';""§4;;'.:‘-.'€Li~}$‘_l .3424, 44 .:E.3§E' ‘; -‘a‘. . 4. .4 :4: , ‘I .~.::‘ 4.144744 , d at. MN K .4, "-9.: 3'3"? WT-‘WIWT :I Iv 3' ‘rII:{L343'14_ .4 I 44.4444. I (M. 3I\-141444.4444 (444:3C JAY“, 4'4. ”chum: X” "Fix-'1;- 4. 3474 I L 3.4... .. 33‘1‘1‘33 ul‘ .{J I. I' my 4 I. ‘I 44444» “If, {£31. I. 44444444 3 v 333%{341‘44'u3m J. ER 4 . ‘1444444444 1. 4:24: .1441 44444 {1.3 4454-44431 14444414.; 444444; 44n4444444$4 14".?in 444mb ‘44‘34544444 44.544 4.244% I 333333-11" 41.} 30144 . «dry; $1} 3&3ng 3:44 ‘f-hui -I | 401% :4. 434344 ,1. (u «a s “gym. ab “ “I‘VM‘h mg 4.454..- 44 , fi‘ Azéi‘pf'z" ‘7- .a, ‘32. 4192;. IV: .. LII-5‘ w:r—J V144 , “4% 4 ”I”? . 3;; 13- ‘S'I‘ff 4' ‘ - .: - igzggfig. 4 .-_4.4._£«:4'.‘—=..;:s42'4 ,3, 444* 1;: _'z. {.443 @4354 3 . III-414 324 ‘ : Eq’ng“ “A; 3314:3411. 3% u ' W33! ‘13 .33 $53.3 33353-44‘3444-441144 L 3in , ;. 9-1:,- .‘4. 44:59:“ ‘13:“; 444444444 44444 3 3 ‘3 3' C. (3333.3 3144*“:3“L..c:4414 .‘LJ'. 9‘113Ili3'sql'13 :44. ~. 4‘: 4 ytg.4u44444‘1444:4 44 4444444444444 ‘31)“ 3‘43. 335%; 31 44 4 w 4 444. 4 ' 5:44 448544 3314:“ 144% 444444: 3"” 3' 451419-44” ‘I4444%443.4 1%..343IHI‘3‘4 'r-IIIE‘I‘I'. ”I %4 4 «WI ‘3': 3 13931333324 341‘ 3323‘ ' Ln 4, “ 5:314. . 31: 4134:4994} 1;; 4445i $344 4444 4. 21%: 444454444 .4444..If4.44m.‘141§41c1 , 4444,44 " .55 n ,4". 333341;? 4444444444 3331' 43344444 15.4 4 44.4.. 31489245.“: 4444444 . 44 444443444: 1.1:! 331i“! v‘L filer“ 1‘ mquhfih'i €6.34; “44414.4 .4 13314.3 4444544 44444444 #4:" "333134: 3133393 L44 3431333?“ ‘afi‘agh 311333333 33333130: %I $43134 «Vi MM: ”444 $4113“ ‘3: 33‘“ IX” 4441:4451” “cm. KIM»: I afi‘fil‘nk‘i" 44 4444: 333% 751%? am"; 3:313: 1442:4444 3.444 4. i5; «‘44: 4 4 . I?“ 3 " 4&3:th M433“ $3?le ‘ 333% x133 ‘ 3i 31‘ 31M in ‘ waififi 41.44“ 44:34 44.44.4144“? “$4444 ' 44:3an“. 4 , ' ‘33.». . “31.44%“1‘ " 1.3533: In . ,, ' 3k“ 3‘. an": 31444434442§§§3W44 % V5 “ {64?}: ‘ fi3‘fia‘4 $1 :24 we fix“ A. :I. u 31443331 ~32‘ 4443:3124- ..444 4 443:” I fivg‘tgg ‘CCI 4 3&3 “‘4 :54‘44 “3%” 3‘ 23:24-33 ‘1‘“ 3;; 3::‘Ef'55‘32 €95 3% 33‘? 3‘14““ “filfi. I I. “$35“ :§:‘2‘:~ 141 ,- .4 443333444“:wa 1344. ($1.14.?!th $444444 mum-u?“ 1‘11in .53. “4 13¢: 44.41.44 3;; 33313 4;}? w. m £44 4344444kw4411444: If?“ 4‘ .. 4 33% 31.44 G 34:. I 4 I444? 4:44:44 I I?"- 3.1,?“ MICHIGAN STATE UNIVERSITY LIBRARIES l ||I||I|I|| I II 3 1293 10664 3574 III This is to certify that the thesis entitled TELEMEDICINE: ACCEPTANCE VARIABLES IN POLICY RESEARCH presented by Stephen John Danowski has been accepted towards fulfillment of the requirements for _£I°A-—degree in -VM. E‘éiefiigazz “Si/aw fin‘ivee’gitg Ex??? 33 a- - rmrasr'x: E km» v”, ' -.,.., Te lecommunications Major professor Gretchen Barbatsis Date ML 0-7639 MS U is an Affirmative Action/Equal Opportunity Institution MSU RETURNING MATERIALS: Place in book drop to remove this checkout from UBRARms . AID-(SIIIL your record. FINES Will be charged if book is returned after the date stamped below. “;"‘ 0 ‘~11. a 18 '3 1 8; fl TELEMEDICINE: ACCEPTANCE VARIABLES IN POLICY RESEARCH BY Stephen John Danowski A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Telecommunications 1985 To my family ii ACKNOWLEDGEMENTS I wish to thank Drs. Gretchen Barbatsis and Linda Kohl for their tireless assistance in every stage of this thesis. Without them it never would have been written, (and I now know the truth behind thisstatement). I also wish to thank, Drs. Lawrence Schwartz and Martin Wong, as well as the entire staff of the Veterans Administration Medical Center, Battle Creek, for their generous support in the data gathering phase of this investigation. Also, I wish to acknowledge my brother and sister-in-law, Drs. James Danowski and Donna Danowski-Olson, for their special enrichment of this project. Other members of my immediate family, as well as friends, new and old, provided aid in the overall accomplishment of this project. I thank you. iii List of Chapter I. II. III. IV. TABLE OF CONTENTS Tables. . . . Introduction . Definition and Overview. Current Underutilization Systems . . Purpose . . . ‘Summary . . . References . . Theoretic Rationale Overview. . . Introduction . Brief Historical Review. Communication Structure Variables Attitude Variables Research Questions Summary . . . References . . Procedures and Methods Overview. . . Introduction . Research Methodology. Statistical Analysis Procedures Research Questions Operationalized Summary . . . References . . Results . Overview. iv 10 12 l4 l4 14 15 16 19 21 23 24 26 26 26 27 4O 42 43 45 46 46 Response Rate . Identification of Isolates. . Overview of Statistical Procedures Analysis of the Summary . . . References . . Chapternotes . V. Discussion . . Overview. . . Discussion of Research Data. . . . Questions. Suggested Direction for Further Investiga- tion . . Other Interested Patterns Which Emerged Strategies for the VAMCBC . . References . . APPENDICES: Appendix A: Appendix B: Appendix C: Appendix D: BIBLIOGRAPHY . . . . Crosstabulation of Net- work Connectedness by Role . . Mean Values for Media 0 Style Variables. Factor Coefficients for Weighting Lickert Items Questionnaire 58 58 59 64 65 67 69 72 79 81 83 95 LIST OF TABLES Primary Communication and Isolate/Non— isolate Status. . . . . . . . . . Kendall's Tau Summary Table of Correlation Coefficients . . . . . . . . . . Crosstabulation of Network Connectedness by Role . . . . . . . . . . . Mean Values for Media Style Variables . . Factor Coefficients for Weighting Lickert Items. . . . . . . . . . . . . vi 48 50 72 79 81 ABSTRACT TELEMEDICINE: ACCEPTANCE VARIABLES IN POLICY RESEARCH BY Stephen John Danowski This investigation into acceptance variables in telemedicine innovation was conducted with 35 subjects from the Veterans Administration Medical Center, Battle Creek. Respondents were members of a medical team prac- tice including physicians, psychologists, nurses, and social workers. This correlational study utilized vari— from empirical research on the communication of innova— tions. These included structural network variables, such as connectedness and integration, along with an exploratory variable called Media Style. The research questions explored relationships between these variables and an individual's preference for the innovation. Some Network Connectedness variables were significantly related to preference for the innovation at the 95% sig- nificance level. Media Style variables were also signi- ficantly related to preference for the innovation. CHAPTER I Introduction Definition and Overview In order to organize a discussion of telemedicine, the: marriage of technology and the practice of medicine, it appears thematically profitable to follow the functional description put forth by Simon in his Science Q; the Artifi- cial (Simon, 1969) regarding any man;made artifice: 1.) a discussion of its purpose or goal; 2.) a description of its inner characteristics; and 3.) a description of the surroun- dings in which it operates. According to Simon, any manufactured thing is molded by the inner and outer environ- ment and is in fact an interface between the two. Simon's model depicts the outer environment controlling the goals of a system, and the goals, in turn, responsible for the cha- racteristics which the system assumes. In the context of the present investigation of telemedicine, the outer environ-ment is considered to be public policy and the inner environment is considered to be the technology itself. The following discussion will utilize Simon's convenient functional description in order first to characterize historical or existing telemedicine systems based on their purpose. Following this their inner environment, the 2 technology itself, will be scrutinized; and finally the outer environment, or public policy dimensions, will be delineated. System Objectives The objectives for which telemedicine systems have been designed range from uses for diagnosis of patients either directly or over a distance, for consultation between health care personnel, for education--either formative or in- service training, or for administrative functions of a health care organization. 8 The primary distinction which may be drawn between the purpose and objectives of any telemedicine system, regard— less of its design components, appears to be whether its intended usage is for direct diagnosis by the physician or care practitioner, or whether its purpose is more generally for communication with the possibility of substituting for hands—on contact by the physician. The distance to be bridged may be a few feet, a city block, or even vast reaches of outer space. In an extreme View formulated by Maxmen (Maxmen, 1976), computer branching theory can be utilized by a computer-telemedicine system for diagnosis and prescription of treatment with the displacement of the phy- sician altogether. 3 Inner Characteristics: the Technology Following Simon's model, the inner characteristics of a telemedicine system are considered for this paper to be the technology itself. The technology of telemedicine ranges from sephisticated and costly machines used for direct diag- nosis and laboratory testing to hybrid telecommunication systems with video, telephone, and computer components. An example of an everyday, yet nonetheless complex, telemedi— cine system is the telephone, a staple of the medical practice for consultation, such as between a physician and pharmacist. Examples of telemedicine systems date as far back as Marconi when ship to shore radio was used for medical emer- gencies at sea (Williams, 1984). Video technology made its entrance into the field of telemedicine in 1959 when Dr. Cecil Wittson bridged a distance of a few feet across a street at the University of Nebraska School of Medicine for the purpose of psychiatric consultation (Wittson, 1965). Since then, video technology has assumed a variety of roles. To cite but a few examples, miniature video cameras are used directly by the physician for orthoscopic surgery. In the last ten years, the proliferation of cable systems in this and other countries has provided a means for health informa- tion to be distributed through a videotex medical database. The ongoing educational needs of the medical profession in some cases are served by professional journals stored on video tape or disc. The AT6 Satellite and attendant video 4 technology has also been used for educational purposes as well as for linking physicians with patients in remote areas for diagnosis (Brown, p. 23). Computers have been used for all housekeeping tasks of the health center, from patient check-in to billing (Brown, p. 13). In addition to analog and digital cable linkages, other means of distributing telemedicine information include microwave, satellites, and cellular radio. Outer Environment: Policy Issues Referring back to Simon's model describing the science of the artificial, the surrounding environment of the tele- medicine system is considered in this paper to be public policy regarding health care, especially telemedicine. This latter point is postulated on the assumption that matters affecting the quality and cost of health care are not a private concern but instead one shared by most of the mem- bers of society. The public policy environment in which telemedicine systems function, at least in the United States, appears to be preoccupied for the most part with the cost and equitable distribution of health care, especially as the pOpulation of rural areas may be impacted (Brown, 1983). Several projects have adequately demonstrated the ability of telemedicine systems to serve these populations with high quality care, but what remains uncertain is whether these programs have been cost effective (Park, p. 32). Another public policy consideration is the feasability of maintaining a nationalized system of health care. It has 5 been suggested that the computerization of medical records offers a means for managing health insurance aspects of such a nationally coordinated system. A secondary policy issue regarding telemedicine is the programming of the many new channels resulting from the proliferation of cable systems and other means of distributing information (the new rela- tive abundance of electronic spectrum channels). The new ability of these means to distribute health care information is tied to a preventative mode of health care in which the physician is not necessarily the center. Elton has summarized the policy issues surrounding telemedicine as follows: Financial (funding implications); Privacy (e.g., access to computerized medical information); Diffusion of Responsibil- ity; Personnel Supply Implications; Compatibility (its advantages versus those of diversity for the sake of compar— ison); Flexibility (e.g., leaving Open options for future adaptation); Differential Impact on Sectors of the Community (with regards to current inequalities of distribution); Common Carrier Issues; Risk of Coupling (risks of failure after society becomes dependent on systems); and Research. Elton states the research policy issue to be identifi- cation of: What existing or future research needs to be given high priority to remove or reduce the uncertainties most significant at the planning level (Elton, p.178-9). Following Simon's model with the outer environment predicting goals of a system, and these goals determining 6 the characteristics of a system, this chapter will go on to specifically discuss research as a powerful policy issue shaping the current state of affairs regarding communica- tions oriented telemedicine systems. Human Factors. Alex Reid addresses the research policy issue raised by Elton in an abstract to telemedicine research funded by the British Government (Reid, 1971). Reid confines himself to a policy perspective "on all fours" with Elton's and states that more than enough is known about the technological capabilities of telemedicine systems.' He places the crucial need for research at the level of "human factors," calling for studies that would combine the laboratory and outside world, focused on typical consumers (Reid, 1971). Technological Capability. Contemporary research originating in the United States did not share Reid's focus on the human factors of telemedi- cine research. Instead, United States government funded research efforts were confined almost exclusively to further elaboration of technological capability variables. The most widely known and earliest of this series of field trials was a bidirectional video connection for consultation between physicians at Massachussetts General Hospital, in Boston with nurses at Logan Airport, also in Boston. It was started by Dr. Kenneth Bird, soon after a tragic plane crash resulted in many unnecesSary deaths because physicians were blocked from the disaster scene by masses of gawkers along the freeway (Park, p.25). The most recent wave of research