INFLUENCE OF SELF CONCEPTS, SIGNIFICANT OTHERS AND INTERACTIVE PATTERNS ON THE PUBLICATION ACTIVITY OF AGRICULTURAL SCIENTISTS Thesis for tho Degree of Ph. D. MICHIGAN STATE UNIVERSITY John S. Murray 1965- {THESIS This is to certify that the thesis entitled INFLUENCE OF SELF CONCEPTS, SIGNIFICANT OTHERS AND INTERACTIVE PATTERNS ON THE PUBLICATION ACTIVITY OF AGRICULTURAL SCIENTISTS presented by John S . Murray has been accepted towards fulfillment of the requirements for Eh.D. degree in Jamminat ion Date November 29. 1965 0-169 L [B P ". R Y M 1‘35“ State Univeigity "15 INFLUENCE OF SELF CONCEPTS, SIGNIFICANT OTHERS AND INTERACTIVE PATTERNS ON THE PUBLICATION ACTIVITY OF AGRICULTURAL SCIENTISTS \ ) B 9)” kafip John S. Murray AN ABSTRACT OF A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Communication 1965 ABSTRACT The individual scientist is the primary disseminator of the knowledge he generates. This study was an attempt to account for variation in certain publication activities among a particular set of researchers. The main criterion variables were publication rate and audience emphasis. Associations be- tween publication activity and evaluations received from both peers and administrators were also examined. The predictor variables, which were derived from sym- bolic interaction theory, were self-concepts, significant others and interactive patterns. Because of the exploratory nature of the study, the relevance of a large set of other variables was also investigated. Data were collected, through personal interviews and self-administered questionnaires, from 104 researchers at the Michigan State Agricultural EXperiment Station. The respond- ents were established researchers who held a PhD degree, obtained at least 50 per cent of their salary from the Experi- ment Station and who had held their position for at least three years. Self-esteem as a researcher was not associated with any other variable used in the study. The results from a self rating scale and an index of discrepancy between ideal John S. Murray researcher and self as a researcher were not related to each other. The lack of predictive power of self-esteem was attri- buted to the restricted effective range among this sample of established researchers. The audiences researchers claimed for their publications seemed to reflect their immediate social situation and inter- active patterns. Self conceptions and significant others were, by themselves, not adequate predictors of publication audience emphasis. Researchers who seldom talked to lay audiences or who claimed only peers as immediate others rarely wrote for lay audiences. Situational and interactive variables were also the determinants of publication rate. It appeared that multiple audiences are associated with multiple publication, which in turn makes for high overall rate. High peer evaluation was associated with orientation toward peers as significant and immediate others and with few talks to lay audiences. The exceptions to these associations were mainly among professors, who were more highly evaluated by peers than were non-professors. Although tenure was the major determinant of percentage salary increment, extent of peer orientation among immediate others was also relevant. Researchers of shorter tenure and exclusive peer orientation generally received higher salary increments. The apparent relevance of the immediate social situation and c evalt gene: nuni; the r ?atti publi John S. Murray and of interactional patterns to publication activity and evaluation was interpreted as indirect confirmation of the general theory. Implications of the study for research, for practice and for the theory were discussed. The implications emphasized the significance of face to face interaction to mediated com- munication. The role-taking capabilities and tendencies of the researcher source are determined by his interactional patterns and eXpressed in his selection of audiences for his publications. INFLUENCE OF SELF CONCEPTS, SIGNIFICANT OTHERS AND INTERACTIVE PATTERNS ON THE PUBLICATION ACTIVITY OF AGRICULTURAL SCIENTISTS By e? by] John ST Murray A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Communication 1965 ACKNOWLEDGMENTS Many people helped me with this study. I wish to thank them all, sincerely and publicly. Dr. Carl J. Couch was my dissertation advisor. He and the other members of my guidance committee: Dr. Hideya Kumata, Chairman; Dr. Mason E. Miller; and Dr. Verling C. Troldahl, provided much helpful advice and assistance. Dr. Lloyd M. Turk, as Director of the Michigan State Agricultural Experiment Station, provided financial support for a pretest, gave permission to interview the station researchers and furnished certain data directly. The Department of Communication at Michigan State University and the Canada Council supported me financially dur- ing the course of the study. A few of the many graduate students who helped in numerous ways were James P. Bebermeyer, John G. Elliott, Robert L. Crom, Albert Talbott and H. Stuart Hawkins. The good will of agricultural scientists at the Uni- versity of Connecticut and Michigan State University who gave their time and attention in interviews literally made the study possible. My wife Marg is to be eSpecially thanked, for not eXpressing discouragement when I did. ii Chap: II Chapter II TABLE OF CONTENTS INTRODUCTION................................... The Research Context........................... Science as Communication....................... Science and Agriculture........................ Role of Agricultural Researcher................ Communication Activities of Researchers........ Statement of Theory............................ Interaction-Role-Taking-Communication...... Role-Self-Othera-Evaluation................ Rationale...................................... Self Concepts and Publication.............. Significant Others and Publication......... Interaction and Publication................ Publication and Evaluation................. HYPOtheseSOOOOOO00.0.0000...OOOOOOOOOOOOOCOOOOO RESEARCH DESIGN................................ The Sample..................................... Operationalization of Variables................ Self-Concept............................... self-Bate“...O...00....OOOOOOOOOOOOOOOOOO. Significant OtherBQQOOO.OOOOOOOOOOOOOOOOOOO iii Page 10 14 18 19 22 26 27 29 31 32 38 38 39 40 42 37 Chapt aaoooo-d-r- D to no. . III C O O " C C O . I I I C C C O ‘ .‘QIOOQCOO‘ none-00.309 Chapter III Immediate 0thers........................... Significance of Immediate 0thers........... Publication Rate........................... Audience Emphasis.......................... Peer Evaluation............................ Administrative Evaluation.................. Interaction with Lay Audiences............. Data callectionOOOO....0.000000000000000000000. FImImS m DISCUSSIONOCOOOOOO.OOOOOOOOOOOOOOO Description Of the SmleOOOOOOOOOOOOOOO00.0... General Characteristics of the Respondentsoo000000000.00000000000000.0000. Description of the Sample in.Terms of the Theoretic variables.................... Self Esteem as a Researcher........... Self Concept as a Basic-Applied ReseaICherOOOOOO...0.0.0...0.0.0.0000. Significant Others.................... Immediate Others...................... Significance of Immediate Others...... Publication Rate...................... Audience Emphasis..................... Peer Evaluation....................... Administrative Evaluation............. Frequency of Interaction‘with my AudienceBOO00.000.00.00...0...0... Tests of Hypotheses................... iv Page 43 43 44 45 46 46 47 48 49 54 54 56 58 61 62 63 65 66 67 48 Chapt IV C O Q I I D ’ C C . . D 0 C . C O I O C O C O I C O O C D O C I OOIIOOOOOOOOOOOOOIOOOOI. COCOCOOOOOOOIOOOQIOIOI 7 .......OI....0...‘.I.II.I....Q-' Chapter Additioml FimimBOOOOOOOOOOOOOOOOOOOOOOOOO0.. Relationships Among the Theoretic variableBOOOOOOOOOOOOOOOOOOOIOOOOOOOOOOOOOO Relationships Between General Character- istics and Theoretic Variables............. IV DISCUSSION, SUMMARY AND CONCLUSIONS............ Discussion..................................... Audience Emphasis.......................... Publication Rate........................... Peer Evaluation............................ Administrative Evaluation.................. Summary........................................ Conclusions.................................... Audience Emphasis.......................... Publication Rate........................... Peer Evaluation............................ Administrative Evaluation.................. Implications for Measurement................... Self-Esteem Index.......................... Significance of Immediate Others........... Basic-Applied Orientation.................. Implications for Research...................... Implications for Practice...................... Implications for the Theory.................... BIBLIwRAPHYOOOOOOO...O...OOOOOOOOOOOOOOCOOOOOO Page 74 74 82 89 89 89 92 94 96 102 104 104 105 105 106 106 106 107 108 108 109 111 113 O C O O 9 o o C O i . o o a n o a O O o o O O a O I o O O I C O I O O I O O O I O 9 o O Q U... Chapt Chapter APPENDICES Appendix A: Appendix B: Appendix C: Appendix D: Appendix E: Appendix F: Appendix G: Interview Schedule............... Self-Administered Questionnaire.. Letter of Legitimation........... Correlation Mattrix for Self Eateem Indexaeooeeooeooeooao Relationships Among the General CharaCterist1C8000000000000...... Relationships Among the Theoretic Variables.............. Relationships Between General Characteristics and Theoretic vari‘b1e80000.0.0.0000...0......O vi Page 10 14 15 16 17 18 ooooooooooooooo Table IIIIIIIIIIIIIIIIII OOOOOOOOOOOOOOOOOOOOOOOO 10. 11 12 13 14 1: 10. 11. 12. 13. 14. 15. LIST OF TABLES Response rate.................................. Summary of general characteristics............. Categorization of administrative units......... Response distributions on self-esteem measures. Distribution of respondents by scores on the self-esteem masuresOOOOOOOOOGCOOO0.00.0.0.0000 Distribution of respondents on basic and on applied orientation measures................... Correlations among basic and applied items..... Percentage of response types to significant Other question...0.00.00.00.000000000GOOOOOOOOO Percentage of response types to immediate Other questionOOOOO0......OOOOOOOOOOOOOOOOOOOOO Distribution of respondents according to the number of immediate others who were also significantOCOO000.00.00.00...OOOOOOOOOOOOOOOOO Distribution of respondents by publication rate.‘0.0..OOOOOOOOOOOOOOOOOOOOOO0.00.00.00.00. Percentage of researchers receiving various numbers of nominations for outstanding researCh...OOOOOOOOOOOOOOOOOOOOOOO0.0.0.0000... Distribution of respondents by salary increment intervals.0.00.....00...OOOOOOOOOOOOOOOOOOOOOO. Orientation to applied research and researcher audience emphasis.OOOOOOOOOOOOOOOOOOOOOO0...... Significant other orientation and researcher audieme mpmsisOOOOOOGOOOOOOOOOOOOOOOOOO0.0.0 vii Page 48 50 53 54 55 56 57 59 61 62 63 65 66 68 69 o 9 u o I O a C O C O ' C O I I O I O I m I O O O C O I O O O O i I O V O D O I ‘ O ‘ I O O C C O I O O O O O O . O C O O C C I C I O O C I O C O I C C Q 9 I V l a a o o a a U . U 0 9 Q 0 Q O O Q 0 O O O I i O I C O O O I O O O O I O i O I C O I C O c O O I l a O m 0 C I O D O o I O I I O C u o o O O 6 Tabl 16. ‘18. 19. 20 Table 16. 17. l8. 19. 20. 21. 22. 23. 24. 26. 27. 28. 29. 30. Immediate other orientation and researcher audience empha8180000000000000000OOOOOOOOOOOOOO Significant other orientation and publication rateBOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO Significance of immediate others and publi- cation rate...’OOOOOOOOCOOOOOOOOOOCOO...0...... Correlations between researcher audience emphasis and evaluation by publication rate categorieBOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO.0... Significance of immediate others with signifi- cant others and with immediate others.......... Frequency of talks to lay audiences with applied research orientation and with immediate Others.OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO Significant other orientationnwith peer evalution.‘00.000000000000000.0000000000000000 Immediate other orientation with publication rate; peer evaluation and administrative evalutionOOOOOOOOOOOOOOOOOOOOOOOOOOOOO00...... Frequency of talks to lay audiences with audience emphasis and publication rate......... Subject matter area and percent salary for research with audience emphasis................ Tenure and administrative evaluation........... Academic rank with peer evaluation............. Subject matter area and applied orientation, frequency of talks to lay audiences and immediate other orientation*with audience ”phasisOOOOOOOOOOCCOOOOOOOOOOOOOOO0.0.0.000... Immediate other orientation and frequency of talks to lay audiences with publication rate... Academic rank and publication rate, significant other orientation, and immediate other orien- tation with peer evaluation.................... viii Page 69 70 71 72 76 78 79 80 81 86 87 91 92 95 a a o o o o a a I O I a o o a u D G 0 O a a I a o a a a O O D I a a C I O c O u a I D I o C I O O I o 0 C O c a I C o o I 0 O o a a m m o O D O O O O I l l Table 31. 32. Table Page 31. Tenure and immediate other orientation with administrative evaluation...................... 98 32. Academic rank and subject matter area with administrative evaluation...................... 100 oooooooooooooooooooooo oooooooooooooooooooooo Figs: Figure 1. LIST OF FIGURES Summary of relationships among the theoretic variables.00....0..0.0.00.0...0.00.00.00.00..0. Summary of relationships between general characteristics and theoretic variables........ Page 75 83 ..................................... OOOOOOOO CE} 19¢ Chapter I Introduction "Dissemination of research findings to both scientific and non-scientific audiences is of increasing concern for research and education administrators, scientists and practitioners. The scientist himself is the key figure in the discovery of new knowledge. But scientists vary in the extent and effectiveness with which they disseminate research results to different audiences or work with intermediaries to do so. "Very little research has been carried on as to what encourages scientists to take initiative in the communication of the results of their re- search, what influences the nature of their communications efforts, or what highly communi- cative scientists are like compared with less comunica tive ones . "Social science findings in other areas suggest that such factors as the perception.a person has of himself as a communicator, the perceptions others have of him in this role, the personality of the communicator, the social system in'which the person communicates, and the kinds of be- havior that are rewarded in that system are correlated with communicative behavior. There is a definite need to apply the findings of such research to a study of scientists and theirgpart in communication of research finding ." The foregoing is excerpted from the minutes of the North Central Regional Mess Communications Committee, Chicago, May, 1963. This committee is not alone in recognizing the problem of and stu ape cat vel 0i ind ing pro stu cul PS] 88. th of scientific communication.1 Several scientific disciplines, e.g., Chemistry, Biology and Psychology, have undertaken large scale communication studies.2 The National Science Foundation has established a special section to promote and support research on the communi- cation of scientific information. Herbert Menzel recently de- veloped a proposal to integrate research findings in the field of scientific and technical communication. Such activity indicates an increasing awareness of the problem and a will- ingness to deal with it on the part of social scientists. The study proposed here is a further attempt to deal with the problem of communication in science. The scientists to be studied are researchers employed by the Michigan State Agri- cultural Experiment Station. The theoretic perspective of the study will be social- psychological, in the tradition of George HerbertiMead.3 It is accepted as given, that scientists engaged in re- search generate scientific information, and that they publish this information for different audiences. Frequency of l Floydiuann and Rensis Likert, "The Need for Research on the Communication of Research Results," Human Organization, 1952, ll, 4, 15-19. . 2 Herbert Menzel, Review of Studies in the Flow of Infor- mation.Among Scientists, Bureau of Applied Social Research, Columbia university, 1960, 60-62. 3Manford H. Kuhn, "Major Trends in Symbolic Interaction Theory in the Past Twentyefive Years," The Sociological pub pub sci: ing: CUT] nat' tin Vir publication.and diversity among the apparent audiences of the published reports will be key concerns in this study. The study is an attempt to account for variability in these publication behaviors in terms of the researcher's con- ceptions of self and others and their interactive patterns. The relationship between source behaviors of researchers and evaluations received will also be investigated. The Research Context Research in the area of science and communication of scientific information.may be classified under several head- ings. The following six categories suggest major emphases of current and past research on this topic. 1. The earliest studies include those dealing with the nature of, and conditions surrounding, scientific creativity. This search for understanding of the creative process con- tinues to receive attention from.many eminent scholars. 2. General studies of scientists, their work and en- vironment, have been treated by Caplow and McGee, Lewis M. Terman, "The Discovery and Encouragement of Exceptional Talent," American Psycholggy, June 1954, 221-230; J. P. Guilford, "Creativity," American Psycholggy, 1950, 2, 444-454; and Homer G. Barnett, Innovation: The Basis of Cultural Change, New York, N.Y.: McGraw Hill, 1953. 5 Theodore Caplow and Reece J} McGee, The Academic ‘Marketplace, New York, N.Y.: Basic Books, Inc., 1958. 1 ma1 I a: bee try bee USE exi SIT SCI Re: 10. PII S’IS’ [Elf—5’ If.” 7 These works have documented the B. Glaser6 and'W. Kornhauser. nature of the scientist role and have identified some of the major types of role conflict and resulting problems. 3. Within specific disciplines intensive studies have 8 been carried on.and are continuing, e.g., psychology, chemis- 10 Much of this work has try,9 and biology and biochemistry. been concerned with the information needs of scientists, their use of information sources, the adequacy and function of the existing information channels and problems of indexing, ab- stracting and translating. 4. The scientific process in specific organizational settings constitutes still another area of research. The 6Barney G. Glaser, "variations in the Importance of Recognition for Scientists' Careers," Social Problems, 1963, 19, 3, Winter, 268-276; also, Organizational~Scientists: Their Professionalfigareers, Indianapolis, Ind.: Bobbs-Merrill, 1964. 7William Rornhauser, Scientists in Industry: Conflict and Accommodation, Berkeley, Calif.: university of California Press, 1962. 8American Psychological Association, Reports of the American Psychological Association's Project on Scientific Information.Exchange in Psycholggy, Vol. 1, December 1963. 9Anselmi. Strauss and Lee Rainwater, The Professional Scientist, Chicago, 111.: Aldine Publishing Co., 1962. 10Herbert Menzel, "Flow of Information on Current De- velopments in Three Scientific Disciplines," Columbia Uni- versity, Bureau of Applied Social Research, Reprint No. 232; also, "The Infonmation Needs of Current Scientific Research," Columbia university, Bureau of Applied Social Research, . Reprint No. 333, 1961; Biological Sciences of Communication Project; George Hashington university, Suite 700, 2000 PM, N.W., washington 6, D.C., 1963 and 1964. of Shepard11 and Pelz12 and Ben-David13 are typical. The main concern in these studies has been the specification of rela- tionships between (1) the organization and administration of science and (2) "scientific productivity" as measured by volume of publication or reputation. 5. Information theory, computers and systems analysis also provide a base for an approach to problems of scientific infonmation. Brownson.14 and the special issue of The American Behavioral Scientist15 describe this area quite adequately. 6. The literature concerning the diffusion of inno- vations includes considerations of the dissemination of 11Herbert A. Shepard, "Nine Dilemas in Industrial Re- search," Administrative Science Quarterly, 1956, 1, 295-309; "Patterns of Organization for Applied Research and Development," Journal of Business, 1956, 22, 53-58; "Superiors and Subordi- nates in Research," Journal of Business, 1956, 22, 261-267. 12Donald C. Pelz, "Some Social Factors Related to Per- formance in.a Research Organization," Administration Science Qggrterly, 1956, 1, 310-325; "Interaction.and Attitudes be- tween Scientists and the Auxilliary Staff: I. Viewpoint of Staff," Administrative Science Qggrterly,1959, 4, 3, December, 321-366; "Interaction and Attitudes between Scientists and the Auxilliary Staff: II. Viewpoint of Scientists," Administrative Science Qggrterly, 1960, 4, 4, March, 410-425.. 13Joseph Ben-David, "Scientific Productivity and Aca- demic Organization in Nineteenth Century Medicine," American Sociological Review, 1960, 2g, 6, December. . 14HelenL. Brownson, "Research on.Randling Scientific Information," Science, 1960, 132, 3440, 30, December, 1922- 1931. 