is relatively minor per published paper. This information could not be systematically quantified in an acceptable manner, but the Taskforce felt that this information in the dataset was advantageous in making decisions, but that a more thorough study is needed.
The Taskforce felt that faculty members are not facing squarely the fact that they often sign away their copyright to their intellectual writings without thought to the detriment to the academic community. Faculty must be circumspect about the use of their intellectual property and more restrictive in waiving copyright to journals which have exceedingly high prices or adverse policies.
D. Implications for Scholarly Communication
Data presented in this study force several conclusions concerning journal practices as they relate to academics, their libraries, and their institutions. There would appear to be a disparity of pricing by select publishers some of which can be localized to specific countries. These facts are indisputable and may not be dismissed as a change in the U.S. dollar value although this is an influence between 1988 and 1994 when the U.S. dollar lost value. The European Community Consumer Price Index went up 7.8% more than the U.S. CPI. The commercial publishers average agricultural price increase was 77.8% for Europe. The society and association titles increase of 33.3% is the only one close to the changes accountable to the dollar and CPI influences. The improvement of the U.S. dollar against other currencies in the past two years has demonstrated only a slowing of price increases from these publishers and countries. The growing practice of canceling heavily-priced journals which academic libraries feel they must accept in order to meet restricted budgets is not acceptable as a long-term solution. The widespread availability of important scientific literature must be improved, not further restricted, and this cannot be done by curtailing access to advanced literature. The following brief sections summarize implications for select participants in this dilemma.
1. The Scholars
There are no identifiable trends in the dissemination of scholarly information which will soon alter the pattern of transfer. Electronic dissemination is coming on strongly, but is unlikely to change the publishing patterns of established journals or scholars in the near future. Scholars must face the question of how they can influence this communication system of which they are primary players.
Academic authors have been criticized for creating a system whereby they write articles for journals which their libraries must buy back for extraordinary prices for the use of colleagues and students. And for their own use. Academic institutions further encourage investments by their faculties as editors of scholarly journals and as reviewers of submitted papers. This distribution system has advantages and works relatively well. It does not work well when the journals are extraordinarily priced. The burden with very expensive journals cannot be assumed interminably by an academic institution.
The choices seem to be these:
- Do nothing about the problem and let the libraries make the adjustments as they must.
- Re-examine the pattern of publishing including frequency and duplication.
- Publish primarily in journals which have acceptable library subscription charges, and a pattern of concern about costs and the dissemination of information.
The academic community has a long tradition of supporting publishing through its professional societies and associations. This is extensive in both the agricultural and biological sciences. This has been done well in nearly all cases, providing about thirty-seven percentage of all journals in these two subjects. As has been demonstrated in the data analysis section, the societies with few exceptions have reasonable rates for academic library subscriptions and recently have expressed concern about their costs and page charges. There is a serious concern by most societal publishers to get their information out at reasonable prices and in timely and authoritative publications.
The choices available to academics should be discussed openly and in forums where some solutions can be agreed to and implemented. Effective solutions will probably only be realized if there are coordinated efforts as a community of scholars.
The publishers of agricultural and biological journals are diverse in governance and economic structure. Governments as publishers of these journals have almost disappeared, and their influence is minor among the major influential titles. University press titles are concentrated with three publishers with near fifteen titles in both agriculture and biology. These are some of the most valuable journals but only one or two titles are pricey. Societal, association and university pressesí subscription costs did not grow dramatically during the period of the study. With one or two exceptions, this group has done very well in keeping costs at a reasonable level for libraries and scholarship. The extraordinary increases of the past 15 years belong to commercial publishers which are easily identifiable from this study. All of them are for-profit corporations with immense capital and what appears to be an expansion of publishing sources which they control. These
five to ten large houses with a heavy concentration of important titles are identified with Germany, the Netherlands, the United Kingdom, and to a lesser extent the United States.
There has been activity by publishers with numerous journals exploring the delivery of journals by electronics. To date, the efforts have not been far reaching and seem to be taking paths not likely to reduce the cost of printed journal subscriptions. Several of these firms have online services offering full text or tables of contents with abstracts, or ordering of articles. All of these services are still expensive for academic libraries and few buy them. The most useful electronic/printed copy model has not yet been widely implemented to the satisfaction of scholars, libraries, and publishers. There also appears to be a consolidation of ownership of many titles, thereby influencing the future direction of electronic, scholarly publishing.
The choices for commercial publishers intending to continue in scholarly communications seem to be these:
- Do nothing about the problem of cost, and let the libraries make the adjustments as they must.
- Re-examine the concept of high subscription costs for institutions.
