Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Bibliometric analysis of reading research journal literature 1976

You don't seem to have a PDF reader installed, try download the pdf

Item Metadata

Download

Media
UBC_1976_A8 B37.pdf
UBC_1976_A8 B37.pdf
UBC_1976_A8 B37.pdf [ 9.7MB ]
Metadata
JSON: 1.0078371.json
JSON-LD: 1.0078371+ld.json
RDF/XML (Pretty): 1.0078371.xml
RDF/JSON: 1.0078371+rdf.json
Turtle: 1.0078371+rdf-turtle.txt
N-Triples: 1.0078371+rdf-ntriples.txt
Citation
1.0078371.ris

Full Text

BIBLIOMETRIC ANALYSIS OF READING RESEARCH JOURNAL LITERATURE by EAVID J. HARNETT B.A., University of Alberta, 1967 B.Ed., University of Alberta, 19 71 A THESIS SUBMITTED IN PARTIAL FULFILHENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ARTS in the Department of READING EDUCATION FACULTY OF EDUCATION We accept t h i s thesis as conforming to the required standard. Adviser and Thesis Supervisor Header UNIVERSITY OP BRITISH COLUMBIA ( September, 1976 (o) David J. Barnett, 1976 In presenting th i s thes is in pa r t i a l fu l f i lment of the requirements for an advanced degree at the Un ivers i ty of B r i t i s h Columbia, I agree that the L ibrary sha l l make it f ree ly ava i l ab le for reference and study. I further agree that permission for extensive copying of th i s thesis for scho lar ly purposes may be granted by the Head of my Department or by his representat ives. It is understood that copying or pub l i ca t ion of th i s thes is fo r f i nanc ia l gain sha l l not be allowed without my writ ten permission. Depa rtment The Univers i ty of B r i t i s h Columbia 2075 Wesbrook Place Vancouver, Canada V6T 1W5 ABSTRACT i i The results from a .biblio metric analysis of reading research journal l i t e r a t u r e are reported i n t h i s i n v e s t i g a t i o n . The major purposes of the study were to: establish a sample of reading research. l i t e r a t u r e ; determine the core journal structure of the sample; describe developmental c h a r a c t e r i s t i c s of the reading research journal l i t e r a t u r e ; and i l l u s t r a t e d i s c i p l i n a r y connections among journals reporting reading research. Summaries from the Annual Summary of Research on Reading (ASHR) for the years 1959, 1964, 1S68 and 1972 — representing the years 1959 to 1972 — provided the l i t e r a t u r e for analysis, 768 (84 percent of the total) of the journal a r t i c l e s appearing in the four summaries were collected and provided the referencing and c i t e d sets of journal t i t l e s . Three major analyses were performed. In the f i r s t , the referencing c o l l e c t i o n of journal a r t i c l e s was described and sets of core journals l i s t e d . Developmental c h a r a c t e r i s t i c s of reading research were described in the second using a number of bibliometric measures including average number of references per a r t i c l e , age of c i t e d materials, type of publication c i t e d , frequency of author s e l f - c i t e s , and patterns of multiple authorship both i n the referencing and c i t e d set of journals. In the t h i r d analysis, two c l u s t e r i n g programs (UBC C-Group and O s i r i s Hiclust) were used to s t a t i s t i c a l l y group the core cited journal t i t l e s . £°.E! J°J2I__1 Structure Core journals were i d e n t i f i e d using three c r i t e r i a : number of a r t i c l e s appearing i n the ASRR, quantity of references produced by the a r t i c l e s , and volume of c i t a t i o n i n the referencing set of journals. For a l l three l i s t s , the most productive journals accounting for 50 and 80 percent of the t o t a l a r t i c l e s , references and c i t a t i o n s i n the two sets of journal t i t l e s are i d e n t i f i e d . The journals i s o l a t e d as the cores for the three l i s t s fellow tha general Pareto d i s t r i b u t i o n , confirming e a r l i e r work by Price (1965), G a r f i e l d (1972) and others, thus demonstrating the predominance of small cores of highly productive journals in the reading research information network. Comparison revealed the three core l i s t s represent subject areas such as reading, growth and development, curriculum, educational research, general education, educational psychology and several areas of psychology. The d i s c i p l i n e d i v e r s i t y of the journal t i t l e s increased markedly with the selection c r i t e r i o n based on volume of c i t a t i o n in the referencing set of journals. £svelo__ental C h a r a c t e r i s t i c ^ Based on the re s u l t s of the study, and comparison with research using other l i t e r a t u r e s , the following developmental c h a r a c t e r i s t i c s for reading emerged. Reading research i s becoming a more scholarly f i e l d using quantity of c i t a t i o n s per a r t i c l e as a c r i t e r i o n . There i s a s l i g h t movement toward a more immediate research front, indicated by age of cited materials, but t h i s i s not strong and the f i e l d s t i l l r e l i e s heavily on archival and near archival resources i n i t s research. A movement toward generation of science-like paradigms may D S developing, based on proportion of s e r i a l and monographic usage, but t h i s i s i v tentative at best and not yet a pronounced trend, Beading research may be becoming more cumulative as indicated by increasing author s e l f - c i t a t i o n . F i n a l l y , based on multiple authorship data, reading research i s d e f i n i t e l y becoming more collaborative. £lssteri.ng of Journal T i t l e s Two s t a t i s t i c a l algorithms, one using c o r r e l a t i o n a l techniques and the other Euclidian distances i n n-dimensional space, were applied to the 36 core cited journal t i t l e s . I n t u i t i v e l y acceptable journal groupings were produced i n the clu s t e r analysis with the two programs generally confirming each other. Ten journal groupings emerged. Three were somewhat ambiguous with the remaining seven i l l u s t r a t i n g strong i n t e r r e l a t i o n s h i p s suggesting the existence of c l u s t e r s of i d e a t i o n a l l y related content among subjects i n journals reporting reading research. Recommendations for further research include; s t a t i s t i c a l analysis of the dispersion of the i d e n t i f i e d core journal l i s t i n g s ; comparison of the core cited' journals with recent issues of the ASBR and ERIC's CIJJ ; further study of author productivity i n reading research; development of a Journal of S e a l l ^ Ij£ortajnt Pagers , analysis of conceptual research fronts in reading research; broader analysis of the extent to which archival sources are used i n reading research; analysis of c i t e d journal t i t l e s which emerged as clu s t e r s to delineate conceptual maps related to reading research; and development of a Reading Research Literature Ci t a t i o n Index based on the annual summary of research. V TABLE OF CONTENTS CHAPTER PAGE LIST OF TABLES ........ - X LIST OF 'FIGURES x i LIST OF APPENDICES , x i i I STATEMENT OF THE PEOB1EM 1 Background And Significance Of The Problem ..... 3 Objectives And Focus Of The Study .............. 8 Limitations 13 Definition Of Terms 16 Summary And Overview 17 II REVIEW OF THE LITERATURE AND CONCEPTUAL FRAMEWORK . 18 Introd uction 18 Growth Of Science And S c i e n t i f i c Literatures ... 21 Citat i o n Analyses 25 Core Journals ............................... 27 Citat i o n s Per A r t i c l e ....................... 29 Age Of Cited Material 30 Type Of Publication ......................... 33 S e l f - C i t a t i o n 35 Multiple Authorship 35 Clusters Of Journal T i t l e s 37 Bibliometric Analysis Of Reading Research Journal Literature 38 Time Sampling 39 Core Journals 40 Citations Per A r t i c l e ....................... 4 1 Age Of Cited Material 4 2 Type Of Publication 42 Sel f - C i t a t i o n s 43 v i TABLE OF CONTENTS CHAPTER PAGE Multiple Authorship 4 3 Cluster Analysis 43 Summary 44 III DESIGN, METHODOLOGY, AND STATISTICAL ANALYSIS ..... 48 Development Of The Journal Literature C o l l e c t i o n 4 9 Description Of The Journal Literature C o l l e c t i o n 50 Determination Of The Developmental Characteristics Of The Coll e c t i o n 51 Citations Per A r t i c l e 51 Age Of Cited Material 51 Type Of Publication . 52 Sel f - C i t a t i o n s 52 Multiple Authorship 52 Cluster Analysis Of The Cited Journals ......... 53 The Clustering Rationale And Selection Of The Two Programs 53 Preparation Of The Cluster Data ............. 57 Reduction Of The Matrix ..................... 59 Adjustment For S e l f - c i t a t i o n ................ 61 Clustering S t a t i s t i c a l Analysis ............. 62 Summary 6 3 IV RESULTS OF THE STUDY 64 Description Of The Referencing Journal Literature C o l l e c t i o n 64 Id e n t i f i c a t i o n Of Core Journals 69 Comparison Of The Core Literature Results ... 87 v i i TABLE OF CONTENTS CHAPTER PAGE Developmental Characteristics Of The Reading Research Journal Literature Collection ......... 91 Citations Per A r t i c l e 91 Age Of Cited Material 91 Type Of Publication 93 S e l f - C i t a t i o n 95 Multiple Authorship 98 Cluster Analysis Results 101 C Group Results 103 Hiclust Results 114 Summary 122 Citations Per A r t i c l e 122 Age Of Cited Material 123 Type Of Publication 124 S e l f - C i t a t i o n 125 Multiple Authorship 125 Cluster Analysis 126 7 0 SUMMARY, CONCLUSIONS AND IMPLICATIONS, FURTHER RESEARCH 127 Summary 127 Conclusions And Implications 128 Core Journal Structure ...................... 128 Developmental Characteristics 131 Clustering Of Journal T i t l e s 136 Further Research ..... 138 BIBLIOGRAPHY 14 2 APPENDICES 149 v i i i LIST OF TABLES TABLE PAGE I AS.S1 FOR FOUR TIME PERIODS ........... . . 66 • II TOTAL ARTICLES AND PERCENT OF JOURNAL ARTICLES, ASRR.. 67 III TOTAL JOURNAL ARTICLES AND PERCENT OF ARTICLES LOCATED, ASRR 6 8 IV JOURNALS BY NUMBER OF ARTICLES IN ASHE IN POUH TIME PERIODS 71 V JOURNALS BY NUMBER OF RE.FEBENCES PRODUCED IN ASRR IN FOUR TIME PERIODS 76 VI JOURNALS BY NUMBER OF CITATIONS RECEIVED FROM AS ER IN FOUS TIME PERIODS 82 VII CORE JOURNALS PRODUCING 50 PERCENT OF AETICLES, REFERENCES, AND CITATIONS 90 VIII AVERAGE NUMBER OF REFERENCES PER ARTICLE . .... 92 IX AGE OF CITED MATERIALS 94 X PERCENTAGE OF CITATION BY TYPE OF PUBLICATION ..... 96 XI FREQUENCY OF AUTHOR SELF-CITES 97 XII FREQUENCIES OF REFERENCING JOURNAL ARTICLES HUH ONE, TWO, THREE, AND FOUR OR MORE AUTHORS FOR FOUR TIME PERIODS 99 XIII FREQUENCIES OF CITED DOCUMENTS WITH ONE, T«fO, THREE, AND FOUR OR MORE AUTHORS FOR SIX TIME PERIODS . . . 100 ix LIST OF FIGOEES FIGUSE PAGE 1 CITED JOURNALS IN SIMPLIFIED TWO-DIMENSIONAL SPACE BEFORE NORMALIZATION 58 2 CITED JOURNALS AFTER NORMALIZATION 60 3 JOURNALS CONTRIBUTING ARTICLES TO ASRR ............ 75 4 ' JOURNALS PRODUCING REFERENCES IN ASRR 80 5 JOURNALS RECEIVING CITATIONS FROM ASRR ............ 88 6 CLUSTER ANALYSIS OF JOURNALS RELATED BY CITATION TO THE FIELD OF READING RESEARCH (C-GROUP METHOD) . 104 7 RELATIVE ERROR AT EACH STEP OF C-GROUP CLUSTER ANALYSIS 105 8 CLUSTER ANALYSIS OF JOURNALS RELATED BY CITATION TO THE FIELD OF READING RESEARCH {HICLUST METHOD) . 115 9 CORRELATION COEFFICIENT AT EACH STEP OF HICLUST CLUSTER ANALYSIS 117 X L I S T O F A P P E N D I C E S A P P E N D I X P A G E A T H E C L U S T E R I N G P R O G R A M S . . . . . . 150 B C O M P L E T E T A B L E S I V , V A N D V I 153 ACKNOWLEDGEMENTS I wish to express my sincere gratitude and appreciation to the following people for th e i r contributions toward the completion of t h i s d i s s e r t a t i o n . Dr. E. G.,Summers, my adviser and thesis supervisor, for his help in i n i t i a l l y defining the problem and for his ideas and enthusiasm at a l l stages of the task. Ha has made the production of this dissertation an invaluable learning experience. I am indebted to him for introducing me to the f i e l d of information science and to the importance of the relat i o n s h i p between knowledge about a subject and knowledge about the l i t e r a t u r e of a subject. Dr. R. D. Chester, Department Chairman, for the i n t e r e s t he took in this thesis and his contribution as Reader. Mr. L. E. Varga, computer programmer, for the many o r i g i n a l programs, the hours of work, and the many suggestions on matters of s t y l e . Ms. S. Jeroski for tracking down certain l i b r a r y materials. F i n a l l y , I would l i k e to thank Mr. J. Nightingale of the Computing Centre for his assistance with the UBC FMT program. 1 CHAPTER I STATEMENT OF THE PROBLEM The published l i t e r a t u r e i n a d i s c i p l i n e provides the vehicle for communication and information transfer among the members of the d i s c i p l i n e . A l i t e r a t u r e i s a body of thought expressed in published writings and represents an archival record that can be c a r e f u l l y analyzed and studied. Many sources make up the print information system of a d i s c i p l i n e , but the journal a r t i c l e constitutes one of the more important a r c h i v a l l i n k s . The major purpose of t h i s study i s to examine a c o l l e c t i o n of journal a r t i c l e s and use the data contained in th e i r bibliographic references to describe some of the emerging c h a r a c t e r i s t i c s of the f i e l d of reading research. The research was conducted i n the t r a d i t i o n of unobtrusive measurement as described in the well-known work of Webb et a l . (1965). In introducing their unobtrusive measures study of bibliographic c i t a t i o n s in behavioral science journals, Parker, Paisley, and Garrett (1967:1) state: The common element in t h i s quite varied, highly pragmatic and usually imaginative research strategy i s the observation and analysis of natural a r t i f a c t s or traces of the phenomenon being studied, rather than measurement or observation that might influence behavior. There i s l i t t l e danger that the behavior being studied i s aty p i c a l (as i n some experiments); no danger that the responses are distorted to f i t the predispositions of the questioner or the ideals of the Note: This d i s s e r t a t i o n was formatted by computer using the UBC FMT documentation program available through the Computing Centre, University of B r i t i s h Columbia. 2 respondent (as i n some observational techniques). There are obvious disadvantages in unobtrusive measurement: natural traces don't always provide di r e c t answers about the behavior the researchers would l i k e to observe, or ask about, or d i r e c t l y manipulate, and the imagination of the researcher i s often taxed i n the more complicated ana l y s i s , and inference processes that the strategy often requires. There are good reasons for examining the c o l l e c t i o n of research l i t e r a t u r e i n the f i e l d of reading. In any d i s c i p l i n e , the unobtrusive written record can be used to i d e n t i f y patterns of growth and development. For example, Kochen (1963) suggests that the primary concern of a new d i s c i p l i n e should be with the conditions under which the growth of knowledge becomes stable. He notes that while knowledge grows i n individuals by "natural learning processes", knowledge grows i n communities by other natural processes, one of which i s s c i e n t i f i c growth and advancement. Patterns i n the growth of knowledge, and i n s c i e n t i f i c change, can be revealed by analysis of the l i t e r a t u r e output of a d i s c i p l i n e . Price (1963), i n what has become a c l a s s i c work, presented a series of unique information measures, based on counts of s c i e n t i s t s and publications, that are useful in establishing the growth c h a r a c t e r i s t i c s of any d i s c i p l i n e . Kuhn (1962) suggests that linkages among bibliographic c i t a t i o n s i n a c o l l e c t i o n of journal l i t e r a t u r e can reveal the communication network which in turn defines subgroups of s c i e n t i s t s who share similar paradigms. Such paradigms are s t i l l emerging in most d i s c i p l i n e s , and patterns in their information bases, p a r t i c u l a r l y i n the cutting edge journal l i t e r a t u r e , reveal these changes. The journal thus plays an important role in the communication network of any d i s c i p l i n e . Analysis of journal 3 a r t i c l e s , which constitute the primary l i t e r a t u r e , can provide insight into growth patterns i n reading research and the in t e r r e l a t i o n s h i p of reading with other d i s c i p l i n e s . The general goals of the present study include: (1) specifying a time period and selecting a set of a r t i c l e s representing that time period to provide a c o l l e c t i o n of referencing reading research journal l i t e r a t u r e for analysis, (2) describing the referencing c o l l e c t i o n of reading research journal l i t e r a t u r e , (3) analyzing the bibliographic references contained i n the referencing c o l l e c t i o n to: -describe selected c h a r a c t e r i s t i c s of the growth and development of the f i e l d of reading research, -determine the patterns of interconnections among journals representing d i s c i p l i n e s as cited i n the referencing c o l l e c t i o n . BACKGROUND AND SIGNIFICANCE OF THE PROBLEM Most of the previously reported research involving bibliometric analysis of the reading research journal l i t e r a t u r e has consisted of tabulating freguency data on selected variables. These data have then been used to i d e n t i f y sets of important journals or to describe changes i n the quantity of l i t e r a t u r e being published for spec i f i e d time periods. This study adds to such research, but another important dimension i s included. Intensive analysis i s made, for the f i r s t time, of the bibliograjjhic references i n the journal a r t i c l e s reporting reading research. The bibliographic reference, whether i t i s found i n a footnote or i n a bibliography, has been used i n a number of ways in research on use of information and information sources. The value of such analysis i s increasingly 4 recognized by other d i s c i p l i n e s and f a l l s under the general l a b e l of " c i t a t i o n analysis" i n the l i t e r a t u r e of information science. Analysis of bibliographic c i t a t i o n s provides data that can be used to pinpoint trends in the growth and development of the f i e l d of reading research. Through use of c l u s t e r i n g algorithms, c i t a t i o n analysis can also reveal patterns of i n t e r d i s c i p l i n a r y relationships among the journals which are cited by the journals which report research in reading. The research methodology of c i t a t i o n analysis i s based on several underlying assumptions. F i r s t , i t i s assumed that i t i s possible to i d e n t i f y a source or set of sources for the references which are representative of the d i s c i p l i n e being studied with respect to d i s t r i b u t i o n and type of c i t a t i o n . Second, i t i s assumed that there i s a positive r e l a t i o n s h i p between an author's use of a piece of material and the c i t a t i o n of that material as a reference item i n an a r t i c l e . That i s , the author i s c i t i n g a l l documents and sources used that are d i r e c t l y related to the work and that c i t a t i o n s are not being made only for the purpose of increasing the number of sources or f o r the prestige value they contribute to the work. Third, i t i s assumed that the cited materials i n an a r t i c l e are conceptually and substantively related to the topic under consideration and that materials from other d i s c i p l i n e s indicate legitimate i n t e r d i s c i p l i n a r y connections with the f i e l d of the referencing a r t i c l e and r e f l e c t the transfer of ideas from d i s c i p l i n e to d i s c i p l i n e . F i n a l l y , i t i s assumed that frequency of c i t a t i o n does have a d i r e c t relationship to the value of the cited material. That i s , a substantial correlation e x i s t s between 5 frequency of c i t a t i o n and the value of the work or journal being cit e d . This i s an appropriate time in the emergence of reading as a d i s c i p l i n e for conducting a broadly based bibl i o m e t r i c study of the primary research - journal l i t e r a t u r e related to the f i e l d . There are several c r i t e r i a which can be applied i n judging whether a f i e l d of study i s a d i s c i p l i n e including growth i n the number of people involved in the d i s c i p l i n e , the role of the d i s c i p l i n e i n the academic structure, the existence of a national learned society, and the strength of the publication program and presence of journals to publish scholarly a c t i v i t y in the f i e l d . The l a s t twenty years have seen steady growth i n the number of people involved i n the f i e l d of reading. The International Reading Association has developed into one of t.he largest and fastest growing learned s o c i e t i e s in the world, and provides a broad base for a c t i v i t i e s i n reading. The Association also includes membership and chapters on an international basis. In terms of the role of reading within the academic structure, a recent IRA study (Guthrie, 1976) l i s t s over 300 colleges and u n i v e r s i t i e s i n Canada and the United States which offe r programs leading to advanced graduate degrees in the f i e l d of reading. The IRA conducts extensive l o c a l , regional and international professional conferences, in addition to i t s yearly massive national conference on reading, and s p e c i a l symposia and i n s t i t u t e s of various kinds. Another v i s i b l e indicator of the status of reading as a d i s c i p l i n e i s the presence of an extensive publishing program including journals 6 such as tha Reading, Teacher, the Journal of Reading, and the Research _ u _ r t e r l _ . Reading also f a l l s within the category ox an emerging research d i s c i p l i n e , Donohue and Karioth (1966), basing t h e i r analysis on work by Berry (1965) and Hoselitz (1961), suggest that three conditions t y p i f y a new emerging d i s c i p l i n e : (1) a set of problems exist which have attracted the attention of several investigators, (2) s u f f i c i e n t data has been c o l l e c t e d to promulgate broad generalizations which focus on common features of the problems under investigation, and (3) the d i s c i p l i n e has attained o f f i c i a l or i n s t i t u t i o n a l recognition. In reading research, the problems, although as yet i l l defined, are ce r t a i n l y recognized by the f i e l d . Data i s widely available but has yet to be cumulated within a conceptual/theoretical framework adequate for broad scale generalization. Reading has attained o f f i c i a l recognition, both i n s t i t u t i o n a l l y and in t e r n a t i o n a l l y , and thus q u a l i f i e s as an emerging research d i s c i p l i n e . Increases i n the volume of research journal l i t e r a t u r e , i n p a r t i c u l a r , are strong indicators of the growth and development of reading as a d i s c i p l i n e . There has been sustained and steadily expanding in t e r e s t in research on reading for the past 50 years. Gray i n 1925 noted the increasing volume of published reading research and organized the f i r s t comprehensive review of l i t e r a t u r e i n the f i e l d . This landmark review covered 4 36 research a r t i c l e s published prior to 1924. Since publication of that f i r s t review, a summary of research related to reading has appeared on an annual basis for the past f i f t y years. 7 Summers (1968) reported a study which analyzed the yearly summaries and tabulated and compared their growth from t h e i r inception through the 1965-66 edi t i o n , A set of important core journals was sp e c i f i e d based on a decade by decade analysis of the volume of a r t i c l e s produced across journals reporting reading research. Summers concluded that the 1965-6 6 summary l i s t e d better than four times as many journals carrying seven times a many a r t i c l e s as the f i r s t annual summary appearing four decades e a r l i e r . This represents a 400 percent increase i n journals reporting reading research and a 7 00 percent increase i n the number of actual a r t i c l e s reported. This approximates an exponential increase and approaches the information "flood" stage as described by L i c k l i d e r i n 1966. Along with t h i s increase i n productivity, Summers also noted an increased d i s c i p l i n e d i v e r s i t y among the t i t l e s of the core journals producing the greatest volume of research a r t i c l e s . His analysis did not extend to the references, i n the a r t i c l e s i n the summaries. Kling (1971) l a t e r underscored the same d i s c i p l i n e d i v e r s i t y noted by Summers based on his own evaluation of substantial quantities of reading research i n the preparation of f i v e comprehensive reviews for the U.S. Office of Education's Targeted Research and Development Program on Reading. Reading research appears to be s h i f t i n g from a p r a c t i c a l l y oriented, educationally based l i t e r a t u r e to a broader i n t e r d i s c i p l i n a r y focus. Examination of the contents of more recent annual summaries of research also reveals t h i s changing pattern i n the type of l i t e r a t u r e reported, and suggests that the d i s c i p l i n a r y 8 base in reading research, i s broadening. With the growth of reading as a d i s c i p l i n e , the increase in volume of reading research reported, and the frequently offered hypothesis that such research i s also broadening i n i t s i n t e r d i s c i p l i n a r y content, the time seemed suitable for conducting a study which examines the archival record and adds to the description of the primary research journal l i t e r a t u r e in the f i e l d , pinpoints some of i t s developmental and growth c h a r a c t e r i s t i c s , sheds l i g h t on the d i s c i p l i n a r y interconnections among the journals c i t e d i n the reading research journal l i t e r a t u r e , and compares reading as a research area with other d i s c i p l i n e s for which bibliometric data are available. OBJECTIVES AND FOCUS OF THE STUDY The i n i t i a l task In the study was to establish a reasonable time period and specify a set of journal a r t i c l e s for analysis. Journals are preferred to monographs or other c i t a t i o n sources because they most adeguately r e f l e c t current p r i o r i t i e s and research fronts and inte r e s t s in the f i e l d . They are the chief medium for the d i f f u s i o n of emerging and developing areas of concern i n a d i s c i p l i n e . In addition, journals are rafereed and thus have undergone professional screening and possess some measure of guality c o n t r o l . The source of references being analyzed i s the most c r i t i c a l factor in any c i t a t i o n study. To adeguately r e f l e c t the ch a r a c t e r i s t i c s of a subject l i t e r a t u r e , the references must be t y p i c a l of the materials being used for research in that p a r t i c u l a r area. A standard approach i n c i t a t i o n analysis i s to 9 f i r s t subjectively generate a l i s t of journals which are thought to be most relevant to the goals of the research. A variation i s to spread t h i s decision across a broader group by using a jury approach i n selecting journals. Once the journals are selected, an appropriate time period i s designated, and the a r t i c l e s produced by the selected set of journals during that period become the data base for the study. Many variations of t h i s approach have also been reported, including selection and use of a single representative journal source. In t h i s study, the decision was made to sample the annual summary of research on reading (ASRR) in developing the data base of journal a r t i c l e s . This provides a sampling from the "leading journals of the area being studied and accurately r e f l e c t s the f i e l d of reading research. The time period was established as approximately 1959 to 1972 and the summaries for the years 1959, 1964, 1968, and 1972 were selected f o r processing. The rationale for these decisions i s presented more f u l l y i n Chapter I I . B a s i c a l l y , i t was f e l t that use of the summaries maximized professional input in establishing the journal data base and i n generating a reasonably representative set of heterogeneous journal a r t i c l e s for analysis. Three major analyses comprise the study. In the f i r s t , information i s generated that i s used i n describing the referencing journals from the ASRR which constitute the data base for the study. This involves determining the t o t a l number of journal a r t i c l e s l i s t e d in the summaries and specifying the a actual number of a r t i c l e s obtained for analysis. The journals producing these a r t i c l e s are then tabulated. This data i s also 10 used to define two sets of core journals for the ASRR based on rank ordered frequency counts of a r t i c l e s and the references contained in those a r t i c l e s . The second analysis focuses on providing data which can be useful i n indicating developmental trends i n reading research as reflected by the journal l i t e r a t u r e . The data are analyzed to i l l u s t r a t e c h a r a c t e r i s t i c s on the following variables: average number of references per journal a r t i c l e , age of ci t e d materials, types of publications c i t e d , frequency of author s e l f - c i t e s , patterns of multiple authorship i n both the referencing and cited materials, and core journals based on the cited materials. A l l the above variables have been found to be of value i n previously reported c i t a t i o n studies for i l l u s t r a t i n g developmental c h a r a c t e r i s t i c s of a body of l i t e r a t u r e . In the t h i r d analysis, the major objective i s to group the cit e d journal t i t l e s contained in the journal a r t i c l e s which appeared i n the ASRR using a s t a t i s t i c a l approach. Cluster analysis i s used for t h i s grouping with the t o t a l body of cited materials for the four time periods conflated into one set of data. In cluster analysis, the journals are grouped based on the way they were used by the authors who wrote the a r t i c l e s rather than ad hoc groupings based on the subjective analysis of referencing journal t i t l e s . Clustering provides a technique f o r automatically dividing a data set into natural groups without previously specifying the groups in any way. The technique i s applied to journals in the following fashion. The contributors to a journal w i l l refer to or c i t e 11 a r t i c l e s in other journals. When the referencing behaviors of the contributors to one journal are pooled and treated c o l l e c t i v e l y , i t becomes possible to examine the . "preference" that the journal has for other journals i n terms of the frequencies with which i t c i t e s them. It i s also possible to examine the varied preference that a journal obtains from a set of journals by examining the varied frequencies with which i t i s cited by them. The frequencies obtained by two or more journals may be compared by examining the s i m i l a r i t i e s in their frequency patterns (examining for co-variance across the referencing set). Journals with s i m i l a r patterns could be grouped or "clustered". Thus, s i m i l a r i t i e s in referencing patterns provide a means of objectively organizing cited journals. The actual mechanics of journal t i t l e grouping by c l u s t e r analysis have been accomplished using a variety of methods and algorithms. A general procedure in clustering studies has been to u t i l i z e more than one method to check r e l i a b i l i t y and compare resu l t s . This i s p a r t i c u l a r l y appropriate i n studies with smaller samples of data. Therefore, two methods of c l u s t e r i n g are applied in t h i s study. For both methods, the input data i s an m x n matrix, the rows (m) and columns (n) are l a b e l l e d with the t i t l e s of c i t e d and referencing journals respectively. The number of columns i s the number of source journals (variables) and the number of rows i s the number of c i t e d journal t i t l e s (data points). The c e l l s contain the number of c i t a t i o n s from each referencing journal to each c i t e d journal. Since the clustering programs group data points based on s i m i l a r i t y of treatment across a number of variables, the cited journal t i t l e s 12 w i l l be clustered based on s i m i l a r i t y of treatment from the source journals. The matrix i s f i r s t reduced to accomodate the program l i m i t a t i o n s and to increase the c e l l density of the matrix, and hence to accomplish the clustering on the more highly c i t e d journals. In one method ( O s i r i s ) , Pearson Product-Moment c o r r e l a t i o n s are calculated for each pair of c i t e d journals based on th e i r c i t a t i o n patterns across the referencing journals. The correlations are used as s i m i l a r i t y measures on which the clu s t e r i n g i s accomplished. In the second method (C-Group), ci t e d journals are considered as points i n n-dimensional space, and the clusters are based on the Euclidean distances between the points. Both programs begin with each cited journal considered as a separate cluster and then group the most s i m i l a r two journals into one larger cluster which takes a pooled score from i t s composite journals. Both programs proceed stepwise u n t i l a l l cited journals are placed i n one f i n a l c l u s t e r . Since the i n i t i a l and f i n a l steps of the algorithms y i e l d no information, and since what i s of i n t e r e s t i s the way. in "which the journal t i t l e s c l u s t e r , the r e s u l t s are i l l u s t r a t e d in a dendogram (or branching tree diagram) which displays the clu s t e r i n g r e s u l t s from the f i r s t step to the l a s t . An error term plotted beside each step indicates the r e l a t i v e jump that the program must make in order to accomplish the next cl u s t e r i n g . This term may be used as a point of entry into the dendogram i f natural clusters do exist in the data. The re s u l t s of such s t a t i s t i c a l grouping of journal t i t l e s can be examined and obvious s i m i l a r i t i e s determined. The 13 descriptive words i n journal t i t l e s contained in a c l u s t e r , as i n , for example, The B r i t i s h Journal of Educational Ps^chologjr , suggest the d i s c i p l i n a r y focus of the cluster. Since a l l the referencing journal a r t i c l e s contributing references were from an annual summary of research on reading, the r e s u l t i n g c l u s t e r s can be examined for t h e i r i n t e r d i s c i p l i n a r y r e l a t i o n s h i p to the f i e l d of reading. LIMITATIONS The following are seen as boundaries and l i m i t a t i o n s to the scope of this study: 1. The selection of the set of source journals w i l l a f f e c t any analysis of c i t a t i o n data. However, i t has been found that random selection of source journals does not provide p a r t i c u l a r l y appropriate c o l l e c t i o n s . Subjective s e l e c t i o n i s s t i l l the best approach. C i t a t i o n studies are usually based on a set of journals designated by professionals i n the f i e l d as being representative of the information network. In t h i s study, the ASEE journal a r t i c l e s were designated as comprising the information base. The journal a r t i c l e s i n the summary, covering the four time periods, were tabulated and the . journals publishing those a r t i c l e s became the set of source journals. This approach seems more than adequate in l i g h t of discussions of source journal selection techniques described i n other recent c i t a t i o n c l u s tering studies. 