15The American BehaviorQILScientists, special issue on Information Retrieval in the Social Sciences: Problems, Programs, and Proposals, edited by Ted Gurr and Hans Panofsky, 1964, 21;, 10, June. . sci sum tre DOD bei 131 pos the Sid‘ tiS' org; beh Shi was pos liu scientific infonmation to non-scientists. These works have been summarized by Rogers16 and Katz, Levin.and Hamiltonl7 but seldom treat the scientist as the source of scientific information for non-scientists. This sample review illustrates the scope of research being conducted in this field and the differing emphases. It is within this research context that the present study is pro- posed. The proposed study is distinctive in that it focuses on the researcher as the source of scientific information, con- siders both scientific and lay audiences, and includes scien- tists from a variety of disciplines who work within a common organizational framework. Science as Communication Symbolic communication enables man to coordinate his behavior with others and thus establish stable relation- ships -- social systems. However, until written communication was achieved, a social institution such as science was not possible. The development of writing expanded the traditional limits of time, space, fidelity and quantity associated with 16 Everett M; Rogers, Diffusion of Innovations, New York, N.Y.: Free Press, 1962. 17Elihu Katz, H. L. Levin and K. Hamilton, "Traditions of Research on the Diffusion of Innovations," American Sociological Review, 1963, 2g, 2, April. V8} an. ti re: of ge fic by verbal communication. In its early stages, while the total amount of knowledge was relatively small, science was rela- tively unified. Learned men could, and some did, address the reports of their observations and reflections to the community of learned men. These early scientists, of an age and area, generally shared a common language. This language was suf- ficient for them to share meaning on a full range of topics. The learned societies of the seventeenth century were formed by menwwith varied interests but with a common language.18 As the sum of human knowledge increased, specialization occurred. With the evolution of disciplines and subdisci- plines, refined languages developed which were unique to their own subject matter.19 While such languages (terminologies) promote clarity and specificity within a discipline, they magnify the difficulties of communicating scientific infor- mation across disciplinary boundaries and to the larger society beyond the scientific community. The communication of scien- tific information to scientists and to the lay public appears essential for continued progress in science and for the betterment of society at large. Problems of communication in science have been accentuated by the current expansion of 18 Warren'Weaver, "Science and the Citizen," Bull. Atomic is!» 1959, 1_3_, 10, December, 361-365. . 9Charles S. Slichter, Science in.a Tavern, Madison, Wis.: university of Wisconsin Press, 1958; and Joseph K. Senior, "The Vernacular of the Laboratory," Phil. Sci., 1958, 25, 3, July, 163-168. . tiq ogJ laul Pli 88 otl CO! or} is CO' ti IE 30 fc ti science, the attempt to accelerate development in the emerging nations, and the attempt to utilize new information storage- retrieval systems. Florian Z’naniecki20 has discussed the role of the scien- tist as (1) an explorer, (2) a theoretician, and (3) a technol- ogist. While the Znaniecki formulation may have utility, its lack of emphasis on communication activities makes it incom- plete. In the first place, the work of a scientist, whether as explorer, theoretician or technologist does not occur in a social vacuum. The scientist must have interacted with others to become a scientist. The conduct of his inquiry is continually influenced by the information he receives from others. Secondly, unless the information a scientist produces is shared with others, there is no science. A "private science" could not be distinguished from any other covert experience. The communication of research findings is a necessary condi- tion for the existence and continuance of science. The cur- rent public support of scientific activity carries with it some obligation for scientists to make their findings available for practical application. From this perspective, communica- tion is an essential, central process in science. Science and Agriculture The land-grant colleges of agriculture were the first 20 Florian Znaniecki, The Social Role of the man of Knowl- edge, New YOrk, N.Y.: Columbia university Press, 1940, p. 14. 1a V8. an' no Sme -f- _—.._-s—— — — T mo: St] the de\ ag: ch. 8t 0t SE E! IN large scale attempt to publicly support science. The intent was to deliberately foster practical technological innovations and to diffuse these innovations to the primary producers, (who formed a large part of the population). This endeavor was notably successful.21 It has resulted in the proportionately smallest agricultural labor force and the least expensive and most abundant food supply for a large population in all history. The science of the colleges of agriculture has not been re- stricted to routine testing and problem-solving.22 Throughout their existence, colleges of agriculture have fostered basic, developmental and applied research in a variety of traditional and newer disciplines. Despite the success of the colleges of agriculture in making and instigating technological and social change, they are themselves now caught up in this change. These colleges are faced with severe competition for students, personnel, monies and clients from a multitude of other agencies. The dual emphasis on basic and applied re- search and the conflicting demands of scientific and lay audi- ences are increasingly troublesome. As diversity increases in both agricultural science and in client demands, the limita- tions of current communication systems grow more apparent. On 21 H. C. Knoblauch, E. M. Law and W. P. Meyer, "State Agri- cultural Experiment Stations. A History of Research.Policy and Procedure," united States Department of Agriculture, Miscel- laneous Publication, No. 904, May, 1962. 2 H. C. Knoblauch, "Basic Research at State Stations," Science, December 11, 1959, 130, 3389, 1639-1640. . EX‘ E181 st: 1m 8“ 9t a: Hc PE a; 10 the one hand, a force of extension specialists and an infor- mation service unit have been established to synthesize and facilitate the flow of scientific information to the public. On the other hand, the office of experiment station editor and interdisciplinary research projects combat, to some extent, the impediments to communication of scientific infor- mation among scientists. It is within this general context that the experiment station scientist conducts his research and enacts his role as information source. Role of Aggicultural Researcher The most striking feature of the role of agricultural experiment station researcher is its breadth and diversity. The scientists who enact this role represent a wide range of subject matters, conduct diverse types of research with a vari- ety of methodologies, and work under many different social arrangements. The disciplines represented in the experiment station include economics, engineering, animal medicine, soil physics, food chemistry, statistics, microbiology, plant pathology and floriculture to mention.a few. JHost of these subjects have obvious relevance to agricultural or at least rural problems. However, the significance of knowledge generated through ex- periment station research is seldom confined to the agricultural application. Furthermore, the theoretic, methodological and em tc in en fr 8C fr he te: in ba do V8 11 [0) ll empirical findings of the "mother disciplines"'must be attended to by experiment station researchers. Research conducted with- in the experiment station encompasses most of the physical sci- ences and parts of the social sciences. The research of experiment station scientists ranges from the strictly controlled chemical experiment to the de- scriptive-historical area resource study. Instruments vary from the spectrograph and.microscope to the mailed question- naire and personal interview. Within.most of the disciplines both highly theoretic (basic) research and strictly limited testing (applied) are carried out. The same individual may be involved simultaneously in several projects representing the basic and applied aspects of a question. Other individuals do research which is only basic or only applied. Considerable variation exists in the activities regarded as research with- in the experiment station. A final area of diversity lies in the immediate social setting of the researcher's work. Some individuals work alone, they cOOperate with no one in the research enterprise. Their research may fit into an overall department program and uti- lize graduate students but it is exclusively theirs from be- ginning to end. Other station researchers do most of their 23 work as members of teams or project groups. The extent of 23Gordon‘W. Blackwell, "Multidisciplinary Team Research," Social Forces, 1955, 33, 4, May, 367-374. , th an al pl of po W8 du SQ. fr SC 12 the autonomy of the individual in choice of project, methods, and coworkers varies by department and over time. Researchers also differ concerning the centrality of their work to a disci- pline. Some work clearly within a single discipline. The work of others is multidisciplinary, even to the extent of dual ap- pointments. Some disciplines are very new, others large and well established. Some researchers have formal non-research duties in administration, teaching or extension. Some re- search is "on contract" or "funded." Such funding may result from the researcher's own initiative or from the efforts of some committee or administration. The individual researcher has much leeway to make his role in the experiment station. iMany options and opportuni- ties are present which enable the scientist to do the kind of research he wants to or is able to within.the experiment station. Should he tire of research or fail at it, most station scientists can move, gradually and with ease, to teaching, extension or administration,‘within the college of agriculture. Many station researchers also have freedom to ‘mowe to other colleges on or off campus or to industry, while retaining or discarding the researcher role. In view of the original self selection, the role flexibility, and the con- tinued potential mobility, the agricultural experiment station researchers should be a fairly happy lot. Lest the foregoing give the impression of anarchy or SC S EN- 81 (4 e) 13 social disorganization in the experiment station, some con- sensual expectations will now be stated. The researcher is expected (1) to have proper credentials - academic degrees, (2) to search for knowledge in.a specified area, (3) to employ scientifically justifiable procedures, and (4) to communicate his findings to others - i.e., to publish. In return for such performance, the researcher may expect from the organization, (1) a title, (2) salary, (3) resources with which to conduct research, and (4) a certain amount of freedom in inquiry and disposition of findings. This role definition is minimal and does not distinguish the agricultural researcher role from that of most other re- searchers.2 However, any more stringent definition seems invariably to exclude individuals who now occupy research positions in the agricultural experiment station. The present- day agricultural researcher is a scientist who, by choice or accident, happens to be working for an agricultural agency. The organization (college of agriculture) is somewhat dis- tinctive in that research, teaching and extension activities are carried out and to some extent integrated. However, the individual researcher is largely free to participate or not in teaching or extension. This scientist has a variety of others‘with whom he may interact and an assortment of audiences 24Gerard DeGre, Science as a Social Institution, new YOrk, N.Y.: Random House, 1955, pp. 21-30. 14 to whom he may direct his written reports. Communication.Activities of Researchers A report from the Case Institute of Technology states: "The data collected on a large representative sample of chemists indicated that the average chemist working in industry seems to spend more time in scientific communication (reading, writing, listening and talking about scientific matters) than in all the rest of h activities concerned directly with science." Of 32.4 hours of scientific activity per week, the chemists spent 16.5 hours in scientific communication, 10.4 hours work- ing with equipment, 3 hours in data treatment and 2.5 hours thinking and planning. The outcome of all this scientific activity, communi- cative and otherwise, is the published report. The written account is the "permanent" public evidence of science. The report is a visible expression of many decisions made by the scientistdwriter. The scientists' choices of topic, language, form and apparent audience may be inferred from his reports. The publication activities of researchers will be em- phasized in this enquiry. Such emphasis does not deny the importance of other communicative activities but regards them 25 Case Institute of Technology, Operations Research Group. An Operations Research study of the scientific activity of chemists. Cleveland, Ohio, November, 1958. as lllllll To mi 0. w. 15 as antecedents or consequences of publication.2 The type of research performed seems clearly related to the type of report published. Scientists who do basic research are likely to use highly technical language in writing journal articles for an audience of other basic researchers. Applied researchers seem more likely to direct non-technical reports to audiences without formal scientific education. The basic- applied distinction.has been discussed by Barney G. Glaser27 and seriously investigated by Norman W. Storer28 and Gerald Gordon.29 Storer used a "three-item Guttman-type scale" to meas- ure "relative interest in doing research which would be im- portant to other scientists or to nonscientists." The former group were regarded as "basic scientists," the latter as "applied scientists." While it is convenient to dichotomize research into basic-applied, some scientists deny the relevance of the 26Gideon Sjoberg, "Science and Changing Publication Patterns," Phil. Sci., 1956, 23, 2, April, 90-96. 27Barney G. Glaser, "The Local-Cosmopolitan Scientist,” Anerican Journal of Sociolm, 1963, LXIX, 3, November, 249-259. 28 Norman W. Storer, "Research Orientations and.Attitudes Toward Teamwork," I.R.E. Transactions of the Professional Groups on Engineering Management, 1962, E.M-9, 1, March. 29 Gerald Gordon, "The Organization Designed to Produce Change." Anworking paper presented at the Seminar on Inno- vative Organization, university of Chicago, April, 1964. 16 30 distinction. Such scientists may label their research "de- velopmental," meaning it has theoretic significance and ap- parent practical implications for some group of nonscientists. Still other scientists conduct both basic and applied research, or maintain that while their research is essentially basic or applied, their findings have relevance to both scientific and nnnscientific audiences. It appears that the relationship be- tween type of research and apparent audiences of reports re- quires further qualification. Publication rate is another commonly considered vari- able. Donald C. Pelz has reported a relationship between scientific performance and the scientist's interactive con- tacts. He found higher performance associated‘with: Frequent (daily) contact with several scien- tific colleagues who, on the average, have been employed in settings different from one's own, who stress values different from one's own, and who tend to work in scientific fields different from.one's own; and at the same time, frequent contact with at least one important 31 colleague who has similar professional values. These findings indicate the relevance of interpersonal com- munication to scientific activities as indexed by publication 30 L. C. Goldberg, F. Baker, and A. H. Ruberstein, "Local- Cosmopolitan; unidimensional or'Multidimensional," American Journal of Sociology, 1965, LXI, 6, May, 704-710.. 1 Donald C. Pelz, "Some Social Factors Related to Per- formance in.a Research Organization," Admin. Science Quart., 17 frequency. Heterogeneity among others regularly interacted ‘with was associated with higher publication rates as long as at least one of these others shared similar values. Pelz implied that all publications considered were directed to scientific audiences but made no direct statement to this effect. 32 In his M.A. thesis, Hawkins reports considerable vari- ability amnng agricultural scientists in publication rates of professional and lay reports. Forty-five percent of his re- spondents produced fewer than two lay reports in.a three year period while only 24% produced fewer than two reports for other scientists. Conversely, while 532 of his respondents authored three-quarters of the professional reports, some 36% produced three-quarters of the lay publications. The mean number of lay publications per respondent was 3.4, and of professional reports 4.1. Approximately 202 of the sample had published 8 or more professional or lay publications during the sample period. Hawkins also related publication type and rate to commitment indexes and to diversity of face-to-face channel usage. He found significant positive correlations between (a) commitment to the immediate organization and rate of lay 32Hawkins, H. Stuart, Influence of Role Conmitments, Perceived Role-Conflict, and Gregariousness on the Publication Behavior of Agricultural Scientists, Michigan State University, M.A. Thesis, 1964. 18 publication, (b) commitment to the profession and rate of pro- fessional publication, and (c) a significant negative corre- lation between rate of lay publication and diversity of face- to-face channel usage. The relationship between rate of pro- fessional publication and diversity of face-to-face channel usage was not significant. This latter finding is in apparent conflict with Pelz findings but different'measurement tech- niques were used and Pelz reported no statistical tests. The studies cited indicate some of the ways scientists differ in publication activity. They also suggest some corre- lates of this variable performance. Statement of Theory Communication is a central concern in the social psy- chology of George Herbert Miead.33 In this orientation the individual human - a self - is regarded as a process of inter- action in which the individual is both subject (I) and object (me). The individual comes to be an object to himself in social interaction via the significant (social) gesture. A language of significant symbols enables the individual to take the attitude (role) of another toward himself. He can thus 33George H. Mead,iMindI Self and Society, Chicago, 111.: University of Chicago Press, 1934; and Anselm Strauss, Th5. Social Psychology_of Georgg_Herbert Mead, Chicago, 111.: University of Chicago Press, 1962. . 19 regard or view himself from the standpoint or perspective of the other, i.e., as an object. But the "others" with whom an individual interacts (communicates) have somewhat different perspectives. By generalizing these various perspectives of others toward himself, the individual acquires a self. The self is maintained and modified through continued interaction. Just as communication is the basic process in the for- mation.and maintenance of the human individual, so it is the fundamental activity in the development of human social systems. The ability to interact symbolically with others is a pre- requisite to coordinated activity, whether it be "just a game" or science or politics. This facility for interaction allows for the continued modification of planned outcomes while the activity is being performed. Through communication, humans can take each other into account reciprocally. The outcomes of such interaction are jointly determined. Stable social relationships are evolved in the course of communicating, by the sharing of expectations of behaviors appropriate to the participants in an activity. Thus social systems are seen to be formed, maintained and modified by communicative interaction among people. Most human activity occurs within the context of established role relationships. Interaction - Role-taking - Communication People affect others and are affected by them through 20 the significance or meaning they assign to the symbols they use. The effective meaning of symbols lies not in their physical configuration but in the significance which users mutually share toward them. Tb communicate effectively, the individual must accurately predict the meaning others will assign a set of symbols. He must share the receiver's meaning for the symbols. The concept of role-taking refers to the process in which an individual takes a standpoint other than his own toward an object or event.34 By role-taking the individual predicts the meanings of symbols to other senders or receivers. In ongoing face-to-face communication, participants frequently assume "same"’meaning. Discrepancies are overcome as inter- action continues. Frequent intense interaction among people enables them to develop consensual significance for a large set of symbols. Where prior interaction is extremely dis- similar for the parties, communication is restricted to an exchange of gestures such as the first meeting of Robinson Crusoe and Friday. Only with continued interaction do they acquire the ability to take the role or attitude of the other toward symbols. With shared meaning for a set of symbols people are able to communicate effectively. 34Ralph M; Turner, "Role-taking: Process versus Con- formity," Chapter 2 in Rose: Human Behavior and Social Processes, Boston: Houghton Mifflin, 1962. 