- Reduce expenses where feasible so as to lower subscription costs.
- Consolidate some journal titles, implement more stringent editing practices, including the reduction of duplication of publication.
There are advantages in maintaining scholarly journals published by commercial firms which must not be lost. But it is clear that the high subscription costs charged by select commercial publishers must be addressed or the transfer of research and instructional information may be greatly impaired.
3. Libraries and Academic Institutions
Libraries are faced with many financial complications which has always been the case. Balancing service to the public with an adequate collection of literature is what library management must do. Library budgets may not have kept abreast of financial growth as in other areas of an academic institution, but they do not seem far in arrears. Several points evident from the data gathered are not new to library administrators although they may have provided some different perspectives. This information will not be well known to most faculties or university presidents and administrators.
What is clear to the librarian from this study is that many of the core journals with immense price tags seem not to be as highly ranked as the price might cause one to conclude. There appears to be a disparity between price and value which is not easily reconciled. The lack of positive correlation in Figure 8 with price and impact factors for fifteen to twenty journals provides some doubts. The solution has often been based on empirical observations resulting in the cancellation of those journals which appear to be far out of financial reach for the benefits accrued. This leads to decisions which probably only extenuate the problem for the publisher.
An academic institution is constrained to provide information and library services for its faculties. Some faculties demand more than others. Library expenditures at a large university campus are enormous and it is reasonable for university administrators to demand economies where they can be realized. The point at which these economies cut into the basic information and library needs of the faculties will help determine the quality and shortcomings of a university. Ownership of the literature and its use are joint functions of university officers, the faculties, and the librarians. This intersection of decisions is under pressure and one of the most obvious and easily observed is that of library journal prices which many believe excessive.
The choices for the academic library and the university administration seem to be these:
- Do nothing about the problem, and make the library adjustments as needed.
- Establish a utility model based on accrued value and subscription cost which can be used automatically as an alerting guide or decision tool.
- Discuss with primary publishers who have extraordinary journal costs the problems involved and seek some alternative solutions to those currently employed, or seek discussions with these publishers through standard professional sources.
- Mount a campaign to apprise university faculties or professional organizations of the need for concerted efforts to reach a better resolution on scholarly publishing and costs to libraries.
- Consider major changes in the ìpublish or perishî syndrome, or in rewards and recognitions.
This journal price study was reviewed by the members of the faculty Task Force, by members of the Mann Library Committee, and some select faculty members in the College of Agriculture and Life Sciences at Cornell University. The report was also sent to the Biometrics Unit of the College of Agriculture and Life Science for review of the statistics and possible alternative answers from the data. The report was returned from the Biometrics Unit with no major comments about errors or omissions but this observation was made: while the study focuses on the most important or core journals within each field, multivariate analysis suggests that the price differentials between commercialized and society published journals can be found within the entire range of journals in the biological and agricultural sciences.
- Cornell University
- June 1998
1Carol Tenopir and Donald W. King, "Setting the Record Straight on Journal Publishing: Myth vs. Reality," Library Journal (March 15, 1996):32-35.
2H. Craig Petersen, "University Libraries and Pricing Practices by Publishers of Scholarly Journals," Research in Higher Education (31) 4 (1990):307-314.
3Kenneth E. Marks, Steven P. Nielsen, H. Craig Petersen, and Peter E. Wagner, "Longitudinal Study of Scientific Journal Prices in a Research Library," College & Research Libraries (52) 2 (March 1991):125-138.
4Barbara Meyers and Janice L. Fleming, "Price Analysis and the Serials Situation: Trying to Solve an Age-old Problem," The Journal of Academic Librarianship (17) 2 (1991):86-92.
5Bruce R. Kingma and Philip B. Eppard, "Journal Price Escalation and the Market for Information: The Librarians Solution," College & Research Libraries (53) 6 (Nov. 1992):523-535.
6Library Journal; June 15 issue yearly.
7Bowker Annual , yearly.
8American Libraries; most recent May 1997.
9(a) Henry H. Barschall, "The Cost-effectiveness of Physics Journals," Physics Today (July 1988):56-59. (b) H.H. Barschall and J.R. Arrington, "Cost of Physics Journals: A Survey," Bulletin of the American Physical Society (33) 7(1988):1437-1447.
10A press release from the American Physical Society dated 28 August 1997 indicated that AIP and APS had been absolved of any wrong-doing and the case against them dropped.
11Allyn Jackson, "The Gordon & Breach Lawsuit Against the AMS," Notices of the American Mathematical Society (1996). 2p.