2. The conclusions based on t h i s study relate only to the reading research journal l i t e r a t u r e base defined as those journals which appeared in the ASEB for the four time periods. I t should be also borne in mind that the four summaries were 14 selected to represent the broader time span from approximately 1959 to 1972. The ASBR does i n f a c t contain only a p a r t i a l (although considerable) sample of the t o t a l universe of journal l i t e r a t u r e reporting reading research for the time span covered. However, i t i s f e l t that, in the opinion of most professionals in the f i e l d , the ASEE would constitute an appropriate representative information base for use i n a c i t a t i o n study of reading research journal l i t e r a t u r e . 3. A t o t a l of 918 journal a r t i c l e s appeared i n the ASRR across the four time periods. Because of l i b r a r y l i m i t a t i o n s , and excessive costs, i t was possible to locate only 768 of the o r i g i n a l l y c i t e d a r t i c l e s . I t i s f e l t that the excluded a r t i c l e s constitute e s s e n t i a l l y a random deletion of material and their contents do not adversly af f e c t the results of the study. The non-locatable material did not concentrate i n any one journal or time period. In fa c t , a good many of the a r t i c l e s appeared in journals with only one c i t a t i o n and would, as a matter of course, have been eliminated from at least the cluster portion of the study. A t o t a l of 84 percent of the o r i g i n a l a r t i c l e s were located and i t can be reasonably argued rhat these in ef f e c t constitute a r e l i a b l e representation of the journal l i t e r a t u r e information network for reading research. 4. The study i s limited by the assumption that the r e s u l t s obtained r e f l e c t the referencing behaviors of researchers rather than the reviewing patterns of those assembling the ASRR. Because of the consistent production of the ASHE for a considerable time period, i t i s f e l t that some confidence can be placed i n the procedures used to i d e n t i f y and obtain journals 15 a r t i c l e s that are t r u l y representative of the f i e l d as a whole. 5. The study i s also limited by the wide variety of referencing habits of various researchers. A l l references are treated here as i f they are given equal weight by researchers, which i s obviously not the case. Further, as noted by Ziman (1970), the c i t a t i o n i s a c a r e f u l l y selected bit of evidence that a researcher uses after the fact to j u s t i f y the place of his or her study in a legitimate and reputable context. As such i t should not be construed as representing the researcher's t o t a l information gathering and usage behavior., 6. For the c l u s t e r analysis, the journal a r t i c l e s were conflated into one source set spanning the four time periods. It was not economically feasible to conduct a cluster study for each time period and compare the results over time. However, such an approach would make an intere s t i n g future analysis. Other studies have indicated considerable s t a b i l i t y for c l u s t e r s over time although the death of eld journals and the b i r t h of new ones does have an eff e c t on the data. The de s c r i p t i v e portion of the study does examine trends over time while the clust e r analysis discusses c h a r a c t e r i s t i c s of the journal l i t e r a t u r e i n terms of the reading research reported for approximately a decade time span. 7. The conclusions drawn from the study are dependent on the assumptions underlying the clustering techniques used and the adjustments to the data made to accommodate the i r models. The clustering techniques were selected a f t e r careful analysis of the problem to be attacked, and the clu s t e r i n g models availa b l e , in an attempt to generate the best "goodness of f i t " 16 between the objectives of the study and the c l u s t e r i n g techniques available. In addition, two techniques were used i n examining the data to negate any adverse e f f e c t s due to use of a single approach. This practice i s becoming common in the f i e l d where the state of the art in cl u s t e r i n g f o r various purposes i s s t i l l emerging. DEFINITION OF TERMS For the purposes of the study the following d e f i n i t i o n s obtain: 1* Cluster Analysis : A method of grouping data points based on t h e i r s i m i l a r i t y of performance across a set of variables without recourse to a preset c l a s s i f i c a t i o n system. 2. Cj.tation : An acknowledgement received from another document. 3» Reference : An endnote or footnote which acknowledges another document. (This d i s t i n c t i o n between "reference" and " c i t a t i o n " i s made to make i t easier to distinguish between such descriptions as "core referencing journals", "core c i t e d journals", "referencing authors", " c i t e d authors", and so on.) 4. Cited Journal : The journal referred to i n the bibliography of another document. 5. _eferencin_ Journal : The journal acknowledging another document. 6. Reading Research Literature : Both a c o l l e c t i o n of journal a r t i c l e s defined by the compilers of the ASRR as related to the f i e l d or d i s c i p l i n e of reading, and the l i t e r a t u r e cited by the a r t i c l e s i n the ASRR, 7« M_Iiometrics : The use of formal, quantitative methods 17 for analysis and management of l i t e r a t u r e s and their contents. SUMMARY AND OVERVIEW Cit a t i o n analysis i s an established research methodology using information from the archival primary l i t e r a t u r e of a d i s c i p l i n e i n order to comment on the structure and development in the d i s c i p l i n e . C i t a t i o n analyses are also well-known and much-used i n b i b l i o m e t r i c s — t h e use of formal quantitative methods for analysis and management of l i t e r a t u r e s and their contents. The present study Is a c i t a t i o n analysis of the primary journal research l i t e r a t u r e in the f i e l d of reading. The purpose of the study Is to generate quantitative data from the primary l i t e r a t u r e which serves as a data base for comments on the structure and growth and development of reading as a research d i s c i p l i n e . The study has three major aspects: (1) the d e f i n i t i o n , c o l l e c t i o n , and computer storage of a representative body of journal references from reading research l i t e r a t u r e , (2) the analysis of those references to provide demographic data and tables, and (3) the measurement of the interconnsctedness of the cited journals by the use of c l u s t e r analysis. Chapter II presents the review of l i t e r a t u r e and the conceptual framework for the study. The design and methodology of the investigation are outlined i n Chapter I I I . Chapter IV presents the analysis of the data and the findings of the study, f i n a l l y , Chapter V presents the summary and conclusions for the investigation, and suggests further research. 18 CHAPTER II REVIEW OF THE LITERATURE AND CONCEPTUAL FRAMEWORK INTRODUCTION The conceptual base for t h i s study derives from a variety of f i e l d s which have been concerned either with describing the a c t i v i t i e s of science as a d i s c i p l i n e or with description and control of the multifold increase of published l i t e r a t u r e . These often overlapping i n t e r e s t s reside in such f i e l d s as the sociology of science (Merton, 1957), the history of science (Price, 1970; Goffman, 1966), the philosophy of science (Kuhn, 1962), and in a r e l a t i v e l y new f i e l d which includes most of the above: information science (Kessler, 1963; G a r f i e l d , 1972; Margolis, 1967; and many others). Change and growth i n science have been measured across a wide variety of variables. A good deal of attention has been paid to s c i e n t i f i c communication i n general and that contained in s c i e n t i f i c papers i n p a r t i c u l a r . A number of c h a r a c t e r i s t i c s of s c i e n t i f i c l i t e r a t u r e have been studied, including, f o r example, dispersion of l i t e r a t u r e , multiple authorship, amount of tablature, number of c i t a t i o n s , and age of cited material [see the excellent reviews by Parker, Paisley, and Garrett (1967) and Donohue (1973)]. As Ziman (1970:77) noted In his a r t i c l e on the growth of information science and the study of information, s c i e n t i f i c communication depends almost e n t i r e l y on i t s fragmentary and derivative 19 primary l i t e r a t u r e . S c i e n t i f i c papers are fragmentary because, according to Ziman, "they are not meant to be f i n a l statements of indisputable truths, but as tiny tentative steps forward, as elements i n a larger scheme." The papers are derivative because they lean heavily on previous research. The study of references i n such papers Is p a r t i c u l a r l y important because i t i s the function of the reference to embed.the referencing document within the conceptual structure of the f i e l d . The c i t a t i o n s r e l a t e the paper to the state cf the art i n the d i s c i p l i n e . Price (1970:6) makes the Important suggestion that a document must be viewed as only a part of the embedding framework. A scholarly publication i s not a piece of information but an expression of the state of a scholar or group of scholars at a particular time. We do not, contrary to superstition, publish a f a c t , a theory, or a finding, but some complex of these. A s c i e n t i f i c paper i s at the same time more and l e s s than a concept or a datum or an hypothesis. If the paper i s an expression of a person or several persons working at the research front, we can t e l l something about the r e l a t i o n s among the people from the papers themselves. Citation studies have focused on the analysis of a number of di f f e r e n t variables contained i n the references i n selected c o l l e c t i o n s of s c i e n t i f i c a r t i c l e s . The re s u l t s have been used to delineate research fronts, i d e n t i f i y i n v i s i b l e colleges, distinguish between "hard" and " s o f t " science, (Price, 1970); id e n t i f y growth patterns of emergent d i s c i p l i n e s (Goffman, 1966); i d e n t i f i y key documents and est a b l i s h such secondary bibliographic services as the Science Cit a t i o n Index and the Social Science Citation Index (Garfield, 1975); and to reveal what may be general trends i n the l i t e r a t u r e of the d i s c i p l i n e , such as the growth of multiple authorship, the number of 20 c i t a t i o n s per a r t i c l e s , and the proportion of c i t a t i o n to journals as opposed to books or monographs. A number of mathematical models have been borrowed, or developed, in attempts to i d e n t i f y and explain mathematical r e g u l a r i t i e s inherent in s c i e n t i f i c l i t e r a t u r e s . For example, Goffman (1966) made use of epidemic theory to explain the spread and growth of the l i t e r a t u r e in mast c e l l research. He found that the mathematical properties of the growth of research on a topic wars si m i l a r to the growth curves observed i n the spread of epidemics, such that there was a period of i n i t i a l rapid growth, a peak, and then a rapid dropping off of frequency of published a r t i c l e s . Other c i t a t i o n studies have made use of Pareto d i s t r i b u t i o n s (Pareto, 1897), Zipf curves (Zipf, 1935, 1949), Lotka's law (Lotka, 1926), and Bradford's law {Bradford, 1934), among many others, to create bibliographic models related to d i s t r i b u t i o n of l i t e r a r y productivity, dispersion of a r t i c l e s on s p e c i f i c topics, and i d e n t i f i c a t i o n of core journals within f i e l d s and subfields, Recent studies, such as that reported by Donohue (1972, 1973), emphasize combining several b i b l i o m e t r i c measures in analyzing c o l l e c t i o n s of l i t e r a t u r e . Computer technology has extended the power of c i t a t i o n analysis by making i t possible to examine the interconnectedness of journals as refle c t e d in large c o l l e c t i o n s of c i t a t i o n s , as i n those contained in a s p e c i f i c journal over a period of time (Broadus, 1952), or those contained in bibliometric tools such as the Science C i t a t i o n Index (Arms and Arms, 1973). The Science £i*aii2fi Index i s probably the most s i g n i f i c a n t application of 21 the bibliographic c i t a t i o n to information r e t r i e v a l . The Index i t s e l f has generated a new group of studies which are beginning to veri f y some of the e a r l i e s t assumptions of c i t a t i o n analysis. Cit a t i o n studies, and bi b l i c m e t r i c studies i n general, were o r i g i n a l l y concentrated in the hard sciences, but with the success of the methods, the techniques have bean increasingly applied i n the s o c i a l sciences as well. Educational l i t e r a t u r e , with i t s sheer volume and dispersion, has long presented problems to those who -would study or control i t . Ci t a t i o n studies using educational l i t e r a t u r e have been reported (Broadus, 1953, 1965; Narin and Garside, 1972), but in the area of reading the closest work has focused only on counts of a r t i c l e s and journal t i t l e s published i n reviews of research. This study extends previous work and broadens the base of analysis using c i t a t i o n data to indicate growth and development trends in reading research l i t e r a t u r e , locate i n t e r d i s c i p l i n a r y relationships, and describe the core journal l i t e r a t u r e related to reading research. The following sections present a selec t i v e review of research on the growth of science and s c i e n t i f i c communication, and on c i t a t i o n analysis, to provide the conceptual framework for the bi b l i c m e t r i c analysis of reading research journal l i t e r a t u r e which follows, GROWTH OF SCIENCE AND SCIENTIFIC LITERATURES This section examines information studies on science as a d i s c i p l i n e and on s c i e n t i f i c communication. The rati o n a l e for such studies i s succinctly put by Nelson and Pollack. (1970) i n the preface to Communication Among Sc i e n t i s t s and Engineers . 22 The assimilation and dissemination of information by s c i e n t i s t s and engineers i s an i n t e g r a l part of th e i r research and development a c t i v i t i e s , Sith the r e a l i z a t i o n on the part of s c i e n t i s t s that they are confronted with an information c r i s i s , much money and e f f o r t has gone into research on s c i e n t i f i c communication and the development of information r e t r i e v a l systems. In fact, Kochen (1970:44) noted that the study of s c i e n t i f i c communication has led to the development of a new science. A new i n t e l l e c t u a l d i s c i p l i n e seems to ae i n the making. I t i s the study of processes by which knowledge grows. It seeks conditions under which such growth i s stable. Knowledge grows in i n d i v i d u a l s by natural learning processes. Knowledge grows in communities by other natural processes, such as s c i e n t i f i c advance and revolution. Kuhn (1962), in h i s well known work on the structure of s c i e n t i f i c revolutions, suggested the p o s s i b i l i t y that f i e l d s (or schools) which have a shared consensus of t h e o r e t i c a l opinion, which he l a b e l l e d a "paradigm," d i f f e r from f i e l d s that do not, in that the former have a community of s p e c i a l i s t s (not necessarily concentrated i n one i n s t i t u t i o n ) who share professional communication. He suggested a research technique whereby a s h i f t i n the d i s t r i b u t i o n of technical l i t e r a t u r e c i t e d i n the footnotes to research reports ought to be studied as a possible index to the occurrence of revolutions. In a recent application of Kuhn's suggestion, Small (1975:34) examined c i t a t i o n patterns i n s c i e n t i f i c l i t e r a t u r e and hypothesized that the actual set and configuration of highly-cited documents i n a specialty provides a concrete representation of a paradigm for that specialty in Kuhn's sense of that term. Small developed maps of s p e c i a l t i e s over four year periods using the c o - c i t a t i o n of highly-cited documents (from 23 five successive Science Citation Indexes , 1970-1974) as a basis for c l u s t e r i n g and multidimensional scaling as i n t e r - c l u s t e r s i m i l a r i t y measures. The results markedly i l l u s t r a t e d the s h i f t s i n the d i s t r i b u t i o n of l i t e r a t u r e that Kuhn predicted. Fully one-third of the s p e c i a l t i e s over the four year period underwent sudden major s h i f t s in the set of c i t e d documents. Based on these cases, we anticipate that s p e c i a l t i e s undergo, on the average, one revolution every 12 years. Similar attempts to describe professional communication and the other methods that s c i e n t i s t s use to gain access to information have inspired a wide variety of research. Parker and Paisley (1966) reviewed the l i t e r a t u r e on the study of information use i n science. They noted seven frequent methods of investigation: questionnaire, interview, and diary studies; participant-observer studies; socicmetric analyses; experiments and quasi-experiments in f i e l d settings; laboratory experiments involving motivation for information seeking and cognitive organization of information; document analyses; and computer- based research. Document analyses have been p a r t i c u l a r l y a t t r a c t i v e to researchers for a number of reasons. Compared to, for example, questionnaire data, document data i s r e l a t i v e l y easy to obtain, i s organized i n various handy and revealing c o l l e c t i o n s across a l l sciences and applied f i e l d s , and e x i s t s in a dated and tabulated archival form. Further, document analyses are a t t r a c t i v e because they may reveal a much more accurate picture of science than can be obtained i n such sources as textbooks and other such secondary derivative l i t e r a t u r e . Carpenter and Narin (1973:425) commented, "There i s an inherent challenge to 24 structuring the [journal] l i t e r a t u r e , since i t may well r e f l e c t the mosaic of s c i e n t i f i c knowledge." The idea that the l i t e r a t u r e can be used to r e f l e c t the mosaic of s c i e n t i f i c knowledge i s also expressed in s l i g h t l y d i f f e r e n t form by Price (1965: 515). It seems clear that in any c l a s s i f i c a t i o n into research-front subjects and taxonomic subjects there w i l l remain a large body of l i t e r a t u r e which i s not completely one or the other. The present discussion suggests that most papers, through c i t a t i o n s , are knit together rather t i g h t l y . The t o t a l research front of science has never, however, been a single row of k n i t t i n g . It i s , instead, divided by dropped s t i t c h e s into quite small segments and s t r i p s . From a study of the c i t a t i o n s of journals by journals I come to the conclusion that most of these s t r i p s correspond to the work of, at most, a few hundred men at any one time. Such s t r i p s represent objectively defined subjects whose description may vary materially from year to year but which remain otherwise an i n t e l l e c t u a l whole. I f one would work out the nature of such s t r i p s , i t might lead to a method for delineating the topography of current s c i e n t i f i c l i t e r a t u r e . With such a topography established, one could perhaps indicate the 'overlap and r e l a t i v e Importance of journals and, indeed, of countries, authors, or i n d i v i d u a l papers by the place they occupied within the map, and by t h e i r degree of s t r a t e g i c centralness within a given s t r i p . Journal c i t a t i o n s provide the most r e a d i l y available data for a test of such methods. Price i s suggesting a number of concepts whxch are central to information science in general and to c i t a t i o n studies i n p a r t i c u l a r . F i r s t , because of i t s s o c i a l role, there i s a good' deal revealed in s c i e n t i f i c l i t e r a t u r e beyond i t s substantive content. Second, s c i e n t i f i c l i t e r a t u r e can be used to map and define the topography of science, and therefore, mathematical r e g u l a r i t i e s can be observed i n the growth of science and i t s l i t e r a t u r e . Third, such studies can be useful i n the i d e n t i f i c a t i o n and r e t r i e v a l of key documents and the i d e n t i f i c a t i o n of core journals. 25 Before turning to a review of c i t a t i o n studies, i t i s worth noting that there i s wide consensus of opinion among information s c i e n t i s t s about the existence of consistent patterns i n many aspects of s c i e n t i f i c communication, Lotka (1926) reported on the frequency d i s t r i b u t i o n of s c i e n t i f i c productivity i n chemistry and physics. Goffman (1966) and Goffman and Newill (1967) attributed epidemiological properties to the spread of ideas through a population of s c i e n t i s t s . Goffman and Warren (1969) found support for Bradford's Law of Dispersion i n l i t e r a t u r e s r e l a t i n g to research i n schistosomiasis and mast c e l l l i t e r a t u r e s . Rapoport (1953) comments on the mathematical theory of information that was born among communication engineers such as Shannon and Weaver (1949), and which relates the behavior of Information (or order) to the law of entropy. Acccording to t h i s theory, "the process of obtaining information i s quantitatively equated to the process of ordering portions of the world." Thus, many information studies have sought to i d e n t i f y and apply the mathematical r e g u l a r i t i e s to be found i n c o l l e c t i o n s of c i t a t i o n data. This study extends such research and examines a variety of growth c h a r a c t e r i s t i c s revealed i n the c i t a t i o n s i n a body of reading research journal l i t e r a t u r e . CITATION ANALYSES Citat i o n analysis has a long and interesting history i n bibliometric research and information science. Carpenter and Narin (1973), i n their review of the history of research on s c i e n t i f i c l i t e r a t u r e , noted that early works by Coles and Eales (1917), and Hulme (1923) attempted to relate publication and 26 author counts to external economic and p o l i t i c a l events. Gross and Gross (1927) were the f i r s t to relate frequency of c i t a t i o n with importance and made recommendations about journal purchase based on c i t a t i o n counts i n the Journal of the Chemical Society. A burst of papers by science l i b r a r i a n s followed . these early works, but with the exception of Bradford's 1934 analysis (which w i l l be discussed l a t e r i n this section), the late 30's and 40's were not productive i n terms of l i t e r a t u r e analyses. Carpenter and Narin (1973:425) commented on the more recent growth. In the 1950's there was a gradual re-emergence of analysis of the l i t e r a t u r e and, as science becomes large in the 1960's, more and more attention was focused on managing the large and rapidly growing s c i e n t i f i c enterprise. Price made use of a number of l i t e r a t u r e counts in devising his macroscopic outlines of the s c i e n t i f i c enterprise. C i t a t i o n counting began to attr a c t more and more attention as a pot e n t i a l means of structuring the s c i e n t i f i c l i t e r a t u r e . As noted in the introduction to t h i s chapter, c i t a t i o n data have been used for a wide variety of analyses, i n c l u d i n g : the evaluation of researchers (Cole and Cole, 1967) and the quality of research (Cole and Cole, 1971); the evaluation of a r t i c l e s (Margolis, 1967) (Garfield, 1975); of journals (Garfield, 1972); the design of information systems (Arms and Arms, 1973); and in understanding c o l l e c t i o n s of l i t e r a t u r e (Donohue, 1973). Within this broad variety of c i t a t i o n l i t e r a t u r e , t h i s review focuses on research which relates most s p e c i f i c a l l y to the goals of the study, and includes representative studies which used c i t a t i o n analysis to examine (1) core journals, (2) c i t a t i o n s per a r t i c l e , (3) age of ci t e d documents, (4) types of cited documents, (5) s e l f - c i t a t i o n , (6) multiple authorship, and (7) clust e r analysis of journal t i t l e s . 27 Core Journals Studies of document dissemination have made use of di s t r i b u t i o n s related to the dispersion, u t i l i z a t i o n , and scatter of recorded knowledge. The exploration of various empirical d i s t r i b u t i o n laws i s based on the desire to provide p r a c t i c a l information systems b u i l t on the basis of some formally observed properties of communication. Bradford (1934) published a study of the dispersion of a r t i c l e s on s p e c i f i c topics. He found that when he ranked periodicals containing a r t i c l e s to a s p e c i f i c topic, then divided the c o l l e c t i o n into three .parts such that each part contained the same number of a r t i c l e s on the to p i c , the number of journals i n each part increased geometrically. This led to the concept that there was a nucleus or core of journals i n a topic, f i e l d , or d i s c i p l i n e . 'Fairthorne (1970:523), In a very complete review of the empirical d i s t r i b u t i o n s used in bibliometric description and prediction, suggested that there may be a number of names f o r es s e n t i a l l y the same function. Different manifestations of the hyperbolic d i s t r i b u t i o n for both discrete and continuous variables are found in many f i e l d s and named variously after Fachner, Zipf, Pareto, Bradford, l i l l i s , Berger- Mandelbrot, and others. But these are names associated with p a r t i c u l a r manifestations of t h i s type of behavior, not names for the type of behavior i t s e l f . Pareto (1897), f o r example, was an economist studying the d i s t r i b u t i o n of wealth, while Zipf (1949) was studying the d i s t r i b u t i o n of words based on the freguency of their occurence in text. These various manifestations of hyperbolic d i s t r i b u t i o n s are referred to by Fairthorne as "Stable Paretian laws" (after Pareto). 