in telemedicine was conducted by NASA in conjuction with the space shuttle program. Confronted with the possibility of providing medical care to persons over a vast distance of outerspace, an ambitious program --STARPAHC--was undertaken. Its main significance in updating prior telemedicine research was in its incorporation of microcomputers into a central role in the care delivery network. The computer was situated at the hospital and was linked to field .units. It was used primarily for: l.) diagnostic and treatment aid for parame- dics in the field; 2.) scheduling patient referral visits and transportation; 3.) drug formulary information; and 4.) Census and nursing orders for patients (Brown, 1983). Dozens of telemedicine studies have filled in the spaces between Logan Airport and STARPAHC; yet each study in turn has omitted "human factor" variables such as acceptance of a telemedicine innovation (Bashshur, 1975). Current Underutilization of Communications Systems It is the premise of this project that policy deter- mines the objectives of a system (Simon, 1969). In regards to telemedicine, it is specifically government funded re- search which has played a significant role in determining the current state of affairs regarding communications—type telemedicine systems: In contrast to the high rate of 8 technological adoption for telemedicine systems used for direct diagnosis or laboratory testing by the physician, communications-type systems which take away hands-on contact by the care practitioner have lagged far behind (Williams, 1984). In fact, there is far more technology in place than is currently being utilized (Williams, 1982; Brown, 1983). We argue that this situation has occurred because of the omission of the important acceptance variables from the funded research initiatives. The absence of these variables appears to have had a differential impact on direct diagno- sis versus communications-type systems. It is reported that the physician will resist any attempt to interfere with a change in hands-on contact with the patient (Brown, 1983). If a trial system calls for such a change, the care practioners' acceptance of the new system would seem to supercede any other technological capability variables in relative importance within the context of the research pro- ject. However, acceptance variables were not included in the projects originating from United States government ini- tiatives, even though the communication science had long underscored the importance of acceptance variables in the innovation process (Rogers 1971). According to Argyris, the failure of policymakers to include a variable such as acceptance derived from the behavioral sciences is not unique to telemedicine. It is the problem of planning and policymakers as system designers in general. Argyris asserts that "the planners 9 in our society have, until recently, been economists and lawyers turned bureaucrats." Argyris further criticizes the policy creation process citing the bureaucrats' failure to consider the implementation stage (Argyris, p. 4). After a system is designed by the planner it is merely turned over to the organization's administrators. Often it is not even utilized. The topdown nature in which directives are issued to subordinates by higherups does not remove the problem of employees accepting and utilizing an innovation. Given the health care practitioners resistance to changes in hands-on procedures, the innovation of telemedicine systems may be expected to introduce even more problems of acceptance than in non-medical organizational settings. Purpose This research goes outside of the strict boundaries of a typical policy study which usually involves legal or economic issues. The purpose of the study is to introduce variables from the communication science in an exploratory setting within the policy research framework established above. This is done in order to assist the policymaking process by uncovering significant relationships between these and other variables which will lead to further inves- tigations relevant to the planning stage of a telecommunica- tions system. A secondary purpose of the research is to apply knowledge about communications networks in the process of innovation in order to develop strategies for the imple- mentation of a system in a hospital in Battle Creek. 10 The reported failure of telemedicine systems creates a loss of potential benefits to society, such as providing service to rural populations. Additionally, there is a loss incurred by society because telecommunication systems, once adopted by an organization or group of organizations, have been shown to accrue additional benefits beyond those origi- nally envisioned by the system designers (Bretz, 1983; Park, 1974). The failure of telemedicine systems to gain acceptance in the medical work place results not only in the primary loss of improving patient care, but also in an Opportunity cost where potential benefits remain masked until widespread acceptance fosters new applications. The research described in this paper was conducted at a Veterans Administration hospital in Battle Creek, Michigan. At the time this research was designed and conducted, the hospital was without a telemedicine system for patient check-in. At the time this thesis was written, hospital administration was in the early implementation stages of such a system. Summary This chapter has set a framework for looking at telemedicine policy issues. A telemedicine system has been described as the interface between the technology of its internal composition and the outer environment of public policy. The discussion has underscored the importance of human factors for research leading to the acceptance of ll telemedicine systems. A specific area that the current exploratory field study will investigate is the set of social system variables known in the behavioral science as structural communication networks. Relationships will be explored between these and other variables in order to lead to further investigations which may be useful to policymakers involved in system design. 12 REFERENCES Argyris, Chris. Intervention‘Theory and Method; 3 Behavioral Science View. Reading: Addison-Wesley Publishing Co., 1970. Bashshur, Rashid; Armstrong, Patricia; and Youssef, Zakhour Telemedicine: ExplOrations in the Use of *Telecommunica- tions for Health Care. Sprifigfield: Charles C. Thomas, 1975. Bretz, R. Media for Interactive Communication. Beverly Hills: Sage Publications, 1983. Brown, J.H.U. Telecommunication for Health Care. Boca Ra- ton: CRC Press, 1982. Elton, Martin C.J. "The Use of Field Trials in Evaluating Telemedicine Systems." Chapter in Telemedicine: Explorations in the use 9f Telecommunications in Health Care, ed. by Rashid Bashshur, Patricia Armstrong, and Zakhour Youssef. Springfield:Charles C. Thomas, 1975. Maxmen, Jerrold S. The Post-Physician Era. New York: Wiley, 1976. Park, B. AB Introduction to Telemedicine: Interactive Tele— vision for Delivery of Health Services. New York: The Alternative MedIa Center at New York University School of Arts, 1974. Reid, Alex New Direction in Telecommunications Research. Bethesda, MD: ERIC Document Reproduction Service, ED 062 803, 1971. Rogers, E. with Shoemaker, F. Communication 9f Innovations. New York: Free Press, 1971. Simon, Herbert A. The Sciences 9f the Artificial. Cambridge: M.I.T. Press, 1969. 13 Williams, Frederick. The Communications Revolution. Beverly Hills: Sage Publications,71982. Williams, Frederick . The New Communications. Belmont: Wadsworth Publishing Co., 1984. Wittson, Cecil L. "Nebraska Initiates Cross-Country TV psychiatry." Educational Screen and Audiovisual Guide 00 (1965) 22-24. CHAPTER II Theoretic Rationale Overview The following chapter will summarize the findings of the general literature review of experimental telemedicine systems reported in Chapter I regarding system goals, tech- nological composition, and public policy issues, especially policy. research. It will review general social system variables from historical innovation research in keeping with a focus on "human factors" in regards to policy research. The discussion will then be narrowed to communi- cation structural network variables, as well as exploratory attitude and media style variables. Introduction The entry of telecommunications technology into the health care setting can cause a great deal of uncertainty for personnel. The critical factor determining the survival of a telemedicine system is acceptance by the physicians and patients (Brown, p.9). Government funded research initia- tives in the United States, however, have failed to look at the acceptance factor in telemedicine system design, focusing instead on technological capability variables and 14 15 shunning those from the behavioral sciences. These research projects, through ommissiOn of acceptance variables, have cast a spotlight on the policy formation phase of telemedi- cine system design and the importance of policymakers becoming sensitive to problems of implementation. Brief Historical Review Historically, the behavioral sciences have held not. only technological variables but alSo social system variables as important within the innovation process. These latter variables encompass a wide range from 'the general influence of the peer network (especially those who have already adopted the new idea), to compatibility (the degree to which an individual perceives a system to fit with her or his experiences), and complexity (the clarity of meaning of the innovation for the individual adopter) (Rogers, 1971). According to Shoemaker, acceptance factors differ from case to case with special differences existing between innova- tions of choice versus those disseminated downward by higherups in a hierarchic organizational structure (Shoemaker, 1971). This is referred to as authority disse— mination of an innovation. This aspect of innovation re- search is studied from time to time yet not consistently. A survey of literature from organizational theory de- picts the employees confronted with technological change in a state of fear. With respect to this, Lawrence draws a distinction between fear of modification of their physical routine (technological variables) and how people expect this 16 change to "alter their established relationships in the organization" (Lawrence, p.639). He states that employees do not resist technical change so much as they resist the anticipated changes in their human relationships. He under- scores the importance of the policymaker becoming sensi- tive to the "specific social arrangements" which employ- ees may believe will be impacted by a technological inno- vation (Lawrence, p.641). The problem of getting policymakers to consider the implementation stage essentially is getting them to look at what knowledge already exists from the behavioral sciences about how innovations become adopted by an individual or organization.. Within the last decade, social system variables, particularly communication structural variables, have moved to the center of these discussions. Communication Structural Variables Lauman has addressed communication structural variables at length in his work regarding the form and substance of urban social structures. He bases his technique of analysis on the individual as a focal anchoring point and considers their links with three other individuals identified as friends. Lauman considers both the homogeneity of the net- work (similiarity in social position) and its connectedness (to what extent do the three friends have friendship rela- tions with each other). This measure of connectedness is then described as radial versus interlocking. 17 Figure I Graphic Depiction 2f Communication Structural " )7 Completely Partially Partially Radial Interlocking Interlocking I Interlocking II Networks V The personal network described by Lauman assumed new prominence when it emerged as an extremely important variable in innovation research, especially originating from the communication science. Connectedness and Integration Two of the main theoretical building blocks upon which communication scholars base their investigations are: a.)connectedness--an index measure of the relationships an organizational member (unit of analysis) has with another organizational member and b.)integration--the degree to which members of an individual's personal communication network are linked to each other, ' i.e. whether the net- work is interlocking or radial (a star pattern) (Rogers, 18 1981). It is important to note that Rogers and others have come to sustitute the term integration for Lauman's connectedness. The connectedness term is then used to refer to some index which quantifies the association between the unit of analysis and the three nodes around it. As stated above, structural communication variables such as connectedness and integration have been shown to be associated with the innovation adoption process. Following. are some research findings which were helpful in formulating research questions for the current project: Medical doctors with higher individual connectedness (more friends and professional associates among other doctors) were more innovative in prescribing a new antibiotic drug. Isolates were markedly less innovative in adopting the new drug than were nonisolates (Coleman, 1966). 1. Earlier adopters are more highly integrated with the social system than later adopters (Rogers, 1971). 2. Radial network individuals use audioconferencing, computerbased "broadcasting," and electronic mail more than non-radials (Danowski, 1983). 3. Earlier adopters have more social participation (higher connectedness) than later adopters (Rogers, 1971). 19 Network Stability Another structural network variable which receives only occasional attention in communication science is the stability of communication networks which may be associated with other variables in an innovation adoption scenario. Stability of a comunication network quite simply is whether the network has the same respondents from time one to time two (Rogers, 1981). The model promulgated by Rogers et al.' was essentially a linear one. The subject over time is depicted proceeding through stages of awareness, interest, evaluation, trial, and adOption ,(Rogers, 1971). The stability variable retains importance therefore in the con- text of this linear model. Attitude Variables As stated above, the linear model of innovation adoption depicts the individual or some other unit of analysis involved in an educational, evaluative process leading to trial and possible adOption. There has been research, however, challenging the linear model (Rice and Case, 1983). This research reports that preexisting attitudes to the innovation and not the amount of time of exposure to it may be associated with amount of utilization and acceptance. Other evidence supporting an association between preexisting attitudes about technology and adoption of an innovation is reported in a study by Dordick of the uses of telecom- munication technology for vocational rehabilitation. 