College and Research Libraries LAWRENCE J. PERK Secondary Publications in Education: A Study of Duplication Duplication of coverage is examined to give librarians a closer ac- quaintance with three indexes found in education-psychology li- braries. Two educational indexel overlap of coverage is studied in terms of editorial policy, growth and size of coverage, cover-to-cover vs. selective indexing, and the subject orientation of the periodicals concerned in order to compare two approaches to the same discipline. One index in psychology is contrasted to the two in education, but the main focus is on the mutual interests of these two disciplines. To WHAT EXTENT do these indexes du- plicate each other? I seem to see the same articles listed." In an education- psychology library the user was refer- ring to Education Index, ERIC's Cur- rent Index to Journals in Education, and Psychological Abstracts. Naturally, the two educational indexes would du- plicate each other but to what degree and how do they differ? Are the mutual interests of education and psychology centered solely on educational psychol- ogy, or do they coincide in other areas? And what are those "other areas"? That is, what are the periodicals and what are the fields supporting these two disci- plines that they represent. To answer these questions this study examines duplication of coverage and interests between these indexing services and sketches a profile of the subject orientation of the periodicals within each service. Hopefully, the resulting picture will provide librarians with a better "feel" for the indexes which will prove useful when guiding the user. Lawrence]. Perk is an information special- ist, Mechanized Information Center, Ohio State University Libraries, Columbus. METHODOLOGY Data organization required a master list of those periodicals indexed, noting (a) which index covered a title, (b) whether the indexing was cover-to-cover or selective, and (c) the subject orien- tation of the publication according to periodical and serial directories, 1- 3 or es- timation based on the title. Totals were simply extracted from this list. The data are based on one point in time, December 197 4. But the indexes had grown considerably in the five pre- ceding years. This growth should be kept in mind when users are scanning more than the latest issues. Figure 1 shows that CIJE (Current Index to Jour- nals in Education) had doubled its cov- erage since its inception in 1969, and PA (Psychological Abstracts) had ex- panded half again as much in the same period. The time up to 1971-72 was one of rapid growth for both but has since tapered off. In contrast, the coverage of EI (Education Index) was one of rela- tive non-growth, and, therefore, its data can be better extended over time. EDUCATION There are five supporting elements in I 221 222 I College & Research Libraries • May 1977 900 800 700 J' PA....,... v I "''Il~ v / CUE / 600 '"0 Q,) 500 >< Q,) '"0 I: -ell 400 "'; / v ~ I -- Total Coverage --- Indexed -I C) ;a Cover-to-cover .S2 300 ,_, Q,) EI Pot ...... 0 200 ,_, Q,) ~ ~ ~--- !' ... -, CI]E - ~ ~-- .. .... __ PA ............ s 100 ::s z 0 1969 1970 1971 1972 1973 1974 Figure 1 Growth of Coverage by Three Bibliographic Services: Education Index ( EI ) , Current Index to Journals in Education ( CI]E), and Psycho- logical Abstracts ( P A) this study-namely, ( 1) editorial policy, ( 2) the overlap and non-overlap of cov- erage, ( 3) several indicators of the core titles, ( 4) the size of coverage, and ( 5) the subject nature of the above. All five are utilized in studying education where two indexes handling the same subject area can be compared. Editorial Policy A brief history can be helpful to identify those editorial decisions that have formed each index's particular characteristics. Education Index (EI) of the Wilson Company has been the index in educa- tion from its inception in 1929 up to the appearance of CI]E in 1969. EI's selection policy is suggested by the W.il- son Indexes Committee (ALA's Refer- ence and Adult Services Division) which has combined periodic in-depth studies with subscriber vote since 1951.4• 5 If this monitored-feedback system has been effective, the titles should represent a carefully tailored list of the most im- portant journals in the field. However, it has responded only modestly to re- quests for greater coverage. And when doing so, it deleted personal author en- tries and certain types of materials from 1961 to 1969.6 Because of this limited coverage in EI and the overall picture of secondary publications in education, the need was more forcefully stated for a new index that could provide better access to the vast amount of literature in the field, provide author entries, and supply some type of abstracting. 