12ARL Serials Prices Project. Report...: A Compilation of Reports Examining the Serials Prices Problem Including Analyses from the Association of Research Libraries "Overview and summary"... Washington, DC: The Association of Research Libraries, 1989. 87p.
13Brookfield, Karen, ed. Scholarly Communication and Serials Prices: Proceedings of a Conference, 11-13 June 1990 sponsored by the Standing Conference of National and University Libraries and the British Library Research and Development Department. London; New York: Bowker-Saur, 1991. 155p.
14 Joe A. Hewitt, "Altered States: Evolution or Revolution in Journal-based Communications?" American Libraries (20) (June 1989):497-499.
15Philip H. Abelson, "Combating High Journal Costs," Science (244) 4909 (June 9, 1989):1125.
16Herbert White, "Scholarly Publication, Academic Libraries, and the Assumption That These Processes are Really Under Management Control," College & Research Libraries (54) 1 (Jan. 1993):293-301. 17 The Literature of the Agricultural Sciences has these individual titles published by Cornell University Press:
Agricultural Economics and Rural Sociology: The Contemporary Core Literature, by Wallace C. Olsen, 1991; The Literature of Agricultural Engineering, Carl W. Hall and Wallace C. Olsen, eds., 1992; The Literature of Animal Science and Health, Wallace C. Olsen, ed., 1993; The Literature of Soil Science, Peter McDonald, ed., 1994; The Contemporary and Historical Literature of Food Science and Human Nutrition, Jennie Brogdon and Wallace C. Olsen, eds., 1995; The Literature of Crop Science, Wallace C. Olsen, ed., 1995; The Literature of Forestry and Agroforestry, Peter McDonald and James Lassoie, eds., 1996.
The set of seven volumes was awarded the Oberly Award for agricultural bibliography by the Association of College and Research Libraries in 1997.
18 "U.S. Periodical Price Index for 1994," by Kathryn Hammell Carpenter and Adrian W. Alexander. American Libraries, May 1994: 450-456.
- Mann Library, Cornell University
- Journal Study April 1997
In the enclosed data sheet, only those titles currently published and available at Cornell were included to accommodate physical access, scanning, and cost information. A total of 174 core journals are covered in the survey of biological journals. The years 1988 and 1994 were used to gather information for each title.
- Type 1988 and 1994: This indicates the type of publisher by this code:
- 1. University; 2. Commercial; 3. Government; 4. Society or Association. Use of letters after nos. signifies if the journal was published for or with another organization: s=society; g=government.
Ads: A Yes indicates that ads are included in the pages of the scholarly text; a blank means that they are not although ads may appear elsewhere, paged separately.
Annual Subscription Price: Cost of the journal for the two years as paid by Mann Library. List price from the journal was used when not available in Mann's records.
Number of Pages: Number of pages published in that subscription year. Un-numbered or separately numbered pages separate from the main text were not counted unless they were a scholarly supplement.
Average Characters per Page: A random choice of pages for sampling was machine-determined after the page counts. Using optical character recognition (OCR) techniques these pages were scanned. The number of characters on each page was totaled and an average number of characters per page was calculated for each year.
Total Characters per Volume or Year: Number of pages multiplied by the average characters per page.
Annual Cost per Page: Annual subscription cost divided by the number of pages.
Annual Cost per 1000 Characters: Annual subscription cost divided by the number of characters in that subscription year, in thousands.
SCI Citations to Journal: As reported by Science Citation Index in its annual Journal Citation Report.
ISI Impact Factor: As computed by Science Citation Index and reported in its Journal Citation Report.
Average Figure Content per Page: Figures (not tables) as a percent of total page space. In the scanning process, graphics were "windowed out" when a page contained both text and figures. Tables and figure captions were retained and scanned as text. The figure space was measured and tallied to determine average figure space as a proportion of the total page.
Page Charges: Charges to authors for publication; in some cases they are processing fees. See the notes field for clarification.
Reprints: Availability of free reprints or if authors have to buy them.
NI: No information found.
GUIDELINES FOR JOURNAL PRICE SURVEY WORKSHEET 1997
Two worksheets are generated for each journal; one page for each survey year. A unique number is assigned to each journal. Data collected for the survey will be noted on the worksheet as collected and eventually transcribe onto a computerized spreadsheet. These guidelines explain how data should be collected and how scanned image/character files are to be managed.
1.Publisher: note who publishes the journal. If the publisher is different from copyright owner, note both publisher and copyright owner.