20 Attitude was considered one of the most complex barriers to surmount in getting people to use the new -technologies. These attitudes were derived from experiences with the tech- nology in everyday life (Dordick, 1978). Rice and Case also suggested the presence of a specific attitude variable they termed as a media style (a personal preference for a speci- fic media over others as a means for day to day commu- nication) (Rice and Case, 1982). There does not appear to be a great deal of information available regarding the attitudes of physicians and other care providers to telemedicine systems. What little infor- mation which has been published are reports from physicians themselves regarding their colleagues. Dr. Kenneth Bird has commented on the resistance to change by physicians accompanying the introduction of a telemedicine system. One department chief of a telemedicine project observed that physicians "called the camera 'Big Brother,‘ covered it with their coats, and tampered with the equipment" (Park, 1974). A possible explanation given for this phenomenon of physicians' problematic attitude towards communication- oriented telemedicine comes from Dr. J. U. Brown: The physician resists any attempt to interfere with direct patient contact; he [she] prefers handwritten records to computer tapes and he [she] prefers not to delegate authority to other members of the health care team. These atti- tudes militate against the use of technology to deliver health services (Brown, p.1). A telemedicine innovation scenario holds many variables: disruption of the doctor-patient relationship; changes in 21 other social relationships, such as between physicians and subordinates; and potential attitude barriers towards the technology itself. Little light has been shed on relation- ships between these variables, however, at least in part due to the apparent lack of telemedicine innovation research with physicians as respondents. The likelihood of obtaining meaningful results from such research without the coopera- tion of physicians is poor and as reported above, physicians for the most part do not have a history of cooperation with the telemedicine research effort. The present research proceeded, however, with the assumption that physicians would respond to this research inquiry. Chapter IV, Results, addresses the actual response rate encountered. Research Questions 1. What communication structures are most likely to foster acceptance of the innovation? Prior research has placed an emphasis on integration variables in association with innovation adoption (Rogers, 1981; Danowski, 1983). 2. Is perceived stability of communication networks from pre- to post-adoption associated with likelihood of adoption? Network stability variables are often overlooked in innovation research (Rogers, 1981). Because of the lack of empirical data surrounding this variable, the stability variable has been included as a means of providing data for 22 further investigation associating it with innovation adoption. 3. Do individuals with radial communication structures perceive technological innovation as an improvement for the general system of care? Research results have portrayed the radial individual as more likely to adOpt computer-mediated messaging systems (Danowski, 1983) than individuals with more interlocking networks. If usage is a measure of perceived benefit, perhaps more radial individuals would perceive a computer- mediated telemedicine system as an improvement to the system of care. 4. Do individuals with interlocking communication structures perceive technological innovation as a worsening of the general system of care? This research question poses the reverse assumption of Question 3 regarding degree of radiality and preference for computer-mediated communication. 5. Are attitudes about media from everyday life asso- ciated with likelihood of adoption? This research question is premised on the findings of Dordick (Dordick, 1978) showing existing attitudes about media to be a significant barrier blocking acceptance of a telecommunications innovation for vocational rehabilitation. There is additional support for posing this question in the findings of Rice (Rice and Case, 1983) regarding the 23 existence of a personal preference for specific media asso- ciated with the amount of usage of a computer messaging system by university faculty. Summary The preceding chapter has presented a historical back- ground for telemedicine innovation research. It has re-A ported findings in the literature about employee resistance to technical change and uncertainty over resulting changes in social relationships. It has introduced‘ the social struc-ural variable--communication networks--which is one of the important concepts used in this research. Further, the theoretical building blocks of connectedness and integration have been introduced. A sampling of constructs derived from the communication sciences regarding innovations and network structure has been reviewed. The new research findings suggesting a link between innovation adoption and prior attitude have been reported as well as special problems of physicians as gatekeepers to innovations which remove hands- on contact in the care practice. Finally, research questions have been formulated and discussed within the context of prior research findings. 24 REFERENCES Bashshur, Rashid; Armstrong, Patricia; and Youssef, Zakhour Telemedicine: Explorations ig the Use pf Telecommunica- tions for Health Care. Springfield: Charles C. Thomas, 1975. Brown, J.H.U. Telecommunication for Health Care. Boca Raton: CRC Press, 1982. Coleman, James S., and others. Medical Innovation: A Diffu- sion Study. New York: Bobs-Merril, 1966. Dordick, H.S., Goldman, R.J. and Hanneman, G.J. "Telecommu- nications and Vocational Rehabilitation: A User's Guide to Breaking the Barriers." Annenberg School of Communi- cation at the University of Southern California, for the Office of Research and Demonstrations Rehabilita- tion Service Administration, Department of Health, Educa- tion and Welfare, July 1978. Danowski, J. A. "Automated Communication Network Analysis." Paper presented to the Information Systems Division of the International Communication Association, Dallas, 1983. Elton, Martin C.J. "The Use of Field Trials in Evaluating Telemedicine Systems." Chapter in Telemedicine: Ex: plorations in the use of Telecommunications in Health Care, ea. by RashId'-——B§§hshur, Patricia Armstrong, and Zakhour Youssef. Springfield: Charles C. Thomas, 1975. Lauman, Edward. Bonds 9f Pluralism: The Form and Substance 9f Urban Social Networks. New York: Wiley, 1973. Lawrence, Paul R. "How to Deal with Resistence to Change." Chapter in Organizational Behavior and Administration, ed. by Lawrence, Paul R., Barnes, Louis B.; and Lorsch, Jay W. Homewood: Irwin, 1976. (Article first published May-June 1954 Harvard Business Review. Park, B. An Introduction to Telemedicine: Interactive Tele- vision for Delivery of Health Services. New York: The Alternative Media CenfEr at New York University School of Arts, 1974. 25 Rice, R. and Case, D. "Electronic Message Systems in the University: A description of Use and Utility.” Journal 2: Communication, 33 (1983) 131151. Rogers, E. with Shoemaker, F. Communication pf Innovations. New York: Free Press, 1971. Rogers, E. with Shoemaker, F. Communication Networks. New York Free Press, 1981. Shoemaker, F. Floyd. Anal sis pf Structural Effects 2g Diffusion. Doctoral D1ssertation, Department of Communi- catibn, Michigan State University, 1971. Williams, Frederick. The Communications Revolution. Beverly Hills: Sage Publications, 1982. CHAPTER III Procedures and Methods Overview The following chapter will briefly review the context' in which the research questions were introduced; describe the research methodology including classificatory variables; detail the methodological assumptions; and finally will restate the research questions in operational form relative to the instrumentation and procedure. Introduction Thus far we have established the importance of the policymaker becoming acquainted with behavioral science knowledge concerning communication networks and the process of innovation in an organization. We have established five research questions concerning structural communication net- works, attitude, and the implementation phase of a telecom- munications innovation. The inference of likelihood of adop— tion of the prOposed innovation rests heavily on the answer to the following question: How do the personnel of the health care organization ascribe a value, positive or nega- tive, to the proposed system. Likelihood of adoption is inferred directly from the degree of favorability towards 26 27 the proposed system. Another dimension of this investi- gation rests outside of the research questions, i.e., to approach the unique challenges of implementation offered by a medical setting which must be overcome if a system is to gain widescale acceptance. Considering the case by case variance of telemedicine systems, specifically what is called for to approach this secondary problem is a field study. It is beyond the scope of the present study to con- duct more than a one-shot case study. However, baseline data will be generated which could facilitate a future study at the site measuring before/after effects of the technological intervention. Beyond that, a study of the entire life cycle of the innovation could be undertaken. The present study, however, seeks to utilize whatever relationships are uncovered between structural network variables, attitude and likelihood of adoption in the context of the studies speci- fic site, the Veterans Administration Medical Center, Battle Creek, in order to develop implementation strategies which may be useful to policymakers at the site. Research Methodology The research methodology selected for the present investigation is a correlational study. Phase one attempts to identify relationships between classificatory variables outlined below. Phase two is descriptive--based on the results of phase one. 28 Research/Variables Network Connectedness. An index measure of the amount of communication between the respondent and other organizational members. Network Connectedness is measured separately for current and projected (a time following adoption of a telemedicine system) patient intake, task and social networks. Network Integration. An index measure of the estimated amount of communica- tion between the organizational members named by the respon- dent. Network Integration is measured separately for cur- rent and projected patient intake, task and social networks. Network Stability. An index measure of the extent to which organizational members named by a respondent remain the same from time one (current) to time two (projected) for patient intake, task and social networks. Media Style. An index measure of repondent's preference for everyday media, such as video, telephone, computer, face-to-face, and written as a means of communication. 29 Value 9: the System. An index measure of respondent's degree of favorability towards the telemedicine system proposed in a scenario of technological innovation. Methodological Assumptions The sociogram derived from the instrumentation detailed below will yield data at the ordinal level. Therefore, nonparametric correlational analysis will be used for analy- sis of the data. Limitations of the Study Correlational Analysis. Correlational analysis identifies simply "what goes with what" in relationships under investigation with no indication of cause and effect models. It is less rigorous than an experimental approach with a primary shortcoming being the lack of a controlled independent variable. A danger of this means of approach to the research questions is that it might uncover spurious relationships with little or no relevance or value. Other relational patterns uncovered may be arbitrary and ambiguous (Isaac, 1971). Faults in the Instrument. The research instrument used in the present study incorporates components such as a sociometric, Lickert 30 items, and a set of semantic differentials. Problems from its application in the present study will be discussed in Chapter V of this work. The current investigation is a one- shot case study. Therefore in order to elicit information regarding variables following adOption of an innovation, it is necessary to ask the respondents for projections along the lines of inquiry followed in the sociometric portion of the instrument. If significant findings should emerge, for example, regarding Network Stability and disposition to the innovation, these results would require validation in a study following the actual technological intervention at the hospital site. Research Site The Veterans Administration Medical Center Battle Creek (VAMCBC) is located in the Fort Custer Military Com- plex, Battle Creek, Michigan. The medical center campus is comprised of 30 buildings housing live-in patients and facilities for outpatient care. The population under investigation at this site are staff members belonging to what is known as the team prac- tice. Individuals from the greater staff populations of physicians, psychologists, nurses, and social workers are assigned to teams (one professional role member per team), which are permanently attached to the various wards of the hospital. The team is involved in the administration of care to each patient from intake to discharge. The VAMBC lent itself as a particularly attractive site for this 31 exploratory field research because at the present time it is without central processing equipment or microcomputers. Yet at the same time, the administration of the medical center was in the discussion stage of introducing computer equip- ment at the facility. One of the applications for which a system would be brought on-line is the collection of demographic data from patients at the intake stage of the care process. (At present, interviews are conducted face-to- face by each member of the intake team). At the same time, the VAMBC has a history of introducing telecommunication innovations. A great deal of technology is in place at the site. This is significant in the context of the present investigation as it provides a basis for exploring atti- tudes towards technology used in everyday experiences as these attitudes may relate to acceptance of a telemedicine system. A one-way closed circuit video system links the buildings which house the medical wards, housekeeping opera- tions and administrative functions of the hospital. Video programming can be originated from the well-equipped media center or can be received by satellite dish. Operations of the medical center appear to rely somewhat heavily on the use of telephones for the completion of day—to-day tasks. Perhaps this is due to the sprawling nature of the medical campus and its decentalization. (Most support functions, e.g. security, are housed in their own buildings, as are the various medical wards). 