7 As a result the Current Index to journals in Education (CI]E) appeared in 1969 as part of the ERIC system. It, too, gathered subscrib- er information and appointed a panel, but this time the goals were less restrict- ed resulting in the following compari- sons. Overlap in Education When two independent services index the same literature, it could be expected that both would include the same titles. Actually, CI]E does duplicate 180 ( 83 percent) of EI' s 218 periodicals, indi- eating substantial agreement on the most useful titles. One difference then lies beyond this overlap of 180. CI]E's coverage of 702 periodicals is more than three times that of EI, further illustrat-· ing its policy to assure "access to impor- tant articles published in periodicals which fall outside the scope of educa- tion-oriented literature."8 In contrast, EI prefers a non-growth coverage. However, while CI]E expands deeply into noneducation-oriented literature, it bypasses thirty-eight periodicals or 17 percent of what EI considers relevant to education. Of these, twenty-eight are in education and, therefore, not merely alternatives for peripheral choices. (See Schorr for a list of the thirty-eight periodicals. ) 9 Core Periodicals Since both services want to include the major periodicals of the field, the core should exist in those chosen by both. However, since the peripheral lit- erature can also overlap, a stronger indi- cator is whether a journal is fully indexed (cover-to-cover) or not. Full- indexing in this study is viewed as an editorial choice that infers total rele- vancy to the subject field. Full-indexing by several services is the ultimate indi- cator of the major journals. EI treats its entire coverage as a core collection. It indexes all 218 periodicals cover-to-cover, inferring that these choices are so relevant to education that every article in them should be acces- sible. C I 1 E fully indexes 159 ( 23 percent), 101 of which overlap with EI. It is these 101 that we can call the hard core, since they are fully indexed by both services. Beyond this level of 101 lies the differ- ence of treatment. CI]E continues with alternative choices for total indexing while preferring partial scanning for titles that EI considers worthy of full accessibility. As noted before, CI]E overlaps 83 percent of EI. But the 101 figure of full-indexing overlap reduces that fig- ure to 46 percent of EI. Therefore, while CI]E does duplicate most of Efs coverage, the type of indexing differs on half of the titles. EI includes those thirty-eight titles totally bypassed by CI]E. These last features should be kept in mind if CI]E appears to subsume Efs coverage and usefulness. It should also be noted that numerous in-volume cross- references, the use of subheadings I sub-subheadings, and a time lag of only one to two months compared to CI]E's four to . five months are some of the features that make EI an easier index to use manually and preferable to cer- tain users.to Secondary Publications I 223 On the other hand, Schorr has cast a doubt over the quality and depth of in- dexing in EI as compared to that of CI]E. In his two examples, EI used only one and no subject descriptors while C I] E used close to its customary aver- age of six per title. His article should be referred to for some additional com- parisons including costs.n .PsYCHOLOGY Compared to education, Psychological Abstracts (P A) fully indexes a smaller portion of its coverage, 138 or 16 per- cent of its 866. As with CI]E, PA relies heavily on an extensive periphery of 84 percent and Figure 1 reviews the some- what similar pattern of rapid growth that expanded each coverage. Of the total overlap, seventy ( 8 percent of P A) are surveyed by all three ser- vices, six of which are the fully-indexed hard core. If one considers any overlap with one or both of the other indexes the