28 Price (196 3) used the Pareto d i s t r i b u t i o n i n examining a variety of indicators of s c i e n t i f i c l i t e r a t u r e growth and d i s t r i b u t i o n . Of p a r t i c u l a r interest here i s his adaptation of the Paretian d i s t r i b u t i o n in a simple square root c a l c u l a t i o n to conclude that although 30,000 journals were estimated to e x i s t , half the reading that was done used only the 170 most popular items. That i s , according to the data then available from the Science Ci t a t i o n Index , i f the c i t a t i o n of a journal i s taken as a measure of i t s use, and i f the journals are rank-ordered in terms of frequency c i t e d (or used), then 170 journals out of the t o t a l of 30,000 w i l l receive 50% of the t o t a l number of c i t a t i o n s . G a r f i e l d (1972:476), using the massive amount of c i t a t i o n date available i n the Science C i t a t i o n Index , did a si m i l a r analysis. He ranked journals by frequency of c i t a t i o n , and found that a small group of 250 journals were named in almost half of the 3.85 mi l l i o n references processed for the SCI in 1969. He concluded that "the predominance of cores of journals i s ubiquitous," and such data provide conclusive support f o r Bradford's bibliographic law. The data reported here demonstrate the predominance of a small group of journals i n the c i t a t i o n network. Indeed the evidence seems so conclusive that I can with confidence generalize Bradford's bibliographic law concerning the concentration and dispersion of the l i t e r a t u r e of i n d i v i d u a l d i s c i p l i n e s and s p e c i a l t i e s . Going beyond Bradford's studies, I can say that a combination of the l i t e r a t u r e of i n d i v i d u a l d i s c i p l i n e s and s p e c i a l t i e s produces a m u l t i d i s c i p l i n a r y core for a l l of science comprising no more than 1000 journals. The e s s e n t i a l m u l t i d i s c i p l i n a r y core could, indeed, be made up of as few as 500 journals i f , for example, one i s attempting to s a t i s f y the needs of l i b r a r i e s i n developing countries. 29 Citations Par A r t i c l e C i t a t i o n habits d i f f e r widely from author to author, but averages across c o l l e c t i o n s of l i t e r a t u r e may y i e l d r e l i a b l e data. Parker, Paisley, and Garrett (1967:62) suggested that as a f i e l d becomes more mature, i t w i l l exhibit an increased number of c i t a t i o n s per a r t i c l e . They found increased average c i t a t i o n rates in eight s o c i a l science journals across the f i f t e e n year span studied, from 8.4 c i t a t i o n s per a r t i c l e i n 1950 to 15.2 in 1965. Tremendous variance i n the c i t a t i o n habits between journals was also noted. For example, Psychological B u l l e t i n had an unusual median of 36.5 c i t a t i o n s per a r t i c l e , American Soc i o l o g i c a l Review and Journal of Abnormal and S o c i a l Psychology had 18.0 and 12.0 c i t a t i o n s per a r t i c l e respectively, and American Documentation had 3.7 c i t a t i o n s per a r t i c l e . They related increases in average c i t a t i o n s per a r t i c l e to emergence of a f i e l d as a research d i s c i p l i n e . It i s a reasonable inference that a small number of c i t a t i o n s indicates an emerging f i e l d , because the number of c i t a t i o n s i s increasing with time across the set of journals, a l l of which represent f i e l d s i n early or intermediate stages of evolution as d i s c i p l i n e s . Price (1970:7-8), in his study of c i t a t i o n measures of hard science, soft science, technology, and nonscience, using Science Citation Index data, reported that there seems to be a slow but steady increase in referencing i n a l l f i e l d s . While notable exceptional documents e x i s t , the amount of referencing i n a paper can be related to the " s c h o l a r l i n e s s " of that document. The etymology of "scholarship" indicates that i t derives from the scholia , the added explanatory footnotes put into school texts, so perhaps i t i s reasonable to i d e n t i f y the amount of such footnoteage and referencing with our i n t u i t i v e idea of 30 "sc h o l a r l i n e s s . " Price noted that the general norm of scholarlrness i s a paper l i s t i n g about ten to twenty-two references. The amount of c i t a t i o n varies greatly within f i e l d s because of the existence of review papers, research papers, and ex cathedra pronouncements by experienced s c i e n t i s t s in unreferenced papers, also technologists, s c i e n t i f i c scholars, and u n s c i e n t i f i c scholars a l l tend to have d i f f e r e n t c i t a t i o n habits. Price warned that amount of referencing by i t s e l f i s not t o t a l l y v a l i d for discriminating science and nonscience, hard and soft. Nevertheless, c i t a t i o n counts across c o l l e c t i o n of a r t i c l e s provide a useful t o o l for analysis of developmental trends and the general l e v e l of scholarship within a d i s c i p l i n e . Price suggests: Scholarliness as I have defined i t may be taken not just as a diagnostic but also as prescriptive f o r a cumulating knowledge system. A2£ of Cited Material •The age of material cited at the time of c i t a t i o n has been used as an indicator of the r e l a t i v e importance a f i e l d assigns to i t s a r c h i v a l l i t e r a t u r e . That i s , median age of c i t a t i o n s can reveal to what extent a f i e l d (or publication, or researcher) favors a recent body of publications over older l i t e r a t u r e . Broadus (1965), i n an analysis of the references used i n the 1950 and 1960 Encyclopedias of Educational Hesearch , found that the median age of cited materials changed from 12 to 7 years over the ten year span of the two books. He suggested that such information should be considered seriously when building l i b r a r y c o l l e c t i o n s of educational materials and attempting to develop a 31 decision structure for information subscriptions in research l i b r a r i e s . The extreme contemporaneity of some s c i e n t i f i c c i t a t i o n f i t s the i n t u i t i v e model that views research as a cumulative process by which new knowledge grows from r e l a t i v e l y recent findings, in pyramid fashion. Cole and Cole (1967) found that papers in physics have a " h a l f - l i f e " of no more than f i v e years; that i s , at least half the c i t a t i o n s i n any year are to work published within the f i v e preceding years. Price (1970:9-10) reviewed a number of measures which could be used to distinguish between science, hard and s o f t , and technology and nonscience. His experimentation and experience with the vast number of references available in the Science C i t a t i o n Index , led him to conclude that age of cited material could serve as the basis for such discriminations. The measure he developed takes the proportion of references dated within the l a s t f i v e years as an indicator of whether a f i e l d (or author, or journal) r e l i e d upon a research front, a general archive, or some combination of the two. Price's index i s a deceptively simple measure. Price f e l t his e a r l i e r work had conclusively shown that there was not a single population of references, but two overlapping ones. On the one hand there was a f a i r l y uniform raiding of the archive of a l l the available l i t e r a t u r e , past and present, with only a slow secular decrease in the usefulness of l i t e r a t u r e as a function of i t s age. Secondly, there was something which I c a l l e d an "Immediacy E f f e c t , " a special hyperactivity of the rather recent l i t e r a t u r e which was s t i l l , so to speak, at the research front. Thus, the index gives a proportion between the research front immediacy and the normal archival use of the l i t e r a t u r e , 32 while measures such as " h a l f - l i f e " simply give a p r o f i l e of the f a l l - o f f in immediacy effect. More s i g n i f i c a n t l y . Price pointed out that his index i s a tool for distinguishing between modes of scholarship which are "cumulative" versus "noncumulative," or " s c i e n t i f i c " versus " n o n s c i e n t i f i c . " . . . i t would seem that t h i s index provides a good diagnostic for the extent to which a subject i s attempting, so to speak, to grow from the skin rather than the body. With a low index one has a aumanistic type of metabolism i n which the scholar has to digest a l l that has gone before, l e t i t mature gently i n the c e l l a r of his wisdcm, and then d i s t i l l forth new words of wisdom about the same sorts of questions. In hard science the positivenass of the knowledge and i t s short term permanence enable one to move through the packed down past while s t i l l a student and then to emerge at the research front where interaction with one's peers i s as important as the storehouse of conventional wisdcm. The thinner the skin of science the more orderly and c r y s t a l l i n e the growth and the more rapid the process. Price was able to give an estimate of the range of values of his index but noted that the growth rate of the l i t e r a t u r e also affected the measure. ...we may note that a l i t e r a t u r e growing at a rate of 5 percent per annum doubles in size i n 13.9 years and contains about 2 2 percent cf a l l that has bean published i n i t s l a s t f i v e years of publication. A f i e l d growing at the most rapid rate experienced of 10 percent per annum, with a doubling time of 6.9 years, has within the l a s t five years about 39 percent of a l l i t s archive. Price's Index (as I might c a l l it) w i l l vary from 22 percent for normal growth to 3 9 percent for most rapid growth for a f i e l d that i s purely ar c h i v a l , raiding a l l the l i t e r a t u r e that has gone before equally, with only a gradual secular decline with aging, and without t h i s special immediacy of an active research front. In work reported since that of Price, G a r f i e l d (1972:476) analyzed the overa l l chronological d i s t r i b u t i o n i n each ed i t i o n cf the Science Ci t a t i o n Index and concluded: 33 An analysis of this d i s t r i b u t i o n has shown that the t y p i c a l c i t e d a r t i c l e i s roost heavily cited during the 2 years aft e r i t s year of publication. (In any given year, 21 to 25 percent of a l l references c i t e a r t i c l e s that are 3 or fewer years old.) Parker, Paisley, and Garrett (1967) found that i n 17 s o c i a l science journals i n four time periods across the 15 years from 1950 to 1965, the percentage of c i t a t i o n s dated 0 to 4 years before the referencing publication was r e l a t i v e l y stable at close to 40 percent and those dated 5 to 9 years were also stable at about 26 percent. In ether words, from 1950 to 1965, the mean percentage of c i t a t i o n s in t h i s sample of s o c i a l science which were less than 10 years old was s l i g h t l y over 65 percent for a l l four time periods recorded. Broadus (197 1) reported that the median age of cited material in education had dropped from 12 to 7 years from 1950 to 1960. llES of £__1_____22 Type of publication simply analyzes whether the reference being cited i s a book, a r t i c l e or other kind of medium. The types of publications referred to i n references vary greatly from d i s c i p l i n e to d i s c i p l i n e . Price (1970) noted that i n the average journal a r t i c l e , 80% of the c i t a t i o n s are to other s e r i a l publications rather than to books, theses, reports, and unpublished work. Parker, Paisley, and Garrett (1967) found 438 of the c i t a t i o n s from 17 selected behavioral science journals were to journals and 3*\% were to books. Lin and Nelson (1969:49) found s t r i k i n g differences between the types of publications in three major s o c i o l o g i c a l journals when compared to those i n a major o p t i c a l journal. They presented a variety of hypotheses to explain the greater proportion of reference to books i n the 34 s o c i o l o g i c a l journals (about 50%) when compared with that i n the o p t i c a l journal (about 15%). Of p a r t i c u l a r i n t e r e s t here.is t h e i r reference to paradigms. Assuming that optics i s an- old, well-established science with a history of paradigms, and sociology i s a younger and less established science without rigorous paradigms, then the r e s u l t s seem to lend support to Kuhn's hypothesis, drawn from a study of the history of science, that d i s c i p l i n e s with paradigms tend to publish their work i n journals, while d i s c i p l i n e s without paradigms tend to publish i n books. Broadus (1971) reviewed a number of c i t a t i o n studies which referred to the form of the publications c i t e d : Martin (1952), in p o l i t i c a l science, found 51.331 of references to "monographs," while i n economics, Livesay (1953) and Mark (1956) reported finding 51.3% and 47.6% to "monographs." Quinn (1951) , Meier (1951), Broadus himself i n 1952 and again in 1967, showed that in sociology 45.21 and 55.6% referred to "monographs" and 53,7% and 61.51 to "non-serials." Sarle (1958) reported that in business administration 32.6% of the references were to "monographs." In education, Broadus (1953,1965) discovered 30.9% and 32.7% of references were to books. Guttsman (1966) found i n general s o c i a l sciences that 43.9% of references were to " t r e a t i s e s " and "monographs," while, i n the same area, Earle and Vickery (1969), found 46.2% of references to books. Broadus (1971:241) noted the d i f f i c u l t i e s inherent i n comparing the studies due to the s l i g h t l y d i f f e r e n t meanings assigned to terms l i k e "book," "monograph," and " s e r i a l . " Nevertheless, he f e l t that some comparison was possible. 35 The proportions take on a meaning when compared with data obtained from c i t a t i o n studies in other d i s c i p l i n e s . Vaughan, i n music, found that 69.5 per cent of the references were to "monographs"; Tucker, in philology, c l a s s i f i e d 60.4 per cent as "monographs." Simonton, in his study of fine arts l i t e r a t u r e as a whole, discovered that 71.4 per cent of the c i t a t i o n s referred to "books." Fussler's analysis for 1939 showed chemists using 5.21 per cent "monographs," while physicists were using 7.75 per cent. Craig found that "books" used i n geology made up 17.5 per cent of the t o t a l references in 1960; 21.5 per cent in 1965. With one exception each of the studies dealing with the l i t e r a t u r e of the humanities showed a higher percentage of c i t a t i o n s r e f e r r i n g to "books" or "monographs" than did any of those in the s o c i a l sciences. Apparently, however, physical s c i e n t i s t s use many more s e r i a l s than monographs, and the contrast with the s o c i a l sciences i s notable. S e l f - C i t a t i o n Comparatively l i t t l e work has been done i n r e l a t i o n to the number of times an author refers to his or her own work. Parker, Paisley, and Garrett (1967:29) found trends toward increased frequency of c i t i n g own work i n eighteen behavioral science journals across four time periods. The percentages increased steadily from 33% i n 1950 to 381 i n 1955, to 44% i n 1960 and to U6% in 1965. These figures might reasonably be interpreted as indi c a t i n g an increased commitment to cumulative research by those working and publishing i n t h i s i n t e r d i s c i p l i n a r y behavioral science area. If that i s so, then the percentage of c i t a t i o n s to own previous work might be taken as indicators of the stage of development of a subfield of science, with higher frequencies interpreted as more cumulative (and, presumably, more theoretical) research. Multiple Authorship Co-authorship i s increasing generally in scxence, and most rapidly in f i e l d s which receive proportionally more economic support. Price (1963) noted the movement toward increased 36 collaboration in data from chemical Abstracts . Clarke (1964) could f i n d no marked trend toward multiple authorship in biomedical writers. The average number of authors per paper in t h i s group remained steady (frcm 1946 on) at about 2.3. Clarke speculated that there may be considerable collaboration variance between dif f e r e n t groups of s c i e n t i s t s due to conditions other than the growth of "Big Science" as suggested by Price. Parker, Paisley, and Garrett (1967) observed that number of authors per a r t i c l e had been shown to correlate with whether or not f i n a n c i a l support for research was reported. In their research on behavioral science journals, they reported that the average number of authors per a r t i c l e was 1,34 with a range of 1.1 to 1.8 across journals. Lin and Nelson (1969) examined the mean number of authors per a r t i c l e in three core s o c i o l o g i c a l journals and found i t to be approximately 1.4. Price (1970) f e l t his research demonstrated that collaboration arises more from economic that i n t e l l e c t u a l dependence. He concluded that the amount of collaborative authorship measures no more than the economic value accorded to each f i e l d by society. These studies contained authorship data based on c o l l e c t i o n s of documents, not on c i t a t i o n data. The only a r t i c l e located which examined j o i n t authorship i n c i t a t i o n data was by Xhignesse and Osgood (1967). They found that 59% of the a r t i c l e s cited were by single authors, and 41% by two or more. Apparently, h i s t o r i c a l analyses of multiple authorhip trends i n c i t a t i o n data are not very common, even though the ar c h i v a l nature of the c i t a t i o n s can permit extensive h i s t o r i c a l research. 37 ________ __ _______ ______ Clustering provides a method of i d e n t i f y i n g groups within c o l l e c t i o n s of objects without recourse to preset c l a s s i f i c a t i o n . This s t a t i s t i c a l method can be used to group together sets of journals with s i m i l a r patterns of c i t a t i o n . The application of clu s t e r i n g techniques to c i t a t i o n data i s r e l a t i v e l y new. Kessler (1963a, 1963b, 1965) used the method of bibliographic coupling to link both a r t i c l e s and journal t i t l e s . He created two step maps which defined the unit of coupling as "one item of reference used by two papers." Xhignesse and Osgood (1967) used the multidimensional scaling of the reci p r o c a l c i t a t i o n among 21 psychology journals to produce clusters of intere s t i n g journals. Their methods produced, i n their words, "several i d e n t i f i a b l e c l u s t e r s of journals." Price and Schiminovich (1S68) created a clustering program based upon the extension of bibliographic coupling through the use of the computer. They applied t h e i r technique to a c o l l e c t i o n of 240 th e o r e t i c a l high energy physics papers. They f e l t t h e i r c l a s s i f i c a t i o n scheme had v a l i d i t y for use i n l i t e r a t u r e dissemination systems. A key study by Carpenter and Narin (1973) made use of two s i m i l a r i t y measures between journals as the basis of t i t l e c l u s t e r i n g . The f i r s t i s a Euclidean distance measure based on a cr o s s - c i t i n g matrix and the other i s a rank c o r r e l a t i o n measure also based on c i t a t i o n s received from other journals to the compared journals. They used their procedure to c l u s t e r 288 highly cited journals i n molecular biology, physics, and 38 chemistry. The resulting c l u s t e r s could be i d e n t i f i e d by national, subject, and subdisciplinary d i v i s i o n s . A research project (DISISS) based at the University of Bath, i n England, attempted to carry out the research necessary for the e f f e c t i v e design of information systems i n the s o c i a l sciences. An excellent working paper by Arms and Arms (1973:10) describes i n d e t a i l t h e i r work on the clustering of journal t i t l e s by c i t a t i o n data. They developed a technique referred to as the Scicon method and compared i t to other approaches and algorithms. Since t h e i r concern was with the development of a technique, t h e i r conclusions are ba s i c a l l y guidelines to be considered in future j o u r n a l - t i t l e c l u s t e r i n g attempts. Arms and Arms main objective in applying cluster analysis to bibliographic data was to provide c r i t e r i a for structuring bibliographic f i l e s . They acknowledged that the data obtained could be used in a number of ways, and of particular interest here i s their comment about structure in r e l a t i o n to emergent d i s c i p l i n e s . Journal t i t l e c lusters provide data on the structure of the primary l i t e r a t u r e i n a d i s c i p l i n e . From t h i s general picture new f i e l d s can be i d e n t i f i e d , either by comparison with e a r l i e r r e s u l t s or because unexpected patterns are displayed. BIBLICMETRIC ANALYSIS OF READING RESEARCH JOURNAL LITERATURE The studies reviewed i n the previous sections provide methodology and a useful conceptual base for organizing a bibliometric analysis of the journal l i t e r a t u r e related to reading research. Such an analysis i s predicated on the assumption that the f i e l d (area of knowledge or learning) of reading can be viewed as a d i s c i p l i n e , in the broad sense of 39 that term, and that i t i s reasonable to analyze and speculate about the written behavior of people working i n the d i s c i p l i n e . The c i t a t i o n s contained within a body of reading research l i t e r a t u r e represent a sample of the written behavior of reading researchers. As such, they constitute a r e l a t i v e l y unobtrusive measure of c i t a t i o n habits and information sources. A time trend sample of reading research journal l i t e r a t u r e , then, in l i g h t of concepts and methodologies from th i s review, could be examined in terms of several bibliometric c h a r a c t e r i s t i c s including: (1) core journal structure, (2) number of c i t a t i o n s per a r t i c l e , (3) ages of cited documents, (4) types of cited documents, (5) proportions of s e l f - c i t a t i o n s , (6) freguencies of multiple authorships, and (7) i n t e r - j o u r n a l structure as revealed by cluster analysis of journal t i t l e s . The following section discusses the rationale for the application of each measure in the analysis of reading research journal l i t e r a t u r e conducted i n thi s study. Time Samj3lijng Samples of journal a r t i c l e s reporting reading research should be selected over time because the study focuses on patterns of growth and change, as exhibited through various bibliometric indices, rather than a single point i n time. In his study of the Annual Summary of Research on Reading (ASRR), Summers (1968) found a steady increase in journals reporting reading research, and in the number of a r t i c l e s reported, u n t i l approximately 1958-1959. From that point onward there was an almost geometric spurt in growth and a sharp increase i n the amount of reading research reported. Thus, the time period 40 sampled should begin with the 1958-1959 spurt i n reported research. In addition, following Small's recent 1975 analysis, i t also seems reasonable to examine growth c h a r a c t e r i s t i c s i n the specialty of reading covering approximately a twelve year span. Thus, the upper l i m i t of the period sampled should be set at 1972. Because of the volume of data involved, i t would not be feas i b l e to analyze each of the annual research summaries (13 in a l l ) for the t o t a l time period. Therefore, four spaced summaries (1958, 1964, 1968, 1972) should be selected to provide a sample to represent gross trends for the t o t a l time period. Small (1975), Arms and Arms (1973), Naria and Carpenter (1972), and Xhignesse and Osgood (1967) a l l found considerable s t a b i l i t y in c i t a t i o n structures over time. Therefore, while the developmental measures for the study should _e selected and compared using the four separate annual summaries, the journal t i t l e c l u s t e r i n g can be accomplished by combining the data across the four i n d i v i d u a l yearly summaries. _2__ Journals Summers (1968) i d e n t i f i e d a core journal structure in reading research based on ranked frequency of number of a r t i c l e s reported annually i n the ASRR. That i s , i n the year 1936/37, the Elementary School Journal was the highest ranking journal because i t contributed 12 of the 79 journal a r t i c l e s reported for that year. In t h i s study, computer coding and storage of the data can f a c i l i t a t e obtaining three types of core journal l i s t s , The f i r s t l i s t s p e c i f i e s the core of journals which produce the bulk of the a r t i c l e s l i s t e d i n the ASRR for the four selected time periods and for the separate periods combined. The second 41 l i s t s p e c i f i e s the core of journals whose a r t i c l e s produced the greatest number of references for each of the four selected periods and for the four separate periods combined. The th i r d l i s t s p e c i f i e s the core of journals most often referenced (or al t e r n a t i v e l y , which received the greatest number of ci t a t i o n s ) for the four selected periods and for the four separate periods combined. Tabulating core journals by three methods can o f f e r a basis for confirmation of Garfield's (1972) statement that the predominance of cores of journals i s indeed "uaiquitous." The re s u l t s can serve to substantiate findings across many d i s c i p l i n e s , and across the Science C i t a t i o n Index i n general, which suggest that a small core of journal t i t l e s account f o r the preponderance of the c i t a t i o n s (and a r t i c l e s and references) in the reported research in the d i s c i p l i n e . The three l i s t s can also be extremely useful in developing a decision structure f o r monitoring the most frequent journals producing reading research based cn number of times cited in the ASRR, number of references produced, or c i t a t i o n s received. Citations Per A r t i c l e As noted previously, i t i s reasonable to i d e n t i f y the number of references i n a paper with our i n t u i t i v e idea of scholarliness. Therefore, a rough index of the scholarliness of reading research can be obtained by calculating the norm of referencing i n reading research papers. Further, since there seems to be a steady increase in the average number of references per a r t i c l e i n a l l f i e l d s , the data can be analyzed using t h i s measure to determine i f trends toward more scholarly 4 2 research are evident i n the l i t e r a t u r e of reading research over the time period covered by the study. ___ 2L Material Reading research i s c e r t a i n l y not a new f i e l d i n the same sense as mast c e l l and laser research. Considerable research has been reported since the early 190 0's. The f i r s t summary i n the 1925 ASRR included 436 research a r t i c l e s r e l a t i n g to reading reported during the l a t e 1800's to June 30, 1925. From then to 1966 alone, the ASRR reported more than 5000 research documents. Thus, an archival l i t e r a t u r e exists spanning approximately a century of reported reading research. Determining the d i s t r i b u t i o n of the ages of c i t e d material in a recent sample from that c o l l e c t i o n , and c a l c u l a t i n g Price's Index, can indicate i f a research front e x i s t s i n reading si m i l a r to other d i s c i p l i n e s and whether or not that research front i s changing i t s fundamental nature across time. ___§. 2_ Publication Broadus (1971) noted that researchers found that the l i t e r a t u r e of the humanities showed a higher percentage of c i t a t i o n s to books or monographs than did the s o c i a l sciences, and that physical s c i e n t i s t s use many more s e r i a l s than monographs, In s t r i k i n g contrast to the s o c i a l sciences. Determining the r e l a t i v e proportions by which reading research l i t e r a t u r e c i t e s books, journals, and other types of publications can permit a comparison of reading with other s o c i a l sciences. Again, through time sampling, comparison and trends over time can be indicated. 43 S e l f - C i t a t i o n s S e l f - c i t a t i o n can indicate a commitment to cumulative research. Determining the proportion of reading research documents which contain at least one s e l f - c i t e to those which do not can permit comparison with r e s u l t s from other s o c i a l science d i s c i p l i n e s , such as those reported by Parker, Paisley, and Garrett (1967). Multiple Authorship Si g n i f i c a n t conclusion oriented educational research, p a r t i c u l a r l y i n North America, i s usually heavily funded by foundations and governmental agencies. Such research i s program or project oriented, involving investigative teams, rather than being pursued by the i n d i v i d u a l researcher operating in i s o l a t i o n . If Price's (1970) conclusions on increases i n multiple authorship are v a l i d , reading research should demonstrate an increase over time i n the average number of authors per paper. This should be p a r t i c u l a r l y true of research conducted in the United States. In addition, growth i n multiple authorship over time can be determined using the authorship of the references as opposed to the authorship of the a r t i c l e s i n the ASRR. Since references are dated, the sample can permit analysis on t h i s p a r t i c u l a r variable on 30 years of prior published reading research l i t e r a t u r e . Cluster Analysis Cluster analysis, as a t o o l for structuring bibliographic data, i s s t i l l i n an exploratory stage, as was demonstrated by the 1973 Bath University study. Journal t i t l e c l u s t e r i n g . 