2. Vol(s).: note volume number(s) that are included in the year. By "year" we mean the time period the journal itself recognizes as one year. Sometimes this period does not correspond completely to the calendar year (i.e. spanning across late fall and early new year may occur). If there is some question as to how many volumes the publisher includes in a year, check information given in the first issue of the annual subscription period. Something like "3 volumes (84-86) will be published in year 1994. Annual subcription price is....." Very occasionally you might have to check the last issue of the preceding year (i.e., 1993) or the first issue of the subsequent year (i.e., 1995) to get this information--or to confirm it if doubt still persists.
Make a note when a journal includes supplements. Usually, though not always, these are included in the journal price and are bound with the regular volume issues for the publishing year.
In rare cases, the number of issues or volumes published per year is highly variable and not noted within the journal. In these cases, set the journal aside and bring it to the attention of the supervisor. Note the number of volumes & issues per year: Note number of volumes, supplements (if any), and number of issues published in that year. Usually, but not always, this is a constant figure across the years. However, also note if combined issues published (e.g., 4 vols. & 4 issues/year; 8 issues in 7 for 1994).
3. Publisher type: select appropriate code as follows:
1 - University or academic press
1s or 1g - University/academic press for society/association or govt. agency (e.g., U of Chicago Press for American Society of Naturalists). In these cases, copyright is usually owned by the society in question.
2 - Commercial press
2s or 2g - Commercial press for society/association or govt. agency (e.g., Springer International for the Federated European Biochemical Societies). In these cases, copyright is usually owned by the society.
3 - Government agency; including U.N. agency (e.g., National Research Council of Canada, U.S. Office of Technology Assessment, International Labor Office)
4 - Society or association (e.g., American Association of Cereal Chemists, American College of Physicians)
4. Annual subscription price: The price paid by Mann Library for each year will be on the worksheets. Also note the price given in first issue of the year for the annual subscription price. If an overseas publication, note airmail price in U.S. dollars if given, otherwise in the currency given.
If institutional or individual subscription price is given, note both, marked as INST and IND.
If society non-member and member prices are given, note both marked as NM and M.
The price that Cornell actually pays for the journal may be different from the price published in the journal.
5. No. pages in year: If pages are cumulatively numbered across issues within one volume, give cumulative number of pages per volume; pages for individual issues are not needed.
If more than one volume is published per year, note on the worksheet the number of pages for each volume. Calculate total for year and note.
If pages are not cumulatively numbered across issues, note number of pages per issue on the worksheet. Calculate total for year and note as well.
Be on the look out for any supplementary issues; often these are not numbered consecutively as part of the regular annual volume. Note this supplementary page total on the worksheet next to page total for volume (e.g., 467pp. + 85pp. in 1 supplement).
6. Page charges and reprints: Usually, page charges and reprint information is given in a section, or a page titled "Notes to Contributors" or "Instructions to Authors." Look for any information given on a) charges made to authors for publishing their articles in the journal (including costs of reproducing graphics) and b) reprints or offprints provided to authors free of charge (note number of reprints/offprints provided). If this information is not indicated in the "Notes to Authors" section, check fine print on inside cover or title page of journal. If no page charge or reprint information is found in either place, note with "NI" (i.e., "no information").
7. Notes: Indicate any special circumstances you run across in collecting the data (e.g., "1994 volume not found in stacks--data given is from 1995;"or: "94 subscript. price not given in journal"; or: "not published in 1994--data given is from 1995"; or: "discrepancy in issue price and subscription price"as noted above).
8. TIF/OCR section:
The overall aim of this project is to determine an average character count per page for two different years (1988 and 1994) of each journal, calculate from this the average cost of the journal per 1000 characters, and then compare these costs across the time period. The cost trend of individual journals can then be compared to average journal cost trends in general.
The estimated average character count per page of each journal will be calculated from the average count we derive from OCR conversions of 10 randomly selected pages of the 1988 and 1994 volumes of each journal respectively. We will determine the random page numbers for each journal with the help of Excel based on the total page numbers issued in the given year (i.e., based on data collected in step 5 above.)
Randomly selected page numbers (10 total) for each journal in an appropriate year will be stapled on to each worksheet as soon as these are generated. For journals that have more than one volume per year and/or non-consecutively numbered pages, the Excel facilitated conversion of random numbers to corresponding journal page numbers will also be attached.
Each randomly selected page of the journal will be scanned once, creating a TIF file. (Textbridge Pro will be used for this step of the scanning process). Each TIF file will then be OCR'd to create an OCR file. (For this step, an older, speedier version of Textbridge will be used). Thus, for each year of each journal 10 TIF image files and 10 corresponding OCR files will be created. Total number of files created for both years of each journal: 40.