32 Respondents Respondents selected for the study were members of patient intake teams of the various wards at the VAMCBC. The substance abuse ward was omitted from the sample as it had an operational structure different from the other wards. Therefore, the wards studied included 12, 7-1, 39-1, 39-2, 143 and 10. With the exception of 10, these wards each had two patient intake teams assigned to it. The total number' of intake teams was 11. Sample Units The patient intake team of the various wards in the hospital provided the sampling unit for this study. The teams are composed of unique and known members from each of the roles described above (psychologist, etc.). Sample Design Department heads from Social Work, Psychology, and Psychiatry acted as informants who identified staff person— nel assigned to each of the teams. The head nurse of each ward named the nurses assigned to the intake teams. As there was a great deal of variance in nursing assignments due to shift and part-time personnel, head nurses were asked: "Which nurses were assigned to team practice on January 18, 1984?" 33 Unit 9: Analysis Although the research instrument is designed to get at the patient intake communication network (as well as other task and social networks), the person-centered sociometric technique described later in this chapter restricts our data analysis to the individual respondent as the unit of analy- sis. This limits generalizability of the results to the population at the VAMCBC. Instrument The survey instrument is comprised of three sections (found in Appendix C). The first section is a sociometric used in numerous studies of structural communication net- works like those by Danowski (1984). The sociometric por- tion of the instrument is focusses upon current communica- tion contacts and also for a time following adoption of a telecom-munications system for patient intake. The scenario supplied to the respondent is as follows: The patient is seated beside an attendant who will input data into a computer file with a keyboard. The attendant asks the patient a battery of questions amassing a complete patient history, and enters responses into the computer. The attendant instructs the computer to sort through these data using existing interview guidelines for patient check— in by departments. Separate history reports of the patient are compiled for the psycholo- gist, physician, social worker, and nurse and will be delivered to each person's office. The second section is a series of Likert items addres- sing a variety of issues suggested by the specific scenario 34 of technological innovation detailed above. The third sec- tion of the instrument is a series of semantic differential items from Osgood (Osgood, 1957) applied to five specific media the respondent is expected to encounter from day to day at the hospital. Network Integration and Network Connectedness variables are Operationalized from responses to the sociometric por- tion of the research instrument. It was adapted from standard person-centered sociometric instruments from commu- nication science. This technique for gathering data on com- munication networks closely follows the technique used by Lauman (1964) who places its origin in the anthropological sciences beginning in the 1940's. The respondent is asked to name the three people with who, he/she talks the most. The duration of the normal contact for each communication link is quantified with a ratio measure (minutes) for current connectedness, and with an interval measure (frequency over standard intervals of time, such as days, weeks, and months) for current integra— tion and all of the projected measures. Lauman did not originate the breakout of communication activities by con— tent areas. Communication scholars added this dimension to the sociometric technique during the 1970's. For the pre- sent study these content areas are: Intake Networks, Task Networks, and Social Networks. Intake procedures are considered separately from other task procedures because the scenario of technological adoption provided the respondent calls for a telemedicine system to facilitate patient ’35 intake. The remaining content area addressed in the socio- metric questions is social contact, referred to as non-task. The respondent estimates how often he/she converses with each of the three people nominated for these content areas and then estimates the amount of communication between each pair of the individuals named. Respondents are asked to make these estimates for the present time and also to make corresponding estimates for a time after introduction of the telecommunications innovation. This is done to elicit infor- mation about perceived change of communication networks which is addressed in the second research question regarding stability from pre- to post-adoption.' The person-centered technique used by Lauman was chosen for the present study because it allows guaranteed ano- nymity. The guaranteed anonymity was required by the human subjects committee. This condition was accepted by the researcher. It was felt that less reluctance by respondents asked to complete questionaires would be encountered with an anonymous instrument. Further, prior research has demon- strated that the person—centered technique allowed adequate reliability for the present study. Lauman sought to measure the reliability of his technique of network construction by interviewing the persons named as friends by study subjects. He reports about a 43% overlap between those which respon- dents reported as friends and these latter individuals' self-reports of friendship bonds. The Likert section attempts to get at the value, from negative to positive, assigned by the respondent to the 36 hypothetical innovation's impact on the current system of health. care at the research site. The resultant value is needed for planned comparisons between communication struc- ture and likelihood of adoption, as well as attitudinal variables towards technology in general. The items were developed for the present research from descriptive data gathered at the health center prior to the research. It was designed to include variables such as administrative bar-' riers, improvement of the general system, improvement of the intake process as well as the ease with which the system would be adapted. It was impossible to conduct a pretest of these items at the present site without contaminating the sample because nearly all professional staff members are involved in the patient intake process. It would be equally undesirable to pretest the Likert items at a similiar insti- tution given the lengthy time investment required by the Veterans Administration's research screening procedure (six months in the present case). The lack of a pretest for the Likert items raises the issue of construct validity. To address the issue the early stages of the data reduction procedure includes a factor analysis of the Likert responses with results reported in Chapter IV. Items which show an acceptable degree of validity will be weighted using factor coeffi- cients and then summed. This procedure should allow a suffi- cient degree of construct validity for the Likert section. The semantic differential section of the study was 37 designed to yield data for the exploratory portion of this research about the role of attitude toward technology as a predictor of the likelihood of acceptance of a telecommuni- cations innovation. It was adapted from the body of litera- ture surrounding the work of Osgood, Tannenbaum and Suci (Osgood, 1957). Adjective pairs selected have been shown to cut across the three factors which can be meaningfully measured with this technique: activity, potency, and evalua- tion (Osgood, 1957). Five media were selected for this portion of the study: telephone, video, face-to-face, writ- ten memo, and computer. For each media, the identical set of adjective pairs were presented to the respondent for rating using a seven-point scale. In the data analysis phase of the investigation a mean will be computed for each media based on the coded responses to the semantic differen- tial adjective pairs. The data base derived from this section, therefore, will be a set of five mean values, one corresponding to each media, for every respondent. The means will then be used in correlational analysis with the value variable. Procedures First contact was made with the population in mid-July, 1983. The researcher met with Dr. Larry Schwartz, outgoing chair of the VAMCBC Joint Research and Development Committee. A proposal for research was drafted and submitted for consideration in late August, 1983. Due to Dr. Schwartz's objection to technical jargon used in the 38 proposal it was withdrawn by the researcher before consideration by the human subjects committee. The proposal was resubmitted in September, 1983, and subsequently approved. In October, 1983, the researcher went before the Joint Research and Development Committee to defend the pro- posal. The committee recommended acceptance of the resear- cher as a non-salaried employee of the Veterans Administra— tion for the conduct of the research. This status was granted to the researcher in December, 1983. Various members of the intake teams under study were members of either of the two committees with reviewed the proposal for this project. Contact with members of the population not on these committees first occurred in January, 1984. The contacts which each of the intake team members leading to their participation in the study occurred during the winter months of 1984, when patient levels are at their peak. Informants at the site refer to these annual peak levels of intake as "the wintering phenomena." (It is reasoned that potential patients spend most of their time outdoors in moderate weather months and do not precipitate intake proceedings by relatives.) Social workers were selected by the researcher for the first wave of survey administration. 39 One of the requirements of the Joint Research and DevelOpment Committee was the conduct of "Informed Consent" interviews with every subject asked to complete a question- naire. Members of the population selected as respondent's were contacted by telephone and an appointment was set. At the meeting resulting from this appointment, the subject was read the "informed consent" document provided by the Vete- rans Administration. In addition, the subject was read the form entitled "Information About the Technology Survey" which contained a brief description of the study including its purpose, the involvement required, guarantees of anony- mity, the voluntary nature of the respondent's participa- tion, and the promise of a brief written report sent. to those respondents indicating an interest in receiving it. The informed consent interviews were usually conducted in groups of two or three for efficiency. When this procedure was completed each subject was handed a survey and asked to complete it at that time. The researcher waited nearby in the office or some other waiting area and was available to collect the surveys after completion. Following survey administration to the population of social workers involved in patient intake, a procedural change was initiated in order to use time more efficiently. Following completion of the informed consent interview, the respondent was handed the survey with a stamped envelope addressed to the researcher. This change was cleared through informal contact with the Joint Research and Deve- lopment Committee. 40 Statistical Analysis Procedures Reduction 9f Network Data Isolate Identification. The first step towards answering the research questions regarding relationships between network structure variables, attitude and favorability towards the innovation is to sepa— rate relative communication isolates from the rest of the respondent pool. In order to achieve this, a primary index describing Network Connectedness was constructed for current patient intake networks as follows: 1. Respondents with less than three contacts are sepa— rated from the others. These are considered extreme iso- lates and further measures of network structures are not computed. 2. For the remaining respondents, the amount of commu- nication with each of the three named contacts is summed. The mean amount of primary communication was calculated. Cases falling below one standard deviation from this amount were coded as isolates. The remaining valid cases were coded as nonisolates. Network Integration. The construction of an integration index for current and projected patient intake, task, and social networks was 41 accomplished as follows: 1. Median splits are performed on each pair of indivi- duals named by the respondent with those pairs above the median coded as one, and those below as O. 2. These three new variables are summed with scores ranging from 0 to 3 which constitutes a 4-point scale. Zero represents a more radial network and 3 a more interlocking one (Danowski, 1984). Network Stability. To answer the research questions regarding the per- ceived stability of the respondent's patient intake, task, and social networks from time one to time two (following technological intervention), an index measure was taken from the subject's response to the questions regarding: a.) any changes in the individuals named for patient intake, task, and social networks from time one to time two; and b.) the number of individuals named at time two different from time one. Media Style In order to answer the research question regarding meaningful relationships between disposition towards media in everyday life and disposition to the specific innovation described in the scenario. Mean values from responses to semantic differential adjective pairs are computed for five different media. 42 Value gt the System All of the research questions concern associations between structural network, and Media Style variables; and likelihood of telemedicine system adoption inferred, from favorability towards the system. In order to measure the latter variable it is necessary to reduce the Likert data into a single variable describing the respondents overall perception of the innovation. To accomplish this, the fol- lowing analytic procedures are followed: 1. The Likert items are factor analyzed. Factor scores from the rotated varimax factor matrix are used to weight each of the Likert variables. 