44 though., can provide a useful method of roughly structuring the primary research l i t e r a t u r e i n the d i s c i p l i n e of reading. At worst, a cluster analysis would present amorphous c l u s t e r s of journal t i t l e s which would not follow any i n t u i t i v e sense of d i s c i p l i n e and subdiscipline patterning. At best, the method would reveal i n t u i t i v e l y related sets of journals ( i . e . general education, educational psychology, psychology, sociology, etc.) which could be used to i d e n t i f y the i n t e r - d i s c i p l i n a r y contributions to the information base i n reading research. SUMMARY The desire to understand, often i d e n t i f i e d as the goal of pure science, and the desire to produce products and to control, often i d e n t i f i e d with technology, are not easily distinguished in information science. For example, one of i t s methods, bibliometric analysis, appears to have grown c o - j o i n t l y from the desire to understand the growth of knowledge and from the need to control an exponentially-growing published l i t e r a t u r e . Over time, philosophers, historians, and s o c i o l o g i s t s of science, and science l i b r a r i a n s have come to share s i m i l a r methods—in p a r t i c u l a r , the technique of c i t a t i o n analysis. The former were concerned with c i t a t i o n s because they contained a written history of the relationships among s c i e n t i s t s , d i s c i p l i n e s , and i n s t i t u t i o n s . Rarely had researchers been presented with such a well-dated and well-tabulated archival record of their subject matter. Science l i b r a r i a n s , of course, perceived that c i t a t i o n s represented a record of the information usage of s c i e n t i s t s , and could, therefore, be r e l i a b l y used to make evaluations and recommendations i n structuring and ordering 45 information delivery systems. This twofold i n t e r e s t i n c i t a t i o n analysis i s i l l u s t r a t e d i n the ways in which the Science Citation Index and the Social Science C i t a t i o n Index are used. The SCI organizes 4 m i l l i o n sources of c i t a t i o n s which have appeared i n 2,600 of the most important s c i e n t i f i c journals since 1961. A similar technique i s used in the SSCI with s o c i a l science journals. Primarily, the SCI i s an information r e t r i e v a l system, but much of the work of such major s c i e n t i f i c historians as Derek DeSolla Price and others i s based upon SCI data. Cit a t i o n research, as with information science in general, i s a r e l a t i v e l y new phenomenon, but ample quality work has been done i n the last decade and a half to establish i t s methodology and test i t s r e l i a b i l i t y and v a l i d i t y . Beading research can be viewed as an emerging d i s c i p l i n e and i s a f i e l d i n which bibliometric methods can be productively applied. Five factors support t h i s contention. F i r s t , an extensive primary journal system exists and has been under good yearly a r c h i v a l control for some time. A number of active professional associations, such as the International Reading Association and the National Reading Conference, exist and knit researchers and research e f f o r t s together through various information e f f o r t s . They also generate an extensive secondary information system which produces major reviews, research summaries and other state-of-the-art products. In addition, reading l i t e r a t u r e i s included i n an extensive information system developed for education as a whole (ERIC). Second, previous bibliometric research such as that 46 conducted by Summers (1968), Mayes (1973), Kling (1971), Raygor et a l . (1974), and by Wsintraub et a l . (1976), have brought the state of the art in the f i e l d of reading to a point just short of extended descriptive analysis and c i t a t i o n analysis. Third, an ample number of c i t a t i o n studies have been conducted i n the sciences and the s o c i a l sciences i n general, and i n education in p a r t i c u l a r , to provide legitimate comparisons based on an array of bibiometric variables. Fourth, the major reviews i n the f i e l d of reading have reached the stage where d i f f i c u l t i e s are being encountered i n selecting and covering the most relevant journal l i t e r a t u r e . Research could provide a better basis for the reviews of the future. F i f t h , perhaps common to educational research as a whole are guestions about the type and qu a l i t y of research i n a f i e l d which i s heavily p r a c t i t i o n e r oriented. The fundamental question asks, "Just how s c i e n t i f i c and i n t e r d i s c i p l i n a r y is.reading research?" Reading has been one of the most heavily research- oriented of the education d i s c i p l i n e s , and a number of indicators, such as the r e l a t i v e l y recent development of a new research-oriented journal ( Reading Research Quarterly ), and the dramatic membership growth of the International Reading Association, may suggest a trend toward increased s c i e n t i f i c awareness in reading research in p a r t i c u l a r , and i n educational research i n general. In summary, the focus of t h i s study i s on the description of selected developmental c h a r a c t e r i s t i c s and patterns of growth and structure within a c o l l e c t i o n of reading research journal 47 l i t e r a t u r e . The conceptual base, design and methodology derive from the review and analysis of selected, related l i t e r a t u r e i n the areas of information science, c i t a t i o n studies, and reading research. The study makes use of a number of measures which s a t i s f y the requirements of e f f e c t i v e c i t a t i o n analysis in answering questions germane to the study of reading research journal l i t e r a t u r e . 48 CHAPTER III DESIGN, METHODOLOGY, AND STATISTICAL ANALYSIS The purposes of the study required that a set of a r t i c l e s be sp e c i f i e d to represent the reading research journal l i t e r a t u r e f or the approximate time period 1959 to 1972. The ASRR for the periods 1959, 1964, 1968, and 1972 were designated as the journal l i t e r a t u r e c o l l e c t i o n s . (At t h i s point i t i s worth re-emphasizing that according to the def i n i t i o n s of t h i s study the terms "reference" and " c i t a t i o n " are di f f e r e n t i a t e d in the following fashion: a reference i s given and a c i t a t i o n i s received.) The a r t i c l e s contained i n these summaries were tabulated and the journals producing them became the referencing set of journals. The references contained in the a r t i c l e s were tabulated and the journals producing them became the c i t e d set of journals. The sections which follow present the design, methodology and s t a t i s t i c a l procedures used i n developing the co l l e c t i o n of journal a r t i c l e s which served as the data base, determining the developmental c h a r a c t e r i s t i c s of the reading research journal l i t e r a t u r e , organizing and applying the clustering methodology, and describing the i n t e r d i s c i p l i n a r y nature of the reading research l i t e r a t u r e . 49 DEVELOPMENT O.F THE JOURNAL LITERATURE COLLECTION This section describes the development of the referencing reading research journal l i t e r a t u r e c o l l e c t i o n . Copies of the ASRR were obtained for the four time periods and photocopies made of the referenced materials for each summary. Since the study was concerned with the journal l i t e r a t u r e , only journal a r t i c l e s i n each of the four summaries were i d e n t i f i e d and prepared for analysis. Each s p e c i f i c reference was glued to a 3 x 5 card. An attempt was made to locate a complete copy of each a r t i c l e and to photocopy the references l i s t e d throughout and at the end of the a r t i c l e . These copies were then stapled to the appropriate journal a r t i c l e card. Some journal a r t i c l e s were not located because the UBC l i b r a r y did not subscribe to some journals and because some journals were either l o s t or on extended loan and could not be retrieved. I t was also not possible to include some s p e c i a l i s t and foreign journals. The next step was to prepare the journal a r t i c l e s and th e i r associated references for computer input and processing using a special coding system (see example i n Appendix). The following data, when available, were keypunched: (1) the source journal (journal i n which the reference appeared), (2) the date of appearance (date of appearance i n the annual summary, e.g. 1959, 1964, 1968, 1972), (3) the ci t e d journal or type of publication i f non-journal (code l e t t e r s were used for journal t i t l e s or i d e n t i f i c a t i o n of books, monographs, newspapers, et c . ) , 50 (4) the date of publication (date contained i n the c i t a t i o n ) , (5) the number of authors, (6) the i d e n t i f i c a t i o n of author s e l f - c i t a t i o n s . Selected data from each reference within the set of c i t i n g a r t i c l e s was then keypunched onto computer cards using the IBM 29 Card Punch. A single card contained data from one reference. A t y p i c a l card contained a code which i d e n t i f i e d the 'source* journal i n which the reference appeared as a reference (that i s , the journal refered to i n the ASRR), the date of the summary in which the source journal appeared (1959,-1964, 1968, 1972), a number which coded the reference to a s p e c i f i c a r t i c l e in the source journal (because a source journal might have 22 or so a r t i c l e s from i t during one year), a code which i d e n t i f i e d the cited journal (journal t i t l e contained in the reference i n the source journal), the date of the reference, the number of authors, and f i n a l l y an indication as to whether the author was ' s e l f - c i t i n g * (referring to one of own works). DESCRIPTION OF THE JOURNAL LITERATURE COLLECTION The referencing set of materials was analyzed and described. Volume of a r t i c l e s and references produced for the four time periods and the number of references actually obtained for analysis were calculated and tabulated. Data were generated to specify a set of core referencing journals rank ordered on the volume of a r t i c l e s in the ASRR and on the volume of references produced. A t h i r d set of core journal data was generated based on the volume of c i t a t i o n s received. The core 51 journals were compared to each other and with the Pareto c h a r a c t e r i s t i c described by Price (1963). DETERMINATION OF THE DEVELOPMENTAL CHARACTERISTICS OF THE COLLECTION Data were developed, based on the referencing and cited materials, to i l l u s t r a t e the c h a r a c t e r i s t i c s of the l i t e r a t u r e using the following variables: average number of c i t a t i o n s per a r t i c l e , age of ci t e d material, types of publications i n the cited materials, author s e l f - c i t e s , and patterns of multiple authorship. Citations Per A r t i c l e The average number of c i t a t i o n s per a r t i c l e was computed for the journals l i s t e d in the referencing set for the four time periods and tabled for comparison. ___ __ _____ Material The age of c i t e d material was measured by subtracting the date of publication of the cited a r t i c l e from the date of appearance i n the ASRR in order to reveal the age of the material at the time i t was referred to by the author of the referencing a r t i c l e . This i s possible because the large majority of documents referred to in the ASRR are works which are published in that year. The r e s u l t s ware scaled into the following intervals:. 0 to 4 years, 5 to 9 years, 10 to 14 years, 15 to 19 years, 20 or more years, and undated. Results were displayed by number i n each category and by the percent that number represented of the t o t a l year's c i t a t i o n s . 52 2123. °f Publication The yearly percentage for the following types of publications occurring in each of the four time periods were calculated and displayed in tables: books, conference r e p o r t s — published, conference reports—unpublished, abstracts, theses, personal communications, unpublished materials, newspapers and magazines, miscellaneous documents, tests and test manuals, and journals (by journal name). Sel f - C i t a t i o n s The percentage of times an author cited his or her own work was calculated and displayed for each time period. This was computed by comparing the number of a r t i c l e s which contained at least one s e l f c i t e to the t o t a l number of a r t i c l e s . These were recorded and displayed as raw numbers and percents by time period. J3iiiii£lS authorship The 1SRR contained two groups of authors—those responsible for the referencing documents and those responsible for the c i t e d documents. Multiple authorship f o r the former group was calculated by recording the number of documents which had one, two, three, four, or f i v e plus authors across the four time periods of the sample. Multiple authorship in the c i t e d set was calculated by recording the number of documents waich had one, two, three, four, and f i v e plus authors across f i v e year i n t e r v a l s between 1940 and 1970 measured by date of publication of the c i t e d document. Documents prior to 1940 were ignored. The r e s u l t s of the two methods were compared. 53 CLUSTER ANALYSIS OF THE CITED JOURNALS This section describes the clustering rationale and the selection of the two programs used; the preparation of the data for s t a t i s t i c a l analysis, including the reduction of the matrix and adjustment for s e l f - c i t a t i o n ; and the analysis of the data using the two clu s t e r i n g techniques. The data from the four time periods of the ASRR were conflated into one set of c i t e d and c i t i n g journals for the cluster analysis. __§. Clustering Rationale and Selection of the Two Programs The clustering technique has been defined as a method of grouping a set of data points without recourse to a pre-set c l a s s i f i c a t i o n . Cluster analysis d i f f e r s from other s t a t i s t i c a l techniques, such as regression or the analysis of variance, i n that i t i s s t i l l going through a process of d e f i n i t i o n . Presently, the term i s applied to a wide variety of approaches which share the common objective of i d e n t i f y i n g groups of points within a given set of data points. In t h i s study, c l u s t e r i n g techniques were applied i n order to determine i f the journal t i t l e s cited i n the a r t i c l e s of the referencing set could somehow be organized into approximate groupings to i l l u s t r a t e the d i s c i p l i n e s being cited i n reading research l i t e r a t u r e . The cluster analysis was not expected to assign hard and fast d i s c i p l i n a r y groupings among c i t e d journal t i t l e s but to suggest the general d i s c i p l i n a r y structure. As mentioned previously, i t i s assumed that when an author c i t e s a reference, the content of that reference has some substantive 54 re l a t i o n s h i p to the topic in the paper being reported. Another c i t a t i o n to a d i f f e r e n t journal i n the same a r t i c l e can be taken as a measure of s i m i l a r i t y between the two cited journals. As large numbers of c i t a t i o n s are taken into account, journals Instead of a r t i c l e s become a more p r a c t i c a l source of s i m i l a r i t y measures among the c i t e d journals. That i s , i f journal A contains a large number of a r t i c l e s which c i t e journal B, then i t can be said that A c i t e s B heavily. If A also c i t e s journal C heavily then the assumption i s that B i s related to C. The r e l a t i o n s h i p between C and B can be measured across a variety of referencing journals, and the results can be used to define groups of related journals based on s i m i l a r i t y measures obtained from the referencing set of journals. Thus, cited journals are formed into groups based on the c i t a t i o n habits of the authors of the a r t i c l e s i n the referencing journals. A h i e r a r c h i c a l clustering program begins by grouping the most highly related journals together, then the next highly related, and so on u n t i l a l l the journals are joined into one large group. Since the end points of the program are meaningless, the program must be interrupted at some point i n order to examine the groups formed at that point. A variety of rationale exist for determining the optimal point of interruption and w i l l be discussed l a t e r . The journal t i t l e s i n a group can be examined to see i f they embrace a s i m i l a r substantive content or exhibit a single d i s c i p l i n a r y focus. Recent Increased interest i n the use of c l u s t e r i n g technigues for the organization of bibliographic material has arisen both from the desire to plan patterns of journal coverage 55 for secondary services and from the desire to explore and map the structure of the primary l i t e r a t u r e i n a d i s c i p l i n e . A review of the l i t e r a t u r e in both these areas produced numerous clu s t e r i n g studies but f a i l e d to reveal any firm guidelines for s e l e c t i n g "the" technique among the wide variety of e x i s t i n g clustering techniques and programs. In f a c t , i n an excellent review of the f i e l d by Arms and Arms (1 973), a number of studies were faulted for t h e i r f a i l u r e to describe i n d e t a i l the methods used i n cl u s t e r i n g . The c l u s t e r i n g technique i s s t i l l being developed, and the e x i s t i n g methods tend to be i d i o s y n c r a t i c to the p a r t i c u l a r study i n which they are used. The reader desiring a broader elaboration of the rationale and methodology related to the c l u s t e r i n g of journal t i t l e s , plus an extensive review of the related l i t e r a t u r e to date, i s referred to the Arms and Arms (1973) study just mentioned. The most widely used approach to journal c l u s t e r i n g makes use of the frequency of c i t a t i o n to a particular journal across a set of referencing journals. Clustering groups data points which consist of measurements on each of a f i x e d set of variables. In journal t i t l e c l u s t e r i n g , the data points are cited journal t i t l e s , the variables are referencing journal t i t l e s , and the measurements consist of frequencies of c i t a t i o n . Although researchers are aware of the dangers of eguating value, use, and frequency of c i t a t i o n , there i s , as Arms and Arms (1973) and others have pointed out, a positive r e l a t i o n s h i p between use and c i t a t i o n . That i s , one can assume that when an author c i t e s an a r t i c l e , the a r t i c l e was germane to the study, and further, that material i s not generally cited haphazardly or 56 following some devious rationale. Two of the more commonly used and comprehensive c l u s t e r i n g methods which followed t h i s pattern were selected f o r use and comparison i n t h i s study. These are l a b e l l e d the UBC--C Group method and the O s i r i s Hiclust method, (descriptions of both programs are given i n the appendix.) In the C Group method, the algorithm places the objects to be clustered into n-dimensional space and forms c l u s t e r s based on the Euclidean distances between the points. In the O s i r i s Hiclust method, the algorithm calculates correlations between the objects and t r e a t s the scores as s i m i l a r i t y measures on which the c l u s t e r i n g i s accomplished. Arms and Arms (1973) provide elaboration on the rationale and methodology underlying each method. If perfect clustering could be obtained, both programs would provide the same res u l t s . This i s seldom the case. If the two methods produced guite d i f f e r e n t hierarchies, t h i s would demonstrate that the concept i s not a clusterable one and that the data are too "noisy" for precise analysis. For purposes of t h i s study, i t was decided to use the UBC C Group program to produce one cl u s t e r hierarchy, the O s i r i s Hiclust program to produce another, and then to compare the two as a r e l i a b i l i t y check on the " c l u s t e r - a b i l i t y " of such a limited sample as the one being used i n t h i s study. (Often, clustering of journal t i t l e s i s accomplished using masses of data such as that contained in a large c o l l e c t i o n of c i t a t i o n s l i k e the SCI.) 57 J?I££3£a.tion Qt. iii® Cluster Data The c i t a t i o n data from the four time periods were arranged into an HI X n matrix where m i s the number of ci t e d journals and n i s the number of referencing journals. A p a r t i c u l a r entry (Eij) represents the number of times journal J referred to journal I. The UBC C Group clustering program w i l l accept an m x n multivariate array and plot the m points in n-dimensional space. The c l u s t e r i n g i s based on the Euclidean distance between the ends of the vectors. The volume that a journal i s ci t e d w i l l a f f e c t the clustering in the following fashion: imagine the n-dimensional space reduced to the two dimensions n1 and n2 as i n Figure 1. I 1 j INSERT FIGURE 1 ABOUT HERE | I Points A, B, C, and D represent the location i n space of various cited journals. The length of the vector i s determined by the volume of c i t a t i o n , but the direction of the vector i s determined by the r e l a t i v e proportion of the c i t a t i o n s received from journals n1 and n 2 . Clustering by Euclidean distance w i l l group journals A and B together and journals C and D together. If the matrix i s normalized so that the vectors are reduced to unit dimension space, as in Figure 2 , the program w i l l c l u s t e r A with C and B with D. This clustering more accurately r e f l e c t s Note: The figures In t h i s d i s s e r t a t i o n were produced on the UBC PLOT program available through the Computing center, University of B r i t i s h Columbia. 58 Figure 1 REFERENCING JOURNAL NI Cited Journals i n S i m p l i f i e d Two-Dimensional Space Before Normalization 59 the c i t a t i o n patterns of journals n1 and n2. I INSERT FIGURE 2 ABOUT HERE | I The normalization of the multivariate array was not necessary with the O s i r i s Hiclust clustering program because t h i s program reduces the multivariate array to an m x m s i m i l a r i t y matrix i n which each entry E i j i s the c o r r e l a t i o n between cited journals I and J. The correlations (Pearson product-moment) were determined by comparing the performance of journals I and J across the referencing set of journals. Correlations are not affected by c i t a t i o n magnitudes. Reduction of the Matrix Clustering programs involve physical l i m i t a t i o n s i n terms of the amount of material that can be handled i n computer manipulation.. Also, the time taken by the algorithm increases rapidly with the number of data points. F i n a l l y , c l u s t ering i s more r e l i a b l e when the data i s as dense as p o s s i b l e — t h a t i s , when many of the c e l l s of the matrix contain some value. It i s common practice in c i t a t i o n c l u s t ering to eliminate those journals with very low c i t i n g or c i t e d frequencies. Therefore, the m x n matrix was reduced i n size (a) to accommodate the input l i m i t s of the clustering programs, (b) to increase the c e l l density of the matrix, and (c) to deal only with the most productive journals In terms of frequency of c i t a t i o n i n both the referencing and the cited sets. Figure 2 60 61 A preliminary cal c u l a t i o n indicated that reducing the number of c i t a t i o n s by one t h i r d would s a t i s f y the above conditions. An optimal reduction i n the size of the c i t a t i o n matrix would result i n the subset of rows and columns of journals for which the average number of ci t a t i o n s per c e l l , or c e l l density, reached a maximum with a pre-determined number of c i t a t i o n s , such as two-thirds of the t o t a l . This represents a complex combinatorial problem because the set of c i t a t i o n s eliminated by the deletion of rows overlaps the set eliminated by the deletion of columns. An optimal solution would locate the pattern of deletion, from a l l possible patterns, i n which the overlap was the large s t . For a large matrix t h i s solution would require an immense amount of computing time. To make the matrix reduction f e a s i b l e , a cruder approach was employed. The core referencing journals responsible for 80 percent of the c i t a t i o n s were retained as columns and the core c i t e d journals receiving 80 percent of the c i t a t i o n s were retained as rows. This should retain approximately two-thirds the t o t a l number of c i t a t i o n s (80 percent of 80 percent equals 64 percent). At t h i s point the average c e l l density was calculated by dividing number of c i t a t i o n s by number of c e l l s , and the r e s u l t s were reported. Adjustment for S e l f - c i t a t i o n Journals tend to c i t e themselves frequently and s e l f - c i t a t i o n s tend to d i s t o r t the c i t a t i o n patterns of the journals. Various methods of adjustment for s e l f c i t a t i o n have been applied in previous research. Rather than reduce the number of s e l f - c i t a t i o n s by a fixed amount, say 25%, i t was decided to reduce them by a proportion r e l a t i v e to the popularity of the 62 journal across the t o t a l set of cited journals. The assumption made was that journal A should prefer i t s e l f over other journals in the same proportion that the t o t a l referencing set prefers A over the t o t a l set of cited journals. The formula that was applied for c e l l adjustment was: [NEW SELF-CITES OF A] TOTAL CITATION OF A TOTAL~CITES~OF~A~ TOTAL~CITATIONS Clustering S t a t i s t i c a l Analysis The prepared data was run using the respective c l u s t e r i n g programs. Both the C Group and the Hiclust programs c l u s t e r the journals i n the step-wise fashion described in the Appendix. The r e s u l t s in each case are conveniently displayed i n branching tree forms known as dendograms. The dendograms i l l u s t r a t e the hierarchy of the c l u s t e r i n g process from the f i r s t step where each journal i s considered to be a separate cluster to the f i n a l step where a l l clusters are f i n a l l y joined into one large cl u s t e r comprised of the entire set of cited journals. The resulting dendograms can be displayed and compared. The dendograms are entered at the appropriate point and interpreted. The appropriate point i s determined by following the error values at each step of the program and entering at the point where there i s a large jump in t h i s figure. Clusters formed beyond that jump are joined with less confidence than the ones before i t . It i s assumed that the c l u s t e r s already formed are r e l a t i v e l y strong, and that i f the programs are revealing si m i l a r c l u s t e r s , their s i m i l a r i t y should be apparent at t h i s stage. The r e s u l t s from the C Group and the Hiclust method can 63 be compared by a pairwise matching of s i m i l a r c l u s t e r s . SUMMARY This chapter has presented the design, methodology and s t a t i s t i c a l procedures used i n developing the c o l l e c t i o n of journal a r t i c l e s which served as the data base, i n determining the developmental c h a r a c t e r i s t i c s of the reading research l i t e r a t u r e , and i n organizing and applying the c l u s t e r i n g methodology in order to obtain an outline of the i n t e r d i s c i p l i n a r y nature of the base for reading research journal l i t e r a t u r e . v 64 CHAPTER IV RESULTS OF THE STUDY T h i s c h a p t e r p re sent s the r e s u l t s of the s t u d y and i s o r g a n i z e d i n t o t h r e e main d i v i s i o n s i n c l u d i n g ; d e s c r i p t i o n o f the r e f e r e n c i n g r e a d i n g r e s e a r c h j o u r n a l l i t e r a t u r e c o l l e c t i o n ; deve lopmenta l c h a r a c t e r i s t i c s of the r e a d i n g r e s e a r c h j o u r n a l l i t e r a t u r e c o l l e c t i o n ; and c l u s t e r a n a l y s i s r e s u l t s . DESCRIPTION OF THE -REFERENCING JOURNAL LITERATURE COLLECTION The f i r s t p r e r e q u i s i t e i n any j o u r n a l c i t a t i o n s tudy i s to deve lop a se t of r e f e r e n c i n g j o u r n a l a r t i c l e s f o r a n a l y s i s . The b i b l i o g r a p h i c i n f o r m a t i o n from four t ime p e r i o d s o f the ASRS was used f o r t h i s purpose . The ASRR has been produced by a number of d i f f e r e n t a u t h o r s , at s e v e r a l d i f f e r e n t l o c a t i o n s , and has been p u b l i s h e d i n a number o f j o u r n a l s throughout i t s h i s t o r y . A complete h i s t o r y and a n a l y s i s of the ASRR from i t i n c e p t i o n through 1966 can be found i n Summers (1968). Subsequent y e a r l y summaries can be a n a l y z e d fo r the p e r i o d 1967 through 1976. The annual summaries f o r the p e r i o d s 1959, 1964, 1968, and 1972 were used as the data base f o r t h i s s t u d y . The annua l summary f o r any one year spans the academic r a t h e r than the c a l e n d e r year and i s u s u a l l y r e f e r r e d to us ing two d a t e s , i . e . , 1958-59. For purposes o f t h i s s t u d y , the summaries are r e f e r r e d to by the l a s t date of the p e r i o d c o v e r e d . Tab le I p re sent s 65 general publication information on the four summaries included in the study, Table II provides data on the number of references contained i n each summary with the proportion of journal a r t i c l e s . i INSERT TABLE I ABOUT HERE J I I INSERT TABLE II ABOUT HERE J I As i n any bibliographic study, d i f f i c u l t i e s were encountered in obtaining a complete set of materials. Time, l i b r a r y a v a i l a b i l i t y , and economics did not allow for c o l l e c t i o n of a l l 918 journal a r t i c l e s l i s t e d for the four time periods. For example, of the 113 journal a r t i c l e s l i s t e d for the 1959 summary, 91 were obtained. In a l l 768 out of 918 a r t i c l e s were included. Table III summarizes the results for a l l four time periods. f I j INSERT TABLE III ABOUT HERE J I I a. 1 I t was f e l t that for purposes of t h i s study, 8 4 percent provided a s u f f i c i e n t data base. The unobtainable journal a r t i c l e s , i n essence, approximated a random deletion and the i r absence would not unduly influence the res u l t s . Analysis of the unobtainable materials indicated that they were largely from 66 TABLE I ASRR for four time periods Date of Author (s) Journal Location Number of ASRR a r t i c l e s 1959 W.S. Gray of University 120 _-.__-li_.2B-- o f Chicago Research 1964 H.M. Robinson, The _______ University 264 S. Heintraub, Teacher of Chicago C.A. Hostetter 196 8 H.M, Robinson, _______ Indiana 376 S. Weintraub, ________ University H.K. Smith _ _ l _ _ e _ _ _ 1972 S. Weintraub, _______ Indiana 302 H.M. Robinson, Research University H.K. Smith, _uarterly G.S. Plessas ~ 67 TABLE II Total A r t i c l e s and Percent of Journal A r t i c l e s , ASHE Year of Number of Number of Percent ASHE A r t i c l e s Journal A r t i c l e s 1959 120 113 94 196 4 26 4 22 2 84 1968 37 6 3 30 88 1972 302 253 84 Total 68 TABLE III Total Journal A r t i c l e s and Percent of A r t i c l e s Located, ASSR Year of Number of Number of Percent Summary A r t i c l e s Cited A r t i c l e s Obtained 1959 113 91 81 1964 222 186 84 1968 330 278 84 1972 253 213 84 Total 69 journals which would have dropped out of the matrix reduction stage. In addition, the materials did not focus in any one of the four time periods. The three occurrences of 6 4 percent were coincidental and did not represent a conscious cut-off point. An a r t i c l e was considered obtained even i f i t contained no > references. The bibliographic data from 768 reading research a r t i c l e s , dispersed over 108 journals, were available for keypunching, computer storage and analysis. The 108 journals became the referencing set of materials. I d e n t i f i c a t i o n of Core Journals One purpose of the study was to provide descriptive data on the reading research journal l i t e r a t u r e c o l l e c t i o n . Part of the description included the delineation of the core journals for the time period 1959 through 1972. In previous research, core l i t e r a t u r e had been designated l a r g e l y on the basis of the number of a r t i c l e s published in the ASRR. In this study, core l i t e r a t u r e was designated on the basis of the number of a r t i c l e s published in the ASRR, the number of references appearing i n the a r t i c l e s published i n the ASRR, and the frequency of journal c i t a t i o n i n the referencing materials. Note that the l a s t set are journals i d e n t i f i e d by the referencing habits of the researchers whose a r t i c l e s appeared in the ASRR. Data were tabulated separately f o r each of the four time periods and then conflated i n designating the three core l i s t i n g s . Price (1963) suggested that the number of large contributing journals w i l l equal the square root of the t o t a l population and can be expected to account f o r approximately 50 percent of the t o t a l a r t i c l e s produced. Price's Pareto 70 ch a r a c t e r i s t i c i s well known and has been used widely i n information studies and provides a general rule of thumb i n ind i c a t i n g the core journal demarcation i n a set of l i t e r a t u r e . The Pareto findings of Price obtain i n the c o l l e c t i o n analyzed here. Table IV presents the data for the 108 referencing journals ranked by quantity of a r t i c l e s produced f o r the t o t a l period. The f i r s t 9 journals account for approximately 50 percent of the t o t a l number of a r t i c l e s . I I INSERT TABLE IV A BOOT HERE Using the Pareto rule of thumb, we would expect 10 journals (square root of 108) to account f o r approximately 50 percent of the a r t i c l e s . The results are very s i m i l a r — a small number of journals predominate i n the reading research l i t e r a t u r e reported in the ASRR and can be designated as core publications on the basis of their carrying a large percentage of the a r t i c l e s published. The nine journals i n t h i s core l i s t i n g include H reading journals, 2 educational research journals, 1 English journal, 1 journalism journal and one general elementary journal, suggesting that the 50 percent of a r t i c l e s produced by the core appear i n a broad variety of journals. Adding the next 25 most productive journals would produce a highly useful set of approximately three dozen core journals accounting for over 80 percent of the a r t i c l e s published in the 108 journals. Table IV presents the ranked l i s t i n g for the journals producing 80 percent of the a r t i c l e s . The data for a l l i J I 71 TABLE IV Journals by Number of A r t i c l e s i n ASRR i n Four Time Periods. (Ranked by Total.) JOORNAL TIME PERIOD CUM, 1959 1964 1968 197 2 TOTAL Journalism Quarterly 26 29 22 86 11.20 The Reading Teacher 22 36 6 67 19.92 Journal of Educational Research 18 8 14 12 52 26.69 Elementary English 11 8 11 39 31.77 Journal of Educational Psychology 13 33 36.07 Journal of Reading 0 25 32 40.23 Elementary School Journal 10 27 43.75 Journal of Developmental Reading 18 0 27 47.27 Journal of Reading Behavior 0 21 21 50.00 Jr.nl of Verbal Learning and Verbal Behavior 16 52. 