21 A role-taking perspective suggests that readingdwriting activity is not a simple extension of vocal face-to-face inter- action. A lack of, or delay in, feedback with mediated com- munication makes the difference. The selection and ordering of vocal gestures is a rapid flowing process whereas the choice and arrangement of written symbols is usually a slow process ‘with halts, retreats, revisions and not infrequently destruc- tion.and a new start. In normal vocal conversation, the role of speaker and listener are not separate; to converse is to perform bodh activities simultaneously. To read is analogous to the enactment of a listener-role in a formal audience sense. The reader (or listener) attends to the stimuli provided; he accepts a passive but willing role - that of respondent. So far as he shares significance for the symbols, the reader's response is determined by the writer's selection and presen- tation of symbols. Reading requires that the role of the writer be adopted but this need not be a "writer-role." To write is a different process. The writer must assume the roles of reader-writer. He may assume his own standpoint and write to himself as in.a private diary or he may role-take with some potential reader(s) and deliberately try to elicit certain responses. While spoken words can not be unsaid, the written word is subject to revision or sup- pression until deliberately transmitted to an audience. Thus interaction is a prerequisite to role-taking which 22 in turn is prerequisite to mediated communication. Mediated comunication follows from unnediated, but is not equivalent to it from the standpoint of interaction or role-taking. Interaction, role-taking and coulnunication are activi- ties which occur within established social systems. These activities, and their interrelationships are crucial to the maintenance and modification of social systems. The distinc- tions between these activities and the relationships among them are more clear when written comunication is emphasized. 331; - Self - Others - Evaluation 5315 is the set of expected behaviors proper to an occupant of a particular position in a social system. The expectations are held and expressed by others enacting recipro- cal roles to the role in question. Merton35 refers to the role set as "that complement of role-relationships in which persons are involved by virtue of occupying a particular social status." Herein, the term role set will be used to refer to the M 9.5.1122. who "legitimately" hold expectations toward the occu- pant of a specific status or position. A role can only be defined in terms of other roles. Thus, role X consists of those behaviors the actor is consensually regarded as owing to actors M, N and O and the behaviors which M, N and O owe in turn to X. 35Robert K. Merton, "The Role-Set: Problems in Socio- logical Theory, " British Journal of Sociology, 1957, 8, 110. 23 The expectations which constitute a role are duties the actor owes to others and the obligations of these others toward him. Typically, the expectations encompass a limited set of be- haviors and situations and a specified number of others. Roles are seldom static; rather, they are constantly modified in the course of continued interaction. Any individual has many roles. Sarbin?6 has even sug- gested that the self consists of the repertoire of roles which the individual has played or could play. From this standpoint, the self conception is the organized (but interactive) cumula- tion of roles the individual sees himself as holding. The stability of this organized pattern of roles varies, as does the salience of a particular role across time and situations. An individual's conception of self in.a given role is an im- portant determinant of his performance or enactment of the role. Reference group concepts have frequently been used to deal with the relationships between an actor and others.37 The employment of the term "group" for such a purpose not only vio- lates its traditional usage but raises troublesome distinctions, 36T. Sarbin, "Role Theory," in Gardner Lindzey: Handbook of Social Psychology, Vol.1, Reading, Mass.: Addison4Wesley, 1954, pp. 223- 258. 37 Manford H. Kuhn, "The Reference Group Reconsidered," The Sociolggical Qggrterly, 1964, 5, 1, 5-22. 24 e.8., reference individual vs. reference group, membership vs. non-membership groups. Turner38 clarified the issue by pro- posing the concept of audience to refer to the actor's percep- tion of others observing his role performance. Such audiences may include persons or groups involved in, or beyond a particu- lar set of role relationships. Three types of others can be identified on the basis of their relationship to the actor. The types are not exclusive in that a specific other may be related to an individual in several or all of these ways. Those involved in reciprocal relationships with a par- ticular role will be termed Relevant Others. Relevant others play roles in the role set. The Siggificant Others of a role player are those regarded as important by the actor, whose expectations are taken into account in the enactment of the role. Some of the significant others of an actor must also be relevant others, but perfect correspondence is unnecessary and unlikely. Immediate Others are persons who are interacted with frequently in the course of performing a role. These are rele- vant necessarily. If relevant others are represented by R, immediate others by I, and significant others by S, the relationships can be ex- plicated as follows. All I must be R. Some S must be R. The 38Ralph E. Turner, "Role Taking, Role Standpoint, and Reference Groups' Behavior," American Journal of Sociology, 1956, _6_]_., 316-328. . """"‘“ 25 other relationships vary according to the type of role and the individual orientation of each actor. Roles differ in the extent to which the expectations of relevant others must be taken into account. Some roles are quite flexible in that the actor has much choice in selecting others to attend to and in establishing the nature and amount of significance to be assigned to each. Rigid roles, on the other hand, are clearly prescribed and deviation beyond the es- tablished norms is little tolerated by the powerful relevant others. However, an actor is seldom obliged to assign signifi- cance to all the so-called relevant others; they are not "equally" relevant. Generally, certain minimal combinations in the assignment of significance (expressed implicitly in role performance) will suffice for an.actor to retain.a role. Couch and.Murray investigated associations between the selection of significant others and administrative evaluation for performers of three different roles. They hypothesized that when surveillance is high and role prescriptions specific, no associations between "significant others" and evaluation can be expected. The acquisition of such roles involves a relatively high commitment to specific behavioral patterns. Subsequently, the performance of these fixed sets of acts provides the criteria for evalu- ation. On the other hand, when surveillance is low and the role expectations flexible, associ- ations between "significant others" and evalu- ation can be anticipated. In these situations, 26 variation in "significant others" affects the way the role is performed and the consequent evaluations. The actor is obliged to define relevant others as the prime audience for his performance, (i.e. - as significant), only ‘where the role is clearly defined and role performance closely monitored by the evaluators. Rationale In this section, the communication situation‘will be re- examined in light of the theory. The synthesis'will provide a basis for a subsequent statement of expected relationships. The social system‘which is the agricultural experiment station can be viewed as a set of interrelated roles. ‘Within this large set are many subsets of related or reciprocal roles. Some of these subsets are project teams, others are based in common methodologies or common types of research, e.g., basic- applied. The roles enacted by experiment station researchers may also be grouped by departments in.the college of agricul- ture. Here the research roles are classed with teaching, ex- tension and administrative roles by commonness of subject matter. The researchers may also relate themselves to scien~ tists elsewhere on campus, on other campuses, with other agen- cies, or to nonscientists beyond the academic community. 39 Carl J. Couch and John S. Murray, "Significant Others and Evaluation," Sociometry, 1964, 21, 4, December, 502-509. 27 These subsets overlap in.many instances. The boundaries of the role-sets are blurred and constantly changing. There is, however, considerable stability to these relationships. They are at least sufficiently stable that the overall system is maintained and that research does get done. This is the context in which the individual station- scientists enact their researcher-role. Through ongoing inter- action, these people plan, finance, and execute an organized search for knowledge. The immediate visible outcome of this activity is the published report. It is the report rather than the day-to-day activities of the researcher which is subject to surveillance by others. Information derived from published reports provides much of the basis for evaluation of role per- formance. While reports of scientific findings have potential sig- nificance for the scientific community and society as a whole, this will be dealt with only implicitly in the proposed study. Rather, an attempt will be made to account for variation in the published material of researchers in terms of their self conceptions and their interactive relationships with others. variation in evaluation‘will be regarded as a consequence of publication.activity. Self Concgpts and Publication The individual brings to the role ideas about himself in the role. He has conceptions of what kind of researcher he 28 is, what is appropriate and inappropriate for him to do, his competence as a researcher, and about the audiences for whom his findings should and do have relevance. The station re- searcher modifies these ideas in the course of continued inter- action. However, these concepts are sufficiently stable to give coherence to the ongoing performance. The individual who regards himself as a competent applied researcher whose find- ings have relevance to farmers is unlikely to suddenly devote his energies to highly theoretic issues. Nevertheless, he may adjust his self-esteem.or redefine his relevant audience (to a subgroup of farmers or another group of lay clients). Such modifications in self-concept may be made while maintaining essentially the same role. Experiment station researchers commonly distinguish be- tween basic and applied research. Applied research includes that which is presumed to have immediate practical relevance to nonscientists. The relevance of so-called basic research to nonscientists is less obvious. A researcher's conception of himself on a basic-applied continuum should be related to his performance as a source of scientific information. The more basic researcher would seem.more likely to restrict his audience to other scientists. Similarly, the more applied re- searchers will probably direct the bulk of their reports to nonscientists. Researchers who define themselves close to midway or who deny the validity of the basic-applied dis- tinction (by defining themselves as high or low on both basic 29 and applied) would be expected to have the greatest hetero- geneity in their types of reports or apparent audiences. Another dimension of self is self-esteem. This is the value an individual places on his own performance of a role. While self-esteem can be regarded as the resultant of evalu- ation by others, it also serves as a determinant of behavior in a role. Since publication is a required behavior of re- searchers and since self-esteem is based in part on adequacy of role performance, a positive relationship between self- esteem and frequency of publication appears necessary. Siggificant Othgrs and Publication People assign significance to others; they care what certain others think of them, An individual's significant others are generally role-specific. For instance, a researcher may regard a coworker as significant to his role as researcher but consider him irrelevant to his husband or father roles. The assignment of significance to others indicates that the assignor acknowledges their expectations in organizing his be- havior. He takes them into account when performing role- relevant acts. Roles differ in the extent to which signifi- cance must be assigned to specific others who enact reciprocal roles. The role of experiment station researcher seems flexi- ble in this regard. However, published research reports are available to others for inspection. Through the report, the researcher 30 presents himself to these others as a researcher. Their opin- ions of him as a researcher are largely based on his writings. To the extent that a researcher's significant others have consensual expectations toward him, his performance should be uniform. Where the expectations of significant others are heterogeneous, some variety can be expected in the performance. A researcher who includes only similar researchers among his significant others is likely to direct all his reports to this audience. On the other hand, several types of reports are expected from.a researcher who admits nonresearchers and non- scientists, as well as other researchers to his set of sig- nificant others.40 Because the same set of findings can be presented from different perspectives and in differing idiom to several audiences, the researcher whose significant others are heterogeneous should publish more frequently. While "double publication" is frowned upon throughout much of the scientific community, it is condoned or encouraged in the college of agriculture. Interaction.and Publicatigg The individual allows Significant Others to influence his behavior. There are, however, other persons he must deal 'with regularly in performing his role. These, Immediate Others, influence an actor's behavior regardless of his intent to be 40 Hawkins, H. Stuart, 22. c_i_§_., pp. 40-42. 31 influenced. His intent, of course, can modify the extent of this influence from inediate others. A researcher may include no Imediate Others among his significant others. In such a case the effect of imediate others on the researchers would be minimal. Another researcher may have as significant others only those who are immediate - with whom he routinely interacts in playing his role. Here the effect of imediate others will be maximal, the researcher allows himself to be influenced by those who are generally available. This type of researcher will likely write less fre- quently (unless writing is an important group norm) because he can comunicate his findings to his significant others without writing. Heavy use of limited distribution mimeo reports is also expected from this type of researcher. As with significant others, heterogeneity among a re- searcher's imediate others will likely be associated with higher publication rates and a greater variety of apparent audiences. This may be a weak association because of the effects of other variables, e.g., self concept and significant others. Nevertheless, the possible effects of mediate others on publication activity should not be disregarded at a prelimi- nary stage of investigation. Publication and Evaluation Role performance is evaluated both formally and infor- mally. Average salary increment is an indication of formal 32 evaluation while reputation.among peers is an informal index. These are not completely independent but are unlikely to be perfectly correlated. Despite the publish or perish adage, it is likely that publication rate will correlate with evaluation only at the extremes; very low publication rates associated‘with low evalu- ation and very high publication rates associated with high evaluation. In the middle range, other factors such as quality of work, and competencies in other areas will likely soften the association between publication and evaluation. W The public evidence of science is publication. Publi- cation forms a large part of the basis upon which a researcher's role performance is evaluated. Self-esteem is an expression of the actor's perception of the adequacy of his performance. Therefore, those who have high self-esteem should be those who play the role well - publish relatively frequently. Hyp. 1 Researchers‘with high self-esteem‘will have higher publication rates than those'with low self-esteem. Role conceptions differ among the actors of a common role. Research reported concerning basic-applied (cosmopolita- 33 41 localite) orientations provides evidence of this variation. The conception an actor has of himself in a role should be ex- pressed in his role performance. One aspect of role perform- ance is the selection of audiences for publications. Self concept as a basic-applied researcher should be reflected in publication audience emphasis. Hyp. 2 The more a researcher is oriented toward basic research, the higher the proportion of publications directed toward other researchers. Hyp. 3 The more a researcher is oriented toward applied research, the lower the proportion of publications directed toward other researchers. It was assumed earlier that "double publication" is ac- ceptable behavior in the college ofagriculture. Researchers who are oriented toward both basic and applied research seem more likely to publish several reports from a particular re- search project. It follows that a dual orientation toward basic and applied research will be associated with high publi- cation rate. Hyp. 4 The more a researcher is oriented toward both basic and applied re- search, the higher will be his publication rate. 41 Storer, Norman W., Sgience and Scientistsign an égri- culturgl_Research Organization: A Sociological Study, Cornell university, Ph.D. Thesis, 1961. Also Hawkins, EL Stuart, 22, cit. . 34 Among actors playing a flexible role, choice of signifi- cant others should be related to role performance. Those who claim a particular class of others as significant are more likely to direct their publications to them as audiences. The same expectation seems applicable to the relationship between orientation toward immediate others and audience emphasis. Hyp. 5 The more peers a researcher claims as significant others, the higher the proportion of publications directed to other researchers. Hyp. 6 The more peers a researcher claims as immediate others, the higher the proportion of publications directed toward other researchers. Researchers who are oriented toward non-peers may pub- lish more easily (and therefore more frequently) through com- mercial, popular or extension channels. Multiple audiences are likely to lead to "multiple publication" (and therefore higher rates) and this seems more likely to occur among researchers who admit some non-peers to their set of significant others. This is the specific rationale for hypothesis 7. Hyp. 7 The more peers a researcher claims as significant others, the lower his publication rate. Immediate others are defined as those who are inter- acted with frequently in the course of role performance. These persons have little need to rely on published output to evaluate 35 an.actor's performance - they have direct and complete evidence. Thus, when an actor regards a high proportion of immediate others as significant, he is likely able to validate his role with less frequent publication. On the other hand, when none of a re- searcher's significant others are immediate, he seemingly must publish to insure that his significant others continue to regard him as a researcher. Hyp. 8 The higher the proportion of immediate others claimed as significant by a re- searcher, the lower his publication rate. The role performance of researchers is publicly expressed in the frequency of publication and in the audiences emphasized. This performance is evaluated formally by administrators and in- formally by local peers. The perspectives of the two sets of evaluators can be expected to differ somewhat because the re- searcher role is a flexible one. Nevertheless, the evaluations are not likely independent; a positive association should be found between the evaluations researchers receive from adminis- trators and from peers. Both peers and administrators can be expected to evaluate positively publication for researcher audiences. A high publication rate‘will likely compensate for a non-researcher audience emphasis. Hyp. 9 Emphasizing other researchers as publi- cation audiences will be positively correlated with administrative evalu- ation, the association being stronger among researchers with below average publication rates. 