total overlap increases to 160 ( ui percent of PA), nineteen of which are fully indexed. Therefore, an ample number of periodicals are relevant to both disciplines. However, there is the possibility that one discipline borrowed totally from the other rather than there being a mutual interest in each other's field. This question reveals the weakness of pure overlap data alone. The follow- ing section tries to offset this factor by identifying the subject orientation of the periodicals involved. SUB]ECf ORIENTATION Tables 1 and 2 are based on the sub- ject assignments in periodical direc- tories. Of the 1,436 periodicals in this study, 84 percent were listed in Ulrich's International Periodicals Directory, Ir- regular Serials and Annuals, or New Ser.ial Titles; 13 percent had subjects added to those titles that seemed repre- sentative; and 3 percent were left with- out subject designations. Of those with subjects, 65 percent had single descrip- 224 I College & Research Libraries • May 1977 tors while each title averaged 1.4 de- scriptors. There are pitfalls in reading the data because many titles belong to several subject categories. The categories cannot be added together without distortion. But in a very general way, one can scan the table and sense the emphases and balance of coverage within each index. Education The "Total" columns of Table 1 rep- resent the total profile for each index in terms of the number of journals in that category and its percentage of the total journals covered in that index. One cannot add categories together, but one may judge a category's strength in relation to the whole and to that of other categories. An examination of Table 2 will give a picture of the num- ber of periodicals that are unique to that index ("a" columns) and those that overlap others (columns b,c). Comparison of the two education in- dexes shows somewhat parallel emphases and representation in each category. There is slightly more stress in CI]E on the periodicals in the social sciences, communication, linguistics, business, and the natural/ physical sciences. These represent the major areas contributing to its peripheral growth. They are only minimally altered when education jour- nals are removed as in the "T-Ed" col- umns, indicating those journals outside the field of education. EI consists most- ly of educationally oriented titles but does have representation in almost every category. Psychology and Education Comparison of CI]E and PA reveals the wide range of interests shared by education and psychology. CI]E scans the periodicals more intensely in linguis- tics, communication, business, and the humanities, and P A depends heavily on TABLE L TOTAL PROFILE FOR EACH INDEX BY SUBJECT ORIENTATION OF THE PERIODICALS Subject EI CIJE Total %3 T-Ed1 Total %'1 T-Ed1 Education 195 89 422 60 (Education & Psychology ) 6 (13 6) (18 3) Psychology 23 10 10 71 10 53 Psychiatry & Neurology 2 1 10 1 Business, Mgt, Economics 14 6 4 77 11 64 Children 9 4 2 23 3 14 Communication 7 3 3 64 9 44 Crime & Law 10 1 7 Ethnic Concerns 3 1 1 15 2 10 Humanities 21 10 6 65 9 43 Linguistics 12 6 82 12 41 Sciences: -General & Technology 10 5 32 5 11 -Health Sciences 13 6 8 30 4 21 -Math & Statistics 5 2 1 14 2 4 -Natural & Physical Sci's 4 2 1 40 6 33 -Social Sciences 8 4 1 78 11 57 Social Services 7 3 7 13 2 11 Other 2 1 1 5 1 3 1) Total minus "Education" category. 2) Total minus "Psychology" category. 3) Percent of the 218 periodicals indexed by EI. 4) Percent of the 702 periodicals indexed by CI]E. 5) Percent of the 866 periodicals indexed by P A. 6) Figures also included in "Education" and "Psychology" categories. PA Total %5 T-Ps2 161 20 120 (41 5) 334 39 164 19 46 5 44 22 3 12 8 1 5 21 2 21 4 0 4 26 3 21 8 1 7 22 3 18 142 16 124 10 1 7 53 6 32 94 11 65 21 2 18 10 1 6 Secondary Publications I 225 TABLE 2 THE NUMBER oF OVERLAPPING AND NoN-OVERLAPPING PERIODICALS BY THEIR SUBJECT ORIENTATION a bl a bl a bl c Subject EI EI/CIJE CIJE CIJE/PA PA PA/EI EI/CIJE/PA Education 26 166 226 91 67 64 61 (Education & Psychology)2 (13) (1) ( 17) (24) {13) (13) Psychology 21 6 66 266 23 21 Psychiatry & Neurology 1 1 2 8 156 1 1 Business, Mgt, Economics 2 12 51 16 31 2 2 Children 9 9 11 11 6 6 Communication 7 57 5 2 5 5 ( Libraries ) s (26) Crime & Law 10 21 Ethnic Concerns 3 12 4 2 2 2 Humanities 7 13 51 3 22 3 2 ( Literature) a (2) ( 1) (18) (3) Linguistics 12 67 6 2 3 3 Sciences: -General & Technology 10 20 4 18 2 2 -Health Sciences 12 8 19 122 10 9 -Math & Statistics 5 7 2 8 -Natural & Physical Sci's 4 35 3 50 2 2 ( Consvr/Envir )8 (21) {1) -Social Sciences 1 8 54 19 75 3 3 (Sociology )8 (3) (24) (14) (44) (3) (3) Social Services 1 6 6 15 Other 2 3 10 1) Includes Column c. 2) Also counted in "Education" and "Psychology" categories. 