08 Journal of Experimental Education 15 54.04 American Educational Research Journal 14 55. 86 Perceptual and Motor S k i l l s 0 12 57.42 Psychological Reports 12 58. 98 TABLE IV (CONTINUED) 72 JOURNAL TIME PERIOD CUM. 1959 1964 1968 1972 TOTAL B r i t i s h Journal of Educational Psychology 1 11 60.42 Reading Research Quarterly 11 61.85 Alberta Journal of Educational Research 10 63. 15 Journal of Personality and Social Psychology 0 10 0 10 64.45 Public Opinion Quarterly 10 65.76 Psychology In The Schools 10 67.06 Exceptional Children 9 68. 23 Educational Research 9 69.40 Journal of The Reading S p e c i a l i s t 9 70.57 Child Development 8 71.61 Educational Leadership 8 72.66 Journal of General Psychology 8 73.70 Education of Vis u a l l y Handicapped 7 74.61 Journal of Applied Psychology 7 75.52 Journal of Advertising Research 7 76.43 Columbia Journalism Review 6 77.21 TABLE IV (CONTINUED) JOURNAL TIME PERIOD COM % 1959 1964 1968 197 2 TOTAL Educational and Psychological Measurement I l l i n o i s School Research 6 7 7 . 9 ' 6 7 8 . 7 : Journal of Abnormal and Social Psychology Journal of Communication 6 7 9 . 5 i 6 80.3< 74 108 journals appears in the Appendix. The r e l a t i v e decrease i n a r t i c l e productivity as the lesser producing journals i n the referencing set are added, and the demarcation l i n e s f o r the 50 and 80 percent points, are also i l l u s t r a t e d graphically i n Figure 3. I | INSERT FIGURE 3 ABOUT HERE The 768 a r t i c l e s dispersed across the 108 journals of the referencing set i n turn l i s t e d 7,642 references. Reference productivity can also be used to designate useful core journals in the AS£B. The references in each journal a r t i c l e were tabulated and summed. Journals were then rank Ordered on the basis of reference frequency. The results appear i n Table V i n order of decreasing references per journal across the t o t a l period. I I | INSERT TABLE V ABOUT HERE j I I L I It should be noted that the number of references contributed for each journal i s related to but not t o t a l l y dependent on the number of a r t i c l e s a particular journal contributes. Some journals produce a small number of a r t i c l e s but these a r t i c l e s have guite high reference counts. The top 12 referencing journals contributed approximately 50 percent (3,941) of the 7,642 references. The 12 journals i n t h i s core include the 9 from the previous l i s t based on quantity Figure 3 1 1 1 1 0.0 20.0 40.0 60.0 80.0 100 CUMULATIVE % OF CONTRIBUTING JOURNALS Journals Contributing A r t i c l e s to ASHE 76 TABLE V Journals by Number of References Produced i n ASRR in Four Time Periods. (Ranked by Total.) JOURNAL TIME PERIOD CUM, 1959 1964 1968 1972 TOTAL Journalism Quarterly 34 170 192 310 706 9.24 The Reading Teacher Elementary English 149 291 44 488 15.62 118 86 92 129 425 21.19 Journal of Educational Research 118 51 100 9 9 368 26.00 Reading Research Quarterly Journal of Educational Psychology Journal of Reading 0 234 134 368 30.82 21 47 142 102 312 34.90 0 224 50 274 38.48 Journal of Experimental Education 69 99 14 49 231 41. 51 Elementary School Journal 11 52 111 34 208 44.23 American Educational Research Journal 0 26 93 81 200 46.85 Journal of Reading Behavior 0 197 197 49.4 2 Educational Research 0 27 96 41 164 51.57 Journal of Developmental Reading Perceptual and Motor S k i l l s 35 129 0 65 93 0 164 53.72 0 158 55.78 TABLE V (CONTINUED) 77 JOURNAL TIME PERIOD CUM. % 1959 1964 1968 1972 TOTAL J r n l of Verbal Learning and Verbal Behavior 47 43 67 157 57.84 Journal of Personality and Social Psychology Psychology In The Schools 0 134 0 134 59.59 0 46 85 131 61.31 Public Opinion Quarterly 44 25 41 110 62.75 B r i t i s h Journal of Educational Psychology Reading World 55 26 16 10 107 64. 15 0 105 105 65.52 Exceptional Children 8 38 5 7 103 66.87 Journal of Abnormal and Social Psychology Psychological Reports 0 88 0 88 68.02 12 14 59 85 69. 13 Education 0 74 0 83 70.22 Journal of Special Education 0 26 56 82 71. 29 Journal of General Psychology AV Communication Review 16 54 0 10 80 72.34 12 38 29 79 73.37 Journal of Genetic Psychology Alberta Journal of Educational Research 79 0 0 0 79 74.40 11 16 16 35 78 75.43 Child Development 0 30 8 39 77 76.43 Young Children 0 77 77 77.44 TABLE V (CONTINUED) JOURNAL TIME PERIOD COM % 1959 1964 1968 1972 TOTAL Journal of Social Psychology Interchange 0 17 55 0 72 78.3 0 68 68 79.2 Journal of Communication 0 15 37 13 65 80.1 American Journal of Sociology 0 61 61 80.9 79 of a r t i c l e s and adds 1 reading research journal and 3 educational research journals suggesting that, when references produced becomes the c r i t e r i o n , the 50 percent of references produced by the core appear i n an even broader array of journals, p a r t i c u l a r l y educational research journals. Table V presents the data for the 34 journals which l i s t e d approximately 80 percent of the t o t a l number of references. Complete data f o r a l l 108 journals are l i s t e d in the Appendix. The data for production of references by referencing journal also approximate Price's Pareto c h a r a c t e r i s t i c . The most frequently cited 10 journals produced approximately 4 7 percent of the t o t a l number of references. Again, a small number of journals predominate and can be designated as core publications on the basis of the number of references produced i n the ASRR. The r e l a t i v e decrease in reference productivity as lesser producing referencing journals are added i s i l l u s t r a t e d in Figure 4, as are the 50 and 80 percent demarcation l i n e s . [ } j INSERT FIGURE 4 ABOUT HERE | I J L J Core journals can also be designated on the basis of the number of c i t a t i o n s received. To develop t h i s data, the frequency of occurrence of journal t i t l e s appearinq i n the 7,642 references of the 108 referencing journals were tabulated. The resul t i n g journals were then rank ordered on the basis of volume of c i t a t i o n s received. This set represents the core journals that the authors of reading a r t i c l e s are referring to most often in t h e i r writing. The a r t i c l e s in the referencing set of Figure 1 Journals Producing References i n ASRR 81 journals produced 3,777 c i t a t i o n s to journals. These were dispersed across 448 separate journal t i t l e s . Table VI presents the re s u l t s and l i s t s the 19 journals responsible for approximately 50 percent (1,895) of the t o t a l c i t a t i o n s , and the 74 journals responsible for 80 percent of the c i t a t i o n s . The complete l i s t i n g f o r a l l 448 journals appears in the Appendix. The 19 journals i n t h i s core add a number of journals to the two previous cores i n psychology, s o c i a l psychology, developmental psychology, verbal learning, and perceptual and motor s k i l l s , suggesting that, when c i t a t i o n s received becomes the c r i t e r i o n , the 50 percent of c i t a t i o n s produced by the core appear i n a very i n t e r d i s c i p l i n a r y and diverse group of journals. I t i s in t e r e s t i n g to note that 3 of the reading journals are not included i n the 50 percent c i t a t i o n s received core although they would appear in the 80 percent core. j INSERT TABLE VI ABOUT HERE Interestingly enough, the Pareto d i s t r i b u t i o n also holds for t h i s set of journals. The most frequently cited 20 journals (square root of 448) produced 50 percent of the t o t a l c i t a t i o n s . The r e l a t i v e decrease in production of c i t a t i o n s as the lesser c i t e d journals are added i s i l l u s t r a t e d in Figure 5 as are the 50 percent and 80 percent demarcation l i n e s , 82 TABLE VI Journals by Number of Citations Received from ASRR in Four Time Periods. (Ranked by Total.) JOURNAL TIME PERIOD CUM, % 19 59 1964 1968 197 2 TOTAL Journal of Educational Psychology Journal of Educational Research Elementary School Journal 39 48 25 31 16 49 75 81 243 6.43 77 4 5 178 11.15 70 26 161 15.41 Journalism Quarterly 32 45 67 146 19.27 Elementary English 19 19 52 40 130 22.72 Journal of Experimental Psychology The Reading Teacher 57 28 4 0 129 26.13 8 15 43 55 121 29.34 Journal of Abnormal and Social Psychology Perceptual and Motor S k i l l s Child Development 8 33 66 10 117 32.43 26 55 86 34.71 0 13 21 46 80 36.83 Journal of Applied Psychology Jr.nl of Verbal Learning and Verbal Behavior Journal of Experimental Education 10 18 36 16 24 13 73 38.76 24 42 69 40.59 6 2 42.2 3 Public Opinion Quarterly 22 16 13 56 43.71 TABLE VI (CONTINUED) JOURNAL- TIME .PERIOD CUM. 1959 1964 1968 1972 TOTAL American Journal of Psychology Psychological B u l l e t i n 24 14 14 55 45.17 28 ' 16 54 46.60 B r i t i s h Journal of Educational Psychology Psychological Review 25 50 47.92 10 10 1 8 44 49. 09 Journal of Genetic Psychology Journal of Personality and Social Psychology Journal of Consulting Psychology Journal of Developmental Reading Journal of Psychology 12 14 10 10 41 50.17 0 35 8 12 18 41 51.26 8 40 52.32 21 1 1 38 53. 32 38 54.33 Education 15 37 55.31 Reading Research Quarterly Exceptional Children 8 29 37 56.29 10 22 34 57. 19 School and Society 10 10 34 58.09 Journal of Reading 0 12 21 33 58.96 Psychological Reports 10 15 32 59. 81 American Journal of Orthopsychiatry 1 12 16 31 60. 63 TABLE VI (CONTINUED) 84 JOURNAL TIME PERIOD COM, % 1959 1964 1968 1972 TOTAL Educational Research 12 5 6 7 30 61.42 B u l l e t i n American Educational 0 0 14 15 29 62.19 Research Journal English Journal 1 5 6 17 29 62.96 Journal of Social 7 6 13 3 29 63.73 Psychology American Psychologist 5 7 15 2 29 64.50 American Journal of 0 3 13 11 27 65.21 Mental Deficiency Educational and 3 1 18 2 24 65.85 Psychological Measurement Journal of Personality 3 7 10 4 24 66.48 Canadian Journal of 0 9 5 8 22 67.06 Psychology B r i t i s h Journal of 4 4 6 7 21 67.62 Psychology Science 0 5 7 9 21 68.18 School Review 3 6 10 2 21 68.73 American So c i o l o g i c a l 1 5 9 4 19 69.23 Review Speech Monographs 3 2 12 2 19 69.74 Teachers College Record 3 4 7 5 19 70.24 Review of Educational 0 0 9 8 17 70.69 Research C a l i f o r n i a Journal of 1 5 3 7 16 71.11 Educational Research TABLE VI (CONTINUED) 85 JOURNAL TIME PERIOD CUM, 1959 1964 1968 197 2 TOTAL Harvard Educational Review 8 16 71,54 Educational Admin and Supervision 1 15 71.94 Jr.nl of the Experimental Analysis of Behavior 15 72.33 Journal of General Psychology 8 15 72.73 Journal of Reading Behavior 0 15 15 73.13 Psychometrika 15 73.52 Psychonomic Science 15 73.92 Genetic Psychology Monographs 14 74. 29 Journal of Speech and Hearing Disorders 14 74.66 Psychology In The Schools 14 75.03 Journal of Education 13 75.38 Journal of Learning D i s a b i l i t i e s 0 13 13 75.72 Acta Psychologica 13 76.07 Journal of Communication 12 76.38 Phi Delta Kappan 12 76.70 Elementary English Review 12 77.02 Editor and Publisher 12 77.34 TABLE VI (CONTINUED) 86 JOURNAL TIME PERIOD CUM, 19 59 1964 1968 1972 TOTAL Social Forces 12 77.65 .educational Research 11 77.95 Archives of Psychology 11 78.24 Jr n l of comparative and Physiological Psychology 11 78.53 Quarterly Journal of Experimental Psychology 8 11 78.82 Journal of C l i n i c a l Psychology 10 79.08 Personnel and Guidance Journal 1 10 79.35 NEA J o u r n a l 10 79.61 Peabcdy ^Journal of Education 1 10 79.88 Perception and Psychophysics 0 10 10 80.14 INSERT FIGURE 5 ABOUT HERE J I .j 87 Comparison of the Core Literature Results The results of the previous section confirm that i t i s possible to establish a minimal nucleus of core journals which are highly productive i n terras of the t o t a l number of a r t i c l e s or references produced in a c o l l e c t i o n . The core concept i s also useful i n establishing the key ci t e d journal t i t l e s in the references of a c o l l e c t i o n . Core t i t l e s can be based on various zones or leve l s or productivity.•A tight set of journals i s produced when a 50 percent factor i s set whereas a broader set of t i t l e s i s generated when the c r i t e r i o n i s set at 80 percent. For purposes of establishing good t i t l e s for reviews of reading research, i t i s f e l t the 80 percent c r i t e r i o n would produce an optimal l i s t i n g . I t i s important to keep i n mind that although the bulk of the important papers i n a specialty w i l l tend to congregate in the key publications, other good papers w i l l be scattered throughout the more general l i t e r a t u r e . This i s p a r t i c u l a r l y true of an emerging d i s c i p l i n e with a diffuse l i t e r a t u r e . Therefore, a l i b e r a l 80 percent c r i t e r i o n has the potential of casting a broad enough net while at the same time s t i l l reducing the labor that would be involved i n adopting a t o t a l 100 percent c r i t e r i o n . Table VII compares the journal t i t l e s producing 5 0 percent of the t o t a l materials based on the number of a r t i c l e s , number of references l i s t e d , and c i t a t i o n volume. Figure 5 20.0 40.0 60.0 80.0 CUMULATIVE % OF RECIP IENT JOURNALS Journals Receiving C i t a t i o n s frcra ASRQ 89 I INSERT TABLE VII ABOUT HERE | I The two l i s t i n g s based on a r t i c l e s and references have many common journals, but the journal base i s nroadened when reference productivity i s added as a c r i t e r i o n f o r i n c l u s i o n . Both should be considered in establishing key journal l i s t s for various purposes, p a r t i c u l a r l y i n working with c o l l e c t i o n s that are research-oriented. There i s usually considerable c o r r e l a t i o n between a r t i c l e and reference productivity but t h i s i s not always the case. For example, on the basis of number of a r t i c l e s produced, the Reading Research _ u _ r t _ r l _ would not have been designated a -key journal, whereas, on the basis of references produced, i t becomes an important journal. Important t i t l e s can be omitted using a r t i c l e productivity as the soie c r i t e r i o n i n designating key journals. The interaction between reading and journals i n other d i s c i p l i n e s i s c l e a r l y indicated i n examining the c i t a t i o n s received column. Journal citedness would seem to o f f e r an important tool i n gaining insight into the manner i n which reading establishes substantive connections with other f i e l d s , p a r t i c u l a r l y in the core l i t e r a t u r e . The cluster analysis, described l a t e r , provides a means of s t a t i s t i c a l l y grouping the most cited journal t i t l e s into d i s c i p l i n e arrays. i 1 1 90 TABLE VII Core Journals Producing 50 Percent of A r t i c l e s , iieferences, and Citations JOURNAL TITLE ART. REF. CIT. Journalism Quarterly X X X Reading Teacher X X X Journal of Educational Research . . X X x Elementary English X X X Journal of Educational Psychology X X X * Journal of Reading X X Elementary School Journal X X X •Journal of Developmental Reading ........... X X Journal of Reading Behavior ................ X X Reading Research Quarterly X Journal of Experimental Education ................ X X American Educational Research Journal ............ X Educational Research X Journal of Experimental Psychology X Journal of Abnormal and Social Psychology X Perceptual and Motor S k i l l s X Child Development X Journal of Applied Psychology .......................... x Journal of Verbal Learning and Verb. Beh. .............. X Public Opinion Quarterly X American Journal of Psychology X Psychological B u l l e t i n X B r i t i s h Journal of Educational Psychology X Psychological Review X Journal of Genetic Psychology .......................... X *The same journal. 91 DEVELOPMENTAL CHARACTERISTICS OF THE READING RESEARCH JOURNAL LITERATURE COLLECTION Citations Per A r t i c l e Another s t a t i s t i c of interest i n r e l a t i o n to measuring the scholarly attributes in a c o l l e c t i o n of l i t e r a t u r e , i s the average number of c i t a t i o n s per a r t i c l e . Table VIII presents the data for each of the four time periods. I t i s important to note that review a r t i c l e s which p u l l together large bibliographies were not included i n the o r i g i n a l l i s t i n g of a r t i c l e s produced in the four issues of the ASRR. | INSERT TABLE VIII ABOUT HERE j I j The steady increase i n c i t a t i o n s from an average of 6.89 per a r t i c l e in 1959 to 12.61 per a r t i c l e in 1972 should be noted. 13§ of Cited Material Price (1970) suggested that age of ci t e d materials provides some clue as to the existence and recency of the research front i n a f i e l d . Table IX presents the results f o r age of cited materials across the four time periods. 92 TABLE VIII Average Number of References per A r t i c l e 1959 YEARS 1964 1968 1972 TOTAL REFERENCES 627 1562 2768 2 685 TOTAL ARTICLES 91 186 278 213 AVERAGE REFERENCES PEE ARTICLE 6. 89 8.40 9. 96 12.61 93 .j 1 INSERT TABLE IX AEOUT HERE I i After the i n i t i a l jump from 12 percent in 1959, the percentage of documents cited less than f i v e years old stays constant at approximately 22 percent of the t o t a l . This figure i s Price's Index (percentage of references to l a s t f i v e years), and indicates that reading research does not make use of a research front, but examination of the other categories indicates a trend toward such a front. There i s a steady decline in the 15-19 year old category (from 12.6 percent to 6.9 percent) and a si m i l a r decline in the 20 year or older category (from 26.2 percent to 13.6 percent). Those ci t e d documents which were r e l a t i v e l y young (5-9 years) showed a steady increase i n importance, growing from 21.1 percent to 40.1 percent of the t o t a l number of c i t a t i o n s . This r e f l e c t s a strong trend toward the c i t a t i o n of a more recent body of l i t e r a t u r e at the expense of an older body represented by those documents which were 15 years or elder at the time of c i t a t i o n . Type of Publication Book and journal c i t a t i o n s usually predominate i n most l i t e r a t u r e c o l l e c t i o n s while other materials receive l e s s frequent emphasis. The breakdown of the c i t a t i o n s into the type of publication i n which they appeared i s presented in percentage form i n Table X. I TABLE IX Age of Cited Materials YEAH AGE IN YEABS 0-4 5-9 10-14 15-19 20 or more 1959 n 81 137 158 82 170 (38) (12.9) (21.8) (25.2) (13.1) (27.1) 1964 345 5G8 261 154 294 (22.1) (32.5) (16.7) ( 9.9) (18. 8) 1968 606 1037 467 262 396 (21.9) (37.5) (16.9) ( 9.5) (14.3 197 2 581 1081 470 186 366 (21.6) (40.3) (17.5) ( 6.9) (13. 6) 95 I 1 j IN5EST TABLE X ABOUT HERE | 1 ! t i While journals account for a r e l a t i v e l y steady 5 0 percent of the c i t a t i o n s , and bocks contribute another 35 percent, some growth has taken place i n the contributions from conference reports (both published and unpublished), abstracts, unpublished l i t e r a t u r e , newspapers and magazines, and tests and test manuals. S e l f - C i t a t i o n An author may s e l f - c i t e i f previous work i s related to present work. In f i e l d s where scholarship i s cumulative, the incidence of related work may, by d e f i n i t i o n , be higher than i n f i e l d s where there are a broad range of topics with which a scholar i s expected to be conversant. Table XI presents the number and percent of a r t i c l e s in the ASRR which contained at least one s e l f - c i t e . I t was decided that i f an author s e l f - c i t e d one or twenty times in one a r t i c l e that i t was s t i l l only one s e l f - c i t i n g document (as opposed to a document which contained no s e l f - c i t e s ) . j , 1 1 1 | INSERT TABLE XI ABOUT HERE I I I Table XI shows obvious growth i n t h i s practice from 30.8 percent i n 1959 to 42.7 percent i n 1972. 96 TABLE X Percentage of C i t a t i o n by Type of Publication PUBLICATION TYPE TIME PERIOD 19 59 1964 1968 1972 TOTAL BOCKS 34. .8 39. 0 34, .9 33. , 0 35. .0 JOURNAL 52, ,6 47.7 49. , 2 49, ,3 49, .0 OTHERS* 12, .6 13. 3 15, ,9 17, ,7 16, .0 * OTHERS includes Conference Reports — Published, Conference Reports -- Unpublished, Abstracts, Theses, Personal Communications, Unpublished Documents, Newspapers/Magazines, Miscellaneous Documents, and Tests/Test Manuals. 97 TABLE XI Frequency of Author Self - C i t e s YEARS 1959 1964 1968 1972 NUMBER OF 91 186 278 213 ARTICLES NUMBER WITH 28 63 114 91 A SELF-CITE PERCENT 30. 8 33.9 41.0 42.7 98 Muljtijgle Authorship Changes i n multiple authorship have also been found to have some relationship with the nature of the research c o l l e c t i o n i n a d i s c i p l i n e . Tables XII and XIII present the number of authors (or multiple authorship) per document for the journals i n the referencing set and f o r the documents i n the c i t e d set. It should be borne in mind that the time span i s greater for c i t e d than referenced multiple authorship because the referenced multiple authors are based on appearance i n the ASRR for an approximate one year range. I ~1 | INSERT TABLE XII ABOUT HERE | f 1 J INSERT TABLE XIII ABOUT HERE | I I L 1 The single-authored documents i n the referencing set show a marked and r e l a t i v e l y dramatic drop from 6 5.93 percent i n 1959 to 50.70 percent i n 1972, while a l l categories representing multiple authorship indicate progressive growth. The growth trends in multiple authorship i n the cited set of documents are even more pronounced as i l l u s t r a t e d i n categories 2, 3 and 4 i n Table XIII. Single-authored documents account for only 56.9 percent of the documents from 1966 to 1970, whereas they accounted for about 7 5 percent i n the period from 1941 to 19 45. The r e s u l t s in Table XII and XIII are comparable when they are matched at points in time. 99 TABLE XII Frequencies (Percentages) of .Referencing Journal A r t i c l e s with One, Two, Three, and Four or More Authors f o r Four Time Periods. YEAR NUMBER OF AUTHORS 1 - 2 3 4 or more 1959 n 60 26 4 1 (6 5.9) (28.6) ( 4.4) ( 1.1) 1964 109 57 11 9 (58.6) (30.6) ( 5.9) ( 4.8) 1968 167 81 23 7 (60.0) (29. 1) ( 8.3) ( 2.5) 1972 108 7 5 23 7 (5 0.7) (35.2) (10.8) ( 3.3) 100 TABLE XIII Frequencies (Percentages) of Cited Documents with One, Two, Three, and Four or More Authors for Six Time Periods. TIME NUMBER OF AUTHORS PERIOD 1 2 3 4 or more 1941-45 n 95 26 6 (%) (74. 3) (20.4) ( 4. 7) 1946-50 176 6 3 5 (71.2) (25.5) ( 2.0) 1951-55 301 113 21 8 (67.9) (25.5) ( 4. 4) ( 1.6) 1956-60 42 0 167 57 23 (62.9) (25.0) ( 8. 5) ( 3. 3) 1961-65 641 297 64 56 (60.5) (28. 1) ( 6.0) ( 5. 1) 1966-70 434 229 71 28 (56.9) (30.0) ( 9.3) ( 3. 5) 101 CLUSTER ANALYSIS RESULTS The o v e r a l l purpose of the cluster analysis study was to take the highest ranked journal t i t l e s (See Table VI i n the Appendix) ci t e d i n the references of the materials published i n the ASRR for the four time periods and determine i f they could be grouped by c l u s t e r analysis to i l l u s t r a t e d i s c i p l i n a r y connections i n the reading research reported. As described in Chapter I I I , c l u s t e r i n g allows one to take the combined c i t a t i o n s for a journal in the c i t e d set and determine how those c i t a t i o n s are distributed across the referencing set of journals. Clustering the set of cited journals i s simply a p a r t i t i o n i n g of the set into mutually exclusive, or exhaustive, groups or c l u s t e r s . The basis for the p a r t i t i o n i n g i s the degree of s i n d l a r i t y with which two c i t e d journals are treated by the referencing set. The UBC C Group programs forms c l u s t e r s based on measures of Euclidean distances in n-dimensional space while the O s i r i s Hiclust method produces clusters based on c o r r e l a t i o n c o e f f i c i e n t s . The o r i g i n a l matrix l i s t i n g a l l c i t e d and referencing journals was 448 by 108. Applying the matrix reduction procedure described i n Chapter III eliminated low- c i t e d and low-referencing materials and resulted i n a matrix of 75 c i t e d and 32 referencing journals. At t h i s point zhe average c e l l density f o r the matrix was one. The data from t h i s matrix were run using the UBC C Group program. A dendogram was produced and provided a v i s u a l display of the clustering hierarchy, The dendogram was analyzed but proved somewhat d i f f i c u l t to 102 int e r p r e t . Entering the hierarchy at that point where the error changed from a steady r i s e to a marked degree produced a t o t a l of 17 clusters of journal t i t l e s . Although some i n t e r e s t i n g and expected combinations of journal t i t l e s emerged within these cl u s t e r s , many anomolies existed and i t was decided that too much noise in the matrix was producing implausible c l u s t e r s . Sunning the data using the O s i r i s Hiclust program corroborated t h i s impression with both hierarchies exhibiting s i m i l a r patterns. In examining some of the ambiguous c l u s t e r s , i t was noted that journals with f a i r l y low c i t a t i o n counts were often clustered, particulary at the l a t e r less r e l i a b l e steps in the program, with journals having very high c i t a t i o n counts. Some journals i n the cited set had less than 10 c i t a t i o n s over the four . time periods of the • ASRR.- In most instances, these groupings appeared to account for the largest ambiguities i n the re s u l t s . Obviously, the general matrix reduction procedures (described i n Chapter III) had not produced a " t i g h t " enough matrix to allow f o r reasonable interpretation of the r e s u l t s . I t was decided to reduce the matrix further by requiring that no ci t e d journal t i t l e enter the analysis with l e s s than 20 c i t a t i o n s over the four time periods of the ASSR. This decision also seemed reasonable in l i g h t cf the fact that the cited materials appeared i n four annual summaries selected to broadly represent the 13 year period 1959 to 1972. Thus, journals with le s s than 20 c i t a t i o n s seemed inappropriate as representatives of the most c i t e d research l i t e r a t u r e . Cut-off points of 30 and 45 c i t a t i o n s were also considered but rejected. A l i m i t a t i o n of 103 20 c i t a t i o n s seemed not too high or too low and assured that the Important most representative t i t l e s would enter the analysis. Useful interpretable data would have been l o s t using a higher c i t a t i o n c r i t e r i o n . _ _____ Results The new clustering r e s u l t s obtained with the UBC C Group program are presented i n dendogram form in Figure 6. r_ 1 I I | INSERT FIGURE 6 ABOUT HERE j 1 I For purposes of i n t e r p r e t a t i o n , the hierarchy i s entered just prior to the point where a substantial change in the error term associated with each l e v e l occurs. (See Figure 7.) f ! \ INSERT FIGURE 7 ABOUT HERE I I i I : I Since the program continues to cluster u n t i l a l l c l u s t e r s eventually unite to form one, there are several jumps In the error term. The jump closest to the approximate number of groups desired for examination i s taken as the entry point. For purposes of t h i s study, a number of groups between 4 and 10 was deemed desirable, so the data were examined for an error jump i n the range between 4 and 10 groups. The error jump indicates that reduction to the next stage of grouping would involve a substantially larger error than was associated aith the previous reduction. The Selection Index i s also helpful in entering the hierarchy. This indicates the r e l a t i v e error increase associated 104 Figure 6 - C •26th STEP American Educational Research Journal Journal of Psychology Journal of Genetic Psychology Journal of Verbal Learning and Verbal Behavior Psychological Reports Psychological B u l l e t i n Journal of Abnormal and S o c i a l Psychology American Psychologist Journal of S o c i a l Psychology American Journal of Psychology Journal of Experimental Psychology Psychological Review Canadian Journal of Psychology Journal of Personality and S o c i a l Psychology Journalism Quarterly Public Opinion Quarterly American Journal of Orthopsychiatry Perceptual and Motor 5 k i l l 3 C h i l d Development B r i t i s h Journal of Educational Psychology Exceptional Children Reading Research Quarterly Education Elementary School Journal Elementary English Journal of Educational Research Journal of Developmental Reading The Reading Teacher Journal of Reading E n g l i s h Journal Journal of Consulting Psychology Journal of Applied Psychology Educutional Research B u l l e t i n Journal of Experimental Education School and Society Journal of Educational Psychology Cluster Analysis of Journals Related by C i t a t i o n to the F i e l d of Reading Research (C-Group method) Figure 7 Relative Error at Each Step of C-Group Cluster Analysis 106 with decreasing the number of groupings by one. The plot of the logarithm cf error terms versus the number of groups provides a useful visual display of the error increase and can also be hel p f u l i n entering the dendogram. Using a l l three i n d i c e s , l e v e l 26, or that point where 10 groups had been formed, seemed the appropriate point for interpretation of the hierarchy. At that point the error jumped markedly in the reduction to 9 groups (see Figure 7 presenting the logarithm of error terms versus the number of groups). Three points should be noted about the obtained clusterings. The f i r s t , i s that the clustering i s performed on the journal t i t l e s c i t e d with the greatest volume i n the a r t i c l e s for the four time periods of the ASRR. Thus, in a sense, the c l u s t e r i n g represents only a d i v i s i o n of the high frequency core cited materials into general d i s c i p l i n e classes. The second point i s that the e a r l i e s t formed c l u s t e r s are r e l a t i v e l y stronger than those formed at l a t e r stages i n the program. F i n a l l y , i t i s of course impossible to completely separate journal t i t l e s into t o t a l l y discrete groups but i t i s possible to divide the set into r e l a t i v e l y meaningful general groupings. No optimal solution i s obtained. As i l l u s t r a t e d in Figure 6, the o v e r a l l hierarchy i n the dendogram at the 26th step (indicated by dotted line) shows 10 clusters containing 36 journal t i t l e s . These 36 core journals account for approximately 65 percent of the t o t a l c i t a t i o n s (See Table VI i n the Appendix). The balance of 35 percent are accounted for by the remaining 412 journal t i t l e s i n the cited set. The journals that are found i n each of the ten groups tend 107 to have similar patterns of c i t a t i o n across the journals i n the referencing set. The 6 journals l i s t e d in Group 1 can be used to i l l u s t r a t e the clustering concept. These journals were cited with the following frequencies i n the referencing set of journals: American .Educational Research Journal cited 29 times Journal of Psychology. c i t e d 38 times _______ of Genetic Psychology . . . . . . . c i t e d 41 times Journal of ______ _sar____ and Verbal Behavior ....7..................cited 69 times Psychological ___ort_ c i t e d 32 times _____________ ________ c i t e d 54 times The clustering program grouped these six journals together on the basis of the fac t that the c i t a t i o n s for each journal exhibited a s i m i l a r pattern across the journals i n the referencing set. In other words, the referencing journals tended to refer to these journals in similar proportions and when they tended to refer to one they tended to refer to the others. This suggests that the 6 journals possess substantive i d e a t i o n a l content that i s in some way s i m i l a r or closely r e l a t e d . This hypothesis i s based on the assumption that authors do not c i t e references haphazardly; they c i t e a r t i c l e s from other journals because those a r t i c l e s somehow relate to or support the conceptual thrust of t h e i r research. Journals exhibiting s i m i l a r s t a t i s t i c a l patterns of c i t a t i o n are thus also i d e a t i o n a l l y related. Examining the nature of the journals clustered together 108 on the basis of pattern of c i t a t i o n , and making comparisons with other journal c l u s t e r s , can i l l u s t r a t e general d i s c i p l i n a r y emphases in the research cited i n reading research a r t i c l e s . The journals i n Group 2 also exhibit s i m i l a r patterns of c i t a t i o n across the referencing set. However, this pattern i s d i f f e r e n t than that established by Group 1 or by any of the remaining 8 groups of journals. This uniqueness holds, i n turn, for a l l 10 groups. A description of each cluster obtained with the C-Group method follows. 