36 Hyp. 10 Emphasizing other researchers as publication audiences will be positively correlated with peer evaluation, the association being stronger emong researchers with below average publication rates. Researchers should be in a better position to judge "quality" of research and can be expected to be less influenced in their evaluations by publication frequency and audience- emphasis than administrators. It is also assumed that peers are more closely associated with other researchers than are administrators and are therefore more likely to rely more on non-publication criteria for their evaluations. Furthermore, it seems that administrators should be more conscious of the political need for publication to lay audiences and evaluate such an.emphasis more favorably than the researcher's own peers. Hyp. 11 Publication rate will be positively correlated with evaluation, the association being stronger for ad- ministrative evaluation than for peer evaluation. Hyp. 12 Emphasizing other researchers as publication audiences will be positively correlated with evalu- ation, the association being stronger for peer evaluation than for administrative evaluation. Chapter 11 Research Design Data were collected at one Agricultural Experiment Station. Personal interviews, self administered questionnaires and organizational records were used. 42 43 Information from the reported research of Pelz, Menzel, 44 45 Storer, and Hawkins, as well as from an exploratory inves- tigation,‘was utilized in selecting and constructing the spe- cific questions. No particular difficulties were encountered in data collection in these several studies. The interviewers in the exploratory study found that: The respondents talked freely and fluently about specific concerns and general issues; they treated each questionnaire item as a scientific question and were quick to point out ambiguitizg. They were most courteous and sincerely helpful. 42Donald C. Pelz, 22, gig, 43 Herbert.Menzel, 22, cit. Storer, Norman W., 22, cit. 45 Hawkins, H. Stuart, 92. gig. 46Carl J. Couch, Mason E. Miller, and John S. Murray, Communicative Behavior of_Ag£;cultural Scientists: A progress report incorporating the results of a pretest conducted at the University of Connecticut, College of Agriculture. Institute for Extension Personnel Development, Michigan State University, East Lansing, Michigan, September, 1964. 37 38 The Sample The sample consisted of 126 non-administrative scientists currently employed by the Michigan State University College of Agriculture who met the following criteria: (1) hold a Ph.D. degree47 or its equivalent, (2) receive at least 502 of their current salary from the agricultural experiment station (research funds), and (3) were hired before July, 1961. The sample was selected because of its proximity, and availability. It is not known to be unique in any significant respect from the remainder of the population of agricultural researchers at land-grant college experiment stations. Operationalization of variables The key variables used in this study were: 1. Self Concept as a Basic-Applied Researcher 2. Self Esteem as a Researcher 3. Orientation toward Peers as Significant Others 4. Orientation toward Peers as Immediate Others 5. Significance of Immediate Others (overlap) 6. Publication Rate 7. Audience-Emphasis of Publications 8. Evaluation by Peers 7This criterion is an educational control. 39 9. Evaluation by Administration 8 10. Frequency of Interaction'with Lay Audiences4 Self-Concept Storer reported a three item "Guttman-like" scale of basic-applied orientation to research.49 The evidence he pre- sented indicated that the scale was internally consistent and an.apparently valid measure. However, the scale was not appro- priate for this study because one item referred specifically to audience preference which is a main criterion variable. Doubts concerning the unidimensionality of the basic- applied concept prompted the attempt to measure self concept as a basic researcher and as an.applied researcher, separately. Each measure consisted of four items‘selected from responses to the pretest question - "Please make ten statements about yourself as a researcher."' The statements selected appeared to represent the relevant dimensions. A typical typical basic orientation statement is: "I try to do research which will make an original contribution to the storehouse of knowledge." The reciprocal applied orientation statement is: "I try to do research which will provide practical solutions to 'real' problems that peOple have." 8 This variable was added after the proposal was sub- mitted. 49 Storer, Norman W., 22, cit. 40 The response categories were (A) Agree-Disagree, and (B) l-Just a little, 2-Quite a bit, 3-A great deal. The categories were assigned values from O to 5. Item.values were summed for each set of four items providing a possible range of O to 20. The two sets of items were intermixed and used consec- utively in the personal interview, (Appendix A, p. 3). The measures just described were developed to check the extent to which basic-applied orientation forms a continuum, and as the measures of Self-Concept. Self-Esteem The measure of self-esteem as a researcher took the following form. A set of nine characteristics of researchers was used by each respondent, first to construct a profile of his "Ideal Researcher" and second to construct an "Actual Self" profile. The characteristics of researchers used for this mea- sure were selected from responses to the open-ended pretest question - "What are your criteria for an ideal researcher?" Some of the-characteristics were: "--keeps up to date in the literature." “--is dedicated to the search for knowledge." This instrument was self-administered. The instruction.and format can be seen in.Appendix B. The following scoring method was used. The value 5 was assigned to the uppermost box (most important), 4 to the two boxes secnnd from top, 3 to the three middle boxes, 2 to the 41 two boxes next to bottom, and l to the lowermost box (least im- portant). Thus if an item appeared in the uppermost box it was scored 5, if in the middle row, 3, if in the lowermost box, 1, etc. .All items were scored in this fashion for the ideal researcher profile and for the self researcher profile. The absolute difference was then calculated between the values ob- tained on each profile for every item. This difference value constituted the item score. These scores were summed across all 9 items and divided by 2(50) to provide an index of self esteem.for each respondent. For methodological interests, a self-rating on "compe- tence as a researcher" was obtained. The respondents were asked to rate themselves from 1 to 9 on competence as a re- searcher in comparison to all the researchers they knew in the discipline. The verbal anchors for this scale were: 1 - among the few least competent 3 - definitely below average 5 - about average 7 - definitely above average 9 - among the few'most competent 50 The procedure used in this index results in only even numbers on total score; dividing by 2 simplied presentation of results. 42 Significant Others The basic item for this measure was: "Whose opinion of you as a researcher is most important to you?" Probes were used in an effort to elicit five names from each respondent. The respondents‘were asked to weigh the significant others named by distributing 15 points among them. The individual responses were coded as peers, superiors and subordinates of the respondents. An index was constructed by counting the number of types of others used by the respondents. The three categories of heterogenity were: (a) one type - all responses either peers or subordinates or superiors; (b) two types; and (c) three types. Several difficulties were encountered with this index. Thirty of thirty-two respondents who named only one type, named only peers, making the category one of peer orientation. Self references could not easily be classified according to this scheme. The index took no account of the weightings used by the respondents to indicate the relative significance of others to them. These considerations were taken into account in de- veloping an alternate approach. The new index of significant others can be regarded as a measure of the gignificance of_peers. The weightings were used to trichotomize the sample. The categories were (a) Peers Only (15 points to peers), (b) Peers Major (8 to 14 points to peers), and (c) Peers Minor (7 or fewer points to peers). This 43 index was regarded as more satisfactory than that first pro- posed and was used throughout the analyses. Immediate Others The researchers were asked to identify five persons fre- quently dealt with in their researcher role. The basic question ‘was: "Who are some of the people you talk to frequently about your research?" Originally it was proposed to index the het- erogenity of immediate others as had been intended with sig- nificant others. This proposal was modified in accordance with the changes made in the operationalization of significant others. The responses were coded as peers or non peers. The index de- velOped reflects the proportion of peers named as immediate others. The three categories were (a) peers only, (b) a majority of peers, and (c) a minority of peers. Significance of Immediate Others Some duplication was expected in the responses to sig- nificant other and immediate other items. The extent of this duplication provided an index of the significance of immediate others. High significance of immediate others was indicated by researchers who listed many same persons in response to the two questions. The values of this index could range from 5 (significance assigned to 5 immediate others) to 0 (signifi- cance assigned to no immediate others). Publication Rate Each researcher was asked to provide a list of his publi- cations during the period September 1, 1961 to September 1, 1964. A simple count of publications constituted the index for publi- cation rate. There are obvious inadequacies in this procedure; it does not reflect differences in single vs. multiple author- ship ncr in length or form (book chapter vs. research note vs. extension pamphlet). Initially, a weighting scheme was pro- posed. After additional discussion, thought and investigation, the weighting proposal was abandoned. No weighting system could be found or conceived which would obviously reflect "amount of individual published output" more adequately than counting the member of reports in'which the individual is credited with au- thorship. The weighting schemes considered seemed only to alter the kinds of distortion present rather than improve the measure. In this investigation, publication rate refers to the number of publications to which the respondent claimed at least partial authorship and provided citations. Audience Emphasis For each publication claimed, the respondents were asked to indicate the primary audience. The instructions and cate- goriss‘were: 45 Main Audience: Circle the one most appropriate: Researchers - own speciality Researchers - own discipline Researchers - other disciplines Teachers Extension Specialists Extension Agents Other non-research Scientists Graduate Students undergraduate Students Administrators Farmers or other non-scientists (Other - please specify) 0 O O O FNHHGPHNUOU» .The original intent was to classify researchers on the basis of the diversity of their claimed audiences. However, other researchers were by far the most heavily emphasized audiences. An alternate approach was used which indexed the extent of researcher audience emphasis. Researcher audience emphasis was defined as specifying researchers (A, B, or C above) as audiences for 901 or more of total publication output. About one-third of the sample met this criterion. Peer Evaluation Each respondent was provided with a list of names of persons in the sample. He was asked: "-dwho are some that, in your opinion, most deserve the reputation of outstanding researcher?" Five nominations were requested. Four differently ordered lists were used to control for any "first page effect." In addition, the respondents were instructed to scan the complete list before 46 making any nomination. The number of nominations received by each respondent was taken as the index of peer evaluation. Mstraflye Evilruation The average percentage salary increment for the preceding three years constituted the index of administrative evaluation. Annual salary increments are made_in July. The method of cal- culation shown below controls for absolute salary level. The method was: A351. Salary as of Aggust'64 - Annual Salary as of June'6l 100 X Annual Salary as of June'6l The values of this measure were used directly for corre- lational purposes. For other purposes, the sample was tri- chotomized on the basis of these values into High, Medium, and Low administrative evaluation. interaction With Lay Audiences this variable was added to the list of theoretic vari- ables. It is a measure of the extent to which the respondents comunicate orally to audiences of non-researchers. The re- spondents were asked, "How frequently do you give talks or lectures to groups of non-scientists, non-students?” Probes were used to elicit the answer in terms of the "number of times of per year." On the basis of the distributionof responses, the sample was divided into low, medium and high frequency 47 categories. The low frequency category included respondents who gave fewer than four talks per year, medium, four to ten talks and high, eleven or more. Data Collection The director of the Experiment Station informed his per- sonnel of the study. His message is presented as Appendix C. Appointments for the personal interviews with the re- searchers were arranged by phone. The interviews took place in the respondents' offices, except for a few instances of shared offices where vacant seminar rooms or laboratories were used. The author conducted all but three of the interviews. These three were conducted by another graduate student in the department of comunication who (like the author) had academic and work experience in a college of agriculture. A self-administered questionnaire (Appendix B) was left with each respondent at the completion of the interview. They were asked to return the completed questionnaires via campus mail. A series of phone calls were made as reminders to those who were slow to return the questionnaires. Data concerning percentage salary increments was ob- tained from administrative records. Chapter III FINDINGS AND DISCUSSION Description of the Sample The original sample consisted of 126 employees of Michigan State University. They held a Ph.D. degree, obtained at least 502 of their salary from the Michigan State Agricul- tural Experiment Station, and had been on staff for at least three years. One individual was added to the sample who re- ceived no salary from the Experiment Station but met the other two criteria. This person was nominated frequently as an out- standing researcher despite the fact that his name did not appear on the list of possible nominees. The disposition of the sample is presented as table 1. Table 1: Response Rate Original sample plus one addition 127 Off campus during interviewing period 8 Refusals ___2 Number intervi ep) 117 Number unusable _]._3_ 104 Final sample 51Reasons for discarding subjects were: 1 had no Ph.D., 2 interview schedules were incomplete, 6 failed to return the self administered questionnaire, 4 self administered question- naires were incomplete. 48 49 One hundred and seventeen persons (92%) were interviewed. Reasonably canplete information was obtained from the 104 (827.) researchers who constituted the final sample. Comparison of the 104 respondents with the remainder of the original sample indicated no consistent differences of rank, department, salary increment or peer evaluation. On the basis of these compari- sons, it appears that the non-respondents were not a particular type or segment of the original sample, and that the respondents are representative of agricultural researchers at Michigan State University. The findings reported here are based on the data col- lected from these 104 respondents, unless otherwise indicated. General Characteristics of the Respondents The intent in this study was to deal only with individ- uals who were well settled or established as researchers. Many of the findings presented in this section are intended to show the extent to which a sample of "mature researchers" was ob- tained. The pertinent data are smarized in table-2. me chronological age of the researchers varied from less than 30 (12) to more than 65 (42), the median age was 45.5. Two-thirds of the sample were between 40 and 54, one quarter were under 40 and about one tenth were older than 55. The bulk of the researchers in the sample would be labelled as "middle- aged." 50 Table 2: Summary of General Characteristics variable Rapge Age 30 - 65 Time 4 - 30 from Ph.D. Time 4 - 47 at M.S.U. Rank 1 50 - 100 Salary Research Number of 0 - 3 other professional positions Alma Mater Mbdian. 45.5 13.3 10.7 Categories under 40 40-49 Over 49 Under 11 11-20 Over 20 under 8 8-14 Over 14 Assistant Professor Associate Professor Professor @' 74.7 under 51 51-79 Over 79 0.67 0 l Mbre than 1 MOSOUO Other Land-Grant Non-Land-Grant N'103 W 25 46 29 1001 35 44 21 1001 33 35 32 1001 11 27 52 100% 37 29 100% less 100% 63 12 1001 51 Approximately one half of the respondents had obtained their Ph.D. degree six to fifteen years ago. None had held doctorates for less than four years, while two researchers had held Ph.D. degrees for more than 30 years. The median time from.Ph.D.‘was 13.3. These findings provide an indication that the respondents meet the criterion of established researchers. Tenure at.M;S.U; ranged from four years (71) to forty- seven years (11), the median was 10.7. More than three quarters of the sample had been employed by Michigan State University for more than six years. The distribution of re- spondents on this measure is a further indication of the extent tO‘WhICh the respondents are established researchers. A.majority (521) of the researchers held the rank of full professor, about a third (37%) were associate professors, the remaining one-tenth were assistant professors. Almost two-thirds (632) of the respondents obtained 70% or more of their salary from the agricultural experiment station. About one-fifth (17%) derived all their salary from this source. The median.research salary was 74.71 for the sample (excluding the added individual). In general, the re- spondents appear to have validated their role as researchers. They have attained higher academic ranks and are still paid, for the most part, to do research. Only six respondents obtained any salary for extension activity. For the remaining 98 respondents, the difference 52 between percentage salary for research and total salary was supplied from general (teaching) funds. There were no three way appointments among the respondents. About one-third (341) of the respondents have held one other professional position, while one-fifth (221) have held more than one. For almost half the sample (44%), their present position is the only one they have held since obtaining their doctorate. The sample seems to represent a stable core of re- searchers associated with the agricultural experiment station. Not only are they long term employees, they have had few other employers. may appear to be career scientists, they have found an acceptable situation and maintained it. Almost nine out of ten respondents held an earned Ph.D. from a land grant college or university. One quarter of the respondents had doctorates from Michigan State University. Eighteen administrative units were represented by the 104 respondents. A unit refers to a department within the College of Agriculture or to the College of Home Economics or to the College of Veterinary Medicine. The variability in number of respondents per administrative unit necessitated some categorization. The categories were developed on the basis of impressions obtained during interviews of the subject matter emphasis of the units. Among the units assigned to the agricultural emphasis category were Animal Husbandry and Agronomy. Biochemistry and Botany were assigned to the 53 category labelled "Conventional Scientific Emphasis" and units such as Food Science and Sociology were assigned to a "Mixed Emphases" category. The number of respondents from each unit and the outcome of the categorization is presented as table 3. Table 3: Categorization of Administrative Units Type of Subject Matter Units Included Percentage Agricultural Emphasis Animal Husbandry (2), Dairy (5), Agricultural Engineering (3), 37 Crops (5), Horticulture (l6), Poultry (2), Soils (6) Mixed Emphasis Agricultural Economics (10), Fisheries and Wildlife (3), Food Science (10), Forest 32 Products (1), Forestry (1), Hana Economics (4), Sociology (4) Conventional Biochemistry (8), Botany (10), Scientific Entomology (l), Veterinary 31 Emphasis Medicine (13) 1007. (104) No units, Horticulture and Veterinary Medicine supplied one quarter (28%) of the sample. Less than three respondents were drawn from Animal Husbandry, Entomology, Forest Products, Forestry and Poultry. The sample represents the diversity of subject matter areas to be expected at a state agricultural experiment station. The variety is such that differences in basic-applied orientation and audiences for publications can be expected. l 54 Description of the Sample in Terms of The Theoretic variables Self Esteem as Researcher Two measures of self esteem were used for comparative purposes. The distribution of responses is shown in table 4. Table 4: Response distributions on self-esteem.measures Score or value (ratigg scale) index b 0 (Law) a 1.02 l 0.01 1.9 2 0.0 4.8 3 1.0 17.3 4 1.0 20.2 5 15.4 27.9 6 7.8 6.7 7 42.3 4.8 8 15.4 9.6 9 (High) 13.5 No Response 3,8 5,8 100.27. * 100.07. M.- (104) (104) The correlations among the nine items of the self esteem index were quite high, the lowest being .76. All items corre- lated‘well with total score, the lowest of these was .82. a Score not present on this measure. bThis distribution was reversed for comparative purposes, actually a 0 score on the index represented high self esteem and was obtained by 10 respondents, etc. * rounding error. "$31.... 