2. The weighted Likert values are then summed. The resulting value is taken to be a measure of likelihood of adoption. Research Questions Operationalized 1. What communication structures are most likely to foster acceptance of the innovation? Are there significant positive or negative correlations between the Network Connectedness index (of current and projected patient intake, task and social networks) and Value of the System? Are there significant positive or negative correlations between the Network Integration index (of current and projected patient intake, task, and social networks) and Value of the System? 2. Is perceived stability of communication networks 43 from pre- to post- adoption associated with likelihood of adoption? Is there a significant positive or negative relation- ship between the Network Stability index (for patient intake, task, and social networks) and Value of the System? 3. Do individuals with radial communication structures perceive technological innovation as an improvement for the general system of care? Is there a significant negative correlation between Network Integration (for current and projected patient in- take, task and social networks) and Value of the System? 4. Do individuals with interlocking communication structures perceive technological innovation as a worsening of the general system of care? Is there a positive significant relationship between Network Integration (for current and projected patient in- take, task, and social networks) and Value of the System? 5. Are attitudes about media from everyday life asso— ciated with likelihood of adOption? Are there significant positive or negative relation- ships between means from semantic differential adjective pairs for five media and Value of the System? Summary This chapter has addressed the research questions out- lined in Chapter Two and discussed how the present study is 44 designed to provide an answer to these questions. The sample units, design, and units of analysis were described as individual members of selected health care teams from the site. Development of the survey instrument was discussed, along with procedures for its administration. Data analytic procedures to be performed on data gathered from the instru- ment were outlined and the research questions were Operationalized in the context of the data analysis proce- dures. 45 REFERENCES Danowski, J. A. "Personal Network Integration: Info- graphic, Psychographic and Demographic Characteris- tics." Paper presented to the Information Systems Division of the International Communication Associa- tion, San Francisco, 1984. Isaac, Stephen, with Michael, William B. Handbook i2 Research and Evaluation. San Diego: R. R. Knapp, 1971. Osgood, Charles; Tannenbaum, Percy; and Suci,George. The Measurement gt Meaning. Urbana: University of Illinois Press, 1957. . CHAPTER IV Results Overview The following chapter will discuss results of the pre- sent investigation into the introduction of a telemedicine innovation at the Veterans Administration Medical Center, Battle Creek. It will detail the response rate; assignment of isolate/nonisolate status to respondent cases; provide an overview of the nonparametric correlation coefficients used; and provide the results of the correlational analysis. Response Rate Fifty "informed consent" interviews were conducted and at the conclusion of each interview a questionnaire was provided to the respondent. Thirty-eight questionnaires were returned for a response rate of 76% overall. Sixteen were received from social workers providing 100% response, nine from nurses for a response rate of 82%, 10 from psycho- logists for a response rate of 91%, and 3 from physicians for a response rate of 25%.1 Of these, 35 were considered valid cases from respondents currently involved in patient intake procedures and comprised the pool of responses sub— jected to data analysis. (Two social workers and one physi- cian reported they were not involved in patient intake and were therefore eliminated.) One case, (a physician) was 46 47 subsequently eliminated because it did not report at least three individuals for communication about patient intake. Identification gt Isolates Thirty-four questionnaires from respondents reporting involvement in patient intake reported at least 3 communica- tion network links for the primary network. These provided the core data base for identification of isolate/nonisolate' status based on the following procedures: the amount of communication with each of the three named contacts for patient intake is summed. (Table 1 presents these primary amounts.) The mean amount of primary communication is then calculated. Cases falling below one standard deviation from this amount were coded as isolates. The remaining valid cases were coded as nonisolates. (Case by case identifica- tion of isolate/nonisolate status, primary amounts of commu- I niCation, the mean value and standard deviation of primary communication can be found in Table 1). Thirty-one cases were coded "1" for nonisolate and included for further analysis. Of these, 2 physicians, 7 psychologists, 9 nurses, and 13 social workers were represented. Three cases were coded "O" as isolates. All three were psychologists. Overview 9t Statistical Procedures As discussed in the previous chapters, relevant items within a factor were identified for construction of an index 48 Table 1 Primary Communication and Isolate/Nonisolate Status Primary 0/1 Case Communication Assignment Number Amount (mins.) 1 60 1 2 35 1 3 50 1 4 135 1 5 2 O 6 75 1 7 50 1 8 25 1 9 18 O 10 6O 1 11 6O 1 12 17 O 13 50 1 14 .....Invalid Case 15 111 1 16 60 l 17 .....Invalid Case 18 ..... Invalid Case 19 30 1 20 75 1 21 60 1 22 100 1 23 60 1 24 180 1 25 21 1 26 30 1 27 35 1 28 25 1 29 45 1 30 65 1 31 35 1 32 35 1 33 40 1 34 ..... Invalid Case 35 45 1 36 35 1 37 75 1 38 55 1 F = 53.43 S.D. = 34.80 49 for the Value of the System variable. This was achieved by summing together the respondent's weighted scores for each item composing the index. (Appendix B contains the values of the rotated varimax factor coefficients used for this purpose.) Index scores were necessary in order to explore variables of the research questions regarding Network Con- nectedness, Network Integration, perceived stability of network relationships, and dispositions towards everyday media revealed by the respondents. Assumptions about the data reduction procedure yielding data representative of ordinal level measurement restrict the current investigation to nonparametric correlational analysis. Kendall's Tau The subprogram Nonpar Corr of SPSS was utilized in this study in order to identify significant correlations between the Value of the System variable and other variables described earlier. Kendall's Tau is a correlation coeffi- cient which describes "the extent to which persons or ob- jects are ordered alike on two variables" (Nunally, 1978). Kendall's Tau is a correlation coefficient useful with data measured at the ordinal level. The Kendall's Tau Summary Table (Table 2) contains information about Value of the System, communication network structure variables, everyday media variables and perceived stability for commu- nication networks from pre— to post-adoption. 50 Table 2 Kendall's Tau Summary Table gt Correlation Coefficients Variable Correlation Significance Coefficient Level Network Connectedness Patient Intake -current .1220 .194 -projected -.2257 .055 Task -current .3681 . .005 -projected .1368 .170 Nontask -current .2093 .071 -projected .1395 .175 Network Integration Patient Intake —current .1417 .160 -projected .0573 .346 Task -current .0831 .279 -projected .0133 .464 Nontask -current .0177 .451 -projected —.O380 .395 Table 2 -continued 51 Kendall's Tau Summary Table gt Correlation Coefficients Variable Correlation Significance Coefficient Level Network Stability Patient Intake -.2217 .103 Task .0585 .326 Nontask Could not be computed Telephone .1067 .233 Video .3847 .006 Face-to-Face .3044 .019 Written Memo .4911 .001 Computer .6381 .001 52 Analysis gt the Data Connectedness Variables Network Connectedness variables for current and projec— ted patient intake, task, and social networks were studied in order to determine whether they could be associated with disposition towards the innovation (positive or negative) as a predictor of likelihood of adoption. Connectedness variables were representative of ordinal scales of measurement. Table 2 provides descriptive statistics of Kendall's Tau correlations showing specific relationships between connectedness variables and the Value of the System variable. Network Connectedness for current patient intake was not significantly related to Value of the System. Network Connectedness for projected patient intake was significantly related (at the 95% significance level) to "value of the system" variable. Connectedness for current task matters were significantly related to the Value of the System variable. Connectedness for projected task matters were not significantly related to the value variable. Connectedness for current and projected nontask matters (social) were not significantly related to Value of the System variable. Integration Variables Integration variables for current and projected patient intake, task, and nontask maters were studied in order to determine what relationships existed between these variables 53 and likelihood of adoption of the proposed innovation. Integration variables are representative of ordinal scales of measurement. Table 2 provides descriptive statistics of Kendall's Tau correlations showing specific relationships between integration and the Value of the Sys- tem variables. Network Integration for current or projected patient intake, current or projected task matters, and cur- rent or projected nontask matters were not significantly' related (at the 95% significance level) to the value variable. Network Stability Variables Network Stability variables for patient intake, task, and social networks were studied in order to explore rela— tionships between the individual's anticipated changes in social relationships after technological adoption with the Value of the System variable. Network Stability variables are representative of ratio scales of measurement. Table 2 provides descriptive sta- tistics of Kendall's Tau correlations showing specific rela- tionships between Network Stability variables and the Value of the System variable. The Network Stability variable for patient intake, task, and nontask matters were not signi- ficantly related to the Value of the System variable at the 95% significance level. 54 Media Style Variables Media Style variables were studied in order to explore whether meaningful relationships exist between disposition towards media in everyday life and attitude toward the inno- vation. Media Style variables are representative of ordinal scales of measurement. The mean value for each of five media per respondent (found in Appendix B) were used for nonpara— metric 'correlation with the Value of the System variable. Table 2 provides descriptive statistics of Kendall's Tau correlations showing specific relationships between Media Style variables and the value variable. The Telephone variable was not significantly related to Value of the System variable. Video (television), Face to Face, Written Memo and Computer were all significantly related to the value variable at the 95% significance level or higher. Summary This chapter has presented results of the current investigation into telemedicine innovation at the Veterans Administration Medical Center, Battle Creek. A response rate of 76% (including an extremely poor response rate by physicians) provided core data used in nonparametric corre- lational analysis between communication structural variables, attitude variables, and the Value of the System variable. Network Connectedness for projected patient in- take networks and current task networks were significantly 55 related to Value of the System at the 95% significance level. Network Integration was not significantly related to Value of the System for current or projected patient intake, task or social networks. Network Stability variables were not significantly related to the value variable. The tele- phone media style variable, one of five media for which attitude measures were collected, was not significantly related to disposition towards the innovation. The remaining Media Style variables, Video, Face-to-Face, Writ- ten Memo, and Computer were significantly related to the Value of the System variable at the 95% significance level. 56 REFERENCES Nunnaly, Jum C. Psychometric Theory. New York: McGraw- Hill, 1978. 57 CHAPTERNOTES 1This study encountered an extremely low response rate from physicians. This is consistent with prior research efforts about telemedicine using physicians as respondents. As stated in Chapter II, physicians resist any effort to alter their relationships with patients. Non-response by physicians is problematic to a telemedicine investigation yet is a form of evidence of the importance of further investigations directed at physician acceptance. CHAPTER V Discussion Overview This thesis has explored telemedicine applications in a field setting. It has been particularly concerned with human communication as an important variable to be addressed in policy research. Two communication structural variables have been measured: Network Connectedness (a measure of the strength of communication relationships between an indivi- dual and a selected group of contacts); and Network Integra- tion (a measure representing the individual and her/his communicants in either a radial pattern or an interlocking network). The study has explored relationships between these communication structural variables and a variable measuring the respondent's disposition towards the telemedi- cine innovation described in a scenario within the research instrument. This latter variable was Operationalized as the sum of subjects' weighted responses to Likert items concer- ning an impending technological innovation. The structural variables were also correlated with subjects' responses to attitudinal measures of telecommunication and other media used from day to day at the hospital. Following is a discussion of the results of these analyses framed within 58 59 the research questions discussed in previous chapters. Discussion gt Research Questions 1. What communication structures are most likely to foster acceptance of the innovation? Prior research offered some evidence that connectedness was positively related to an individual's use of a medical innovation. Results of the current study offer some support for this earlier finding. The higher the Network Connected- ness for projected patient intake and current task networks, the more favorable the disposition towards the- innovation. It appears to be that personnel who report a high degree of communication--or talking--with other staff members about task matters hold a positive view towards the innovation. An alternative explanation to this finding is that personnel engrossed with task matters may have a more favorable disposition towards any authority-disseminated innovation (what in informal terminology might be called a "gung-ho" military orientation) and the relationship between connectedness and preference for a telemedicine innovation may be a spurious relationship confounded by a preference for the mode of dissemination. Network Integration in either a more radial or interlocking pattern did not show significant correlations with favorable disposition towards the innovation. The integration variable is addressed separately in the discussion of research questions 3 and 4. 