Pages with figures, other graphics or some advertisements on them will require windowing these graphic features out when converting to OCR files, although any corresponding text (i.e., figure
captions) should be included in the OCR. Figures and non-advertisement grapics are to be measured with a ruler and their size noted on the worksheet for the corresponding page. The size of the journal page should also be noted at the bottom of the worksheet. These numbers are used to calculate the average amount of space on each journal page which is devoted to graphics.
Ads should be windowed out. No measuring is necessary. If an ad takes up a full page, no need to scan the page. Simply created an empty, appropriately named (see below) file using Simpletext for that page. Also be sure to make a note on the worksheet.
9. Managing TIF/OCR files:
The final phase of the data collection process involves calcuating average character counts for the years 1988 and 1994 of each journal via UNIX, using the character files created for each page. This step is the responsibility of the preservation librarian. To do its job accurately, the UNIX script depends on working with consistently and correctly named folders and files.
Two folders per journal will be created using an abbreviated journal name and the unique number assigned to it, as noted at the top of its worksheets. The abbreviation should be 8 characters long preceded by the journals assigned number. One folder will contain scanned image (i.e., TIF) files, the other will contain "character converted" (i.e., OCR) files for the given journal.
Example: 236annutmetTIF will be the folder name for the TIF files created for Annals of Nutrition & Metabolism; 237annutmet will be the folder name for the OCR files for the same journal
As noted in the preceding section, one TIF file will be created for each page and saved into the TIF folder for the appropriate volume--each journals TIF folders should thus have 20 files. TIF files will not have to be titled manually; the title automatically assigned by Textbridge will do.
Once the 10 selected pages for each year of a given journal have been scanned and filed into an appropriately titled folder, character conversion for that journal can proceed. As each file is "character converted" a separate, "OCR" file is created. Again, 20 OCR files will be generated for each journal (10 for each year of the survey). Titles for these files need to be given manually using the following schema:
and so on through
for the first 10 files corresponding to pages scanned for 1988, then beginning again with
and so on through
for the next 10 files, corresponding to pages scanned for 1994.
All 20 files should be stored in the corresponding OCR folder for that journal, i.e., 237annutmet.
Exceptions to the 10-TIF-and-10-OCR per journal year rule will occur in some cases. More specifically, when:
1) Randomly selected page number falls on page that has both vertically and horizontally oriented type. This is often the case with full page graphs. In this case, two TIF images will have to be created, then OCRd and merged into 1 file using Simpletext. TIF files for this case should be distinguished appropriately.
2) Randomly selected page number falls on a full page of advertisements, classifieds or blank page. In this case, no need to create a TIF image, simply create a blank OCR file in Simpletext and name appropriately. Again, make note on the worksheet.
10. Last but not least....: If you have any doubts or questions about a particular journal volume, leave the whole journal aside and bring it to the attention of a supervisor. Consistency in data collection is essential, so asking when in doubt is essential.
Sample Datasheet sent to a Publisher:
Journals in the Biological Sciences published by Cambridge University Press
Folder Call Number Title Publisher in 1994 Country Type Type Ads Annual Subscription Price Number of Pages Average characters per page Total characters per volume Annual Cost per 1000 character Annual Cost per Page SCI Citations to Journal ISI Impact Factor Average figure content per page Page charge Reprints Folder NOTES No. 88 94 1988 1994 % chg 1988 1994 % chg 1988 1994 % chg 1988 1994 % chg 1988 1994 Rnk 94 % chg 1988 1994 Rnk 94 % chg 1988 1994 % chg 1988 1994 % chg 1988 1994 % chg 1988 1994 1994 No. 26 QH301. G32 (PR) Genetical research. Cambridge University Press UK 1 1 $192 $279 45% 507 463 -9% 2659.13 3540.63 33% 1348176.38 1639309.38 22% $0.142 $0.170 39 20% $0.379 $0.603 29 59% 1380 1582 15% 1.47 1.94 32% 6.67% 8.22% 23.34% NI NI 50 free 57 QK564.B86 European Journal of Phycology. Cambridge UP for the British Phycological Society UK 1s 2s $140 $195 39% 405 283 -30% 2577.10 2617.15 2% 1043725.50 740653.45 -29% $0.134 $0.263 17 96% $0.346 $0.689 24 99% 549 476 -13% 1.03 0.72 -30% 10.24% 22.32% 118.06% NI None 50 free Formerly British Phycological J. Copyright owned by British Phycological Society. (1994) Pg chgs levied for pp. 17+; color illust. costs.
Journals Included in Studies
Data on Agricultural Titles
Data on Biological Titles