55 The correlation matrix is shown as Appendix D. mess results suggest that the items tapped the same dimension. Hires-quarters of the respondents rated themselves as definitely above average (7, 8 or 9) on the nine-point self rating scale. The mean response value was 6.97, the median 7.45. On the self esteem index, two-thirds of the respondents fell in the middle score range (3, 4 or 5). On this measure the mean score was 4.61, the median 4.77. Despite the apparent high level of esteem in the sample according to the self rating scores, it was hoped that the two measures would be closely associated. Table 5 indicates that this was not the case. he overall chi-square for this dis- tribution did not approach statistical significance. The ex- pected trends were simply not evident. Table 5: Distribution of respondents by scores on the self-esteem measures. Self Esteem by Rating Scale Hi Med Lo (8-9) (7) (14} Self Esteem Hi (5-8) 25.02 23.32 16.02 Index Med (3-4) 53.6 48.8 56.0 Lo (0-2) 21.4 27,9 28,0 100 .02 100 .02 100 .02 (28) (43) (25) Chi-square - 1.06, 4 d.f.; p greater than .05, Wo alternative test. 56 This was the first of a series of disappointments with the concept and/or measurement of self-esteem as researcher. Not only were the measures not related to each other, neither *was associated with any other variable. The results presented in the following pages concerning self-esteem were obtained using the index. Similar non significant results from the rating scale are not presented. Sglf Concept as a Basic-Applied Researcher Separate measures were used to assess the researchers concept of himself as a basic and as an applied researcher. All respondents completed the four items of each measure. The dis- tribution of respondents by score is presented in table 6 for both measures . Table 6: Distribution of respondents on basic and on applied orientation measures. Score Interval Percentage on Basic Percentage on.§pplied 0- 2 0.02 4.82 3- 4 0.0 3.8 5- 6 0.0 1.9 7- 8 0.0 6.7 9-10 2.9 19.2 11-12 5.8 11.5 13-14 13.5 11.3 15-16 26.9 18.3 17-18 35.6 9.6 19-20 15,4 * 7,7 100.12 99.82 * (104) (104) Range - 9 - 20 0 - 20 Mean I 16.15 12.21 Median - 16.56 12.72 *rounding errors 57 The inter-item and item-total correlations for the basic and applied measures are presented as table 7. Table 7: Correlations hong Basic and Applied Items Basic Items 29 3o 32 Total 27" -.19 +.36 -.12 .26 29 +.o4 +.04 .69 30 +.16 .68 32 .34 new 31 33 34 Total 288 +.51 +. 26 +.44 .72 31 +.29 +.53 . 77 33 +.39 .65 34 .80 8 Identification numbers are those used in the questionnaire, appendix A, p. 3. The items used to form the measure of basic orientation did "hang together" sufficiently to form a scale. Several items were'negatively correlated with each other, most of the corre- 1ations were low. In comparison, the items of the applied orientation mea- sure were positively associated with each other and all contri- 58 buted satisfactorily to total score. he cause(s) of the unexpected outcomes is not inanedi- ately obvious. he items were selected from responses to a single pretest item to represent two distinct and likely oppo- site dimensions. Poor item selection could certainly account for the findings but they could also result from different degrees of consensus among the respondents as to the nature of basic and of applied research. Possibly the respondents hold a widely shared meaning for applied research but several dif- ferent meanings for basic. Impressions gained during the interviews would support this possibility. his speculation suggests that (a) the concept of basic-applied research may not be undimensional, and (b) the concept of basic research may also not be unidhensional. Exploration of such possibilities is beyond the scope of this investigation. It was decided to use only the measure of applied ori- entation in subsequent analyses. his measure was regarded as having an acceptable degree of internal consistency and a satisfactory distribution. he sample was dichotomized on the basis of applied orientation scores to produce a low applied orientation category (ll-50) and a high applied orientation category (NI-54) . Significant Others he researchers were asked to identify significant others by the question, "Whose opinion of you as researcher 59 is important to you?" Every individual in the sample made some response to this item. he types of responses are smarised in table 8. Table 8: Percentage of response types to Significant Other question. Response Type W Some Self Reference 5.82 Only Groups or Categories Med 5.8 Groups or Categories and Individuals Mined 5.8 Only Individuals Named 82,7 m 100 . 12 (104) *rounding error Of the 86 respondents who named only other individuals as significant; 80 named five others, three named four and three named two. Five of the six respondents who named both individuals and groups made five responses each, two named one group and four individuals, twa named two groups and three individuals, and one named four groups and one individual. The remaining individual who gave a combination response named one group and two individuals. None of the six respondents who named only groups made five responses, four named three groups, one named two, and one named only one. “No of the six respond- ents who made self references claimed no-one else's opinion of them as researchers mattered, mo also named other individuals 60 and the remaining two named groups in addition to self. The individual responses were coded as, peers or non- peers. he weightings assigned by the respondents to indicate the relative significance of the others named, were used to categorize the respondents according to the extent of their orientation toward peers. hirty respondents assign signifi- cance only to peers, 39 were classed as having a major peer orientation (they assigned eight to 14 out of 15 points to peers) and the remaining 35 were placed in a minor peer ori- entation category. his index was used throughout the analyses as the measure of significant others. Mute Others he data for this measure were collected by asking the researchers, "Who are some of the people you talk frequently to about your research?" Ninety-nine respondents gave five responses, as requested, to this question, mo gave two re- sponses and one gave only one response. Ninety-seven respond- ents named only individuals; of these, 95 named five, one named three and one named one. No respondents named only groups, both named two groups. hree of the five respondents who named both groups and individuals gave four individuals and one group, one named one individual and two groups and one named one individual and four groups as imediate others. hese data are sumnarized in table 9. 61 Table 9: Percentage of response types to imediate other question. Response HE Percentage Only Groups or Categories Named 1.92 Groups or Categories and Individuals Named 4.8 Only Individuals Named 93,3 100 .02 (104) As in the case of significant others, the individual responses were coded as peer or non-peer. he proportion of peers named as inediate others provided the basis for the de- velopment of three categories of respondents: peers only (31) ; peers majority (50); peers minority (23). Significance of I_g__nediate Others his measure indexes the amount of overlap between sig- nificant others and inmediate others. he possible range is from 0 to 5, the actual range was 0 to 4. Seven individuals were dropped from this measure; they had given a self-only response to the significant other question or a combination (groups and individuals) response to the inmediate other question. he distribution of researchers on this variable is presented in table 10. 62 Table 10: Distribution of Respondents according to the nunber of Imediate Others who were also Significant. Nunber Imediate also Significant Percentage 0 30.92 1 26.8 2 24.7 3 13.4 4 4,1 I 99.92 (97) *rounding error Almost one-third of the sample named no imediate others significant, about one quarter named one, another one quarter named Wo and the remaining one fifth named three or four im- mediate others as significant. he researchers in this sample showed considerable variability in the extent to which they regarded their innediate others as significant. he measure turned out to be no more predictive than that of self esteem. Publication Rate his measure is simply the number of publications to which an individual claimed authorship during a three year period. he nunber ranged from a low of zero (12) to a high of 56 (12). Table 11 is a smary of the number of researchers authoring different numbers of publications. About one third (322) authored seven or fewer, another third (392) authored 63 between eight and 14, while the rate for the final third (292) ranged from 15 to 56 publications. The median number of publi- cations was 10.75, the mean was 13.3. The 10 highest producers authored more than 13 times as many publications as the 13 lowest producers during the three year period. Table 11: Distribution of respondents by publication rate. Number of publicatiqgg Percentage 0 - 2 4.82 3 - 5 14.4 6 - 8 16.3 9 - 11 17.3 12 - 14 18.3 15 - 17 6.7 18 - 20 3.8 21 - 23 7.7 24 - 26 1.0 27 or more 93,32. (104) ’ *rounding error Audience Emphasis Each publication claimed by the respondents was classi- fied as having or not having a researcher audience. As far as 64 possible, the respondent's indication of publication audience was used. Seventy respondents indicated one and only one audience for each publication. Twenty-seven respondents made at least one multiple response and seven failed to specify audiences for any publication. The bulk of the mltiple re- sponses were within or beyond the three categories of re- searchers. he remaining multiple responses and citations from the seven respondents who did not specify any audiences were classified according to the apparent form of the publi- cation. For example, journal articles were considered to be for researcher audiences and consnsrcial publications for non- researchers. Doubtful cases were coded as not for researchers. The percentage of publications directed to researcher audiences was then calculated for each respondent. he thirty-three respondents who directed 902 or more of their publications to other researchers were classified as having a researcher audience euphasis, the remaining seventy as having a non-researcher audience emphasis. he one respondent who was dropped from this measure had no publications. Peer Evaluation Each respondent was provided with a list of names of those persons who constituted the original salmle and asked to nominate five whom he regarded as outstanding researchers. Ninety-nine respondents made five nominations, two. made two and three refused to answer. he number of nominations received by 65 each respondent constitutes the measure of peer evaluation. The number of researchers nominated is presented in table 12. Table 12: Percentage of Researchers Raceiving Various Numbers of Nominations for Outstanding Researcher. Number of Nominations Percentage 0 30.82 1 19.2 2 11.5 3 11.5 4 6.7 5 or me 332% (104) *rounding error he number of nominations received ranged from 0 (312) to 39 (12). he mean number was 3.8, the Indian 1.5. The dis- tribution was severely skewed, only one tenth of the respondents received more than 10 nominations. Aiministrat ive Evaluation his measure is the percentage salary increment received by each respondent over the previous three years. The incre- ments ranged from 1.12 (12) to 18.92 (12). he mean increment was 7.82, the nedian 7.22. Data were not available for three respondents. Table 13 presents the distribution of respondents 'nr.ry 66 by salary increment intervals of three percent. About one quarter (252) of the respondents received increments of more than 102, one tenth received 42 or less. More than one-third of the sample (372) received increments of 4.1 to 72 and 872 of the sample received between«4.l to 132 increments. Table 13: Distribution of Respondents by Salary Increment Intervals. Salary Interval Percentage 1.1 - 4.0 9.92 4.1 - 7.0 36.6 7.1 - 10.0 27.7 10.1 - 13.0 21.8 13.1 - 16.0 2.0 16.1 - 19.0 _2_:_Q_ 100.02 (101) The distribution on this measure makes it quite accept- able as a criterion. Frequency of Interaction wigh Lay,Andiences The respondents were asked how often they gave talks to audiences of non-researchers, non-students. Low, medium and high frequency categories were made considering the overall distribution. Forty-nine respondents'who reported fewer than four talks per year to lay audiences‘were placed in the low fre- quency category; 27 who reported four to 10 talks were classified 67 as moderate frequency and the remaining 28 (11 or more talks) as high frequency. Tests of Hypotheses ‘A significant positive correlation between self-esteem index scores and publication rate was predicted in hypothesis 1. The Pearson Product Moment correlation for 98 respondents did not approach significance.51 Hypothesis 1 was not con- firmed. The second hypothesis stated that researchers‘with a high orientation toward basic research would emphasize other researchers as audiences. This hypothesis could not be tested . because of the lack of internal consistency in the measure of orientation to basic research. Hypothesis 3 predicted a negative association between orientation toward applied research and proportion of publi- cations to researcher-audiences. The relevant distribution is presented as table 14. r- -.07; an r-.21 was required for significance at .05 level, two tailed test. 68 Table 14: Orientation to applied research and researcher audience emphasis. Orientation to applied research Lo Hi Proportion of 902 or more 42.9 22.2 publications to less than 902 57.1 27.8 researchers 100 .02 100.02 (49) (54) Chi-square I 4.12; l d.f.; p less than .05, two alternative test, with Yates' correction. Researchers classified as having a low degree of ori- entation to applied research were more likely to emphasize other researcher audiences than were those with a high ap- plied research orientation. Hypothesis 3 was confirmed. A dual orientation to basic and applied research was predicted to be associated with high publication rate in hypothesis 4. The inadequacy of the measure of basic ori- entation made it impossible to test this hypothesis. he expectation that orientation toward peers as sig- nificant others would be positively associated with aspha- sizing researcher-audiences was stated as hypothesis 5. he data used to test this hypothesis are shown in table 15. l . .1 {I}: . . f . .r. _ . e r r . L o _ .. _ . 1 r1 . . . . . . n. . c 6 . 1 e s , _ 6 . 7 . t \1 I . H A . . n 6 u c . . . l y . 1 ~‘ r _ n x a 1 ‘ . p w. . . e I. n n 1 1 I V r . _ O O— a e P , r . L . . . r , a .1 . k 1 P! a 4 I C e . e _ . a. . . c . . \ a . , , u e r a y e o r /I r r o .. . a l o r . . n _ _ . . ,. . n a [I H. 1u_~ ... . ... V ‘1‘ 69 Table 15: Significant other orientation and researcher audience emphasis. Significant other orientation Peers Peers Peers Only Major Minor Proportion of 902 or more 46.7 30.8 20.6 publications less than 902 53.3 69.2 79.4 to researchers 100.02 100.02 100.02 (30) (39) (34) Chi-square '- 5.03; 2 d.f.; p greater than .05, two alternative test. mile the trends were in the predicted direction, the chi-square test was not significant. Hypothesis 5 was not confirmed. Hypothesis 6 was a restatement of hypothesis 5 in the case of orientation toward inmediate others. he data are presented in table 16. Table 16: Innediate other orientation and researcher audience emphasis. Imediate other orientation Peers Peers Peers Only Majority Minority Proportion of 902 or more 51.6 26.5 17.4 publications less than 902 48.4 23.5 82,6 100.02 100.02 100.02 (31) (49) (23) Chi-square - 8.42; 2 d.f.; p less than .05, two alternative test. As predicted, the more peers a researcher claimed as inediate others, the more likely he was to direct 902 or more ('\ ,. f" .1 70 of his publications to researcher audiences. he results con- firm hypothesis 6. he confirmation of this hypothesis and lack of confirmation of hypothesis 5 may be due to measurement differences or may be an indication that innediate (inter- active) others is a more powerful predictor of audience em- phasis than is significant others. Evidence to support the latter possibility will be presented later. Hypothesis 7 stated that the more peers a researcher claimed as significant others, the lower his publication‘rate. he data presented in table 17 show no evidence of the pre- dicted association. Hypothesis 7 was not confirmed. Table 17: Significant other orientation and publication rate. Significant other orientation Peers Peers Peers Only Major Minor Publication Lo ( 0- 7) 30.0 41.0 22.9 Hi (15 or more) 36,7 25,6 * 25,7 100.02 99.92 100.02 (30) (39) ‘ (35) *rounding error Chi-square - 4.79; 4 d.f.; p greater than .05, two alternative test. It was predicted in hypothesis 8 that the more immediate others a researcher claimed as significant, the lower his publi- cation rate. he relevant distribution is presented as table 18. 71 Table 18: Significance of imediate others and publication rate. Number of imediate others who were also significant 0 1 - 2 3 - 4 Publication Lo ( 0- 7) 43.3 26.0 35.3 Rate Med ( 8-14) 33.3 42.0 41.2 Hi (15 or more) 23,3 * 32,0 23,5 99.92 100.02 100.02 *rounding error Chi-square I 2.76; 4 d.f.; p greater than .05, two alternative test. The chi-square test of these data was not statistically significant. However, there appeared to be some tendency for researchers who had one or two immediate others as significant to have higher publication rates. This possibility was ex- plored by means of an analysis of variance. The test for homogenity of variances indicated that the assumption of homo- geneous variances was not tensble.52 However, the analysis of variance was still conducted, following Norton's interpre- tation of non-homogeneous variances as summarized by Lindquist.53 he analysis of variance indicated that the differences in publication rate between the categories of "significance of 52F max I 5.59; K I 3, n I 30 and 50; p greater than .01. 53 Lindquist, E., Design and Analysis of Experiments, Boston: Houghton Mifflin, 1953, pp. 78-86. 72 immediate others" were not statistically significant.54 Both the chi-square test and the analysis of variance failed to demonstrate the presence of an association between significance of immediate others and publication rate. Hypothesis 8 was not confirmed. Hypotheses 9 and 10 stated that the correlations be- .tween researcher audience emphasis and evaluation from admin- istration (Hy. 9) and from peers (Hy. 10) would be higher for researchers with low publication.rates than for those'with high rates. The sample was dichotomised using a median cut on publication rate. Point biserial correlations‘were then calculated between researcher audience emphasis and peer eval- nation and between researcher audience emphasis and adminis- trative evaluation for the high and low publication rate cate- gories. These correlations are shown in table 19. Table 19: Correlations between.researcher audience emphasis and evaluation by publication rate categories. Publication.Rate £9. 11.1. Hy. 9 Researcher-audience emphasis x. +312 +502 administrative evaluation Hy. 10 Researcher-audience emphasis x +.19 +330 peer evaluation . r I .29; 45 d.f.; necessary for significance at .05, two tailed test. 54 F I 1.02; d.f. I 2 and 94; F.95 I 3.11. 73 Neither of the correlations used in.hypothesis 9 were statistically significant and no test of the differences be- tween.them was made. In hypothesis 10, the correlation be- tween researcher-audience emphasis and peer evaluationuwas statistically significant. However, the direction of the dif- ference between the correlations was opposite to that hypoth- esized. Hypotheses 9 and 10 were not confirmed. Hypothesis 11 stated that the correlation between pub- lication rate and administrative evaluation‘would be larger than that between publication rate and peer evaluation. The Pearson product moment correlation between publication rate and administrative evaluationuwas +315 and between publica— tion rate and peer evaluation'was +341.55 The direction of the difference between the correlations was opposite to that hypothesized and no test of the hypothesis was made. Hypoth- esis 11 was not confirmed. Hypothesis 12 predicted that the correlations between researcher audience emphasis and evaluation'would be stronger for peer evaluation.than.for administrative evaluation. Point biserial correlations were used. The values of the corre- lationnwas +318 in the case of peer evaluation.and +305 with administrative evaluation. The trend of the difference was that predicted but the correlations were not statistically 55r I .21; 90 d.f.; necessary for significance at .05, two tailed test. 74 significant so no difference test was used. Hypothesis 12 was not confirmed. The objectives of this study were to search for and ex- amine certain relationships between self-other-interaction variables and publication activity; and between.publication activity and evaluation. But the process of research is seldom straightward. In science, the characteristics of the material, the appropriate- ness of methods, and the explanations of outcomes are un- certain. They must be constantly questioned and examined. Variables and relationships at the periphery of the area of study may have profound implications for the interpretation of results. These considerations prompted the systematic investi- gation of relationships among all the variables used. These additional findings have been organized into the following two sections: relationships among the theoretic variables and interrelationships between general.characteris- tics and theoretic variables. Additional Findings Relationships amopg the theoretic variables The results of the chi-square tests of these relation- ships are shown in appendix F. The associations which were statistically significant are represented by the lines in 75 figure 1. The figure is intended only as a summary of the re- lationships; the absence of lines between variables indicates that the associationnwas not statistically significant at the .05 level. Figure 1. Summary of relationships among the theoretic variables. Independent variables Criterion variables Self Esteem Applied Orientation Researcher Audience Emphasis "“'Significant Others ' blication Rate }__Significance of Imedia Peer Evaluation L‘——-Immediate Others ' Talks to Lay Audience Administrative Evaluation Relationships among the independent variables will be presented and discussed first, followed by the relationships between independent and criterion variables. he manner in which "significance of innediate others" was related to the twa variables from which it was constructed is shown in table 20. 