5£PJJ£ 1 Journals American Educational Research Journal Journal of Psychology Journal of Genetic Psychology Journal of Verbal Learning and Verbal Behavior Psychological Reports Psychological B u l l e t i n Two main trees emerged i n t h i s group. Psychological Reports and Psychological B u l l e t i n clustered r e l a t i v e l y early and l a t e r joined with the Journal of Vernal Learning; and Verbal Behayior to form one tree. The American Educational Research Journal and the Journal of Psychology clustered at a somewhat l a t e r point and eventually joined with the Journal of Genetic Psychology to form the second tree. These two trees then merged to form Group 1. Group 2 Journals Journal of Abnormal and Social Psychology American Psychologist Journal of Social Psychology The Journal of Abnormal and Social Psychology joined r e l a t i v e l y early with the American Psychologist and eventually clustered with the Journal of Social Psychology to form Group 2. 109 Group 3 Journals American Journal of Psychology Journal of Experimental Psychology Psychological Review Canadian Journal of Psychology The American Journal of Psychology joined very e a r l y with the Journal of Experimental Psychology to form a c l u s t e r . This cl u s t e r l a t e r joined with Psychological Review . This three journal cluster eventually joined with the Canadian Journal of Psychology to complete Group 3. Group 4 Journals Journal of Personality and Social Psychology Journalism Quarterly Public Opinion Quarterly The Journalism Quarterly joined the Public Opinion Quarterly r e l a t i v e l y l a t e and then merged with the Journal of P§£Sonality and Social Psychology to complete Group 4. Croup 5 Journals American Journal of Orthopsychiatry Perceptual and Motor S k i l l s Child Development The American Journal of Orthopsychiatry and Perceptual and Motor S k i l l s clustered early and joined with Child Development r e l a t i v e l y early to form th i s group. S^oup 6 Journals B r i t i s h Journal of Educational Psychology The B r i t i s h Journal of Educational Psychology had not joined any other journals by the 26th step and was therefore considered i n a c l u s t e r by i t s e l f . 110 Grout- 7 ________ Exceptional Children Heading Research Quarterly ___________ ________ and the Reading ________ _________ merged at the 18th step of the program and formed c l u s t e r 7. __2__ _ Journals Education Elementary School Journal Elementary English Journal of Educational Research Journal of Developmental Reading The Reading Teacher Journal of Reading Two main trees, with two sub-clusters, emerged in t h i s group. Education and Elementary School Journal clustered very early. Elementary English and the Journal of Educational Research also clustered early. These two sub-trees then joined to form one main tree. The Journal of Developmental _______ clustered early with the _______ Teacher . This c l u s t e r l a t e r joined with the Journal of .Reading to form the second main tree. (It should be noted that the Journal of Deyelo_menta1 Reading and the Journal of _______ are the same publication. The fa c t that the program clustered the two together i s considerable support for the s t a b i l i t y and r e l i a b i l i t y of the clusters.) the two main trees then joined to complete Group 8. _____ _ ________ English Journal Journal of Consulting Psychology These two journals cluster at a l a t e stage to form Group 9. 111 GrP-U-E IQL Jo j i rna l s Journal of Applied Psychology Educational Research B u l l e t i n Journal of Experimental Education School and Society Journal of Educational Psychology Two main trees form Group 1 0 . The Journal of Applied Psychology- a n ^ the Educational Research B u l l e t i n clustered very early to form one tree. A three journal tree consisting of the ^QMIiSa.l of Experimental Education , School and Society and the Journal of Educational Psychology also cluster very early to form the second tree. These later merge to form Group 1 0 . , Group 1 appears to take into account features of general experimental psychology with that l i t e r a t u r e cited i n r e l a t i o n with facets of educational research and verbal learning and verbal behavior. Group 2 exhibits a strong flav o r of abnormal and s o c i a l psychology. Group 3 appears to take a more general psychology focus with two national psychology journals of a general nature cited i n r e l a t i o n with a review and general experimental publication i n psychology. Group 4 assumes a s o c i a l psychology flav o r with journal t i t l e s in these areas cited i n r e l a t i o n with personality and s o c i a l psychology. Group 5 has journal t i t l e s representing c h i l d development and behavior problems giving the cluster a strong growth/development f l a v o r . In Group 6 the B r i t i s h journal stands alone. Group 7 i s a rather unigue cl u s t e r with a journal related to exceptionality and a research-oriented reading journal, Group 8 contains a strong reading and language focus c i t e d i n r e l a t i o n with general education, curriculum development, methodology, and educational research. Group 9 i s d i f f i c u l t to interpret. The two journals 112 joined l a t e in the program but combine eventually with the general education and applied psychology groups. The journal t i t l e s i n Group 10 have a strong applied psychology and educational psychology and research focus. Some subjective interpretation of the 10 journal clu s t e r s i s possible. I t should be borne in mind that c l u s t e r analyses do not lend themselves to s p e c i f i c conclusions even though hard data are analyzed. The cited journal t i t l e s represent a wide variety of d i s c i p l i n a r y a c t i v i t y related to reading research, and the problem i s to provide labels to represent the focus of each c l u s t e r . One p o s s i b i l i t y was to c l a s s i f y the journal t i t l e s according to the subject scheme found in Ulrlch's International __________ Directory (1S74). This method would place t i t l e s in very broad categories such as Education, Children and Youth, Journalism, P o l i t i c a l Science, Psychology and the l i k e . These categories are too broad and allow for l i t t l e d i f f e r e n t i a t i o n . An alternative method i s followed which consists of subjectively discussing c l u s t e r s based on statements from journal e d i t o r i a l policy, f a m i l i a r i t y with the journal, and examination of copies of the journals themselves. The C-Group clustering results suggest that the d i s c i p l i n a r y p r o f i l e of the most freguently cited journals appearing in the reading research a r t i c l e s referenced i n the ASSR includes one c l u s t e r of journals representing reading, general education, elementary school curriculum, elementary language development, and educational research. This could be l a b e l l e d the reading and education c l u s t e r . These journals tend to c l u s t e r early, and thus rather strongly, i n Group 8. The 113 Group 10 journals represent a strong educational flavor and were among the e a r l i e s t journals to cluster. The Group 5 journals represent a strong growth and development f l a v o r c l u s t e r i n g r e l a t i v e l y early i n the program. Groups 6, 7, and 9 are somewhat ambiguous and t h e i r co-occurrence and c i t a t i o n patterns across the referencing set are d i f f i c u l t to interpret. However, i f Groups 6 and 7 are considered later in the c l u s t e r i n g program (after the 26th step) i n r e l a t i o n to Group 5, a reasonable grouping of journals results — although t h i s i s not a p a r t i c u l a r l y strong grouping. The remaining four groups (Groups 1,2,3, and 4) are heavily psychological i n t h e i r o r i e n t a t i o n and include groupings of journals related to general psychology, verbal learning, abnormal and s o c i a l psychology, and experimental psychology. Some journals may appear to be i n unusual relationships within groups. For example, the American Educational Hesearch Journal_s position with the psychological journals i n Group 1 may appear to be odd. However, the arrangement of journals in each x c l u s t e r i s based on aggregate a r t i c l e s c i t e d , and i s content as well as s t a t i s t i c a l l y oriented. Previous c l u s t e r i n g studies have confirmed that seme apparently odd journal t i t l e placings can e x i s t without a l t e r i n g the general " f l a v o r " of the cluster as a whole. In f a c t , i f the a r t i c l e s cited within a cluster were actually examined (a rather awesome task), i t would most l i k e l y be revealed that they hold together quite well i n terms of ideational content. It must be borne in mind that the t i t l e of a journal may also be a poor r e f l e c t i o n of the actual contents of the a r t i c l e s therein. The fact that the journal was 114 cited in similar fashion with the other journals i s a stronger rationale for i t s location in a cluster than location based purely on journal t i t l e . As G a r f i e l d (1973) points out, the paper i t s e l f , and not the location in a journal, determines the impact of i t s c i t a t i o n . l i S l i S i Results The reduced matrix data for the core journals was also run using the O s i r i s Hiclust program. The Hi c l u s t program uses the Pearson Product-Moment c o r r e l a t i o n as a s i m i l a r i t y measure between c i t e d journals, but produces a step-wise c l u s t e r i n g s i m i l a r to the C-Group method. The dendogram which graphs the l e f t to right progress of the cl u s t e r i n g i s shown i n Figure 8. -j I INSERT FIGURE 3 ABOUT HERE I I A cor r e l a t i o n term plotted for each step in d i c a t e s where the program could be interrupted and the r e s u l t s to that point examined. The rationale for the use of the c o r r e l a t i o n term i s the same as that for the use of the error term in the. C-Group method. A large correlation drop occurred after the 26th step where 10 clusters had already formed. This drop can be seen graphically displayed in Figure 9. The re s u l t s of both the C- Group and the Hiclust programs indicated that the r e l a t i v e "strength" of the cl u s t e r i n g diminished markedly after the 26th step. This can be seen by comparing Figures 7 and 9. 115 Figure 8 •C _ B r i t i s h Journal of Educational Psychology Journal of Verbal Learning and Verbal Behavior Child Development American Journal of Orthopsychiatry Perceptual and Motor S k i l l s Exceptional Children Reading Research Quarterly Education Elementary English Journal of Educational Research Journal of Reading Journal of Developmental Reading Reading Teacher Journal of Educational Psychology Journal of Experimental Education School and Society Elementary School Journal Educational Research B u l l e t i n Journal of Applied Psychology Psychological Reports English Journal Journal of Consulting Psychology Journal of Personality and S o c i a l Psychology Journalism Quarterly Public Opinion Quarterly Journal of S o c i a l Psychology American Psychologist Journal of Abnormal and S o c i a l Psychology Psychological E u l l e t i n Canadian Journal of Psychology American Journal of Psychology Journal of Experimental Psychology American Educational Research B u l l e t i n Journal of Psychology Journal of Genetic Psychology Psychological Review 26th STEP Cluster Analysis of Journals Related by C i t a t i o n to the F i e l d of Reading Research (Hiclust nethod) 116 I I INSERT FIGURE 9 ABOUT HERE A description of each of the groupings formed at the 26th step of the Hiclust program follows. Group 1 Journals B r i t i s h Journal of Educational Psychology This journal had not joined any of the other journals at the 26th step so i s considered to be i n a cluster by i t s e l f . Croup 2 Journals Journal of Verbal learning and Verbal Behavior Child Development American Journal of Orthopsychiatry Perceptual and Motor S k i l l s £hi.ld. Development , the American Journal of Orthopsychiatry , and Perceptual and Motor S k i l l s clustered r e l a t i v e l y early and were joined l a t e r by the Journal of Verbal Learning, and Verbal Behavior . Croup 3 Journals Exceptional Children Reading Research Quarterly These journals clustered together at the 18th step -in the program, just as they had i n the C-Group program. Croup it Journals C o r r e l a t i o n C o e f f i c i e n t at Each Step of H i c l u s t Cluster Analysis 118 Education Elementary English Journal of Educational Research Journal of Reading Journal of Developmental Reading The Reading Teacher Journal of Educational Psychology Journal of Experimental Education School and Society Elementary School Journal Educational Research B u l l e t i n Four d i s t i n c t subgroups clustered early within t h i s c o l l e c t i o n of obviously education-related journals. Education joined Elementary English and the Journal of Educational Research to form the f i r s t . The second contained three reading journals; the Journal of Beading , the Journal of Developmental Reading , and the Reading Teacher . (Again note that the Journal 2-f. ££velopmental Reading and the Journal of Reading are the same publication.) The t h i r d subgroup contained the Journal of I d e a t i o n a l Psychology , the Journal of Experimental Education , and School and Society . Elementary School Journal and Educational Research B u l l e t i n made up the f i n a l subgroup. The large cluster i t s e l f contains only journals which refer to education or educational subjects such as reading, exceptionality, and elementary English. Of the t o t a l 36 journals clustered only four d i s t i n c t l y education-related journals do not appear i n th i s c l u s t e r : the Reading Research Quarterly , the B r i t i s h Journal of Educational Psychology > the English Journal , and the American Educational Research Journal . In the following clusters there are 18 journals, 14 of which relate d i r e c t l y to psychology. 1 1 9 _ Jg_£M_§ Journal of Applied Psychology Psychological Reports These journals clustered r e l a t i v e l y early at the 13th step, Indicating a r e l a t i v e l y strong s i m i l a r i t y of treatment by the referencing set, but were placed i n completely d i f f e r e n t c l u s t e r s i n the C-Group method. Group 6 Journals English Journal Journal of Consulting Psychology This unigue clustering together of the English Journal and the Journal of Consulting Psychology also took place at the same step (20th) as i n the C-Group method. Grou_ 7 Journals Journal of Personality and Social Psychology Journalism Quarterly Public Opinion Quarterly The two quarterlies clustered f i r s t and were joined l a t e r by the Journal of Personality and S o c i a l Psychology . _____ 8 Journals Journal of Social Psychology American Psychologist Journal of Abnormal and Social Psychology Psychological B u l l e t i n Two of the journals here refer to s o c i a l psychology, and three, the American Psychologist , the Journal of Abnormal and Social Psychology , and the Psychological ________ are a l l American Psychological Association publications. 120 Group 9 Journals Canadian Journal of Psychology American Journal of Psychology Journal of Experimental Psyche logy Cxoujo 1,0 Journals American Educational Research Journal Journal of Psychology Journal of Genetic Psychology Psychological Review The results of the Hiclust program generally confirm those of the C-Group program. 10 clust e r s were formed by each program. I t i s interesting to note that the C-Group clusters which were judged to be somewhat ambiguous (Groups 7, 8 and 9) were exactly replicated i n the Hiclust program as groups 1, 3 and 6. Group 4 i n the C-Group program contains the same journals as Group 7 i n the Hiclust program. Except for the s h i f t i n g of some journals, and the combining of two groups, the res u l t s obtained by the two programs are quite s i m i l a r . Group 1 i n the C-Group program i s quite similar to group 10 in the Hiclust program with the exception that the Journal of Verbal learning and verbal Behavior and Psychological B u l l e t i n have s h i f t e d to other groups. Group 2 i n the C-Group method i s the same as Group 9 i n the Hiclust program with the exception that Psychological Review has been omitted from the Hiclust Group 9. Group 5 i n the C-Group program i s the some as Group 2 in the Hiclust program with the exception that the Journal of Verbal Learning and Verbal Behavior i s added to the Hiclust Group 2. Group 8 and Group 10 i n the C-Group program are e s s e n t i a l l y combined into one group, Group 4, i n the Hiclust program with the exception that the Journal of Applied Psychology from the C- 121 Group program Group 10 i s omitted. The fact that s i m i l a r distinguishable groups sere produced by both programs, and the fact that the groups can be i d e n t i f i e d with subject f i e l d s such as, for example reading and s o c i a l psychology, suggests that there i s a clusterable structure within the ci t e d journal sat and that journal c i t a t i o n data are ammenable to analysis and interpretation using c l u s t e r i n g techniques. The clustering hierarchies represent what could be termed a "network of c r o s s - d i s c i p l i n a r y journals related to reading research." Authors tend to c i t e other reading publications but their c i t a t i o n s also flow to a number of core publications which represent general d i s c i p l i n a r y groupings. The strongest f l a v o r of the groupings Is education and psychology with other areas, such as s o c i a l psychology and exceptionality, included as noted previously. The authors also c i t e a broad array of journal t i t l e s which are strongly related to the core l i t e r a t u r e i n substantive terms, but whose volume of c i t a t i o n does not qualify for i n c l u s i o n i n that group. These 414 t i t l e s did not meet the minimum 20 c i t a t i o n c r i t e r i a established for inclusion i n the core matrix. Thus, some even more diffuse d i s c i p l i n e areas are also represented in reading research. The reading research l i t e r a t u r e appears to focus across a range of d i s c i p l i n e s , both within the general educational l i t e r a t u r e and outside. The cluster analysis technique reveals the general d i s c i p l i n a r y p r o f i l e i n somewhat but not t o t a l l y systematic patterns. The cited l i t e r a t u r e analysis confirms the broad nature of reading research l i t e r a t u r e . The l i t e r a t u r e i s representative of the 122 pattern that could be expected i n an emerging f i e l d where a central tight core i s beginning to emerge but the o v e r a l l pattern i s s t i l l cross d i s c i p l i n a r y c i t a t i o n to a wide variety of l i t e r a t u r e , SUEMABY This chapter has presented the analysis of the data and the findings of the study. Of the 1062 documents reviewed i n the ASRR across the four time periods selected f o r t h i s study, 918 were journal a r t i c l e s . Of the 918 journal a r t i c l e s , 768 (84 percent) were obtained and deemed a s u f f i c i e n t data base for the purposes Core journal t i t l e s were i d e n t i f i e d in three ways: by rank ordering journal t i t l e s by a r t i c l e s in the ASRR, by t o t a l references produced by those a r t i c l e s , and by c i t a t i o n s received from the referencing a r t i c l e s . In each case, results followed a Pareto-like d i s t r i b u t i o n such that of the 10 8 referencing journals, nine journals accounted for 50 percent of the a r t i c l e s in the ASRR, and 34 journals accounted for 80 percent. In the production of references, 12 journals accounted for 50 percent of the 7,642 references produced by the 768 a r t i c l e s i n the 108 journals, and, again, 34 journals accounted for 80 percent. The th i r d measure of core journals, by c i t a t i o n s received, indicated 19 of the 448 journal t i t l e s were recepient of 50 percent of the 1,895 c i t a t i o n s to journals, and 74 journal t i t l e s were recepient of 80 percent. _________ ___ _______ The average number of c i t a t i o n s per a r t i c l e was found to 123 grow steadily from 6,8 9 to 12.61 across the 13 year i n t e r v a l of the study. This indicates that scholarliness (as defined by Price, 1970) i n the f i e l d of reading has been moving closer to the norm for science in general, which i s a paper with from 10 to 22 references. These re s u l t s p a r a l l e l the findings of Parker, Paisley, and Garrett (1967). In eight s o c i a l science journals (psychology and sociology) they found an average of 8.4 c i t a t i o n s per a r t i c l e i n 1950, 9.4 in 1955, 15.2 in 1960, and 15.2, again, i n 1965. 1.2® cf Cited Material The use of Price's Index provides a c r i t e r i o n for assessing the r e l a t i v e recency of journal l i t e r a t u r e c i t a t i o n i n reading research. Price suggested that the larger the Index, the more probable the existence of a research front i n a f i e l d . The results of t h i s study suggest steady growth i n the number of ci t e d documents which are less that 10 years old across the decade studied with the percentage increase moving from 34.7 percent i n 1959 to 61.9 percent i n 1972. This provides support for the notion that reading i s moving rapidly toward le s s reliance on archival sources and more on the c i t a t i o n of recent research l i t e r a t u r e and may indicate the existence of a reading research front that has recency c h a r a c t e r i s t i c s similar to other more " s c i e n t i f i c " d i s c i p l i n e s . However, Parker, Paisley, and Garrett (1967) found that while 70 percent of the cited material in 17 science journals i n 1965 was less than 10 years o l d , the 0 to 4 time period accounts f o r the bulk (41 percent) of the material. In reading, the 0 to 4 period accounts for only 21 percent of the c i t e d material. In addition, Price (1970) 124 estimated that i n a normally growing f i e l d , i f 22 percent of cited documents were les s than 5 years old, the f i e l d was purely archival and without an active research front, Comparisons with these data suggest an emergent possible research front in reading which i s s t i l l more archival than s o c i a l science data i n general and c e r t a i n l y more archival than hard science. T__e of Publication The journal l i t e r a t u r e i n reading cited other journal l i t e r a t u r e approximately 50 percent of the txme, books 35 percent of the time, and other types of documents about 16 percent. These r e s u l t s are comparable with the findings of Broadus (1971) that i n education i n general, 30.9 percent of the c i t a t i o n s were to books in 1953, and 32.7 percent i n 1965. Parker, Paisley, and Garrett (1967) found 43 percent of the c i t a t i o n s from 17 selected behavioral science journals were to journals and 31 percent were to books. However, Price (1970) reported that, in the SCI, the average journal a r t i c l e referred to other s e r i a l publications 80 percent of the time. Lin and Nelson (1969), i n comparing the 50 percent reference to books i n s o c i o l o g i c a l journals with the 15 percent rate in an o p t i c a l journal, concluded that d i s c i p l i n e s with paradigms tend to publish t h e i r work i n journals while d i s c i p l i n e s without tend to publish in books. When compared with these r e s u l t s , apparently reading researchers r e l y upon journal and book l i t e r a t u r e to somewhat the same extent as most s o c i a l s c i e n t i s t s , but considerably l e s s than do physical s c i e n t i s t s . 125 S e l f - C i t a t i o n Taking the proportion of author s e l f - c i t a t i o n as an indicator of the existence of cumulative scholarship within a f i e l d , the finding that s e l f - c i t a t i o n has grown markedly i n reading research from 30.8 percent i n 1959 to 42.7 percent i n 1972 suggests that research in reading may be becoming more cumulative. These findings are very si m i l a r to what Parker, Paisley, and Garrett (1967) found i n 17 behavioral science journals where percentages of s e l f - c i t a t i o n increased from 33 percent i n 1950 to 46 percent in 1965. Multiple Authorship According to Price (1970) and others, multiple authorship may be related to the amount of f i n a n c i a l support available within a f i e l d . Parker, Paisley, and Garrett, (1967) reported that the average number of authors per a r t i c l e in the s o c i a l sciences was 1.34. Lin and Nelson (1969) reported a very s i m i l a r 1.4 average in s o c i o l o g i c a l journals. The re s u l t s here suggest, that i n reading, there was a steady decline in the single- authored document from approximately 75 percent of the t o t a l in the early f o r t i e s to about 55 percent i n the la t e s i x t i e s . The averages for reading based on the referencing a r t i c l e s were found to be 1.40 authors per document i n 1959, 1.59 i n 1964, 1.53 i n 1968, and 1.66 in 1972. Si m i l a r l y , the multiple- authorship i n the cited documents increased s t e a d i l y as the number of authors per a r t i c l e grew from 1.29 during the Second World War to 1.59 in the period from 1966 to 1970. This suggests a trend i n reading to more collaborative work, and may indicate that f i n a n c i a l support of project-type a c t i v i t y i n reading 126 research Is becoming more c h a r a c t e r i s t i c of the f i e l d . Cluster Analysis The cluster analyses by the UBC C-Group and the O s i r i s Hiclust methods yielded similar though not i d e n t i c a l groupings of journal t i t l e s . The general p r o f i l e revealed a group of education related journal which clustered early and contained within i t s e l f a tight subcluster of reading journals. The other clusters relate l a r g e l y to exceptionality, growth and development, educational psychology and psychology. 127 CHAPTER V SUMMARY, CONCLUSIONS AND IMPLICATIONS, FURTHER RESEARCH SUMMARY The l i t e r a t u r e of reading research can be defined as a body of thought expressed in published writings and represents an archival record that can be measured and analyzed. This investigation i s based on the concept that studying the research journal l i t e r a t u r e can reveal much about reading as a d i s c i p l i n e i n addition to i t s substantive content. The major purposes of the study were to; establish a sample of reading research l i t e r a t u r e ; determine the core structure of the sample of reading research journals; describe developmental trends in the reading research l i t e r a t u r e ; and i l l u s t r a t e d i s c i p l i n a r y connections among journals reporting reading research. The conceptual base for the study, and the bibliometric measures used, derive from research in the f i e l d of information science. The Annual Summary of Research on Reading, published yearly since 1925, was selected to represent the reading research journal l i t e r a t u r e . That publication i s an annual summary of research, produced by the International Reading Association, and appears in the IRA journal, the Research _ u a r t e r l _ _ The ASRR provided two sets of journal t i t l e s for analysis, those t i t l e s which constitute the referencing set of journals and those t i t l e s which are cited i n the a r t i c l e s of the referencing set of journals. 128 The ASRR i s the result of an integrated program devoted to i d e n t i f y i n g and summarizing reading research on an annual basis. I t represents a preselected set of materials chosen by professionals i n reading. The summaries for the years 1959, 1964, 1968 and 1 9 7 2 — representing the years 1959 to 1972 — provided the l i t e r a t u r e for analysis. 