60 2. Is the perceived stability of communication net- works from pre- to post-adoption associated with likelihood of adoption? Prior research has suggested that employees confronted with technological innovation may fear a change in their interpersonal relationships which would result in a resis- tance to a technological innovation. Results of the present study did not uncover a relationship between expec- ted change in whom a respondent talks to most and disposi- tion towards the innovation. Possible explanation for these results may be related to prior research showing that the introduction of an audioconferencing system actually in- creased the amount of frequency of face-to-face communica- tion between employees (Bretz, 1983). One respondent added descriptive data to the present study by commenting that a new computer system would foster more interpersonal communi- cation because of inadvertent "bugs" in the system or proce— dure. The implication is that additional or existing levels of face—to-face contact would be necessary even after the launch of a system designed to decrease such contact in order to clarify printed outputs of the new system. Another possible explanations for this is that the survey instru- ments requirement of projected communication patterns does not adequately operationalize a stability variable intended to measure a change from time one to time two. The following two research questions are combined for the sake of discussion: 61 3. Do individuals with radial communication structures perceive the technological innovation as an improvement for the general system of care at the VAMCBC? 4. Do individuals with more interlocking communication structures perceive the technological innovation as a worsening of the general system of care at the VAMCBC? Prior research has suggested that the more radial the individual's communication network, the more they tend to use computer-based message systems. Based on these prior findings it would be expected that the more radial the individual's communication network, the more favorable the response to the innovation. As stated in Chapter III this would be expressed by a strong negative tau coefficient from the nonparametric correlation of integration variables with the value variables. The present study was absolutely without support for a relationship between the degree of radiality and preference for computer-mediated communica— tion. In the context of prior literature on such computer— mediated innovation, this is a startling finding. One possible explanation for the absence of the relationship might be that the scenario used in the instrument had a confounding effect on the expected relationship between increasing radiality and a subject's preference for compu- ter-mediated communication, yet this would not be considered a flaw in the context of the present investigation. Early innovation research indicated authority innovation had an entirely different set of assumptions from that of 62 innovations of choice. Therefore the present study specifi- cally utilized a scenario which did not give the respondents a choice between using the new system or not. However, in studies cited earlier subjects had been given a choice to use the innovation or not to use it. It may be suggested that confronted with authority-disseminated innovations, a more radial individual's positive feelings for an innovation are confounded by a dislike for the authority mode of disse-‘ mination. Additional grounding for this view of the results is organizational theory which generally advocates partici- patory models of management. This interpretation would be supported further by a finding cited earlier that "radial network individuals are more autonomous with higher internal locus of control" (Danowski, 1975) and "place a higher value on independence" (Danowski, 1983). A study by Schomish of farmer's voting patterns with respect to a merger of agri- cultural cooperatives (Schomish, 1983) revealed that farmers who were not given a choice in the matter had a more nega- tive attitude towards the proposed change than did those who were given a choice. This underscores the undiminished importance of communication scholars including authority- dissemination variables in their research inquiries for the sake of generalizability of the research findings to organi- zational settings where often the innovation is not of choice but one mandated by a superior to a subordinate. An alternative explanation of the surprising lack of relation- ship between increasing radiality and favorability towards 63 the system is that an unintentional confounding effect may reside in the scenario which represented the innovation in the survey instrument. The stated final product of the new computer-assisted procedure for patient check-in was refer- red to specifically as a computer printout which would be delivered to the respondent's office. There is some evidence which suggests that radial network individuals use printed media less than nonradials (Danowski and Van Engen, 1983). Radials may be expected to prefer interpersonal contact to the computer printout. 5. Are attitudes about media from everyday life associated with likelihood of adoption? Prior research has depicted the innovation process as a linear model. As the amount of time an individual spends with an innovation increases, it was expected that the likelihood of adoption increases. This was thought to be a result of gaining knowledge about the innovation and, in effect, learning to like it. However, recent research hinted that a powerful variable of predisposition was more at the heart of the innovation outcome. This recent re- search shows existing negative attitudes about media from everyday life act as barriers to adOption of similiar inno- vations. Results of the present study do not refute this finding. Respondents' feelings about the telephone were not significantly related to a positive feeling about the inno- vation. However, positive feelings about video, written memos, face—to-face contact, and computers were 64 significantly related to positive feelings about the pro- posed innovation. At face value, the telemedicine system combines elements of video (the crt), written (the printed output), and computer media into a unified whole. It is not surprising, therefore, that the more favorable attitude to- wards similiar media from day-to-day experiences, the more positive the attitude towards the innovation. Suggested Direction for Further Investigation The present study uncovered relationships between high individual connectedness and favorability towards the pro- posed innovation. At the same time, strong anticipated relationships between radial network patterns and favor- ability towards the innovation were not uncovered. It is not clear, however ,if the surprise in the results is attri- butable to peculiarities of a medical setting for this type of communication network research; unintentional confoun- ding by the survey instrument with representation of the innovation with a scenario assuming a printed output; or an extremely important indication that the lack of authority- elements in prior studies showing more radial network indi- viduals preferring computer-mediated communication limits their generalizability to situations where the innovation is one of pure choice and not disseminated from the top down through a hierarchy. This suggests that future research efforts of relationships between structural communication networks and preference for computer-mediated communication must reintroduce the authority-innovation variable as more 65 general communication of innovation studies had. Further, the peculiarities of innovation in a hospital setting should be scrutinized as there may be a relationship between con- nectedness and innovation in this setting more so than these variables in a non-medical setting. Upon completion of the present investigation several directions for future research emerge in the context of “things this researcher would have done differently:" The analysis of relationships between attitudes towards media from day-to-day experiences and disposition towards the innovation may have been enhanced with semantic differential adjective pairs Nunally developed as a familiarity factor. The need for this aspect of the measurement of attitude towards the telemedicine scenario was pointed out by two respondents who noted beneath the set of adjective pairs for the Computer variable an indication that they did not feel qualified to respond, having never dealt with one. In order to study Network Stability, there simply may be no substitute for a longitudinal study. Con- fronted with the predictable lack of response to survey questionaires by physicians, a study of telemedicine perhaps should include descriptive interviews with physicians to offset limitations imposed on generalizability of the pre— sent study imposed by nonrespondent bias. Other Interestigg Patterns Which Emerged Having amply noted the caveat that approximations of changes in network structure from time one to time two are 66 based on a projection by the respondent as to behavior in the future following adoption of a hypothetical telemedicine innovation, it is interesting to survey the connectedness and integration variables by role noting changes from time one to time two. (Appendix A contains results arrayed as cross-tabulations of role by Network Integration and Network Connectedness). For Network Integration from time one to time two, social workers saw themselves becoming less radial for patient intake, but nurses and psychologists projected communication structures that were more radial for patient intake. For task matters, again social ‘workers saw themselves becoming less radial from time one to time two, yet nurses and psychologists expected little or no change. It was for social communication that social workers, nurses and psychologists came into agreement that they would become less radial following adoption of the telemedicine innovation. This apparent finding echoes Lawrence's view that employees do not consider salient features of the technological innovation itself, but rather what changes adoption may make on their social relationships in the organization. Since there appeared to be traces of a pat- tern for Network Integration and Network Connectedness variables considered for respondents role by role, these variables were subsequently subjected to nonparametric cor- relational analysis with the Value of the System variable using a select-if convention for separation by role. For social workers, relationships between increasing radiality and favorability to the innovation were in the proper 67 direction for social networks yet not at a significant level. Again, however, as in the large group analysis, task connectedness was significantly related to favorability to the innovation. Nurses' projected task and social Network Integration were also in the proper direction for a relationship between increasing radiality and favorability to the system, yet not at a significant level. Increasing Network Connectedness was not significantly related to favorability to the system for nurses, however, (a pattern not in keeping with the larger findings). The only signifi- cant results of connectedness and integration data run with the value variable for psychologists was for projected con- nectedness (increasing projected Network Connectedness asso- ciated with higher favorability to the system). Any conclu- sions to be drawn from this analysis of variables by role would be that connectedness is apparently the strongest structural variable associated with disposition towards a telemedicine innovation disseminated in an authority fashion. Also it should be noted that role by role, there are indications that integration may be associated with favorability in a hospital setting. As noted above, the correlation coefficients were at least in the proper direc- tion to support such a claim. Strategies for the VAMCBC As previously stated, the secondary goal of this re- search is to assist the administrators of the VAMCBC 68 planning a telecommunicationsIcomputer system for the hospi- tal. One means of developing targeted implementation stra— tegies for the VAMCBC would be to run T-tests of the Value of‘ the System variable data to look for differences in attitude between professional groups at the hospital. Addi- tionally, because of the expected concern over changes in social relationships which might ensue following adoption of a telemedicine system, focus group discussions would be advisable as to the specific social issues to which staff members are sensitive. Such discussions are most appropriate at the early planning stages of the innovation.' A practical interpretation of the emergence of the connectedness variable is that word of mouth concerning the innovation would be expected to be positive (the more the respondent reported communicating to others about task matters, the more favorable their disposition to the innovation). There- fore, it would appear profitable to disseminate information about the innovation in staff meeting settings with a good portion of time devoted to discussion. Attitudes about day— to-day media at the hospital appear to be related to atti- tudes about the innovation. If resistance to the innovation is to be addressed, it might best be done so in the context of separate system components (video screen, versus key- board, versus computer data base, etc.) in order to isolate objections and thereby overcome them. 69 REFERENCES Bretz, R. Media for Interactive Communication. Beverly Hills: Sage Publications, 1983. Danowski, James A. "An Information Theory of Communication Functions: A Focus on Informational Aging." Unpublished doctoral dissertation. East Lansing, Michigan: Michigan State University, 1975. Danowski, James A. "Automated Communication Network Analysis." Paper presented to the Information Systems Division of the International Communication Association, Dallas, 1983. Danowski, James A. and Van Engen, M. "Readership of Employee Publications as a Function ' of Interpersonal Communication." Monograph, School of Journalism, University of Wisconsin, 1983. Schomish, T.P. The Effects gt Personal Communication Networks 22 Members Voting tg g Merger gt Agricultural COOperatives. Unpublished doctoral dissertation, University of Wisconsin, 1983. 