76 Table 20: Significance of imediate others with significant others and with inanediate others. Nunber of imediate others also significant 0. 1 2+ Significant Peers exclusively 46. 7 3.8 34.1 Other Peers major 23. 2 57.7 34.1 Orientation Peers minor _3_O_,_ 0 38,5 31,7 * 100 .02 100 .02 99.92 (30) (26) (41) ‘ Chi-square I 13.84; 4 d.f.; p. less than .05, No alternative test Imediate Peers only 46. 7 15 .4 29 .3 Other Peers majority 46. 7 61 .5 46.3 Orientation Peers minority 6,7 * 23.1 24,4 100.12 100.02 100.07. (30) (26) (41) Chi-square I 10.84; 4 d.f.; p less than .05, mo alternative test. *round ing error Respondents who named only peers as significant and as imediate were likely to have either none or mo or more im- mediate others as significant, almost none fell in the middle category. hose who named only researchers as immediate others were also likely to have little or considerable overlap be- tween significant and inlnediate others. Almost mo-thirds of the respondents who named a minority peers as innediate others claimed at least two of them as significant. No explanation of these outcomes was imediately obvious. Further thought and inspection of individual interview schedules 77 suggested that the variable "significance of innediate others" was partially confounded by the location of peers. Researchers who named only non-local peers as significant generally did not claim these same peers as inediate others while those who named local peers as significant others frequently named some of them as imediate others. Rather than a "clean" index of significance of imediate others, the variable seems also to reflect the location of a specific category of significant others - namely peers. he utter lack of predictive power of this variable appears attributable to inadequate operational- ization and/ or conceptualization. Respondents who frequently gave talks to lay audiences were more likely to have a high orientation toward applied research and to claim non-peers as innediate others. hese trends can be observed in table 21. hese results are consistent with the theory in that self definition in a role was demonstrated to be directly re- lated to role relevant interactive patterns. he findings could be stated as follows: role performance (frequency of talks to lay audiences) is a function of self definitiOn (applied research orientation) and interactive patterns (proportion immediate others who are peers). he negative association between orientation toward applied research and researcher audience emphasis was pre- dicted in hypothesis 3 and was discussed earlier. 78 Table 21: Frequency of talks to lay audiences with applied research orientation and with immediate others. Talks per year to lay audiences 0-3 4-10 _1__1_4; Applied Lo 61.2 37.0 35.7 Orientation Hi 38.8 63,0 64,3 100.02 100.02 100.02 049) (27) (28) Chi-square I 6.41; 2 d.f.; p less than .05, two alternative test. Immediate Peers only 42.9 29.6 7.1 Other Peers majority 42.9 59.3 46.4 Orientation Peers minority 14,3 * 11,1 46,4 100.12 100.07. 99.97.* Chi-square I 18.94; 4 d.f.; p less than..05, two alternative test. *rounding error Several hypotheses concerned relationships between ori- entation toward peers as significant others and publication activity. None of these hypotheses were confirmed. However, researchers who assigned high significance to peers were more frequently nominated as outstanding researchers by peers. This distribution is presented as table 22. 79 Table 22: Significant other orientation with peer evaluation. Significant other orientation Peers Peers Peers Exclusively Major Minor Number of O 10 .0 35 . 9 42. 9 Nominations 1-2 40.0 23. 1 31 .4 Received 3 (+) 50 ,0 41 ,0 25 , 7 100 .02 100 .02 100 .02 (30) (39) (35) Chi-square I 10.37; 4 d.f.; p less than .05, twa alternative test. Half of the researchers who assigned significance ex- clusively to peers received high peer evaluation whereas only one-quarter who assigned minor significance to peers were this highly evaluated by them. he relationship is even more evi- dent in the low peer evaluation (o nominations) category. he imediate others variable was associated with all four criterion variables. he higher the proportion of peers a researcher named as imaediate others, the more likely he was to (a) emphasize researcher audiences (Hypothesis 6, confirmed and discussed earlier), (b) have lower publication rates, (c) be highly evaluated by peers, and (d) be highly evaluated by administrators. he distributions for the last three of these associations are shown in table 23. 80 Table 23: Inmediate other orientation with publication rate; peer evaluation and administrative evaluation. Innediate other orientation Peers Peers Peers Only Majority Minority Publication Lo 67.7 46.0 30.4 Rate (Median Cut) Hi 32,3 54,0 69.6 100.07. 100.07. 100.02 (31) (50) (23) Chi-square I 7.70; 2 d.f.; p less than .05, two alterative test. Number of O 32.3 22.0 47 .8 Nominations 1-2 12.9 44.0 26.1 Received 3 (+) 54 ,8 34 ,O 26 , 1 100 .02 100 .02 100 .02 (31) (50) (23) Chi-square I 13.01; 4 d.f.; p less than .05, two alternative test. Percentage Salary Increment Lo and Med 48 .4 75.5 81.0 H1 51 6 24 ,5 19 ,0 100 .02 100 .02 100 .02 (31) (49) (21) Chi-square I 8.41; 2 d.f.; p less than .05, two alternative test. Interactive pattern, indexed as the proportion of peers named as inediate others, was the most useful independent vari- able in the study, in terms of its ability to predict both publication activity and evaluation. he associations between immediate others and evaluation are probably direct because the associations between publication activity and evaluation generally 81 fell short of statistical significance. That the proportion of peers named as immediate others was negatively associated‘with frequency of talks to lay audi- ences was established earlier in this section. In.table 24 it can.be seen that researchers who frequently gave talks to lay audiences were likely to emphasise non-researcher publication audiences and to have high publication rates. These findings are consistent with those presented earlier and suggest that frequency of talks to lay audiences indexes the reciprocal interactive pattern of proportion of peers named as immediate others. Table 24: Frequency of talks to lay audiences with audience emphasis and publication rate. Talks per year to lay audiences 0 - 3 4 - 10 11 (+) Percentage of 902 or more 52.1 22.2 7.1 Publications to less than.902 47.9 77,8 92.9 Researchers 100.02 100.02 100.02 (48) (27) (28) Chi-square - 18.02; 2 d.f.; p less than .05, twO alternative test. Publication 0 - 7 49.0 25.9 7.1 Rate 8 - 14 32.7 37.0 53.6 15(+) 18,4 * 37,0 39,3 100.12 99.92 100.02 (49) (27) (28) Chi-square I 15.80; 2 d.f.; p less than..05, two alternative test. *rounding error 82 Half of the researchers who had low frequency of talks to lay audiences emphasized researcher-audiences and had low publication rates, compared to less than.102 of those who had high frequency of talks to lay audiences. These results and those presented with respect to immediate others indicate that researchers who interact frequently with nondresearchers are more likely to write for them and to have high publication rates. The findings presented in this section demonstrate the relevance of interactive patterns to publication activity and evaluation. Immediate others appeared as a key independent variable. Relationships between General Characteristics and Theoretic variables Despite the seeming consistency of the findings in the previous section, the possibility that the results were influ- enced by the general characteristics was explored. The chi- square values and degrees of freedom for the tests appear in appendix G. Only associations which were statistically sig- nificant'will be discussed. These are sketched in figure 2. It shows that every general characteristic was associated‘with at least one theoretic variable. The lines represent associ- ations which were statistically significant, no line indicates absence of statistically significant chi-square. "I 83 Figure 2. Summary of relationships between general characteristics and theoretic variables. General Characteristics Theoretic variables number of other positions Self-Esteem Alma Mater Applied Orientation Subject matter emphasis Significance of Immediate Others 2 salary (research Significant Others 7. salary (extension) Immediate Others Tenure \‘..“‘~ :lks to Lay Audiences "searcher Audience Emphasis Publication.Rate Administrative Evaluation Peer Evaluation Because of the large number of statistically significant associations, only those concerning the theoretic criterion variables‘will be discussed. Some of the relationships among general characteristics (appendix E) and between general characteristics and theoretic independent variables‘will be referred to for interpretation. Both subject matter emphasis and percentage salary for research were associated researcher audience emphasis. Re- searchers from convention scientific subject matter areas and those who obtained a high percentage of their salary for 84 research were more likely to emphasize researcher audiences. (These were frequently the same individuals; the chi-square be- tween subject matter emphasis and percentage salary for re- search was statistically significant.) The distributions are shown in table 25. Table 25: Subject matter area and percent salary for research with audience emphasis. Subject matter area Agriculture Science Mixed 1 publications 901 or more 15 .4 67.7 18.2 to researchers less than 901 84,6 32,3 81,8 100 .01 100 .01 100 .01 (39) (31) (33) Chi-square - 26.04; 2 d.f.; p less than .05, two alternative test. 1 Salary research 29? 51-79 80 m 1 publications 901 or more 23.1 25.7 51.7 to researchers less than 901 76,9 74,3 48.3 100 .01 100 .01 100 .01 (39) (35) (29) Chi-square - 7.25; 2 d.f.; p less than .05, two alternative test. @the individual who received no salary from the experiment station was included in this category. 85 Such findings could be easily anticipated, but subject matter emphasis was also associated‘with significant other orientation, immediate other orientation.and frequency of talks to lay audiences. The last two of these variables have already been shown to be associated with researcher audience emphasis. Very few respondents from the agricultural areas named only peers as immediate others or had a high frequency of talks to lay audiences compared to those from the science or mixed subject matter areas. The patterns of interaction corresponded to audience emphasis in agricultural and science but not in.the mixed subject matter areas. However the re- spondents from the mixed areas were unique in their orienta- tion toward significant others; only 121 assigned significance exclusively to peers compared to one-third in agricultural and almost one-half in science subject matter areas. Despite their relatively infrequent face to face inter- action with non-peers, respondents from mixed areas express the significance they assign to these audiences by writing for them. Respondents from agricultural areas direct some of their publications to lay audiences almost regardless of the claimed significance of these audiences to them. The re- searchers in the science subject matter areas seem to have the consistent arrangement; few of them interact with, assign sig- nificance to, or write for non-researchers. The immediate social situation (indexed here as subject 86 matter area) appears to influence audience emphasis in con- junction with interaction patterns and significant other ori- entations. The interdependence among this set of variables may explain the lack of confirmation of several hypotheses concerning audience emphasis. The relationships involved are apparently more canplex than those hypothesized. None of the general characteristics was associated with publication rate at a statistically significant level. Three "age variables" (tenure at M.S.U.. time from Ph.D., and chronological age) were negatively associated with administrative evaluation. These three were positively associ- ated with each other, thus only the association between tenure at 11.8.0. and administrative evaluation is presented in table 26. Table 26: Tenure and administrative evaluation. Tenure 0 - 7 8 - 14 15 + Percentage Lo 23 . 8 l9 . l 54 .5 Salary Med 33 . 3 40 .4 33 . 3 Increment Hi 42 , 9 40 .4 * 12 , 2 100 .01 99 . 91 100 .01 (21) (47) (33) Chi-square - 14.47; 4 d.f.; p less than .05, mo alternative test. 7*rounding error Isl lpupJ. 87 The general tend is for percentage salary increment to decrease with longer tenure. This may reflect stronger compe- tition for younger men or may indicate that administrative evaluation is reflected more by absolute rather than percentage salary incrmnent. The latter possibility could not be investi- gated because absolute salary information is considered privi- leged by administration. Academic rank was positively associated with peer evalu- ation and was the only general characteristic related to this criterion. The distribution is presented as table 27. Table 27: Academic rank with peer evaluation. Rank Non-professor Professor Ember of 0 50 .0 13 .0 Nominations 1 -2 32 .0 29 . 6 Received 3 (+) 18 .0 5 7 ,4 100 .01 100 .01 (50) (54) Chi-square - 22.11; 2 d.f.; p less than .05, two alternative test. This finding suggests that the older researchers have proven themselves and have had sufficient time to achieve public recognition. The fact that rank was positively associ- ated with all three "age variables" supports this interpre- tation. However, none of the age variables were associated with peer evaluation at a statistically significant level. 88 Alternate interpretations would be that higher rank is granted to those who come to be consensually defined as out- standing by their peers or that researchers rely on a public criterion such as rank when asked to evaluate their peers. No data were collected'which could be used to further investigate these speculations. The findings in this section showed the relevance of subject matter area to audience emphasis, tenure to»adminis- trative evaluation (as percentage salary increment) and academic rank to peer evaluation. In the discussion section of the following chapter, both general characteristics and theoretic variables will be used in a reexamination of the hypothesized associations. The intent is to assess the relative importance of various variables aspredictors of publication.activity and evaluation. Chapter IV DISCUSSION, SMARY AND CONCLIBIOIB Discussion Findings already presented demonstrate that variation in publication activity and evaluation can not generally be accounted for with a single predictor. For the most part, these criterion variables were found to be associated with several theoretic variables as well as with several so called general characteristics. The hypotheses, most of which were statements of simple relationships, were generally not con- firmed. Even where the predictions were confirmed for the whole sample (hypotheses 2 and 6) subclasses were located where the trends did not appear or were reversed. The purpose of this section is to explore and discuss more complex relationships involving publication and evalu- ation, in an attempt to account for observed variations. A series of three way tables will be presented as evidence and to facilitate discussion. The section is comprised of four parts. In the first two, publication activity will be con- sidered; in the latter two, evaluation. Audience Rumbas is Respondents who directed 901 or more of their 89 90 publications to other researchers have been shown to be more likely to (a) have a low orientation toward applied research, (b) name only peers as imediate others, (c) give more than three talks per year to lay audiences, and (d) to represent conventional scientific subject matter areas. In table 28, subject matter is used in conjunction with each of the other three variables to ascertain their joint effects on audience emphasis. Most respondents from agricultural and mixed subject matter areas emphasised non-researcher audiences while those from science areas emphasized researcher audiences. The orientation of the researchers toward applied research seemed to have influence on these trends except in the case of agri- cultural and mixed areas, where a high orientation toward applied research was almost invariably associated emphasizing non-researcher audiences. High frequency of talks to lay audiences was closely associated with non-researcher audience emphasis across all subject matter areas. However, low frequency was associated with researcher emphasis only among respondents fron the science areas. Similarly, naming only peers as immediate others was associated with researcher-audience emphasis only in the sci- ence areas while naming a minority of peers as inediate others was associated with non-researcher-audience only in agricultural —-='—“" 91 3 8a 83 A3 83 S: as 38 Go sol...l 2: .863 .8. 2: no. as 3.2: so. 2: no.2: so. 2: No.2: o. 2: «.2 o. 8 o oJlnI o 3 o 8 a“: n Jnolm o.on «8 can :3 «cocoa-32. 2. o. o 98 o. S c. on 98 o. 8 me n. m o6... 3 «8 33:33.5 e um fl». .olm. um um mm um H mm "noaufieoauo 35.. 333g 5 5 a: 3V 3 as as so 33 No.83 No.2: .86.: No.83 3.83 so. 83 so. 2: No.8” 852 III :1! o. o. o...8_..a.. 5.8 «.3 case.-. o 2: 8 H 3 «J. m ZS a8 55 83 22833... 3 o6 «.2 R: 0.2 o6 8 a. n we: «.2 3 Rom 83:33.... a .34 21c mud Hum. 2-.» mud 1m again 3 «managers a: 8 8% use 3:3 3: as Sc 35 $8 33 No.83 no.8” .863 so. 8” so. o2 No.9: mama: Mania- .o...o_.c_. _. o a o «Jim: t: «8 use 83 83888." 8 E «.8 98 9: o. a 93 38 no nos 8333321 a E 3 E on E 3 “Engages 32.5 more: oooouom onsuaoowuw< coa< nouns: uuonnsm .maeanaao oncogene nu«3,ao«uouauauo nonuo escapees“ mam «consumes and on exam» mo hosoooouw .aoaueuaoauo voqaeem moo mono nausea noonasm "mu manna 92 and mixed areas. The findings suggest that there is no obviously best pre- dictor of audience emphasis among these variables. Each one seems to have its greatest effect under certain specific con- ditions of the others. Audience emphasis is not a simple ex- pression of self-concept as an applied researcher or of orien- tation towards others (hypothesised in Hy. 2, 3, and 5). It depends also on interactive patterns (Hy. 6) and on the inne- diate social situation. Publication Rate The joint influence of imediate other orientation and frequency of talks to lay audiences on publication is shown in table 29. Naming non-peers as inediate others and giving talks frequently to lay audiences are associated with high publication rate. Several categories were collapsed to better illustrate the relationship. Table 29: Imediate other orientation and frequency of talks to lay audiences with publication rate. Imediate other orientation Peers Peers Peers Number of Only Majority Minority talks per year: 0-10 111 0-10 1.1; 0-10 11;: Publication 0-7 48.3 50.0 35.1 0.0 40.0 7.7 Rate 8(+) 51.7 50,0 .9 100,0 60,0 92.3 100.07. 100.07. 100.01 100.07. 100.07. 100.07. (29) (2) (37) (13) (10) (13) —_.., b—A 93 Inspection of the citations provided by the respondents indicates that "double publication” is at least partly respon- sible for these findings. However, the extent to which "mul- tiple publication" is reflected in overall publication rate could not be accurately detenmined from the citations. The inspection did suggest that multiple publication was more common among respondents who named non-peers as inediate and frequently gave talks to lay audiences than.mmong those who named only peers as inediate and seldom gave talks to lay audiences. To some extent, it appears that publication rate is a function of immediate other orientation and frequency of interaction*with lay audiences and that this relationship may be, in.part, attributable to the number of reports (to dif- ferent audiences) a researcher makes from a particular project. Respondents who frequently interact with lay audiences are apparently more likely to see their research findings as rel- evant to both clients and other researchers. The results indicate that interactive patterns were the only predictors of publication.rate. However, these "inter- action" variables were not independent of the immediate social situation (subject matter area). Thus publication rate could be regarded as the outcome of continuous interplay among a number of social factors, although less obviously so than audi- ence emphasis. Peer Evaluation A basic assmption of hypotheses 9 and 10 was that re- searcher audience emphasis would be positively associated with peer evaluation. Evidence presented earlier indicated that the assunption was not tenable. Peer evaluation was however shown to be positively associated with (a) rank, (b) publication rate, (c) orientation toward peers as significant others, and (d) ori- entation toward peers as inediate others. Table I) shows the effect of rank on peer evaluation in conjunction with each of the other three variables. Host of the non-professors who had a low publication rate received no nominations as outstanding researchers while only half of those with a high publication rate were never nominated. Publication rate appeared to have little effect on peer evaluation. Among non-professors, respondents who assigned signifi- cance exclusively to peers were highly evaluated by peers while those who assigned minor significance to peers were lowly eval- uated by them. These trends were not evident among professors. The more non-professors were oriented toward peers as significant others, the more highly they were evaluated by peers. The trends were sharp and consistent. No such trends appeared among professors. Non-professors who named a minority of peers as inne- diate others were very seldom highly evaluated by peers. The 95 88 88 so.ooH No.oo~ came _o.me o.om o.mn o.n o.o« .um ”mm Amv Amuv No.2: «at? nunm . Aaav Ache so.ooa so.ooa a m. w. 1 °1 “’1 Q0 0 Int-IN Fan-n n H “VD“ HMQ m .m .0 MIN .Sum + tn H nomaomoum Aeav No. Go." Aaav so.ooH .ea .o .mu noun 8| Amnv .so.ooH h I o “one so.ooH nouns nouns—Sue AHNV Amav so.ooa sue.ooa nllnlo . mu m . m +n coaraooom Mann QeHn NIH Qgflflaflgz H.mm ~.no o «o noeaez HM mm " coauouaoauo uofio ou Soon—EH Aauv Aeav so.ooa asn.ooH Nowd OoQN A+VM QflFflGOGM a.o¢ m.ua o we engage HM MM «QOfiuau—UHRO ROSUO udafimdflwflm Amuv A333 .862 .8. 