768 of the journal a r t i c l e s published (84 percent of the total) in the four summaries were col l e c t e d and analyzed. Three major analyses were performed using the journal l i t e r a t u r e . In the f i r s t , the referencing c o l l e c t i o n of journal a r t i c l e s was described and sets of core journals generated. Developmental trends i n reading research were determined i n the second analysis using a number of bibliometric measures applied in the study of other l i t e r a t u r e s including average number of references per journal a r t i c l e , age of c i t e d materials, type of publication c i t e d , frequency of author s e l f - c i t e s , and patterns of multiple authorship both in the referencing and c i t e d set of journals. In the t h i r d analysis, two clustering programs were used to s t a t i s t i c a l l y group the core cited journal t i t l e s and i l l u s t r a t e the d i s c i p l i n a r y p r o f i l e of reading research based on l i t e r a t u r e use habits of authors. CONCLUSIONS AND IMPLICATIONS The conclusions and implications for the study are discussed under the headings of core journal structure, developmental c h a r a c t e r i s t i c s , and clustering of journal t i t l e s . Core Journal Structure The results of the study demonstrate that i t i s possible to 129 view reading research production using the core journal concept. Core journal l i s t s were i d e n t i f i e d using three fa m i l i a r indices: number of a r t i c l e s appearing in the ASRR, guantity of references produced by those a r t i c l e s , and volume of c i t a t i o n i n the referencing set of journals. For a l l three l i s t s , the most productive journals accounting for 50 and 80 percent of the t o t a l a r t i c l e s , references and c i t a t i o n s were i d e n t i f i e d . These results confirm those reported by Garfield (1972) i n working with the l i t e r a t u r e from the large data base of the SCI. The journals i s o l a t e d as the cores for the three l i s t s follow the expected general Pareto d i s t r i b u t i o n , reported i n core journal research using other l i t e r a t u r e s , demonstrating the predominance of a small number of journals as primary nodes i n the reading information network i n terms of a r t i c l e , reference, and c i t a t i o n production. The predominance of cores of journals i s indeed ubiquitous. The journals i n the three core l i s t s represent diverse subject areas such as reading, general education, educational research, growth and development, curriculum, educational psychology and various areas of psychology. The subject d i v e r s i t y i s even more pronounced when the t o t a l journal l i s t s based on the three c r i t e r i o n are examined. The 768 a r t i c l e s i n the referencing materials from the four years of the ASRR produced a t o t a l of 108 journals and the a r t i c l e s i n those journals i n turn generated 7,64 2 references. The 108 journal t i t l e s markedly expand the subject d i v e r s i t y represented by the core l i s t i n g s . This d i v e r s i t y i s expanded even more when the emphasis s h i f t s to ccunting the journals cited in the 130 references. The increase i s fo u r f o l d producing a l i s t of 448 cited journal t i t l e s . These are the t i t l e s referred to by the authors who wrote the 768 a r t i c l e s in the ASRR for the four time periods. If appears that l i t e r a l l y every subject area i n education, and a considerable number of subject areas outside education, i s i n some way related to and speculating about research in reading. However, the t o t a l s i g n i f i c a n c e of the wide d i s c i p l i n a r y spread of t i t l e s c i t e d by reading researchers has yet to be determined. In the context of t h i s study, i t i s taken as a strong indicator of m u l t i d i s c i p l i n a r y a c t i v i t y . Decisions with respect to core journal i d e n t i f i c a t i o n have considerable relevance in developing an annual summary of reading research. Most core l i s t s have been developed using a r t i c l e productivity as the c r i t e r i o n f o r monitoring journals, or, as i s the case i n some limited journal c o l l e c t i o n s , reference counts. Using cited journal t i t l e s as the c r i t e r i o n adds an important dimension to core i d e n t i f i c a t i o n . Theoretically, a hypothetical number of reading research a r t i c l e s exist "out there" for any one year. Leaving aside the f a c t that most summaries now have to be sel e c t i v e and l i m i t t h e i r size because of economics, the good summary i s the one which comes closest to i d e n t i f y i n g and presenting a l l the relevant l i t e r a t u r e for a p a r t i c u l a r year. Considering those journals cited by authors of reading research a r t i c l e s as candidate journals for monitoring would add considerable power to the surveillance network fo r an annual summary. Power would be added because, i n using c i t e d journals as a c r i t e r i o n , the network relationships become conceptual i n nature and not just 131 quantitative as i s the case when the c r i t e r i o n i s either a r t i c l e or reference productivity. In addition, such procedures would also enhance the power because the network of c i t e d journal t i t l e s , as demonstrated by the r e s u l t s of t h i s study, w i l l considerably broaden the d i s c i p l i n a r y base of journal t i t l e s which then become candidates for systematic a r t i c l e monitoring. _____________ Characteristics Price (1970) suggested that a subject grows from the body and the skin; the body, or archival record, consisting of monographic sources, and the skin, or cutting edge record, consisting largely of the research journal l i t e r a t u r e . Analysis of the c i t a t i o n s appearing i n the 768 journal a r t i c l e s published in the four annual summaries suggest several important growth c h a r a c t e r i s t i c s i n the f i e l d of reading research. Studies of various l i t e r a t u r e s suggest that there i s a slow but steady increase i n referencing i n a l l f i e l d s and that a general norm of s c h c l a r l i n e s s can be taken as 10 to 22 references per document. It i s recognized that even with a long l i s t of references some a r t i c l e s lack the s c h o l a r l i n e s s that might be bestowed by bibliographic c i t a t i o n s . Quantity should not be d i r e c t l y equated with quality and the notion that the d i s t i n c t i o n between scholarly and non-scholarly work hinqes on presence or absence of references i s a controversial one. However, bibliographic backscratching and excessive s e l f - c i t a t i o n can be considered as l i m i t i n g assumptions i n such counts. Such practices have been shown to have minimal e f f e c t s when macro conclusions across l i t e r a t u r e s are being made rather than l i m i t e d conclusions based on small c o l l e c t i o n s or 132 i n d i v i d u a l works. The increased frequency of referencing in reading research a r t i c l e s i s obvious from the data on average number of c i t a t i o n s per document. The average has grown steadily from 6.89 i n 1959 to 12.61 i n 1972. The 10 to 22 c i t a t i o n s per document norm figure may be somewhat influenced by inclusion of review a r t i c l e s and unreferenced papers. None of the former and few of the l a t t e r were included i n t h i s study making the data for reading research a harder estimate of actual recent c i t a t i o n volume and placing the subject well within the scholarly range. It has been suggested that c i t a t i o n data serve both a prescriptive and a diagnostic function; prescriptive i n that a d i s c i p l i n e must exhibit c i t a t i o n data to be scholarly and diagnostic i n that the presence of such data indicates general scholarliness. Reading research i s d e f i n i t e l y moving to the generalized scholarly norm for science and i s very close to data for other s o c i a l sciences, sociology and psychology i n p a r t i c u l a r . These findings suggest reading research emerged as a more mature d i s c i p l i n e over the period 1959 to 1972. I t can be argued that reading research i s now i n an intermediate rather than an early stage of scholarliness using average number of references per paper as the c r i t e r i o n . Another c h a r a c t e r i s t i c by which materials used i n reading research were analyzed, i s their age at the time of c i t a t i o n . New knowledge usually grows from recent findings and age of cited materials r e f l e c t s the extent to which a subject favors a recent over an older body of l i t e r a t u r e . I t has been suggested that i f 22 percent of cited documents are less than 5 years old 133 the subject tends to be archival and lacks an active research front i n the sense of the hard sciences. A 39 percent figure Indicates a subject i s in a period of rapid growth. Although the reading results show a steady increase i n the number of documents 10 years of age or less (moving from 34.7 percent i n 1959 to 61.9 percent i n 1972), the evidence does not suggest a science-like research front or immediacy e f f e c t i n i t s research l i t e r a t u r e . For other s o c i a l science data, 4 0 percent of the material was dated 4 years prior to the referencing publication and 26 percent dated within 5 to 9 years. For reading research, 21 percent of the materials were dated 0 to 4 years. There was a steady increase in materials dated 5 to 9 years with a subsequent decrease in older materials. I t can be concluded that there may be a beginning research front i n reading, based on recency of c i t e d materials, but i t s t i l l tends to be more archival than s o c i a l science data i n general and far more archival than research i n the hard sciences. It i s also inter e s t i n g to view t h i s data i n r e l a t i o n to that discussed previously for c i t a t i o n s per document. While the average number of c i t a t i o n s per a r t i c l e i s r i s i n g i n reported reading research, the age of cited material has also been decreasing although the age range i s s t i l l not within that of science a r t i c l e s in general. Reading research i s broadening i n the d i s c i p l i n e l i t e r a t u r e i t relates to, with the t o t a l picture also suggesting c i t a t i o n of more recent materials. Publication type simply indicates whether the reference being cited i s a book, journal or other type of medium. 134 V a r i a b i l i t y i n publication type depends on the d i s c i p l i n e examined, with the humanities and s o c i a l sciences using more monographic materials, while i n hard science, s e r i a l s dominate over monographic usage. For the s o c i a l sciences in general, monographic c i t a t i o n varies from 40 to 55 percent, while i n the sciences, some studies report 80 percent of the references in the average journal a r t i c l e to be to other journals. For reading research, references tc books and s e r i a l l i t e r a t u r e held steady across the four time periods at 35 and 50 percent with some growth taking place in c i t a t i o n of other source. The res u l t s suggest that, i n terms of form of publication, reading research tends to rely less on book and more on s e r i a l publications than some of the other s o c i a l sciences. It has been noted in the sciences that d i s c i p l i n e s with established paradigms tend to publish i n journals while those with l e s s established paradigms rely more on books. The data indicate that there may be some beginning movement toward established paradigms i n reading research, based on the somewhat higher use of s e r i a l over monographic c i t a t i o n s . This must be offered as a tentative conclusion and would have to be supplemented by content analysis of reading research l i t e r a t u r e to determine the existence of legitimate research paradigms. Although there has been l i t t l e work previously reported in t h i s area, degree of s e l f c i t a t i o n may provide some i n d i c a t i o n of commitment to cumulative research and i n d i v i d u a l author productivity. (It i s recognized that s e l f c i t a t i o n could also measure mediocrity or parochialism.) Data based on s o c i a l science journals suggest a trend toward increasing s e l f 135 c i t a t i o n . The figures are roughly paralleled by the reading data. Self c i t e s of 30.8 and 4 2.7 percent were reported i n 19 59 and 1972 suggesting soma increase in commitment to cumulative work. The area of Individual productivity in reading research should be followed up as a possible measure of the research emphasis i n the f i e l d . Clemente (1973) suggests that one of the few areas of consensus i n productivity studies has been that of publication output. Price (1963) also emphasizes t h i s point. Co-authorship i s reported to be increasing in most f i e l d s and i s an indicator of collaborative e f f o r t and the extent to which f i n a n c i a l support i s available for broader project type research. Sociological research provides the only comparison based on c i t a t i o n data and indicates 59 percent of a r t i c l e s with single authors and 41 percent with two or more authors. In reading research, single authorship dropped from 65.93 percent in 1959 to 50.70 percent i n 1972 In the referencing set of journals. In the cited set, the increase i n multiple authorship i s even more dramatic with single authors dropping from 74.8 percent in 1941-45 to 56.9 percent i n 196 6-70. In both sets of data, the drop i n single authorship a r t i c l e s i s accompanied by ri s e s i n multiple authorship, p a r t i c u l a r l y in two-authored papers. The data from t h i s study lend strong support to the conclusion that collaborative work i n reading research increased i n the period 1959 to 1972 and i n the wider time range 1941 to 1970. It i s l i k e l y the increase resulted from broader funding which enabled researchers to mount more comprehensive projects. This could be checked by examining the co-authorad reports i n the 1SEH for information on source of funding. 136 In summary, several conclusions can be offered, with respect to developmental c h a r a c t e r i s t i c s of reading research, based on the sample of journal l i t e r a t u r e derived from the four time periods of the ASEE. Reading research i s emerging as a more scholarly f i e l d , using quantity of c i t a t i o n s as a c r i t e r i o n . There i s a s l i g h t movement toward a more immediate research front, indicated by age of ci t e d materials, but t h i s i s not strong and the f i e l d s t i l l r e l i e s heavily on arc h i v a l and near archival resources i n i t s research. A movement toward generation of science-like paradigms may be developing in reading research, based on proportion of s e r i a l and monographic usage, but this i s tentative at best and not yet a pronounced trend. Some limited data suggest reading research may be becoming more cumulative, as indicated by increasing author s e l f c i t a t i o n i n journal a r t i c l e s . f i n a l l y , reading research i s d e f i n i t e l y becoming more collaborative based on the increasing proportions of multiple authorship, across a wide time range, in both the referencing and cited sets of journals. ClMSiSlisa of Journal T i t l e s Two s t a t i s t i c a l algorithms, one using c o r r e l a t i o n a l techniques and the other Euclidean distance, were applied to the cited set of journal t i t l e s . Matrix reduction techniques produced a l i s t i n g of 36 core t i t l e s for analysis. The t i t l e s accounted for 65 percent of the t o t a l c i t a t i o n s to journals i n the referencing set of 108 journals. It should be borne i n mind that in clu s t e r i n g , c i t e d a r t i c l e s are pooled to represent t h e i r respective journals and the program assigns journals to groups on the basis of s i m i l a r i t y i n the manner i n which their c i t a t i o n 137 axe d i s t r i b u t e d across the referencing set of journals. Thus, journals i n the f i n a l groupings exhibit conceptual as well as s t a t i s t i c a l relationships. I n t u i t i v e l y acceptable journal groupings were produced i n the cluster analysis with the two programs generally confirming each other. It can be concluded that c l u s t e r analysis i s a viable technigue for i l l u s t r a t i n g relationships i n the primary l i t e r a t u r e using c i t a t i o n data from the reading research journal l i t e r a t u r e . Ten journal groups emerged. Three were deemed somewhat ambiguous, consisting of single journals or pairs emerging late i n the c l u s t e r i n g programs. Seven groups i l l u s t r a t e d strong i n t e r r e l a t i o n s h i p s among journal t i t l e s suggesting the existence of s i m i l a r Ideational content among the journals. The journals cited by reading researchers reported i n the ASER can be described as belonging to several d i s c i p l i n e s or subjects. One group, which can be categorized as the l i t e r a t u r e most d i r e c t l y devoted to the subject of reading, incorporates journals from reading, general education, curriculum, elementary English, and educational research. The remaining six groupings can be categorized as those which, while devoted more s p e c i f i c a l l y to other subjects, also publish research related to reading. These six grouping embrace areas such as growth and development, exceptionality, educational psychology, and several branches of psychology. As indicated previously, the most frequently cited journals, and the balance makinq up the c i t e d set, provide useful input i n developing summaries of research. If one 138 intended to 'keep abreast of the most productive journals i n reading, a good m u l t i d i s c i p l i n a r y journal c o l l e c t i o n could probably be b u i l t on about four dozen t i t l e s . The c l u s t e r groups of journal t i t l e s also provide useful d i f f e r e n t i a t i o n for searching the reading l i t e r a t u r e . Those interested i n topics closely related to the subject of reading and i t s educational implications would monitor some groupings while those interested i n reading as i t might be related to other subjects or d i s c i p l i n e s would monitor other groupings. The most useful reading-related journals are usually well known in the f i e l d but i t i s d i f f i c u l t to decide which journals from the vast array available i n other subjects r e l a t e to reading i n a more direc t way. A times-cited l i s t , based on research a r t i c l e s reported in an acceptable summary l i k e the ASHE, can be very h e l p f u l i n making such decisions. FURTHER RESEARCH Several areas for further research are suggested i n the bibliometric analysis of reading research journal l i t e r a t u r e . The study of the three core l i s t s generated nuclei of core journals. Following the Pareto concept, these core journals represented 50 percent of the t o t a l a r t i c l e s , references and c i t a t i o n s produced. I t would be intere s t i n g to u t i l i z e Bradford's (1934, 1948, 1953) bi b l i o g r a p h i c a l law concerning concentration and dispersion of i n d i v i d u a l l i t e r a t u r e s and further analyze the l i s t s i n terms of the s t a t i s t i c a l dispersion of the journals beyond the cores. Succeeding sub-cores of important journals would most l i k e l y be revealed. P a r t i t i o n i n g studies of other l i t e r a t u r e s , by Goffman and Warren (1969) and 139 Leimkuhler (1967), provide a s t a t i s t i c a l methodology f o r such an analysis. It has been suggested i n these and other more recent sources (Donohue, 1973; Murphy, 1973; Pope, 1975), that dispersion analysis should be applied in many areas to generate c r o s s - d i s c i p l i n e comparison data. It would also be useful to generate a study t.o compare the 448 journal t i t l e s i n the ci t e d set with several recent editions of the A SEE to determine the degree to which such t i t l e s are now being included on a yearly basis. The ci t e d t i t l e s should also be compared with those appearing i n EEIC's Current Index to ________ i _ _______2_ (now indexing approximately 700 - journals) to determine the overlap between the general education journal l i t e r a t u r e and that being cited by reading researchers. Study of author productivity i n reading research could be prof i t a b l y extended by conducting research along the l i n e s suggested by Voos (1S74). He points out that Lotka (1926) made the i n i t i a l examination of author productivity and suggested the factor for predicting the number of papers i n a f i e l d to be 1/(n to the power of 2) of the number of authors writing only one paper (If 100 authors wrote only one paper, 25 would write two, 11 three, etc . ) . Price (1965) also suggests the harder the science the greater the probability that authors w i l l publish multiple papers. This factor provides another method f o r ranking research productivity i n various d i s c i p l i n e s along the continuum from hard to soft sciences. Voos found the author fa c t o r for information science to be 1/(n to the power of 3.5). His work provides a good model for such an analysis i n that previous methods are modified to u t i l i z e Chi square as the computational 140 method allowing one to theorize what the author publication volume would have to be to move a d i s c i p l i n e along the science non-science continuum. Extent of c i t a t i o n of one's work by others should also be examined, using reading research, to provide another indicator of research productivity. Analysis of cited journal a r t i c l e s could also yield a l i s t i n g of the a r t i c l e s which have had the most impact in reading research. These could be compiled into what Price (1965) termed a Journal of Really Important Papers. Analyses of reading research a r t i c l e s that have made a difference have been reported in the past. However, the a r t i c l e s have been subjectively selected and c i t a t i o n data have not entered such decisions. I d e n t i f i c a t i o n of nodal papers in the c i t a t i o n network could also aid i n determining i f Price's (1965) t i g h t l y knit groups of papers representing research fronts have been prevalent in reading research, Cawkell's (1974) research provides a good methodological base here. Analysis of c i t a t i o n data could also shed l i g h t on the actual proportions of ar c h i v a l and immediate papers which are ci t e d i n reading research to supplement the more gross age of ci t e d papers data. The cluster portion of the study grouped journals together on the basis of s i m i l a r i t i e s i n the i r patterns of c i t a t i o n by reading researchers. Analysis should s h i f t to the i n d i v i d u a l a r t i c l e s which are represented i n the various strong c l u s t e r groups. It would be intere s t i n g to see i f viable conceptual maps would emerge which i l l u s t r a t e the most important topics and issues i n reading research. F i n a l l y , one very useful project could be developed using 141 c i t a t i o n data. On an annual basis, the c i t a t i o n s i n a l l the research a r t i c l e s making up the ASRR could be organized i n a leading Research Literature C i t a t i o n Indeju The value o f such' indexes have been proven i n other subjects. Tnrough computer technology, a large yearly data base of such c i t a t i o n material could be developed i n a cost e f f e c t i v e fashion. BIBLIOGRAPHY 143 BIBLIOGRAPHY Arms, C. B., and Arms, W. Y., Clustering of Journal T i t l e s _2 ________ _____ Re_2__ on £_______2__ _2£_x _______ ____ _.2______2_x ___ ___________ _________ Design of Information Systems i n the Social _ci_nc_s_ Working _____ No_ T__ London, England: Bath University of Technology, 1S73. (ED 082 754) Berry, J . , "It's a Wise Child." Library Journal t November 1965, 4274. Bradford, S. C., "Sources of Information on S p e c i f i c Subjects." _ _ _ _ _ _ _ _ _ _ j _ (London), 1934, 137, 85-86. Bradford, S. C., _2_________on_ London, England: Crosby, Lockwood and Son Ltd.7~1948 (edition 1), 1953 (edition 2). Broadus, R. N., "An Analysis of the Literature Cited i n the ___£____ S o c i o l o g i c a l Review ." American S o c i o l o g i c a l __X____ 1952, 17, 355-357, Broadus, B. N., "The Literature of Educational Research." School and Society_ January 1953, 77, 8-10. Broadus, B. N., "An Analysis of References Used i n the 1960 Encyclopedia of Educational Research." Journal 2_ _______2__1 _____£__x March 1965, 58, 330-332. Broadus, R. N., "A Cit a t i o n Study for Sociology*" American _2__2_2____x February 1967, 2, 19-20. Broadus, R. N., "The Literature of the Social Sciences: a Survey of Citation Studies." International s o c i a l Science Journal_ 1971, XXIII, no. 2. Carpenter, H. P., and Narin, F., "Clustering of Journal T i t l e s . " _2_____ 2_ ___ American Society for Information Scienc__ November-December" 19737 2 8," 25-4 3 5, Cawkell, A. E., "Search Strategy, Construction and Use of Ci t a t i o n Networks, with a S o c i o - S c i e n t i f i c Example: Amorphous Semi-Conductors and S. R. Ovshinsky." Journal of the American Society for Information Science_ 1974, 25, 123-130. " " — - Clarke, B. L., "Multiple Authorship Trends i n S c i e n t i f i c Papers." ________ February 1964, 143, 822-824. Clemente, F., "Measuring So c i o l o g i c a l Productivity: A Review and a Proposal." The American Sociologist^ 1972, 7, 7. Cole, F. J., and Eales, N. B., "The History of Comparative Anatomy." Science Progress_ 1917, 11, 578-596. 144 Cola, J . , and Cole, S., "Measuring the Quality of S o c i o l o g i c a l Research; Problems In the Dse of the Science C i t a t i o n Index," American Sociologist^ February 1971, 6, 23-29. Cole, S., and Cole, J. R., " S c i e n t i f i c Output and Recognition: A Study in the Operation of the Reward System i n Science," _M£__an Soci o l o g i c a l Reyi_w_ 1967, 32, 377-390. Donohue, J. C., "A Bibliometric Analysis of Certain Information Science L i t e r a t u r e . " American Society for Information S_i_ace_ September-October 1972, 23, 313-317. Donohue, J. C., 0_d___t_nding .Scientific L i t e r a t u r e _ A Bibliometric A__roach_ Cambridge, Massachusetts; MIT Press, 1973. Donohue, J. C., and Karioth, N. E., "Coming of Age i n Academe— Information Science at 21," American Documentation_ 1966, 17, 117-119, Earle, P., and Vickery, B., "Social Science Literature Use i n the UK as Indicated by Citations." Journal of _?_______ation_ June 1969, 25, 123-141. Fairthorne, R., A., "Empirical Hyperbolic Distributions (Bradford-Zlpf-Mandelbrot) for Bibiometric Descriptions and Predictions." In T. Saracavic (Ed.), Introduction to Information Science_ London, England: R. R. Bowker Company, 1970, "521-534. G a r f i e l d , E., " C i t a t i o n Analysis as a Tool i n Journal Evaluation," Science_ November 1972, 178, 471-479. G a r f i e l d , E., " C i t a t i o n Impact Depends Upon the Paper, Not the Journal, Current Contents x 1973, 5, 5-6. G a r f i e l d , E., "Highly Cited A r t i c l e s . 19. Human Psychology and Behavior." Current Contents^ May 1975, 18, 4-11. Goffman, William, "Mathematical Approach to the Spread of S c i e n t i f i c Ideas—the History of Mast C e l l Research." Nature^ October 1966, 212. Goffman, W., and Newill, V. A., "Communication and Epidemic Processes." Proceedings of the Royal Society, 1967, A, 298, 316-334. Goffman, W., and Warren, K. S., "Dispersion of Papers Among Journals Based on a Mathematical Analysis of Two Diverse Medical L i t e r a t u r e s . " _ature_ March 1969, 221, 953-955. Gray, W. S., "Summary of Investigations Relating to Reading." Supplementary Educational Monogra_hs_ Chicago: University of Chicago Press, 1925. Gross, P. L. k., and Gross, E. M., "College L i b r a r i e s and 145 Chemical Education." Science,, 1927, 66, 385-389. Guthrie, J. T. (Ed.) , Graduate Programs and 'Faculty in Reading.. Newark, Delaware: International Reading Association, 1976, Guttsman, 8. L., "The Literature of the Social Sciences and Provision for Research in Them." Journal of Documentation^ September 1966, 22, 186-194. Hoselitz, B., Readerjs Guide to the Social Sciences^ Glencoe, I l l i n o i s : Free Press, 1961. Hulme, E. W., S t a t i s t i c a l Bibii23£l£lil i n Relation to the Growth of Modern C i v i l i z a t i o n . London, England: Grafton, 1923. Johnson, R., "An Algorithm for Successive P a r t i t i o n s of Sets of Points." Psychcmetrika x 1967, 32, 241-254. Kessler, M. M., "Bibliographic Coupling Extended in Time: Ten Case Histories." Information Storage and R e t r i e v a l t 1963a, 1, 169-187. Kessler, M. M., "Bibliographic Coupling Between S c i e n t i f i c Papers." American Documentation f January 1963b, 14, 10-25. Kessler, M. M., "Comparison of the Results of Bibliographic Coupling and Analytic Subject Indexing." American Documentation^ July 1965, 16, No. 3, 223-226. Kling, M. , "Background and Development of the Literature Search: Targeted Research and Development Program in Reading." In F. B. Davis (Ed.), Final Report.: The Literature of Research i S Raading with Emphasis on Models., Office of Education, National Center for Educational Research and Development, United States Department of Health, Education, and Welfare, Washington, D. C., 1971. Kochen, M., " S t a b i l i t y in the Growth of Knowledge." In T. Saracevic (Ed.) , Intro duct ion to Inf2ISation Science^. London, England: R. R. Bowker Company, 1970, 44-55. Kuhn, T., The Structure of S c i e n t i f i c Seyolutionsj. Chicago: University of Chicago Press, 1962. Leimkuhler, F, F., "The Bradford D i s t r i b u t i o n . " Journal of Documentation^. 1967, 23, 197-207. L i c k l i d e r , J . C. R., "A Crux i n S c i e n t i f i c and Technical Communication," American Psychologj.st A 1966, 21, 1 044-1051. Lin, N., and Nelson, C. E., "Bibliographic Reference Patterns i n Core S o c i o l o g i c a l Journals." The American S o c i o l o g i s t ^ February 1969, 4, 47-50. Livesay, M. J., "Characteristics of the Literature Used by Authors of Books in the F i e l d of Economics." Unpublished 146 Master's Dissertation, University of Chicago, 1 9 ^ 3 . Lotka, A. J., "The Frequency Di s t r i b u t i o n of s c i e n t i f i c Productivity," Journal of the Washington Academy of Sciences^ 1 9 2 6 , 167 5 1 7 - 3 2 3 . Margolis, J., " C i t a t i o n Indexing and Evaluation of S c i e n t i f i c Papers." Science*. March 1 9 6 7 , 1 5 5 , 1 2 1 3 - 1 2 1 9 . Mark, F . M., "Characteristics of the Literature Used by Contributors to American and English Economic Journals." Unpublished Master's Dissertation, University of Chicago, 1 9 5 6 . Martin, G. P., "Characteristics of the Literature Used by Authors of Books cn P o l i t i c a l Topics." Unpublished Master's Dissertation, University of Chicago, 1 9 5 2 . Mayes, B., "A Comparison of T i t l e Words and Journal Citations from the Literature of Beading." Unpublished Doctoral Dissertation, Rutgers University, 1 9 7 3 . Meier, E. L., "Characteristics of the Literature Used by Contributors of American Soci o l o g i c a l Journals," Unpublished Master's Dissertation, University of Chicago, 1951. Merton, B, K., " P r i o r i t i e s in S c i e n t i f i c Discovery: A Chapter in the Sociology of Science." American S o c i o l o g i c a l Review x 1 9 5 7 , 2 2 , 6 3 5 - 6 5 9 . Murphy, L. J., "Lotka 1s Las in the Humanities." Journal of the American Society for Information Science^ 1 9 7 3 , 2 4 , 4 6 1 - 4 6 2 . Narin, F., and Garside, D,, "Journal Relationships i n Special Education." Exceptional Children^ 1 9 7 2 , 3 8 , 6 9 5 - 7 0 4 . Nelson, C. E., and Pollack, D. K., (Eds.) Communication Among Sc i e n t i s t s and 3agine^rs x Lexington, Massachusetts: Heath Lexington Books, 1 9 7 0 . Pareto, V., Cours d^Economie Pclitigue^. Vol 2 , Section 3 , Lausanne, 1 8 9 7 . Parker, E. B., and Paisley, W. J . , "Research for Psychologists at the Interface of the S c i e n t i s t and His Information System." American Psychologist x November 1 9 6 6 , 2 1 , no. 1 1 , 1061 - 1 0 7 l T Parker, E. B., Paisley, W. J,, and Garrett, R., Bibliographic Citations as S B g g t r u s i y e Measures of S c i e n t i f i c Communication^ In s t i t u t e for Communication Research, Stanford University, October, 1 9 6 7 . Patterson, J. M., and Whitaker, B. A., Hierarchical Grouping 147 Analysis with Optional Contiguity Constraint^ Vancouver, B r i t i s h Columbia: Computing Centre, University of B r i t i s h Columbia, 197 3. Pope, A., "Bradford's Law and the Periodical Literature of Information Science." Journal of the American Society f o r Information Science^ 1975, 26,~207-213. Price, D. J. de S., L i t t l e Science^ Big Science. New York: Columbia University Press, 1963. ~ ~ ~ Price, D. J. de S., "Networks of S c i e n t i f i c Papers." Science^ July 1965, 149, 510-515. Price, D. J . de S., "Citation Measures of Hard Science, Soft Science, Technology, and Nonscience." In C. E. Nelson and D. K. Pollack (Eds.), Communication Among S c i e n t i s t s and Engineers^ Lexington, Massachusetts: Heath Lexington Books, 1970, 3-2 2. Price, N., and Schiminovich, S., "A Clustering Experiment; F i r s t Step Towards A Computer-Generated C l a s s i f i c a t i o n Scheme." Information Storage and Retrieval^. 