7O APPENDICES 71 APPENDI X A 72 I .a.._¢.v¢~ or. c..~..o oz. .o=-> nope“ co...~o~.ca oz. Loren one v.0 « o oco:<>~ou£0 ocuu..! 0.00— 5.0— 5.0— o.o~ b.0v ~auoh o oo. n.um n.o n on o.n~ Lc.o._. on n n o v_ ”cfiaaoo "m 0. ~ LL 0 cE=LoO o 0 on o o.on 0.00 o o 0 on ~ 0 . o _ 5.3.1.: m . o o _ c.3333 oinv o 0 fun o_nv me. can mv— k n o o v 2332:; r. n _ N 1 2332133.. 0 o o.nn v.vv .«fl —_" v,vv m on o o o n v equzz a. _ _ v m ouuaz 53 n no n.o A... 3.22.. «.3 o n 2.. 72 $.20: ~_ ~ v u n _q_oom ...L n o n m _-.oom n N _ o n ~ L o «neck .oc_xoo—uo.c~ o. _-_v-mv ..om tom _..oe .oc.xoo_uo.c_ o. _-_v-mv .01 tom maul: co:-uo~:: «csoo nape: coLL-coIFJ «csou exquc~ tic—«am vouooao...“ ox-ur: .coLCad «cf—:30 . «3 film. 3“»qude xlflmhqdl- film §-4..Llnmajlnlmd —< .3: 73 .>~«~¢Lvau ox. hounoum 0£~ .0:~n> nova“ coaumuoo.C« 0:. L020” ozk .0.02 0.00" 5.00 5.0" 5 on 0 0m Loves 0.00. o.- 0.~. 0 "v 0 mm -.o& on .u n 0 0 cE:~oU «m 5 e m1 5 £63.00 0 0 0 00. o o 0 0.00" N o o 0 ~ 5.0.33.1 ~ o o o N :23um o.0~ 0.0— o.0~ 0 0m 0.0m 0 0,mv 0 mN A ~ _ ~ N 3:030:03; k m o m ~ 333210.31 0.00 0.n~ n.~_ n.~_ v.00 ~.- _ L. m NN 0 v N a a o«~32 o v N a m onusz 0.00 0.0— n on 5 5 uox~03 5.5 v.0. «.00 5.5 noxuol 0“ n ~ 0 _ "1.00m m" _ N 0 L "-aoow 0 ~ _ 0 m m a o —~.ok Aoc.xoo_uo.c~ 0. L-‘vamv .«0m Sou _u.o& Auc.xoo~uo.c~ 0. _n_v~mv «om Rom nova“ cow«-»0o«c~ «c300 novc_ no.0.»uo«:_ 0:300 sunk pouooqosm xmak «couuuu «How NM co.vn~000c~ xhoxuoz dd co.~a~=n-nnoLU n ~< o~£~e 74 .5.L.aLv-u Lo.-ou0 o£« .o:_¢> mopc— COLLuuqoac. or. .020” one «.02 o.oo_ a .v o oN o N_ _ o_ . _..oe 0.00. a w" _ or 0 am a a. _..oe an n. o v n cE:_oo in e n a or :E:_oo o o.om o o om o o o o 00. ~ 0 a o 1 caao.u>zm N o o o N c..o.nx;m o.~v o.~c n.e_ o n.4, o,m~ m vi a ~« A n n a o .uaooLoaoaum k a a a m cmloorozosmm 8.00 o o . as m on m on m mm o o o o o I .:32 o n m m o ouraz v.n_ n.om «.mm ..m~ uoxuoz o o m.mm A Le toxuoz n2 ~ n m n anaoom m. o o n m _._uom m N a o m N a o _..oe luc_xoo.uo.c~ 0. _.3v.mv ..om tom _..oe Auc.xoo_uo.c~ 0. _..v.m. .01 tom novc~ cofivau00.c_ «£300 nova“ cofi.u~ow.c~ «c300 ~¢Loom vouooq0um ~u—oom «nouuau WHNm NM co--»00.c~ ~¢_oow «Couuau xuotuoz «0 coddu~=nnammoLU o _< o_nu9 75 .nucvouCDEOO UOEIC oouza 3:: «covcoan0u £003.03 .o~.:oo v0auoaou 0£~ N0.m0»6 0:. .0=_-> x0vcu mn0cpcao0ccou 0:. h0£o.£ 0:5 .0002 o.oo~ 4.0— v.o_ .o. o._¢ _..oe 0,00" 0 mm .NN o oN o N. a..oe .n o o N :Ezaoo in N_ o v cannoo o 0 .on 0.0m o o oo_ o N o o a c._o.nazm N o N o c..o..sgm o oN o.oN .v. o.mN N.NN er 0 mN m.¢_ N N N N .uaoo_ozosnm N m N _ Lalooiorusum o N.NN .NN o.mm m.nm NN m mm a "a o o N m 0a.:z o m N _ aucaz m on v n“ .n_ n mm h3:03 n.0m on «.m_ 4 ma toxroz n" v N n _..oom m_ n N N Hauoom n N m 1 o _~.oe A;u_x 0. rods ..00 tom _a.oe .zu.x o. rosy .om tom x0uc~ un0cv0.o0ccoo «c300 x0pc~ nm0c00.00CCOU .czoo 0xa~cH «:0NN-m $0000.0um «coauam «couusu 0~om Hm unocp0No0ccoU xuoxgoz Md coaun~sna.mmoNU . N< win-s 76 uuCIUwCJEfl—OU UOECC 00»:. van «:0vcoan00 £00300: .o-Ncoo p0NNon0N one Nouuouo 0:. .0:-> n0uc. unocuouooccoo 0:. N0:0.£ 0;? .0002 _ cOLNapuonno ocNNNNI o.ooH N.NN .oN 5.0N .oN -.oe o oo. n.nm o.aN m nN N a _..oe on o“ o cE:_oo an a. o m m cannoo 0.0m o .00 o o on o 0 on N _ o c..oN.Ncm N o a o 1 c.101nsgm n.v_ .N. m.._ .mN o No c.0N o o,NN 5 N _ ...uo_ozoanm k m N o N «u_oo~ogoxmm o.nN n.N_ Ni 5.00 N.NN a a" o m o _ onuaz a o N a o amuaz e n" .mN N.o¢ m1 Noxuoz v n_ m on 0 mm o Noxuoz n" N o .naoom N_ N v N o Lauoom m a n N a o ~..os .501: o. 30.0 ..om tom 1..oe Anna: 0. :oac .00 tom H0vc_ «nonvovooccoo “£300 novc~ «00Ep0«00cco0 “:300 x009 30.00.0Nm xnuk «:0uuso 0~mm Hm unocvodooccou xNotuoz flfl Coqua~nnuaunouo a m( 0—nnh 77 00N£N vcu «covconnou c00l~0a «O'Ncoo $0NN010N 0:. you-0N0 0£¢ .03—00 nov:« unocvouo0ccoo 0£« «Caodcafifioo vofiqc Nogofiz 0:9 .0qu . I anoi.u>u0uno ocNuuNI H n mcoaua>uonno madam": o.oo_ N.NN .oa o.on 0 ON _-.oe o co" o.oN N.NN N 0N N.NN _..oe oN c N. N asaroo ON 0 N N N cE=_oo o.on o o On o.oN o o 0 on N a o 1 c..o..>zm N a o o — ca.o..Ncm o .v. «.NN o.NN N.¢_ o N.Ne N N. N o v N .NLNONozoNam N . o N N .miooiozoxam N.NN ._N v... ..1_ N.NN N.NN e .4 o o N v _ onusz a N N v o onusz o .NN N.NN N o" Noxuoz N.N o co N.N N NN .oxuoz N. o N N _..oom N" a c _ v _qroom N _ o N N _ o Nazca Ago.: 0. no.3 ..00 tom 1..oe Azo_x o“ rose .Vom tom x0uc_ 000:00.00ccoo «c300 u0vC~ un0cv0~00CCoU «c300 ~0woow acouuso Nugoom “cohuau odom NM unocuouo0ccou xuotaoz No co“«-:£-ummohu a N4 eig.p 78 APPENDI X B 79 n03~a> 0:NNN_E n0.mu_ch :0: 0.02 00.0 n M v0.0 I M mo.n n W 05.v u M mm v a W 5 n o n 0 Na 0 v v v 0 mm m 0 v e v LN o v n o 0 5n 0 n w m 0 ON 0 n n v v on v n n n 0 o— n v n n v on m w 0 m 0 0— n n o n 0 mm c m m N n w" v v v v v «0 n m c 0 n m— n v 0 n v .0 n n o 0 0 ~— 0 0 0 0 0 on o n 5 o o 0" n o 0 n 0 ON 0 v n m n 00 m n e v v on w v 0 v 0 50 n v n n v 5N v n 0 n m 00 n n m n n on v v n e 0 v0 0 n n o n n~ 5 o 5 0 5 mo 0 w m o 0 cm v e v v v no 0 v e v e mN v n 0 n e no 0E0: ounhloy N02E32 0E0: 0oouao~ N0nE32 Nousafiov couafiul locum o0vu> 0coza0~0k 0m40 yousnEoU :0«~«u3 locum owpfi> 0coza0~05 ou-U uo~nafluu> oaxum «L002 NON m03-> :00: m manna. 80 APPEND I X C 81 APPENDIX C Table C1 Factor Coefficients for Weighting Lickert Items Factor Variable Coefficient 1. Smooth Implementation .59399 2. Changes in Routine -.02565' 3. Administrative Barriers .21969 4. Attitude Barriers .45649 5. Improve Patient Intake .88899 6. Quick Implementation .60103 7. Bugs in the System .35993 8. Improve Administrative Efficiency .69433 9. Improve Service to Patients .81791 Note.Factor Coefficients from Rotated Varimax Factor Matrix. 82 APPENDIX D 8S3 Directions Please complete this questionnaire as carefully as possible. Your responses are Important to the project. Information you provide will be held by the researcher in strictest confidence. Please complete each auestion--do not leave any questions blank. Althou n this questionnaire may appear lengthy. it will go more quickly than it looks ecause it contains a lot of graphic content. 1. Are you involved in PATIENT INTAKE and DIA N SIS? (Please check one) Yes No Please check one box below which corresponds to your professional role. [:1 psychologist 1:] physician 1:] social worker nurse if you are a nurse. are you presently assigned to a team EZ] practice? (Please check one) Yes No Please think about the three people you communicate with most in the hospital about PATIENT HISTORY. INTAKE PROCESSES. and DIAGNOSIS. Write the professional roles of these people in the SQUARE. DIAMOND. AND'TRIAN below. e.g. nurse. physician. psychologist. social worker. or other . (If other. please indicate a professional role.) INTAKE NETWORK PART I @I _ ABBREVIATIONS NURSE PS PSYCHOLOGIST PH PHYSICIAN . SW SOCIAL WORKER Other Now. estimate as best you can how many minutes you usually spend communicating with each person about the typical INCONING PATIENT. 2. How many minutes do you usually spend communicating with the person in the SQUARE about a typical patient intake. minutes per patient 3. How many minutes do you usually spend communicating with the person in the DIAMOND about a typical patient intake? minutes per patient u. How man minutes do you usually Spend communicating with the person in the TRIANGL about a typical patient intake? minutes per patient The next three questions ask you to estimate how often each of the three people you have selected communicate wit EACH OTHER about the typical INCOMING PATIENT. 5. How many minutes do you estimate SQUARE SPENDS communicating with DIAMOND about a typical patient intake? minutes per patient 6. How many minutes do you estimate SQUARE Spends communicating with TRIANGLE about a typical patient intake? minutes per patient £34 7. How many minutes do you estimate DIAMOND and TRIANGLE communicate with each other about a typical patient intake? minutes per patient Please read the following scenario describing a system of patient intake which is dif- ferent procedurally from that which currently exists. This system is based on the use of a computer for patient check-in. ___ Scenario The patient is seated beside an attendant who will input data into a computer file with a keyboard. The attendant asks the patient a battery of auestions amassing a complete patient history. and enters res onses into the computer. The attendant Instructs the computer to sort through t ese data usin existing interview guide- ines for patient check-in by departments. Separate h star reports of the patient are compiled for the psycholo ist. physician. socia worker. and nurse and will be delivered to each person's of ice. Assume that your hospital has implemented the data management system described above. Think about the three people you expect you will communicate with most in the hospital about PATI:NT HISTORY. IN AKE PROCESSES. and DIAGNOSI fo lowin our hos ital's adoHtion 0‘ tfiTs 3 Stem. r e e professiona ro es 0 ese pe0p e n e UARE. . and TRIANGEE fielow. e.g. nurse. physician. psychologist. social worker or oghgr) . (If ”other”. please indicate a professional INTAKE NETWORK PART II w _ ABBREVIATIONS NU NURSE PS PSYCHOLOGAIST PH PHYSICIAN I sw SOCIAL WORKER "‘ Other 8. Are these the same people you listed above in iNTAKE NETWORK PART I (Please check only one box below.) KI] No [I] Yes (If yes. please go to No. 9) (If no. please check only one below) 3 are different 2 are different I is different Now. estimate as best you can how many m with each person about the typical lNCOM of this system. inutes ou PREDICT you will spend communicating ING PAT ENT following your hospital’s adoptIan 9. How many minutes do you predict you will spend communicating with SQUARE about a typical patient inta e? minutes per patient 10. How many minutes do you predict you will spend communicating with DIAMOND about a typical patient inta e? minutes per patient 11. How many minutes do you predict you will spend communicating with TRIANGLE about a typical patient inta e? minutes per patient 855 The next three auestions ask you to estimate how often you PREDICT each of the three pepple you have named will spend communicating with EACH OTHER about a typical PATIENT INTAKE followinggyour hospital’s adoption of this system. ‘ 12. How many minutes do you estimate SQUARE will spend communicating with DIAMOND about a typical patient intake? minutes per patient 13. How many minutes do you estimate SQUARE will spend communicating with TRIANGLE about a typical patient intake? minutes per patient 14. How many minutes do you estimate DIAMOND will spend communicating with TRIANGLE about a typical patient intake? minutes per patient Next. please think about the three people you communicate with most in the hospital about ASK matters regardless of the particular task. Who are the three pepple you communicate with most. Write their professional roles below. Later. you will be asked about non-task communication separately. TASK NETWORK PART I YO T’ ABBREVIATIONS NURSE PS PSYCHOLOSIST SW SOCIAL WORKER Other Now. estimate how often you normally communicate with each person about TASK matters. I. How often do you usually communicate with SQUARE about task matters? (Please check only one box below.) [I] several times per day [:3 daily .[:3 once or twice per week 1:] once or twice per month D less often 2. How often do you usually communicate with DIAMOND about task matters? (Please check only one box below.) [:Z] several times per day [:j] daily 1::3 once or twice per week [::3 once or twice per month [II] less often 3. How often do you usually communicate with TRIANGLE about task matters? (Please check only one box below.) [:j] several times per day £::] daily [:2] once or twice per week [II] once or twice per month [::] less often Next. estimate as best you can how often each of these pepple communicates with EACH OTHER about TASK matters. A. How often do you estimate SQUARE communicates with DIAMOND about task matters? (Please check only one box below.) [:Z] several times per day [:Zj daily [:Z] once or twice per week [Z] once or twice per month [:Z] less often [:1 never E36 5. How often do you estimate SQUARE communicates with TRIANGLE about task matters? (Please check only one box below.) [:1 several times per day 1:] daily III] once or twice per week 1:] once or twice per month [:1 less often III] never 6. How often do you estimate DIAMOND communicates with TRIANGLE about task matters? (Please check only one box below.) ‘EIj several times per day [I] daily [I] once or twice per week [I] once or twice per month [:l less often III] never Please read again the following scenario describing a system of patient intake which is different procedurally from that which currently exists. This system is based on the use of a computer for patient check—in: ___Scenarlo The patient is seated beside an attendant who will input data into a computer file with a keyboard. The attendant asks the patient a battery of auestions amassing a complete patient history. and enters res onses into the computer. The attendant instructs the computer to sort through t ese data usin existing interview guide- lines for patient check-in by departments. Separate h sto reports of the patient are compiled for the psychologist. physician. socia worker. and nurse and will be delivered to each person's of ice. Assume that your hospital has implemented the data mana ement system described above. Think about the three people you expect you will commun cote with most in the hospital about TASK matters following IQQI hosgital's adoBfiign Qf th}z system. Write the profes- sional roles of these peop e n e . . an elow. TASK NETWORK PART II yo __ ABBREVIATIONS NURSE PS PSYCHOLgGIST SW SOCIAL WORKER Ufher 7. Are these the same people you listed above in TASK NETWORK PART I? (Please check only one box below.) [I] No [I] Yes (If yes. please go to No. 8) (If no. please check only one below) 3 are different 2 are different I is different Now. estimate as best you can how often you PREDICT you will communicate with each person about TASK matters following your hospItal’s adoption of this system. 