8H “mum-ml .walmwl +m voauooom 3.9“ n.en ~-H uaonu.aaaoz n.on o.o o no nonasz “mule. Mud “ouch doggone—em wommomouenmoz downs—Seas noon 5.; souuwuoouao nonuo commune-5 use .aowuoumouuo Henna uaooawfimwam .oueu cannons—a use Jae...— oqaooeom «on 03mm. 96 association did not hold for professors. These data show that the associations between peer eval- uation and publication rate, significant others and imediate others were present only among non-professors. these variables do not seem to account for variation in peer evaluation of professors. Non-professors who wish to be highly evaluated by peers would seem well advised to restrict their interaction and assigment of significance to peers and to generate a high publication rate. The ease with which this may be done is likely influenced by subject matter area, because this was associated with both significant and innediate other orienta- tions. Administrative Evaluation In previous sections, the lack of dependence of per- centage salary increment on publication activity was demon- strated, contrary to the asstmptions implicit in hypotheses 11 and 12. The associations bemeen salary increment, the "age" variables and orientation toward innediate others were discussed. The question was raised regarding the utility of percentage salary increment as an index of administrative evaluation in this social system, in light of its close asso- ciation with tenure. Use of the publication variables in 3 way tables in conjunction with imediate other orientation failed to provide clear evidence of the relevance of publication activities to 97 salary increment. In a few instances, there seemed to be some slight tendency for high rate and researcher audience emphasis to be associated with higher salary increments. The table dealing with tenure and orientation toward imediate others displayed much more distinct trends. The distributions are presented in table 31. Percentage salary increment decreases as length of tenure increases. In the middle and long tenure categories, part of the variation in percentage salary increment appears attributable to the extent of orientation towards peers as imediate others. In the middle tenure category, more than half the respondents who fell in the high percentage salary class named only peers as imediate others. Among respondents employed 15 or more years 801 of those who named a minority of peers as inediate others received low percentage salary increments compared to 501 or less in the other categories (of imediate other orientation) . Orientation toward peers as imediate others appears to compensate at least to some extent for increased temre. Apparently the orientation need not be reflected in publi- cation activities. Respondents from the various subject matter areas differed in orientation toward imediate others. It seemed possible that the apparent influence of orientation toward others on salary increment could be as well explained in terms of preferential administrative treatment of (or 98 83 25 AS 8.2: 3.2: 3.2: o.o« ~23 n.mm 9.8 «23 o.on um om 32:: + In ,4 5 $3 $5 No. 00.— No. 8." 80. con o. mu 0. mu m _ mm mama m.n¢ 0.8 mama «.mu no am cm 3:! .Sum obsess 5 so 85 3.2: .82: No.2: n u co m u an 0. ca .E uses—sauna «.mm m. am can ea: 28:3. o. o c. nu c. on 3 omegaoouom HM 11M .mM «noun—3:320 .ll uoauo synapses—H To 50.33725 «aquauumaaqavo 5.!» aoaumamoauo .350 313055 one shaman. «Hm 0.3mm. 99 emphasis on) certain subject matter areas. The distribution for rank (associated with 3 age variables), subject matter areas and salary increment can.be examined in table 32. There appears to be a clear tendency for nondprofessors from science and mixed areas to receive higher percentage in- crements than from agriculture. On the other hand, professors from agriculture appear more likely to receive low percentage increments than those from the other areas. The available data provide no firm basis for inferring whether these appar- ent trends are a deliberate administrative emphasis or merely an unconscious response to "non-normative" orientations. The 3dway table of subject matter, immediate others and salary (not presented) provided some evidence that immediate other orientation is related to salary in agriculture and science but not in the mixed subject matter areas. In any event, it seems clear that percentage salary increment is a measure of more than‘age or tenure and need not be abandoned as a mea- sure of administrative evaluation. The best predictors of (percentage salary increment'were immediate other orientation and length of tenure but each of these were related to several other variables. *** Self esteem and significance of immediate others were not predictive of publication rate (hypotheses l and 7) even when other variables were used as controls. The confounding 100 AmHV .263 .34: o.mm m.hN vouflz nouns wmamssoaa Aa:v AwHV AnHV Am:v Acme No.oo: aa:.oo: so.oo: so.oo: so.oo: «.3 2.3 «.3 «.2 0.3 E ”.5338: m.nn m.au o.o~ m.om o.n¢ ea: muu:um «.mu e.nn h.o« e.n: o.om on unaueoonom ouaoaom enouflaownm« mead: oonoaom ousuasoummfi «some nuance uooaosm Hommomoum nommomoumueoz acne .coqumsao>o o>auouumacuaom nu«s_uone hounds uoonoaa use anon oaaoomo< “an manna 101 of overlap with other variables seems responsible for its failure as a predictor. The only explanation that can be offered for the complete lack of performance by self-esteem, is that the generally high level of self esteem in this sample56 of established researchers, makes the variable irrelevant. 56Self-esteemhas been found to be a useful predictor (or correlate of) role performance in other studies, e.g. Morris Rosenberg, Society and the Adolescent Self-Image, Princeton university Press, Princeton, New Jersey, 1965, pp. 206-239. Summary This study was exploratory in several respects. The central topic - publication activity - has been little ex- amined. The subtopic - relevance of publication activity to evaluation - has been researched even less. A frequent criticism of symbolic action theory is that its utility has not been.demonstrated for empirical purposes. Social scientists who have conducted empirical studies in this tradition.have generally emphasized self and/or other vari- ables to the virtually exclusion of the interaction concept. In this study, these several classes of variables were employed together in an attempt to sort out their relative utility and redundancy. Although industrial scientists have been.recipients of considerable attention from social scientists in recent years, researchers at agricultural experiment stations have been.vir- tually ignored. For these several reasons, there was little background methodology innwhich to base the study. So certain.method- ological innovations'were developed and a few methodological comparisons were built into the study. These innovations were made‘with awareness of the high failure rate of such attempts. Because of relative virginity (most everything is 102 103 relative these days) of the area, the hypotheses were re- stricted, for the most part, to apparently obvious and simple relationships. Despite intentional conservation in formu- lating the hypotheses, most failed to be confirmed at statis- tically significant levels. Such a high failure rate could readily be attributable to inadequacies of (a) conceptual- ization, (b) operationalization and measurement and (c) control. Reanalysis was undertaken to locate the variables asso- ciated with each criterion.and to explore more complex rela- tionships than those originally hypothesized. The reanalysis indicated that two interaction variables (number of peers talked to frequently about research and number of talks per year to lay audiences) had major significance in accounting for variation in publication activity (rate and audience emphasis). The relevance of the social situation (indexed as subject matter area) and social status (rank) to publication behavior was also demonstrated. The interactive effect of both classes of variables was explored and it was concluded that both sets must be used simultaneously in order to Obtain.a useful understanding of the process of source be- havior. The dependence of peer evaluation on total role per- formance was inferred from its associations with a set of mixed variables. Two of the most powerful predictors of peer evaluation.appeared to be immediate other orientation and rank. 104 It was also demonstrated that administrative evaluation - indexed as percentage salary increment - was a function of interactive patterns as well as of tenure. A considerably larger study is required to confirm the assertions made in this smary. The small sample size of the present study prevented the use of standard statistical tests of the more complex relationships. Gonclus ions Audience Eghas is Researcher-audience emphasis is most likely to be found among researchers who: -are located in conventional science subject matter areas; do not define themselves as applied re- searchers; claim mainly peers as imediate others; and seldom if ever give talks to lay audiences. The extent of the in- fluence of any of these variables on audience emphasis ap- peared to be determined by the particular values of the others. The audiences to which researchers direct their publi- cations are a function of their situation, their status and their interactive patterns. The findings strongly suggest that self conceptions and orientation toward others are, by themselves, not adequate predictors of audience emphasis. In this study some of the complexities were uncovered and the relative importance of certain variables to audience emphasis was speculated upon. 105 Publication Rate In this study, high overall publication rate was posi- tively associated with the number of talks to lay audiences and to the proportion of non-peers named as immediate others. These interactive patterns were found more frequently among respondents from agricultural and mixed subject matter areas than among those from conventional science areas. Situation, and interactional patterns appear to deter- mine the number of audiences available to or perceived by the researcher. Multiple audiences are associated with "multiple publication" which tends to make for high publication rate. Overall publication rate was not predictive of either peer evaluation or percentage salary increment. The number of journal articles and book chapters and the manber of "other" types of publications might be more useful indexes of publi- cation rate for predicting evaluation. Peer Evaluation: What constitutes an "outstanding researcher” seems to have many facets. Interactional pat- terns, rank and significant other orientation appear to be components of total performance reflected in peer evaluation. Generally, high peer evaluation seems to depend on assigning significance to peers and acting out this orientation in all things. Interacting with or writing for clients is apparently not expected of (or found among) outstanding researchers. Most professors were highly evaluated regardless of the 106 proportion of peers named as immediate others or the signifi- cance assigned to peers. hang non-professors, respondents who were oriented toward peers as significant others and as imediate others were more highly evaluated by peers. Possibly the use of higher cutting points on peer evaluation categories for professors would uncover the same relationships as were found among non-professors . Administrative Evaluation Although temre was the major determinant of percentage salary increment, orientation toward immediate others accounted for many of the exceptions to the association between tenure and salary increments. Short tenure and orientation toward peers as immediate others were associated with high salary in- crements. Publication activity seemed to be of little direct relevance to administrative evaluation. Questions for Measurement Self-Esteem Index: The attempt to develop a measure of self-esteem was successful from the standpoint of internal con- sistency of the scale items. But it is silly to praise a meas- ure which was not associated with other variables. Work on this problem could proceed on several fronts: use of the scale in its present form with a more heterogeneous sample; develop- ment of a system of item weightings (from total sample item 107 ranks) to refine the measure. Several of these options could be explored from the existing data. A decision to pursue any or all of these alternatives would assume a positive answer to the question: is it worth it? Such an answer is not obvious in view of the apparent predictive power of other variables. Further efforts with the concept of self esteem among agri- cultural researchers could be justified on.methodologica1 or theoretic grounds.57 Significance of Immediate Others: Some measure of the relative significance of immediate others should be a useful predictor of role performance. It might also have utility in specifying the flexibility of specific roles and as distin- guishing characteristic of social systems. The performance (or lack of performance) the measure used in this study does not prove the impotence of the concept. Three alternate ap- proaches immediately seem plausible: investigation of over- lap within specific categories of others i.e. peers, nonspeers; consideration of the frequency of interaction in the operation- alization of the concept; and utilization of (a) category of others, (b) location of others and (c) frequency of inter- action in pursuit of localite-cosmopolite typologies. The existing data would be sufficient for exploration of these possibilities. Until such exploration.has been done, the 57Rosenberg, Morris, 92. gig. 108 concept of significance of imediate others should perhaps not be dismissed. Basic-Applied Orientation: The work done in this study ‘was a beginning step toward the explication of the basic-applied distinction. Considering the overall results, it might be in- ferred that the measurement of basic-applied orientation may not have great utility. Interactive patterns and situational variables are more easily observable (therefore measurable) and appear to be better predictors of publication activity and evaluation. Scaling basic-applied.may simply not be worth the trouble.58 Questions for Research 4A set of interactive patterns which were not considered in this study, but which may have considerable relevance are those involving administrative personnel. Possibly the inter- active patterns of administrators (or those they allow or foster among their staffs) would account for some differences herein labelled "subject matter area". It might also be interesting to develop a rank/tenure index of administrative evaluation. Such an index might be used to evaluate adminis- trative evaluation. Such an index might be used to evaluate 58 Goldberg, L. C. et 3;, 22, c t. Gordon, Gerald, 22, c t. 109 administrative performance as well as that of researchers. A larger study is needed to rigorously test and explore the many associations and relationships uncovered or suggested in this study. If a less restrictive criterion were used for identifying respondents, sociometric techniques might be use- fully employed. A.larger study would necessarily involve the investigation of researchers at several colleges of agricul- ture. This would allow comparisons of interactive patterns between institutions and between conventional administrative units. The consistency of interactive patterns within spe- cific disciplines and across the other classifications could be explored. A.further research implicatinn is the general relevance of the primary group and interactional variables in the study of communicative behavior. It should not be necessary to con- tinually rediscover the primary group and interaction as crucial variables. Investigations of the relationships be- tween face to face communication and reading-writing behavior should be especially productive. The need for medhodological refinements of the inter- actionist approadh was clearly demonstrated in this study. Hopefully, the attempts made here will encourage others to try. Igllications for Practice Despite the small sample and absence of statistical 110 verification, tentative inferences can be drawn from this study for several categories of people. Administrators in the college of agriculture may wish to reflect on the fact that a high proportion of their estab- lished researchers (the core of the knowledge generating unit) are oriented toward peers and seldom write for or talk to client audience. They may also wish to consider that a client orientation appears to be associated with low peer evaluation and in some situations with lower salary increments. They may also be interested in the fact that the 14.8.0. graduates they hire almost invariably express a dual orientation toward basic and applied research but that whether the researchers publish for both client and researcher audiences depends largely on their interactive patterns. The researchers interviewed may be interested in exam- ining their own claimed orientations and actual performances in lidxt of the findings. Researchers who are considering employment in the college of agriculture might ask whether the situation they are contemplating will in fact permit them to express their personal orientations and still provide the evaluations they wish to receive. Extension personnel may ask - how can these findings help me get more information from more researchers? The answer would seem to be - get them involved, talk to them, invite them 111 to meetings, let them know you need them and appreciate their efforts. It might also be wise to reserve judgment as to whether they've "got anything to offer" because it may take some time for them to understand your concerns and then.see their findings from this standpoint. In summary, we would en- courage extension workers to be patient, to be available and to be persistent. The researcher who interacts with you is decreasing his chances of being nominated as an."outstanding researcher". Implications for the Theory According to Mead, interaction is the basic social process. The key to the development, maintenance and modifi- cation of shared expectations is role-taking. Interaction is prerequisite to role-taking. The resulting shared expectations form the social mattrix'which allows for the continuance of interactive (social) behavior. In recent years, interactionists have focused in- creasingly on dimensions of self and others (reference rela- tionships). Perhaps the focus of attention has been unneces- sarily restricted in the struggle for operationalization and measurement. Perhaps too, modest success in measurement prompts a researcher to expanded efforts to demonstrate the worth of (what is by then) his concept and instrument. In any event, the broader perspective was gradually 112 discarded and the more central concept of interaction.