3) Also counted in above category. the health sciences. But beyond these areas, both indexes show an interest in every category. Within the format of the data, edu- cational psychology can only be nar- rowed to those titles classified by both "education" and "psychology" subject headings. According to Table 1, more of these can be found in PA, while Table 2 shows that all but one periodi- cal can be found in P A (Table 2, CIJE "a" column). Therefore, without con- sidering full or partial indexing as a factor, P A should give access to virtual- ly every educational psychology journal covered in these indexes. The tables show that a mutual inter- est in each other's discipline does exist. However, by more than a two-to-one ratio (Table 1), P A appears to scan more periodicals in education than do EI and Cl]E for psychology. This in- cludes an oddity in P A's unique cover- age (Table 2, P A ~~a'' column) where sixty-seven educational journals are in- dexed by P A and ignored by the educa- tional indexes. Of these, twenty-six also concern psychology, but that still leaves forty-one. PA's international orienta- tion accounts for twenty-nine foreign language journals, but again that leaves thirty-eight unique educational titles in psychology. Consequently, to scan all of the educational journals in these in- dexes, one must use all three in order to have · access to those sixty-seven unique to P A and the twenty-six found solely in EI. On the other hand, one must use Cl]E for only six to eight unique titles in psychology and EI for one more in psychiatry. SUMMARY AND CoNCLUSIONs The intent of this study is to provide data that would give librarians a closer acquaintance with these indexes. The 226 I College & Research Libraries • May 1977 data are reduced to figures and tables which can be scanned according to spe- cific interests. In sum, the overlap pro- vides a series of comparisons which con- sider full and partial indexing, the growth and size of coverage, and the subject nature of the periodicals con- cerned. Comparison of the educational indexes reveals that the C I] E scans three times as many journals as does EI. How- ever, it does not subsume EI completely, and the two differ on choices for full indexing. Each still has advantages for particular users. In fact, for access to all of the edu- cational periodicals covered, each of the indexes must be used. On the other hand, almost all psychology and psychi- atry titles of the three are accessible through P A alone. While P A scans more periodicals in education than do EI and CI]E in psychology, the two have a somewhat similar breadth of interests beyond educational psychology. REFERENCES I. Ulrich's International Periodicals ]jirectory, 15th ed. (New York: Bowker, 1974). 2. Irregular Serials and Annuals: An Interna- tional Directory, 3d ed. (New York: Bow- ker, 1974)-? 3. New Serial Titles (Washington, D.C.: Li- brary of Congress ) . 4. Edwin B. Colburn, ''The Wilson Periodical Indexes: A General Statement," MLA Quarterly 30:92-95 (June 1969). 5. James Benson, "The Education Index," MLA Quarterly 30:118-23 (June 1969). 6. Arvid }. Burke and Mary A. Burke, Docu- mentation in Education (New York: Teachers College Press, 1967), p.283-95. 7. Saul Herner, et al., Study of Periodicals and Serials in Education. Final Report (Be- thesda, Md.: ERIC Document Reproduc- tion Service, ED 017 747, 1968). 8. As noted in "Preface" to vol. 1, Current In- dex to Journals in Education (New York: CCM Information Corp., 1970), p.vi. 9. Alan Schorr, "Education Index and Current Index to Journals in Education: Do We Really Need Both?" Journal of Academic Librarianship 2:135-36 (July 1976). 10. Lawrence }. Perk, Review of Cl]E and Education Index, "Education" section, in Serials Review 1:40 (Jan./June 1975). 11. Alan Schorr, "Education Index and Cur- rent Index to Journals in Education."