1968, 4, 271-280. Quinn, E. W. , "Characteristics of the Literature Used by the Authors of Books i n the F i e l d of Sociology." Unpublished Master's Dissertation, University of Chicago, 1951. Rapoport, A., "What i s Information?" ETC: A Review of General Semantics^ 1953, 10, No. 4. ~ - - - Raygor, A. I., Browning, W, G., Kagan, C. E., M i l l e r , P., and Schmelzer, R., "1973 Review of Research in College-Adult Reading." In P. L. Nacke (Ed.), Interaction^. Research and Practice for College-Adult Reading;. Twenty-Third Yearbook of the National leading Conference x Clemson, South Carolina; National Reading Conference,~Inc., 1974, 23-88. Sarle, S . G,, "Characteristics of the Literature Used by Authors of Journal A r t i c l e s i n Business Administration." Unpublished Master's Dissertation, University of North Carolina, 1958. Science C i t a t i o n Index x Philadelphia: Institute for S c i e n t i f i c Information, 325 Chestnut street. Social Science C i t a t i o n Index x Philadelphia; I n s t i t u t e f o r S c i e n t i f i c Information, 325, Chestnut street. Shannon, C. E., and Weaver, W. , The Mathematical Theory, of Communication^ Urbana, 19 49. Small, fl. G. , "A Citation Model for S c i e n t i f i c S p e c i a l i t i e s . " Proceedings of the American Society for Information Iii§nce_e~1975, 12, 3 2-33. 148 Summers, E. G. , "Storing and Searching Reading Research by Computer." B u l l e t i n of the School of __________ Indiana ___________ July 1968, 44, No, 47 49-117. ________ _____________ ---iodicaIs Directory. 15th Edition, New york: Bowker, 1974. ~" ~ " ~ Voos, H., "Lotka and Information Science." Journal of the American Society, for Information Science^ 1974, 25, 270- 272. Ward, J. H., "Hierarchical Grouping to Optimize an Objective Function." Journal of the American S t a t i s t i c a l Association 196 3, 58, 236-2447 - — — - Webb, E. J., Campbell, B. T., Schwartz, R. D., and Sechrest, L. , £_£________ ________! ______£____ Research i n the Social Sciences. Chicago; Rand McNally S Co., 1966. Weintraut, S., Robinson, H. M,, Smith, H. K., Plessas, G. P., Roser, N. L., and Howls, M. , "Summary of Investigations Related to Reading, July 1, 1974 to June 30, 1975." Reading _ _ _ _ _ _ _ _ _________x 1975-1976, XI, 22 3-563, Xhlgnesse, L. V., and Osgood, C. E. , "Bibliographic C i t a t i o n Characteristics of the Psychological Journal Network in 1950 and 1960." American Psychologist. September 1967, 22, 778-791. Ziman, J. M., "Information, Communication, Knowledge." In T. Saracevic (Ed.), Introduction to Information Science. London, England: R. R. Bowkar Company, 1970, 76-83. Zipf, G. K., Psychobiolcgy of Language. Boston, Massachusetts: Houghton M i f f l i n Company, 1935. Zipf, G. K., Human Behavior and the _________ of Least E f f o r t . Cambridge, Massachusetts: Addison Wesley, 1949. 149 APPENDICES 150 APPENDIX A THE CLUSTERING PROGRAMS 151 THE CLUSTEEIHG PfiOGHAMS Two hier a r c h i c a l c l u s tering programs were used i n t h i s study. Arms and Arms (1973:21) describe the step-wise procedure which produces the re s u l t s shown i n the dendograms (see Figures 6 and 8, Chapter IV). The most common family of h i e r a r c h i c a l methods can be described as follows. (a) Input the n(n - 1 ) / 2 s i m i l a r i t i e s between the n points to be clustered. (b) Consider each point as a separate cluster. (c) Choose the two closest clusters p and q and merge them into a new cluster k which replaces p and q. (d) Compute the distance of clus t e r k from any cluster s, as a function of the distance of p and q from s. (e) Heturn to step (c) . The OBC C-Group method makes use of a s i m i l a r i t y p r o f i l e as the basis for determining i n t e r - c l u s t e r distance, Patterson and Whitaker (1971:3) , who wrote the program, describe the process i n UBC s t a t i s t i c a l package manual.' The c r i t e r i o n to determine which pair i s to be combined i s established on the basis of _______ s i m i l a r i t y where the t o t a l within-group variation i s the (value-reflecting) function minimally increased at each step i n the process. The algorithm used i n the program i s that of J. H. Ward as described i n "Hierarchical Grouping to Optimize an Objective Function," Journal of the American S t a t i s t i c a l Association. 1963, 58, 236-244. The O s i r i s Hiclust program i s described by an un-named author or authors i n the O s i r i s II manual. The method i s designed to produce successive p a r t i t i o n s of a set of points, c a l l e d c l u s t e r i n g s , by combining c l u s t e r s at one l e v e l to y i e l d a smaller number of clust e r s at the next l e v e l . The input data in t h i s program consists of a correlation matrix which gives the degree of proximity of pairs of points where the points are variables. 152 The algorithm used in the Hiclust program i s that proposed by B . Johnson i n Psjchcmetrika , 1967, 32, 241-254, 153 APPENDIX B TABLE IV Journals fay Number of A r t i c l e s in ASRR in Four Time Periods. (Ranked by Total.) TABLE V Journals by Number of References Produced i n ASRR in Four Time Periods. (Ranked by Total.) TABLE VI Journals by Number of Citations Received from ASRR i n Four Time Periods. (Ranked by Total.) 154 TABLE IV Journals by Number of A r t i c l e s in ASRR i n Four Time Periods, (Ranked by Total.) •JOURNAL TIME PERIOD CUM, 1959 1964 1968 1972 TOTAL Journalism Quarterly 26 29 22 86 11.20 The Reading Teacher 2 2 36 67 19.92 Journal of Educational Research 18 8 14 12 52 26.69 Elementary English 11 3 1 1 .39 31. 77 Journal of Educational Psychology 13 33 36. 07 Journal of Reading 25 32 40.23 Elementary School Journal 10 27 43.75 Journal of Developmental Reading 18 0 27 47.27 Journal of Reading Behavior 0 21 21 50.00 Jr.nl of Verbal Learning and Verbal Behavior 16 52.0 8 Journal of Experimental Education 15 54.04 American Educational Research Journal 14 55. 86 Perceptual and Motor S k i l l s 0 12 57. 42 Psychological Reports 12 58. 9 8 TABLE IV (CONTINUED) 15.5 JOURNAL TIME PERIOD CUM. 1959 1964 1968 1972 TOTAL B r i t i s h Journal of Educational Psychology Reading Research Quarterly Alberta Journal of Educational Research 1 11 60.42 11 61.85 10 63. 15 Journal of Personality and Social Psychology Public Opinion Quarterly 0 10 0 10 64.45 10 65.76 Psychology In The Schools 10 67.06 Exceptional Children .9 68.23 Educational Research 9 69.40 Journal of The Reading S p e c i a l i s t Child Development 9 70.57 8 71.61 Educational Leadership 8 72.66 Journal of General Psychology Education of Visually Handicapped 8 73.70 7 74.61 Journal of Applied Psychology Journal of Advertising Research 7 75.52 7 76.4 3 Columbia Journalism Review 6 77.21 TABLE IV (CONTINUED) 156 JOURNAL TIME PERIOD 1959 1964 1968 197 2 TOTAL CUM. % Educational and Psychological Measurement I l l i n o i s School Research 6 77.99 6 78.78 Journal of Abnormal and Social Psychology Journal of Communication 6 79.56 6 80. 34 Journal of Genetic Psychology Journal of Psychology Journal of Social Psychology Journal of Experimental Psychology Journal of Negro Education Journal of Special Education Reading World AV Communication Review 0 1 6 81.12 6 81.90 6 82.68 5 83.33 5 83.98 5 84.64 5 85.29 4 85.81 American Journal of Mental Deficiency B r i t i s h Journal of Psychology C a l i f o r n i a Journal of Educational Research Developmental Psychology 4 86. 33 4 86. 85 4 87. 37 4 87.89 Education 4 88.41 TABLE IV (CONTINUED) 157 JOURNAL TIME PERIOD CUM, % 1959 1964 1968 1972 TOTAL Australian Journal of Education 3 88.80 American Journal of Psychology 3 89.19 Childhood Education 3 89. 58 English Journal 3 89.97 Journal of Speech and Hearing Research Optometric Weekly 3 90.36 3 90.76 Research In The Teaching of English American Journal of Orthopsychiatry American Journal of Sociology Academic Therapy 3 91,15 2 91.41 2 91.67 2 91.93 Cognitive Psychology 2 92. 19 Journal of Abnormal Psychology Journal of Speech and Hearing Disorders 2 92.45 2 92.71 Kansas Studies In Education 2 92.97 Library Quarterly 2 93. 23 Phi Delta Kappan 2 93. 49 Science 2 93.75 TABLE I? (CONTINUED) 158 JOURNAL TIME PERIOD CUM. % 1959 1964 1968 1972 TOTAL Sccicmetry School Review 2 94.01 2 94.27 Training School B u l l e t i n 2 94.53 Xoung Children 2 94.79 Adult Education Journal American Journal of Ophthalmology American Sociological Review Bu l l e t i n of The Orton Society College Composition and Communication C a l i f o r n i a English Journal Canadian Journal of Psychology Educational Admin and Supervision Educational Broadcasting Review Educational Review Gifted Child Quarterly Genetic Psychology Monographs 0 94.92 95. 05 95. 18 95.31 95. 44 95.57 95.70 95.83 95. 96 96.09 96. 22 96.35 TABLE IV (CONTINUED) 159 JOURNAL TIME PERIOD CUM. % 1959 1964 1968 1972 TOTAL Harvard Educational Review Interchange I l l i n o i s English B u l l e t i n Journal of The American Medical Association Journal of Applied Social Psychology Journal of C l i n i c a l Psychology Journal of Education J r n l of the Experimental Analysis of Behavior Journal of Experimental Social Psychology Journal of Learning D i s a b i l i t i e s Journal of Reading D i s a b i l i t i e s Journal of Typographical Research NEA Research B u l l e t i n New Zealand Journal of Educational Studies Ontario Journal of Educational Research Personnel and Guidance Journal Programmed Learning TABLE IV (CONTINUED) 160 JOURNAL TIME PERIOD CUM, 1959 1964 1968 197 2 TOTAL Reading School and Community Science Education School L i b r a r i e s Speech Monographs School and Society School Science and Mathematics University of Kansas B u l l e t i n of Education University of Queensland Papers Vocational Guidance Quarterly Wilson Library B u l l e t i n 161 TABLE V Journals by Number of References Produced in ASRR i n Four Time Periods. (Ranked by Total.) JOURNIL TIME PERIOD COM, 1959 1964 1968 1972 TOTAL Journalism Quarterly 34 170 192 310 706 9.24 The Reading Teacher 149 291 44 488 15.62 Elementary English 118 86 92 12 9 4 25 21.19 Journal of Educational Research 1 18 51 100 99 368 26. 00 Reading Research Quarterly Journal of Educational Psychology Journal of Reading 0 234 134 368 30.82 21 47 142 102 312 34.90 0 224 50 274 38.48 Journal of Experimental Education 69 99 14 49 231 41 .51 Elementary School Journal 11 52 111 34 208 44.23 American Educational Research Journal 0 26 93 81 200 46.85 Journal of Reading Behavior 0 197 197 49.42 Educational Research 0 27 96 4 1 164 51.57 Journal of Developmental Reading Perceptual and Motor S k i l l s 35 129 0 65 93 0 164 53.72 0 158 55.78 TABLE V (CONTINUED) 162 JOURNAL TIME PERIOD CUM, 1959 1964 1968 1972 TOTAL J r n l of Verbal Learning and Verbal Behavior Journal of Personality and Social Psychology Psychology In The Schools 47 43 67 157 57.84 0 134 0 134 59.59 0 46 8 5 131 61.31 Public Opinion Quarterly 44 25 41 110 62.75 B r i t i s h Journal of Educational Psychology Reading World 55 26 16 10 107 64.15 0 10 5 105 65. 52 Exceptional Children 8 38 57 103 66.87 Journal of Abnormal and Social Psychology Psychological Reports 0 88 0 8 8 68,02 12 14 59 85 69, 13 Education 0 74 0 83 70.22 Journal of Special Education 0 26 56 82 71.29 Journal of General Psychology 16 54 0 10 80 72.34 AV Communication Review 12 38 2 9 79 73.37 Journal of Genetic Psychology 79 0 79 74.40 Alberta Journal of Educational Research 11 16 16 3 5 78 75.43 Child Development 0 30 8 3 9 77 76.43 Young Children 0 77 77 77.44 TABLE V (CONTINUED) 163 JOURNAL TIME PERIOD CUM, 1959 1964 1968 1972 TOTAL Journal of Social Psychology Interchange 0 17 55 0 72 78.38 0 68 68 79.27 Journal of Communication 0 15 37 13 65 80.12 American Journal of Sociology 0 61 61 80.92 Journal of Experimental Psychology 0 29 3 2 61 81.72 Developmental Psychology 0 55 55 82.44 B r i t i s h Journal of Psychology I l l i n o i s School Research 0 54 54 83. 15 0 53 0 53 83. 84 Journal of Applied Psychology 23 14 10 50 84,49 Journal of Speech and Hearing Research 0 14 34 0 48 85.12 American Journal of Mental Deficiency 0 13 33 46 85.72 English Journal 41 46 86.33 Journal of Psychology 0 46 0 46 86.93 Education of Visually Handicapped 0 44 44 87.50 American Journal of Orthopsychiatry Educational Leadership 0 22 0 19 41 88.04 0 15 24 39 88.55 TABLE V (CONTINUED) 164 JOURNAL TIME PERIOD 1959 1964 1968 197 2 TOTAL CUM. % Journal of The Reading S p e c i a l i s t American Journal of Psychology Genetic Psychology Monographs C a l i f o r n i a Journal of Educational Research Educational and Psychological Measurement Journal of Negro Education Science 0 38 0 38 89.05 0 25 0 1 1 36 89. 52 0 34 0 0 34 89. 96 0 0 30 1 31 90. 37 0 0 23 7 30 90. 76 0 0 2 2 8 30 91. 15 0 22 29 91.53 Training School B u l l e t i n 0 29 0 29 91. 91 Australian Journal of Education Journal of Advertising Research Wilson Library B u l l e t i n 18 0 28 0 10 28 92.28 28 92.65 0 2 8 28 93.01 Cognitive Psychology 0 27 27 93.37 Educational Review 0 27 27 93.72 Socicmetry 0 26 0 26 94.06 Adult Education Journal 0 25 25 94.39 Childhood Education 2 2 25 94.71 Journal of Learning D i s a b i l i t i e s 0 23 23 95.01 TABLE V (CONTINUED) 165 JOURNAL TIME PERIOD" CUM, 19 59 1964 1968 197 2 TOTAL American Sociological Review Academic Therapy 0 21 0 21 95.29 0 20 20 95.55 New Zealand Journal of Educational Studies Optometric Weekly 0 20 0 20 0 20 95.81 0 20 96.07 Research In The Teaching of English Journal of Abnormal Psychology University of Queensland Papers Gifted Child Quarterly 0 20 20 96,34 0 18 96,57 0 18 0 17 0 18 96.81 0 17 ' 97.03 Journal of Speech and Hearing Disorders Canadian Journal of Psychology Journal of The American Medical Association Ontario Journal of Educational Research B u l l e t i n of The Orton Society Journal of Applied S o c i a l Psychology J r n l of the Experimental Analysis of Behavior Programmed Learning 17 0 17 97.25 0 15 0 15 0 12 11 15 15 97.45 0 15 97.64 0 15 97.84 0 12 98.00 11 11 98.14 0 11 98.29 11 11 98.43 Speech Monographs 0 10 0 10 98.56 TABLE V (CONTINUED) 166 JGURNAL TIME PERIOD CO M, 1959 1964 1968 1972 TOTAL C a l i f o r n i a English Journal 8 98.67 Journal of C l i n i c a l Psychology 8 98.77 Journal of Education 8 98.87 Journal of Experimental Social Psychology 8 98.93 School Review 8 99.08 Library Quarterly 7 99.18 NEA Research B u l l e t i n 7 99.27 Reading 7 99.36 Vocational Guidance Quarterly College Composition and Communication 6 99.44 5 99. 50 Journal of Reading D i s a b i l i t i e s 5 99. 57 Journal of Typographical Research 5 99.63 Kansas Studies In Education 5 99.70 Personnel and Guidance Journal 5 99. 76 School Science and Mathematics 5 99.83 Harvard Educational Review 4 99. 88 TABLE V (CONTINUED) 167 JOURNAL TIME PERIOD CUM, 1959 1964 1968 1972 TOTAL I l l i n o i s English B u l l e t i n 4 99.93 University of Kansas Bu l l e t i n of Education 2 99.96 American Journal of Ophthalmology Educational Broadcasting Review 1 99,97 1 99.99 Phi Delta Kappan 1 100,00 Columbia Journalism Review 0 100.00 Educational Admin and Supervision 0 100.00 School and Community 0 100.00 Science Education 0 100.00 School L i b r a r i e s 0 100.00 School and Society 0 100.00 168 2ABLE VI Journals by Number of Citations Received from ASRR i n Four Time Periods. (Banked by Total.) J G O R N A L TIME PERIOD CUM. % 1959 1964 1968 1972 TOTAL' Journal of Educational Psychology 39 48 75 81 243 6.43 Journal of Educational Research 25 31 77 4 5 178 11. 15 Elementary School Journal 16 49 70 26 161 15.41 Journalism Quarterly 32 45 67 146 19.27 Elementary English 19 19 52 40 130 22.72 Journal of Experimental Psychology 57 28 4 0 129 26. 13 The Reading Teacher 15 43 55 1 21 29.34 Journal of Abnormal and Social Psychology 8 33 66 10 117 32.43 Perceptual and Motor S k i l l s 26 55 86 34.71 Child Development 13 21 46 80 36.83 Journal of Applied Psychology 10 18 73 38.76 J r n l of Verbal Learning and Verbal Behavior 24 42 69 40.59 Journal of Experimental Education 16 24 13 62 42.23 Public Opinion Quarterly 22 16 13 56 43.71 TABLE VI (CONTINUED) 169 JOURNAL TIME PERIOD CUM, 1959 1964 1968 1972 TOTAL American Journal of Psychology Psychological B u l l e t i n 24 14 14 55 45.17 28 16 54 46.60 B r i t i s h Journal of Educational Psychology Psychological Review 25 50 47.92 6 10 10 18 44 49.09 Journal of Genetic Psychology 14 10 10 41 50.17 Journal of Personality and So c i a l Psychology Journal of Consulting Psychology Journal of Developmental Reading Journal of Psychology 12 0 35 8 12 18 41 51.26 8 40 52.32 21 11 38 53.32 38 54.33 Education 15 37 55. 31 Reading Research Quarterly Exceptional Children 8 2 9 37 56.29 10 22 34 57.19 School and Society 10 10 34 58.09 Journal of Reading 0 12 21 33 58. 96 Psychological Reports 10 15 32 59. 81 American Journal of Orthopsychiatry 12 16 31 60.63 TABLE VI (CONTINUED) 170 JOURNAL TIME PERIOD CUM, 1959 1964 1968 1972 TOTAL Educational Research B u l l e t i n 12 30 61.42 American Educational Research Journal 0 14 15 29 62.19 English Journal 17 29 62.96 Journal of Social Psychology American Psychologist 13 15 29 63.73 29 64.50 American Journal of Mental Deficiency Educational and Psychological Measurement Journal of Personality 13 11 27 65.21 1 18 7 10 24 65.85 24 66.48 Canadian Journal of Psychology B r i t i s h Journal of Psychology Science 8 22 67. 06 21 67.62 21 68.18 School Review 10 21 68.73 American Soci o l o g i c a l Review 19 69.23 Speech Monographs 12 19 69.74 Teachers College Record 19 70.24 Review of Educational Research 8 17 70.69 C a l i f o r n i a Journal of Educational Research 16 71.11 TABLE VI (CONTINUED) 171 JOURNAL TIME PERIOD CUM, 1959 1964 1968 197 2 TOTAL Harvard Educational Review 8 16 71.54 Educational Admin and Supervision Jr.nl of the Experimental Analysis of Behavior Journal of General Psychology Journal of Reading Behavior Psychometrika 1 15 71.94 15 72. 33 8 15 72.73 0 15 15 73.13 15 73.52 Psyc.honom.ic Science 15 73. 92 Genetic Psychology Monographs Journal of Speech and Hearing Disorders Psychology In The Schools Journal of Education 14 74.29 14 74. 66 14 75.03 13 75,38 Journal of Learning D i s a b i l i t i e s 0 13 13 75.72 Acta Psychologlca 13 76.07 Journal of Communication 12 76.38 Phi Delta Kappan 12 76.70 Elementary English Review 12 77.02 Editor and Publisher 12 77.34 172 JOURNAL TIME PERIOD 1959 1964 1968 1972 TOTAL CUM, 3? /a Social Forces 12 77. 65 Educational Research 11 77.95 Archives of Psychology 11 78.24 J r n l of Comparative and Physiological Psychology 11 78.53 Quarterly Journal of Experimental Psychology 8 11 78.82 Journal of C l i n i c a l Psychology 10 79.08 Personnel and Guidance Journal 1 10 79.35 NEA Journal 10 79. 61 Peabody Journal of Education 1 10 79. 88 Perception and Psychophysics 0 10 10 80.14 AV Communication Review 9 80.38 American Journal of Sociology Educational Leadership 9 80.62 9 80.86 Journal of Speech and Hearing Research 9 81.10 Sociometry 9 81.3.3 Research In The Teaching of English 8 81.55 TABLE VI (CONTINUED) 173 JGORNAL TIME PERIOD CUM, % 1959 1964 1968 1972 TOTAL American P o l i t i c a l Science Review 8 81,76 Arithmetic Teacher 8 81.97 Bel l System Technical Journal 8 82.18 Journal of Experimental Child Psychology National Elementary P r i n c i p a l 8 82.39 8 82.61 Childhood Education 7 82.79 Journal of Special Education 7 82.98 Australian Journal of Psychology 7 83. 16 Biometrika 7 83.35 Brain 7 83.53 Journal of Programmed Instruction 7 83.72 Journal of Teacher Education 7 83. 90 Merrill-Palmer Quarterly of Behavior and Devel 7 84.09 Psychological Record 7 84. 27 Alberta Journal of Educational Research 6 84.43 Journal of Advertising Research 6 84.59 Journal of Experimental Social Psychology 6 84.75 TABLE VI (CONTINUED) 174 JOURNAL TIME PERIOD CUM, 19 59 1964 1968 197 2 TOTAL Journal of Negro Education 6 84.91 Ontario Journal of Educational Research 6 85. 07 College English 6 85.23 Chicago Schools Journal 6 85.39 Human Relations 6 85.54 Journal of Broadcasting 6 85.70 Journal of School Psychology Journal of Social Issues 6 85,86 6 86.02 Language and Speech 6 86.18 American Journal of Ophthalmology Developmental Psychology 5 86.31 5 86.44 Behavior Research and Therapy Education Digest 5 86. 58 5 86.71 The High School Journal 5 86. 84 Information and Control 5 86.97 Journal of American S t a t i s t i c a l Association 5 87.11 Language 5 87. 24 TABLE VI (CONTINUED) 175 JOURNAL TIME PERIOD CUM. 1959 1964 1968 1972 TOTAL Rural Sociology 5 87.37 Educational Review 4 87. 48 Journal of Abnormal Psychology Journal of The Reading S p e c i a l i s t Library Quarterly 4 87.58 4 87.69 4 87.79 NEA Research B u l l e t i n 4 87.90 School Science and Mathematics 4 88.01 Vocational Guidance Quarterly American Journal of Optometry American Journal of Psychiatry Archives Otolaryng. 4 88. 11 4 88.22 4 88.32 4 88.43 Clearing House 4 88. 54 Journal of Consulting and C l i n i c a l Psychology Journal of Experimental Pedagogy Journal of Experimental Research In Personality Mental Retardation 4 88.64 0 '4 88.75 4 88. 85 4 88.96 TABLE VI (CONTINUED) 176 JOURNAL TIME PERIOD CUM. % 1959 1964 1968 197 2 TOTAL Psychological C l i n i c 4 89.07 Pediatric Seminary 4 89. 17 Q u i l l 4 89. 28 University of Iowa Studies 4 89.38 Women: A Journal of Liberation 4 89.49 Word 4 89. 59 Academic Therapy 3 89.67 Columbia Journalism Review 3 89.75 Journal of The American Medical Association 3 89.83 Wilson Library B u l l e t i n 3 89.91 Annals of the Amrcn Acad of P o i i and Soci Science 3 89. 99 American Documentation 3 90. 07 Annals of Mathematical S t a t i s t i c s 3 90.15 Archives of Neurology 3 90.23 American S c i e n t i s t 3 90.31 Academic Therapy Quarterly 3 90. 39 Behavioral Science 3 90.47 TABLE VI (CONTINUED) 177 JOURNAL TIME PERIOD CUM, 1959 1964 1968 1972 TOTAL Catholic Education Review 3 90.55 Comprehensive Psychiatry 3 90.63 Gazette 3 90.71 IPI Report 3 90.79 Journal of Higher Education 3 90.87 Journal of Nervous and Mental Diseases 3 90. 95 Journal of Pediatrics 3 91.02 Journal of Research and Development In Education 3 91.10 Lancet 3 91.18 Media/Scope 3 91.26 Monthly Labour Review 3 91.34 Nature 3 91.42 The Nations Schools 3 91.50 Pediatrics 3 91.58 Psychologische Forschung 3 91.66 Reading Horizons 3 91.74 TABLE VI (CONTINUED) 178 JOURNAL TIME PERIOD CUM, 1959 1964 1968 1972 TOTAL Bu l l e t i n of The Orton Society Cognitive Psychology- Gifted Child Quarterly 0 0 0 2 2 9 1 . 7 9 2 91.85 2 91.90 New Zealand Journal of Educational Studies Optometric weekly Young Children Archive Fur Die Gesamti Psychologie AMA Archives of Ophthalmology American Teacher Archives of Neurological Psychiatry American Journal of Physics Barnards Journal of Education Behavior Therapy 2 91.95 2 92.00 2 92.OS 2 92. 11 2 92. 16 2 92.22 2 92. 27 2 92.32 2 92.37 2 92.43 Cerebral Palsy B u l l e t i n 2 92.48 Contemporary Psychology Child Studies 2 92. 53 2 92.59 Central States Speech Journal 2 92.64 TABLE VI (CONTINUED) 179 JOURNAL TIME PERIOD CUM. 1959 1964 1968 1972 TOTAL Ergonomics 2 92.69 Foundations of Language 2 92.75 Grassroots Editor 2 92. 80 Group Psychotherapy 2 92.85 Grade Teacher 2 92. 90 Human Factors 2 92.96 Ideas Educational 2 93.01 International Journal of American Li n g u i s t i c s Instructor 2 93.06 2 93. 12 Journal of Applied Behavior Analysis Journal of Counseling Psychology Journal of Mental Science 2 93. 17 2 93. 22 2 93. 28 Journal of Optometric Society of America Journal of Physiology 2 93.33 2 93. 38 Journal of Research In Crime and Delinquency The Journalism Educator Journal of Research In S t a t i s t i c a l Sociology 0 0 0 0 0 0 2 2 2 93.43 2 93.49 2 93. 54 TAELE VI (CONTINUED) 180 JOURNAL TIME PERIOD CUM, 1959 1964 1968 1972 TOTAL Journal of Speech Disorders Kybarnetic Mental Hygiene Midwest Journal Modern Language Journal Neurology Neuropsychologia New York State Education 2 93.59 2 93. 65 2 93.70 2 93. 75 2 93. 80 2 93.86 2 93. 91 2 93.96 The Packet 2 94.02 Proceed of the American Philosophical Society Personality Personnel Psycholinguistic Monographs Psychiatric Neurology Pedagogical Seminary and J r n l of Genetic Psych Sociology of Education 2 94. 07 2 94.12 2 94.18 2 94, 23 2 94. 28 2 94.33 2 94. 39 School Library Journal 2 94.44 TABLE VI (CONTINUED) 181 JOURNAL TIME PERIOD 1959 1964 1968 1972 TOTAL CUM. Social Work Teachers Forum 0 0 2 94.49 2 94.55 University of Michigan Schl and Educ B u l l e t i n US News and World Report 2 94.60 2 94. 65 Volta Review 2 94.70 Australian Journal of Education College Composition and Communication Education of Visually Handicapped I l l i n o i s School Research Kansas Studies In Education Programmed Learning Reading Science Education University of Kansas Bul l e t i n of Education American Anthropology American Annals of The Deaf 1 0 94.7 3 94.76 94. 78 94. 81 94. 84 94. 86 94. 89 94.92 94. 94 94,97 95.00 TABLE VI (CONTINUED) 182 JOURNAL TIME PERIOD CUM. % 19 59 1964 1968 197 2 TOTAL Advertising and S e l l i n g Amrcn Assoc of University Professors B u l l e t i n Acta Sociologica Adult Education Archives cf General Psychiatry American Journal of Diseases of Children American Journal of Human Genetics American Journal of Public Health American Journal of Physiological Medicine American Speech Annal of Otology, Rhinology, and larynology Arch. Psychol. Geneva Acta Psychiatrica Neurologica Supplementum Acta Psychiatry and Neurology Scandanavia Annee Psychologigue Assoc for Research i n Nerv and Ment Disorders American S t a t i s t i c s Association Journal 0 0 TABLE VI (CONTINUED) 183 JOURNAL TIME PERIOD 1959 1964 1968 1972 TOTAL CUM. % American School Board Journal Asha Administrative Science Quarterly At l a n t i c B u l l e t i n Academic De Medicine Baltimore B u l l e t i n of Education Business Education Forum Business Education World Berkeley Journal of Sociology B r i t i s h Journal of S t a t i s t i c a l Psychology B r i t i s h Medical Journal B u l l e t i n of The School of Education C a l i f o r n i a Education The Canadian Psychologist Catholic Educator Catholic Schools Journal Council For Basic Education B u l l e t i n 0 0 0 1 0 0 0 0 1 95.47 95.50 95.53 95.55 95. 58 95.60 95.63 95.66 95.68 95.71 95.74 95.76 95.79 95. 82 95.84 95.87 95.90 TABLE VI (CONTINUED) 184 JOURNAL TIME PERIOD CUM, 1959 1964 1968 1972 TOTAL Catholic Counsellor C a l i f o r n i a Journal of Elementary Education Canadian J r n l of Econo and P o l i t i c a l Science The Commonweal Canadian Medical Association Journal Community Mental Health Journal Computer Journal Cortex College Press Review Child Psychology C a l i f o r n i a Quarterly of Secondary Education C r i s i s C i v i l Bights Digest The Chronicle of Higher Education Comparative Studies In Society and History Columbia Univ Contrib to Philos and Psychol Credit World 0 0 0 1 0 0 1 0 TABLE VI (CONTINUED) 185 JOURNAL TIME PERIOD 1959 1964 1968 1972 TOTAL CUM, Of Daedalus Developmental Medicine and Child Neurology Durham Research Review Early Education Ebony EEG In C l i n i c a l Neurophysiology Educational Forum Educational Horizons Educational Method Educational Screen Elementary School Guidance and Counseling Elementary School Teacher English Teacher Education and Urban Society Experimental Education The F l o r i d a F l Reporter 0 0 0 0 .96.37 96. 40 96.43 96.45 96.48 96. 51 96.53 96.56 96.58 96.61 96.64 96.66 96.69 96.72 96.74 96.77 186 JOURNAL TIME PERIOD CUM, 1959 1964 1968 1972 TOTAL F o l i a Phoniatrica Florida. Reading Quarterly Guild Reporter Human Development Health Education Journal Human Forces Improving College and University Teaching I l l i n o i s Education International Journal for the Eductn of the Blind I n d u s t r i a l Medicine In d u s t r i a l Management Review International Review of Education ITA Foundation Report Journal of ACM Journal of The Acoustical Society of America Journal of The American Optometric Association Journal of Chronic D i s a b i l i t i e s 96. 80 96. 82 96.85 96.88 96. 90 96.93 96. 96 96.98 97.01 97.03 97.06 97.09 97. 11 97.14 97. 17 97. 19 97.22 TABLE VI (CONTINUED) 187 JOURNAL TIME PERIOD CUM, 1959 1964 1968 197 2 TOTAL Journal of Child Psychiatry Journal of Child Psychology and Psychiatry Junior College Journal Journal of C o n f l i c t Resolution Journal of Exceptional Children Journal of Educational Measurement Journal of Educational Sociology Journal of Farm Economics Journal of Gerontian Journal of Geography Journal of Health Physical Education and Journal of Individual Psychology Journal of Juvenile Research Journal of L i n g u i s t i c s Journal of Mathematics Journal of Marketing Journal of Motor Behavior 0 0 0 0 1 0 0 1 188 JOURNAL TIME PERIOD CUM, % 1959 1964 1968 197 2 TOTAL Journal of Neurophysiology J r n l of Neurological and Neurosurgical Psychiatry Journal of Psycholinguistic Research J r n l of Proj Techniques and Personality Assessmt Journal of Rehabilitation Journal of The Royal Society of The Arts Journal of Research In Science Teaching Journal of The Royal S t a t i s t i c a l Society Journal of Sociology 0 0 -0 0 0 Journal of Secondary Education Journal of Social Research Los Angeles Educational Research B u l l e t i n Land Economics Library Trends The Library Journal Language Learning TABLE VI (CONTINUED) 189 JOURNAL TIME PERIOD CUM, 1959 1964 1968 1972 TOTAL Multivariate Behavior Research Mind Minnesota Reading Quarterly National Association of Secondary Sch Pri n c i p a l s Negro Digest New Education National Education Journal Negro Educational Review Nervenarzt New Outlook The New Outlook For The Blind New Republic New Research In Education National Review Overview Oceanic L i n g u i s t i c s Occassional Papers 0 0 0 0 0 0 .0 0 0 0 0 1 0 0 0 0 0 0 98. 12 98. 15 98. 17 98.20 98. 23 98.25 98. 28 98.31 98,33 98.36 98. 38 98.41 98.44 98.46 98.49 98. 52 98.54 TABLE VI (CONTINUED) 190 JOURNAL TIME PERIOD CUM, 1959 1964 1968 1972 TOTAL Ohio State University Edctnl Research B u l l e t i n Progressive Education Personnel Journal Philosophic Studien Programmed Instruction Pittsburgh School Psychological Optics Philosophy of Science Psycholopharmacologia Pedagogical Seminary Psychoanalytic Studies of The Child Psychiatry Et Neurologie Psychologia Wychowawcza Psychiatry The Reporter Review of Economics and S t a t i s t i c s Reading Improvement 98.57 98. 60 98.62 98.65 98.68 98.70 98. 73 98. 76 98.78 98. 81 98. 84 98. 86 98. 89 98. 91 98,94 98.97 98.99 TABLE VI (CONTINUED) 191 JOURNAL TIME PERIOD CUM, 1959 1964 1968 1972 TOTAL Review Philosophigue Research Quarterly S c i e n t i f i c American School Management Search Sociologische Gids The Syracruse Journalist School L i f e Social Science Social Science and Medicine Sociology and Social Research Social Psychology and Psychiatry Spelling Progress B u l l e t i n Studies In Public Communication Special Education Speech Pathology and Therapy 99.02 99.05 99. 07 99. 10 99. 13 99.15 99. 18 99.21 99. 23 99. 26 99. 29 99.31 99. 34 99.36 99.39 99.42 TABLE VI (CONTINDED) 192 JOURNAL TIME PERIOD 19 59 1964 1968 197 2 TOTAL CUM, % School and Safety Sight Saving Review Speech Teacher Studies In Reading Supra Survey Synthese Teachers College Contrib to Education Todays Education Teaching Exceptional Children Teachers World University of C a l i f o r n i a Public Education Urban Education Unischool Vanderbilt Law Review wEA Journal Wisconsin English Journal 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 99.44 99. 47 99.50 99.52 99. 55 99. 58 99. 60 99.6 3 99.66 99.68 99.71 99.74 99. 76 99.79 99. 81 99. 84 99.87 TABLE VI (CONTINUED) 193 JOURNAL TIME PERIOD CUM, 1959 1964 1968 1972 TOTAL Wiener Klinische Woche.nschr.ift Western P o l i t i c a l Quarterly Yale Psychological Studies Z e i t s c h r i f t Fur Padagogike Psychologie Z e i t s c h r i f t Fur Klinische Medizin 99.89 99.92 9 9 . 9 5 9 9 . 9 7 100.00

Cite

Citation Scheme:

    

Usage Statistics

Country Views Downloads
Japan 7 0
France 1 0
United States 1 0
City Views Downloads
Tokyo 7 0
Unknown 1 0
Ashburn 1 0

{[{ mDataHeader[type] }]} {[{ month[type] }]} {[{ tData[type] }]}

Share

Share to:

Comment

Related Items