(which is less obviously amenable to measurement) was neglected. In this study an.attempt was made to use self, other and inter- actional variables together in the face of measurement and methodological difficulties. While the main efforts at meas- urement were devoted to self conceptions, the crude measures of interaction still allowed this concept to emerge as the single best predictor of publication activity and evaluation. In the discussion of role-taking, interaction and come munication (p. l9-20), it was stated that a writer must assume the standpoint of the reader toward the symbols in order to communicate‘with the reader. It is implicit in the statement that an individual must have interacted with members of the audience in order to take their role whennwriting for them. In this study, it was found that researchers write for those they interact‘with frequently concerning the topic, and almost invariably fail to write for audiences that they do not inter- act with concerning their research. This is interpreted as strong though indirect confir- mation of the theory. More specifically it is a reminder of the relevance of interaction and role-taking to studies of mediated communication. BIBLIOGRAPHY Merican Behavioral Scientist, special issue on Information Retrieval in the Social Sciences: Problems, Programs and Proposals, edited by Ted Gurr and Hans Panofsky, VII, 10 (June, 1964). Merican Psychological Association, Reports of the American Psychological Association's Project on Scientific Information Exchggge in Psychology, I (December, 1963). Barnett, Homer G. _Innovation: The Basis of Cultural Change, New York: McGraw Hill, 1953. Ben-David, Joseph. Scientific Productivity and Academic Organization in Nineteenth Century Medicine. American Sociological ReviewJ _2_5_, 6 (December, 1960). Blackwell, Gordon W. Multidisciplinary Team Reaearch. Social Forces, 1:1, 4 (1955), 367-374. Brownson, Helen 1.. Research on Handling Scientific Informa- tion. Science, £33, 3440, 30 (December), 1960. Caplow, Theodore, and Reece J. McGee. The Academic Marketplace. New York: Basic Books, Inc., 1958. Case Institute of Technology, Operations Research Group. Cleveland, November, 1958. Couch, Carl J., and John S. Murray. Significant Others and Evaluation. Sociometry, 21, 4 (December, 1964), 502‘5090 , Mason E. Miller, and John S. Murray. Comunicative Behavior of Agricultural Scientists. Institute for Extension Personnel Development, Michigan State University, East Lansing, Michigan, September, 1964. I DeGre , Gerard. Science as a Social Institution. New York: Random House, 1955. Glaser, Barney G. The Local-Cosmopolitan Scientist. Anerican Journal of Sociology, LXIX, 3 (November, 1963), 249-259 113 114 Glaser, Barney G. Variations in the Importance of Recognition for Scientists' Careers. Social Problems, 19, 3 (Winter, 1963), 268-276. Goldberg, L. C., F. Baker, and A. H. Ruberstein. (Local- Cosmopolitan: Unidimension or Multidimensional. American Journal of Sociology) LXX, 6 (May, 1965), 704-710. Gordon, Gerald. The Organization.Designed to Produce Change. A'working paper presented at the Seminar on Innovative Organization, University of Chicago, April, 1964. Guilford, J. P. Creativity. American Psychology, 2, (1950), 444-454. Hawkins, H. Stuart. Influence of Role Commitments, Perceived Role-Conflict, and Gregariousness on the Publication Behavior of Agricultural Scientists. MQA. Thesis, Michigan State University, 1964. Katz, Elihu, M. L. Levin and H. Hamilton. Traditions of Research on the Diffusion of Innovations. American Sociological Review, gg, 2 (April, 1963). Knoblauch, I. C. Basic Research at State Stations. Science, 130, 3389, (December 11, 1959), 1639-1640. , E. M; Law and W. P. Meyer. State ricultural Eyperiment Stations. United States Department of Agriculture, Miscellaneous Publication No. 904, May, 1962. Kornhauser, William. Scientists in Industry: Conflict and Accommodation, Berkeley, Calif.: University of California Press, 1962. Kuhn, Manford H. lMajor Trends in Symbolic Interaction Theory in the Past Twenty-five Years. The Sociological Erterly, g, 1 (1964), 61-84. , The Reference Group Reconsidered. The Sociological arterIZ, i, 1 (1964), 5'22. Mann, Floyd, and Rensis Likert. The Need for Research on the Communication of Research Results. Human Organization, Mead, George H. Mind, Self and Society, Chicago: University of Chicago Press, 1934. ‘ _ l O \ 115 Menzel, Herbert. Flow of Infonmation on Current Developments in Three Scientific Disciplines. Columbia University, Bureau of Applied Social Research, Reprint No. 232. . The Information Needs of Current Scientific Research. Columbia university, Bureau of Applied Social Research, Reprint No. 333, 1961. Merton, Robert K. The Role-Set: Problems in Sociological Theory. British Journal of Sociology, 8, (1957), 106-120. Pelz, Donald C. Some Social Factors Related to Performance in a Research Organization. Administrative Science garterly, I (1956), 310-325. . Interaction and Attitudes between Scientists and the Auxilliary Staff: I. Vinoint of Staff. Administrative Science Brterly, 4, 3 (December, 1959), 321-366. . Interaction and Attitudes between Scientists and the Auxilliary Staff: II. Viewpoint of Scientists. Administrative ngrterly, 4, 4 (March, 1960), 410-425. Rogers, Everett M. Diffusion of Innovations. New York: Free Press, 1962. Rosenberg, Morris. Society and the Adolescent Self Image, Princeton University Press, Princeton, New Jersey, 1965. Sarbin, T. Role Theory, in Gardner LindzeL: Handbook of Social Psychology, I, 223-258. Reading, Mass.: Addison- Wesley, 1954. Senior, Joseph K. The Vernacular of the Laboratory. Phil. S_c_i_.. _22, 3 (July, 1958), 163-168. Shepard, Herbert A. Nine Dilemas in Industrial Research. Administrative Science Qarterle, I (1956), 295-309. . Patterns of Organization for Applied Research and Development. Journal of Business, _2_9_ (1956), 53-58. . Superiors and Subordinates in Research. Journal of Business, 22 (1956), 261-267. Sjoberg, Gideon. Science and Changing Publication Patterns. P1111: SCI... _2_§’ 2 (April, 1956), 90-96. '—r~:.— __x.v—‘ 116 Slichter, Charles S. Science in a Tavern, Madison, Wisc.: University of Wisconsin Press, 1958. Stewart, D., and T. Hout. A Social-Psychological Theory of the Authoritarian Personality. Merican Journal of Sociology, 65, 3 (1959), 274-279. Storer, Norman W. Research Orientations and Attitudes Toward Teamwork. LLB. Transactions of the Professional Groups on Engi_neerigg Management, E.M-9, 1 (March, 1962). . Science and Scientists in an Agricultural Research Organization: A Sociological Study. Cornell University, Ph.D. Thesis, 1961. Strauss, Anselm L. The Social Psychology of George Herbert Mead. Chicago: University of Chicago Press, 1962. , and Lee Rainwater. The Professional Scientist. Chicago: Aldine Publishing Co., 1962. Terman, Lewis M. The Discovery and Encouragement of Excep- tional Talent. Amrican Psychology, (June, 1954), 221-2300 Turner, Ralph M. Role-Taking: Process versus Conformity. Rose: Human Behavior and Social Processes. Boston: Houghton Mifflin, 1962. . Role-Taking, Role Standpoint, and Reference Groups Behavior. A__merican Journal of Sociology, §_]_._ (1956), 316-328. Walker, Helen M., and Joseph Lev. Statistical Inference. New York: Holt, Rinehart and Wins ton, 1953. Weaver, Warren. Science and the Citizen. Bull, Atomic Sci., 1;, 10 (December, 1959), 361-365. Znaniecki, Florian. The Social Role of the Man of Knowledgg. New York: Coltmbia University Press, 1940. . ‘l'ullllllllllil .I$AI. . 9 0‘ a . _ _ . r. e A , . _ u a o I O I as o- .\ 0 A t . I . . . \ ( au .1 . - . . , . . a . ‘ C e a a _ . C ( _ . Ix . I _ , a O x , C . ( O Appendix A (Interview Schedule) SCIENTIST COMMUNICATION STUDY Project Number (1-3) Phase Number ( 4 ) Department and Subject Number (5-8) Locale: Building ' Room Time : Hour Date (9-11) * a a First of all, a few vital statistics.... Q 12. How old are you? 1 - under 30 6 - 50-54 2 - 30-34 7 - 55-59 3 - 35-39 8 - 60-64 4 - 40-44 9 - 65 and over 5 - 45-49 0 - Q 13. What is your academic rank? . Assistant Professor . Associate Professor . Professor «known- Q 14. What is your subject matter specialty?...that is...when you introduce yourself to persons in your discipline, how do you define yourself? Q 15. When did you obtain your doctorate? . ~.r.—- f-"‘»‘" m Q 16. Q 17. Q 18. Q 19. Q 20. Q 21. Q 22. Where did you obtain your doctorate? 1. M.S.U. 2. Other Land-Grant 3. Other 4. Did you have professional employment before coming to this station? 1. Yes 2. No If YES, ask: Q 18 If NO, skip to: Q 27. What kind of work did you have before coming to this station? 1. Research 2. Teaching 3. Research-Teaching 4. (Other) With what type of agency‘were you employed? 1. Other Experiment Station 2. Industry 3. U.S.D.A. 4. College of Agriculture 5. (Other) How long did you work there? 1. less than 2 years 2. more than 2 years, less than 5 years 3. more than 5 years, less than 10 years 4. more than 10 years Did you have other professional employment prior to the position just discussed? 1. Yes 2. No If YES, ask Q 23. If NO, skip to Q 37. What kind of work was it? 1. Research 2. Teaching 3. Research-Teaching 4. Other Q 23. With what type of agency were you employed? 1. Other Experiment Station 2. Industry 3. U.S.D.A.__ 4. College of Agriculture 5. Other Q 24. How long did you work there? 1 less than 2 years 2. more than 2 years, less than 5 years 3. more than 5 years, less than 10 years 4. more than.10 years Q 25. Did you have other professional employment prior to the position just discussed? 1. Yes 2. No If YES, ask Q 26. If NO, skip to Q 27. Q 26. How many other professional jobs have you held? *** Now I'm going to read some statements that Agricultural Experiment Station Researchers have made about themselves and their research. PROVIDE CARD A Agree - Disagree 1 - just a little 2 - quite a bit 3 - a great deal For each statement I'd like you to indicate how you feel about it, that is, how true it is in your own.particu1ar case. When I read each statement, first, tell me whether...in general... you agree or disagree with it...and than.tell me a number...one, two, or three...to indicate how much you agree or disagree. Q 27. I try to do research which will make an original contribution to the storehouse of knowledge. Q 28. The research I do may directly affect a large number of people in a way important to their welfare. Q 29. It would bother me to have other scientists in my field (specialty) regard my research as insignificant or inconsequential. ,_ .Olv" Q 30. My research deals with fundamental scientific questions. Q 31. I try to do research which will provide practical solutions to "real" problems that people have. Q 32. The research I do Lg regarded as important by other scientists in my field (specialty). Q 33. It would bother me to have nonscientists regard my research as impractical or ivory towerish. Q 34. My research deals with innediate problems of Michigan producers-processors-consumers. *** The next questions concern those persons who are important to you as a researcher. I would like you to tell me who some of these persons are....and then a little bit about them. *** Q 35. Whose opinion of you as a researcher is important to you? (request 5 names) 1. Now consider that the amount of importance of these persons to you as a researcher is equivalent to 15 points. Distribute or assign the 15 points among these persons according to their relative importance. *** (The names provided are now written at the top of "Significant Other" sheets, and the remaining information solicited for each.) *** Q 36. Q 37. Q 38. Now, considering all the researchers you know in your discipline... PROVIDE CARD B How do you rate yourself on competence as a researcher? What number? Where 1 means among the few least competent and 9 means among the few'most competent? *** Now a few questions about the writing you do as a researcher. *** First, considering all the things you do as a researcher, how do you feel about writing research reports? Dislike most of all Dislike somewhat Not different from other activities Like somewhat Like most of all Again, considering all the things you do as a researcher, how difficult do you find the writing of reports? Most difficult activity of all Somewhat difficult Not different from other activities Somewhat easy Easiest activity of all *** Now let's consider for a moment the set of persons you deal with most frequently in your work as a researcher. *** Q 39. Q‘40. Q‘41. Q 42. Who are some of the people you talk to frequently about your research? (elicit five names) 1. *** (The names provided are now written at the top of "InInediate Other" sheets, and the remaining information solicited for each.) How frequently do you give talks or lectures to groups of non- scientists (non-students)? No. .per year Which professional journals do you subscribe to? If more than 5, give total number . Which non-professional periodicals (farm magazines, trade maga- zines, popular magazines) do you read regularly (read more than two-thirds of the issues). 5. If more than 5, give total number . 0 e a a sssssssssss Ara... LE 1 Q«43. In all groups of people, some gain a reputation for superior performance. Here is a partial list of the researchers at this Experiment Station; who are some, that in your opinion, most deserve the reputation of outstanding researcher? (list ) l. I. 0. 8. Subject Number: Other Named: (as Sig. too: Y N ) What does he do? (elicit primary work role) Research Administration Teaching Other w W Extension If a scientist, is he of your ---- discipline Y N Specialization Y N Where does he work? This Dept. U.S.D.A. 'mis Campus Other Campus Other What is the basis of your relationship‘with this person? How come you talk to him more than to others? Graduate Student Technician Conduct Related Research Supervisor Share Facilities Colleague Other What aspects of your research do you talk about with this person? General Direction Findings Methodology Application to practice How frequently do you communicate with this person? (talk to, phone, write) More than daily Twice a week Daily Weekly Every other day Less than weekly A. Appendix B (Self Administered Questionnaire) S.T.I.T. If you had to choose between reporting your research in.a popular publication where laymen would see it and perhaps use your find- ings, gg.reporting it in,a scientific journal, which would you prefer? . Definitely popular publication Somewhat popular publication I can't make up my mind Somewhat scientific journal Definitely scientific journal an- '-. How many brothers and sisters did you have who were: a) older than you ? b) younger than you ? How important to you in a job is the chance to serve people -- to help solve their problems? Of utmost importance Very important Somewhat important Not very important Unimpor tant Where did you live until you finished high school? Farm Town (less than 1,000 people) Town (more than 1,000, less than 50,000) City Gmore than 50,000) If it ever came to a choice between working on the practical problems of agriculture (problems important to the farmer and processor), g£.contributing to the development of a body of scientific knowledge, which would you prefer to do? Definitely practical problems Somewhat practical problems I can't make up my mind Somewhat scientific knowledge Definitely scientific knowledge 10 Subj. No. Here are some characteristics that have been attributed to good re- searchers. Read over the list. Then, select that characteristic which, in your opinion, is most important for an ideal researcher to have. Enter the corresponding number in the uppermost box. Then indicate the next two most important characteristics by entering the appropriate numbers in the second row of boxes. Then reverse your perspective, and select the least important characteristic for an ideal researcher and enter the number in the lowest box. Fill in the two remaining rows by selecting the next two least important charac- teristics from the remaining five and entering these numbers in the second bottom row of boxes, and then.enter the remaining three numbers in the middle row. It may help to strike out each statement after it has been used. The Ideal Researcher: 1. - keeps up to date in the literature. 2. - is dedicated to the search for knowledge. 3. - sees the other fellows point of view, most is easy to work‘wiflh. important to an ideal 4. - willingly accepts long hours and hard researcher work. 5. - has an ability to choose the important I: [3 questions. D E] D 6. - is fair in exchanging ideas and criticism with others. DEC] 7. - is analytical in that he is able to least sort out relevant relationships. important to an ideal 8. - uses sound methodology. is accurate and research precise in data collection and analysis. 9. - writes up findings with clarity, and with sufficient but not too much detail. S.-R. SUbJ. N0. Now using the same characteristics and the same procedure, indicate how these characteristics apply to Ygg_as a researcher. As before, select first those which apply most, then those which apply least and finally fill in the middle row. H a I I keep up to date in the literature in.my field. 2. - I am dedicated to the search for knowledge. 3. - I see the other fellow's points of view and applies 3 an easy to work‘with. most to me 4. - I am hard working and willing to accept long hours. 5. - I have the ability to choose important 1[:] questions. 6. - Item fair in exchanging ideas and criticism with others. 7. - I an analytical in that I am able to sort out relevant relationships. 8. - I use sound methodology. am precise and applies accurate in data collection and analysis. least tom 9. - I write up my findings with clarity and with sufficient but not too much detail. 0H4 mun-mu H.- I ;:"_——"___.=.d""'—.._ "_— 13 Publications Sept. 1/61 - Sept. 1/64 Subject Number : ***** Title Citation: (usual additional information provided for reference purposes) Authorship: (check one): Sole Author Form: Circle the ggg_most appropriate: Book Chap. in Book Journal Article Exper. Sta. Bulletin Extension Bulletin . Extension Pamphlet . Extension Folder KHFHHE QMMPOU> N. (other - please specify) Coauthors . Conference Proceedings . Popular Magazine Commercial Publication Department Mimeo Other Mimeo Press Release Main Audience: Circle the 22g.most appropriate: A. Researchers - own speciality G. B. Researchers - own discipline H. C. Researchers - other disciplines I. D. Teachers J. E. Extension Specialists K. F . Extension Agents 5" (other - please specify) Other non-research scientists Graduate Students undergraduate Students Administrators Farmers or other non- scientists gm..-’ L. Appendix C Letter of Legitmation February 15, 1965 Memorandum TO: Department Chairmen, Agricultural Experiment Station FROM: L. M; Turk, Director SUBJECT: Study on Communication of Research Finds Jack Murray, a Graduate Student in the Department of Communi- cation under the direction of Dr. Carl Couch of the Institute for Extension Personnel DevelOpment and Dr. H. Kumata of the Department of Communication, has planned a study to be carried out among Experi- ment Station Personnel dealing with communication of research find- ings and social-psychological factors. It is hoped your staff members will cooperate in this study. Only those staff persons who have the PhD and are 50% or more on Experiment Station time will be contacted. The Administrationnwill not know who does or does not respond. All data will be exclusively confidential to the researchers. The reports derived from this study‘will not concern answers given by particular individuals but will deal only with group characteristics. Appointments for the interviews will be arranged in advance, to suit the schedules of the individuals involved. LMT Ila 14 Appendix D Correlation*'mattrix for self-esteem index Item** 2 3 4 5 6 7 8 9 Total 1 .86 .83 .86 .80 .85 .79 .79 .78 .85 2 .84 .85 .78 .86 .82 .82 .80 .86 3 .78 .78 .84 .77 .77 .76 .82 4 .79 .86 .81 .82 .81 .86 5 .81 .87 .89 .89 .94 6 .81 .82 .82 .86 7 .93 .90 .96 8 .92 .96 9 .95 *These are pearson-product moment correlations. **The item numbers correspond to those shown in.appendix B. 15 .umou o>euoououao osu .Ho>oH no. as usuoqmwowam osHo> ouaaoauwco a .eouoouu mo moouwoo osu om Hosea sou .osHo> ouuamonweo on» ma ouswum Home: use .Haoo some on Anv N om.o sowmmouxo humane N o N ego.ua wN.m souoomou humane N a N a mmoeuumom uoauo NN.¢ om.o an.N mo panama a H N N oe.o «N.o Nm.o sN¢.o noumz.maa< a N o a . N e¢.o mw.o mm.N NH.N om.H sundae a N a a N a on.m «0.0 ma.o on.HH Hm.m ca.mm .n.nm « a # sown saga a N o a N o a 33 36 3." men 2.0 1.2:} 8.: one a N fl N N H N N N aw.o Ho.o wN.H mm.m mo.w an.ee No.n¢ Ho.0N goon em .1 as as. Human: _ooaoeouxm . couoomom meowuamom Home: ounces .n.nm sud uoofiosm muonom N muonum N noouo uaa< scum says mo .02 Anymowuoauouuouuto anyone» use moose mmaomoowuoaom m 383%. 16 e mm.N a e eo.m mN.N N N No.me mo.o k. a e ow.ne em.N k. a a me.» «N.o a N em.we He.m a a a a NN.N Ne.m N N see c ea H e e oe.o mm.a mooooNoo< uooaouomu ooauosuo>m masseuse Non muoaom 9. 3:3 .uuou o>auaououao osu .Nopoa no. no udooemaomqo os~o> ouasouuNeu a .souooum mo aoouwoo ecu «N nosoa sou .osHop sensuouNco ecu aN shaman Home: sou .Hfioo some aH any N mn.a a N on.e mn.n a N a NH.n mn.N nn.N use a N an.eavN oo.me me.m os.N .1 k k. a N a aNm.oH mo.m me.e N e N am.n «Ne.e Nm.e e N e me.o eo.o ao.n some some masseuse aoNuuuNNasm onmouoowuo> oquouoonu sou wcoao newnmsowuoaom om.n eeawo>o a sensuaae ma.¢ N Nw.N c wa.¢ auscuo «unenoaau uaooemasmam unsouo oaoaouosa muoaum moauosao>m noon «Noonmam masseuse comm soNuooNHoom moauo>o uuoouo ouoNnossH muflnuc uauoauaamam N condone nN.N -oauun a o soouam eo.N NN.n umaom soaeaaa -oauqn 17 on.am k. a oo.n e mN.o N eoo oN a mo.m o Nm.n a ow.m a a no.0“ a N mw.m o ~H.H nouns: soonssm k. oonsouNm meadow N H no.5 N mN.N N on.o a oo.m N on.H N mm.~ N nn.o N om.H H om.o N 0N.N L‘- .». A. "I' 'l t l I ‘\ a NH.oN a o 05.0 o o¢.¢ N nN.N a a No.m a wn.o e m¢.N a oo.N N do.N a NN.m souoomom museum N Ase .uoou o>auoauoudo osu .Ho>oa no. as unoonNawNm osao> manomenwno a .aooouum mo moouwoo on» u« Hosea sou .osHoe owoswaneco sou we shaman women one .HHoo some an any a QN.N a Nn.m e mo.N N eo.N a mN.N a oo.NH a NN.H a m¢.N N oo.~ a No.0 «soauaaom Hoouo mo .02 moannauo> oquuuooou one oouumuuouoouooo Housman cassava mmNomsoquoNum N wH.o N mN.o N ow.¢ H mm.o N mw.H N mm.HH .1 N mn.a N Hw.a H oo.o N mN.H noun: mad< e o¢.¢ a no.¢a * a NN.N N «N.o a Nc.N o no.0 o wn.n a NN.¢ N mm.N a o¢.¢ ousmoa o neesoaae a om.a a wm.mH u? a HH.w N mm.o a Ho.m a ww.N a om.N a mo.m N mn.o o ao.u CQOE some «sea a wm.n a ¢¢.ma a No.m N no.0 a HH.m a mo.H a mm.H a *NN.oH N nu.o a mN.N owe N da.m N ¢¢.¢ N «as NN a mm.o N nm.m N on.¢ N mm.~ N m¢.m N No.0 N Nw.m xamm mouooqoso has on exams udoaouumH meadow moauoo~o>m woom neaoomau ooooNon< ovum sowuooNHosm mo~uo>o muonuo omnivanH muscuo usaoameamam eoaaaa< uoNaum assumm umnom 18 ”'Tfiifllfflfillilflifl‘lflmi’flfliijfijljlifiwflilifflflfimfl