8. How often do you predict you will communicate with SQUARE about task matters? (Please check only one box below.) [II] several times per day IZZJ daily [:Z] once or twice per week [II] once or twice per month [I] less often 8'7 9. How often do you predict you will communicate with DIAMOND about task matters? (Please check only one box below.) [I] several times per day [I] daily [II] once or twice per week [II] once or twice per month 1 I less often 10. How often do you predict you will communicate with TRIANGLE about task matters? (Please check only one box below.) [I] several times per day [I] daily .EIJ once or twice per week [[I] once or twice per month [II] less often The next three auestions ask you to estimate how often you PREDICT each of the three peo 1e you have named will Spend communicating with EACH OTHER about TASK matters fol owing your hospitaI's adOptIon of this system. 11. How often do you estimate SQUARE will communicate with DIAMOND about task matters? (Please check only one box below.) [II] several times per day [II] daily [II] once or twice per week [I] once or twice per month [I] less often [I] never 12. How often do you estimate SQUARE will communicate with TRIANGLE about task matters? (Please check only one box below.) . [:1 several times per day [II] daily [:1 once .or twice per week [II] once or twice per month [I] less often [I] never 13. How often do you estimate DIAMOND will communicate with TRIANGLE about task matters? (Please check only one box below.) [II] several times per day [II] daily [II] once or twice per week [II] once or twice per month [II] less often [II] never FinallfiéNplease think about the three people you communicate with most in the hospital about -TASK matters. Who are the three people you communicate with most? Write their profeSSIonal roles below. NON-TASK NETWORK PART I ABBREVIATIONS YO r- NURSE Ps PSYCHOLOGIST PHYSICIAN sw SOCIAL WORKER Other Now. estimate how often you usually communicate with each person about NON-TASK matters. 1. How often do you usually communicate with SQUARE about non-task matters? (Please check only one box below.) [II] several times per day [II] daily [II] once or twice per week [I] once or twice per month [I] less Often 8E3 2. How often do you usually communicate with DIAMOND about non-task matters? (Please check only one box below.) [II] several times per day [II] daily III] once or twice per week [II] once or twice per month [II] less often 3. How often do you usually communicate with TRIANGLE about non-task matters? (Please check only one box below.) [II] several times per day [:2] daily [II] once or twice per week [II] once or twice per month [II] less often Next. estimate as best you can how often each of these people communicates with EACH OTHER about NON-TASK matters. A. How often do you estimate SQUARE communicates with DIAMOND about non-task matters? (Please check only one box below.) [II] several times per day [II] daily [II] once or twice per week [II] once or twice per month [II] less often [II] never 5. How often do you estimate SQUARE communicates with TRIANGLE about non-task matters? (Please check only one box below.) 1:] several times per day [:3 daily I:' once or twice per week :1 once or twice per month 1:] less often [3 'never 6. How often do you estimate DIAMOND communicates with TRIANGLE about non-task matters? (Please check only one box below.) [II] several times per day [2:] daily [2:] once or twice per week [II] once or twice per month [II] less often [II] never Please read the following scenario describing a system of patient intake which is dif- ferent procedurally from that which currently eXIsts. This system is based on the use of a computer for patient check—in: ___Scenarlo The patient is seated beside an attendant who will input data into a computer file with a keyboard. The attendant asks the patient a battery of auestions amassing a complete patient history. and enters responses into the computer. The attendant instructs the computer to sort through these data using existing interview guide- lines for patient check-in by departments. Separate h story reports of the patient are compiled for the psycholo ist. physician. social worker. and nurse and will be delivered to each person’s of ice. Assume that your hospital has implemented the data management system described above. Think about the three pepple you expect you will talk with most in the hospital about NON-TASK matters followin our has ital's adoption of this s stem. Write the profes- sional roles of these people in the SQUARE. DIAMOND. anO IRIANGIE below. NON-TASK NETWORK PART II ABBREVIATIONS YO NURSE PS PSYCHOLOGIST PH PHYSICIAN SW SOCIAL WORKER Other £39 7. Are these the same people you listed above in NON-TASK NETWORK PART I? (Please check only one box below.) [II] No [II] Yes (If yes. please go to No. 8) (If no. please check only one below) 3 are different 2 are different I is different Now. estimate as best you can how often you PREDICT ou will communicate with each person about NON-TASK matters followinggyour hospitg 's adoption of thls system. 8. How often do you predict you will communicate with SQUARE about non-task matters? (Please check only one box below.) [II] several times per day [II] daily [II] once or twice per week [II] once or twice per month [II] less often 9. How often do you predict you will communicate with DIAMOND about non—task matters? (Please check only one box below.) [II] several times per day [:2] daily [2:] once or twice per week [II] once or twice per month [I] less often 10. How often do you predict you will communicate with TRIANGLE about non—task matters? (Please check only one box below.) I I several times per day [II] daily [II] once or twice per week [II] once or twice per month [II] less often The next three auestions ask you to estimate how often you PREDICT each of the three people you have named will spend communicating with EACH OTHER about NON-TASK matters following your hospital’s adaption of this system. 11. How often do you estimate SQUARE will communicate with DIAMOND about non-task matters? (Please check only one box below.) . [3 several times per day [:1 daily [:1 once or twice per week El once or twice per month :1 less often 12. How often do you estimate SQUARE will communicate with TRIANGLE about non-task matters? (Please check only one box below.) [I] several times per day [I] daily [I] once or twice per week [II] once or twice per month [II] less often 13. How often do you estimate DIAMOND will communicate with TRIANGLE about non-task matters? (Please check only one box below.) [II] several times per day [II] daily [:I] once or twice per week [II] once or twice per month [II] less often Please read for a final time the following scenario describing a system ofTRatient intake which is different procedurally from that which currently exists. is system 15 based on the use of a computer for patient check-In: 9C) SCOHOI‘IO T'— The patient is seated beside an attendant who will input data into a computer file with a keyboard. The attendant asks the patient a battery of auestions amassing a complete patient history. and enters responses into the computer. The attendant Instructs the computer to sort through t ese data usin existing interview guide- lines for patient check-in by departments. Separate h star reports of the patient are compiled for the psycholo ist. physician. socia worker. and nurse and will be delivered to each person's of ice. ate whether you STRONGLY AGREE. AGREE. are a ic EE with the statement made. For the followin nine items p le se ind UNDECIDED. DISAG EE. or STRONGLY DISAGR l. The proposed system would be implemented smoothly. SA A U D SD 2. Agéeg tge proposed system is implemented. changes in routine procedures will be U re . SA A U D SD 3. There will be administrative barriers to adoption. SA A U D SD A. There will be individual attitude barriers to adoption. SA A U D SD 5. The proposed system will improve patient intake. SA A U D . SD 6. The proposed system will be Implemented auickly. SA A U D SD 7. There will be "bugs” in the prooosed system. SA A U D SD 8. The prooosed system will improve administrative efficiency. SA A U D SD 9. The proposed system will improve service to patients. SA A U D SD 9]. Listed below are pairs of adjectives often used to describe pepple or things. As shown in the box. the spaces indicate the degree to which something is related to one or the other ends of a scale described by these adjectives. very auite only eaually only auite very closely closely slightly related slightly closely closely related related related to fair related related related to fair to fair to fair or unfair to unfair to unfair to unfair fOIr: : : : : : z : unfair Please rate TELEPHONE COMMUNICATION according to the Idegree to which it is related to one or the other ends of each of the scales elow. In making our iadgements. please mark in the middle of the spaces. check only one space for eac sca e. and do not omit gnylscalesth Please make judgements according to your feelings about telephone communi- a on n e TELEPHONE COMMUNICATION IN VA 1. fair _: _: _: _: _: _: __: unfair weak _: _: _: __: _: _: _: Strong worthless ____; ____: ____: ____: ____: ____: ____: vaanble pleasant ____: ____: ____: ____: ____: ___I; ____: unpleasant COTIDIete _: ___: _: _: _: _: _: Incomplete bad _: _: _: _: _: _: _: 900d safe ____: ____: ____: ____: ____: ____: ____: dangerous humorous ____: ____: ____: _.__= ____: ____: ____: serious Clean _: _: _: _: _: _: _: dirty tasteless _: _: _: _: _: _: ___~: tCISterI contemporary ____: ____: ____: ____: ____: ____: ____: noncontemporary positive ___: _: _: _: ___: _: _: negative soft _: _: _: _: _: _: _: hard Optimistic : : : : : : : pessimistic harmful : : : : : : : beneficial 592 Please rate TELEVISION according to the degree to which it is related to one or the other ends of each of the scales below. 2. fair weak worthless pleasant complete bad safe humorous clean tasteless contemporary positive soft optimistic harmful Please rate FACE-TO-FACE to one or the other ends 3. fair weak worthless pleasant complete bad safe humorous clean tasteless contemporary positive soft Optimistic harmful TELEVISION APPPLICATIONS IN VA U _- FACE-TO-FACE COMMUNICATION IN COMMUNICATION according to the of each of the scales below. VA degree to which unfair strong valuable unpleasant incomplete 900d dangerous serious dirty tasteful noncontemporary negative hard pessimistic beneficial it is related unfair strong valuable unpleasant incomplete 900d dangerous serious dirty tasteful noncontemporary negative hard pessimistic beneficial fi 533 Please rate WRITTEN MEMOS according to the degree to which it is related to one or the other ends of each of the scales below. WRITTEN MEMO COMMUNICATION IN VA A. fair ___: ___: _fi: ___: _: __: ___: unfair weak ____: ____: ____: ____ ____ ____. ____ strong worthless ____. ____, __I_, ____ [____ ____. ____ valuable pleasant _____ ____ ____. ____ ____ ____. ____ unpleasant complete ____ ____; ____ ____ ____ ____: ____ incomplete bad E ___: ‘ _5 __ ‘i \ 900d safe ____ ____: ____ .____ .I___ ____: ____ dangerous humorous ____: ____: ____: ____. ____. ____, ____. serious clean .____ .____. .____ ____ .____ ____ ____ dirty tasteless ____ ____. ____ ____ ____ ____. ____. tasteful contemporary ____ ____: ____ ____ ____. ____ noncontemporary positive ____ ____ ___. ____ ____ ____: ____ negative 80ft E __‘i a _ _‘ \: .5 hard Optimistic _____ ____; ____ ____ ____. ___I: ____ pessimistic harmful ____ ____: ____ ____ ____ ____: ____ beneficial Please rate COMPUTER according to the degree to which it is related to one or the ca e other ends of each of the s 5 below. computer 5. fair ____: ____: ____: ____: ____: ____: ____: unfair weak ____: ____: ____: ____: ____: ____: ____: strong worthless ____: ____; ____: _I__: ____; ____: ____: valuable pleasant ____: ____: ____: .____: ____; ____: ____: unpleasant complete ____ ____ ____ ____: ____ ____ ____ incomplete bad ___: ‘: ___: __: “z _‘2 “1 900d safe ____: ____: ____: ____: ____: ____: ____: dangerous humorous ____: ____: ____: ____: ____: ____: ____= serious clean ___I: ____: ____: ____: ____: ____: ___I: dirty tasteless ____: ____; ____: ____: ____: ____r ____: tasteful contemporary _5: x: a: _‘z I_: 5: ‘1 noncontemporary positive “__: E: k: a: 5: ~53 ~: negative soft ____: ____: ____: .____: ____: ____: ____: hard optimistic \: _h: ___: 5: ‘: __: “1 DeSSImIStIC harmful - : : - : ; ; beneficial I I ~ 5.. ——~ ‘ ~—.— ‘ *— Thank you for your cooperation In this important project. 94 B I BL I OGRAPHY BIBLIOGRAPHY Argyris, Chris. Intervention Theory and Method; 3 Behavioral Science View. Reading: Addison-Wesley Publishing Co., 1970. Bashshur, Rashid; Armstrong, Patricia; and Youssef, Zakhour Telemedicine: Explorations in the Use of Telecommunica- tions for Health Care. Sprifigfield: Charles C. Thomas: 1975. Bretz, R. Media for Interactive Communication. Beverly Hills: Sage Publications, 1983. Brown, J.H.U. Telecommunication for Health Care. Boca Ra- ton: CRC Press, 1982. Coleman, James S., and others. Medical Innovation: A Diffu- sion Study. New York: Bobs-Merril, 1966. Dordick, H.S., Goldman, R.J. and Hanneman, G.J. "Telecommu~ nications and Vocational Rehabilitation: A User's Guide to Breaking the Barriers." Annenberg School of Communi- cation at the University of Southern California , for the Office of Research and Demonstrations Rehabilita- tion Service Administration, Department of Health, Educa- tion and Welfare, July 1978. Danowski, James A. "An Information Theory of Communication Functions: A Focus on Informational Aging." Unpublished doctoral dissertation. East Lansing, Michigan: Michigan State University, 1975. Danowski, James A. "Automated Communication Network Analysis." Paper presented to the Information Systems Division of the International Communication Association, Dallas, 1983. Danowski, J.A. "Personal Network Integration: Infographic, Psychographic and Demographic Characteristics," paper presented to the Information Systems Division of the International Communication Association. San Francisco, 1984. Danowski, James A. and Van Engen, M. "Readership of Employee Publications as a Function of Interpersonal Communication." Monograph, School of Journalism, University of Wisconsin, 1983. 95 96 _ Elton, Martin C.J. "The Use of Field Trials in Evaluating Telemedicine Systems." Chapter in Telemedicine: Explorations in the use of Telecommunications In Health Care, ed. by Rashid Bashshur, Patricia Armstrong, and Zakhour Youssef. Springfield:Charles C. Thomas, 1975. Isaac, Stephen , with Michael, William B. Handbook in Research and Evaluation. San Diego: R. R. Knapp, 1971. Lauman, Edward. Bonds 2: Pluralism: The form and Substance of Urban Social Networks. New York: Wiley, 1973. Lawrence, Paul R. "How to Deal with Resistence to Change." Chapter in Organizational Behavior and Administration, ed. by Lawrence, Paul R., Barnes, Louis B.; and Lorsch, Jay W. Homewood: Irwin, 1976. (Article first published May-June 1954 Harvard Business Review. Maxmen, Jerrold S. The Post-Physician Era. New York: Wiley, 1976. ' Nunnaly, Jum C. Psychometric Theory. New York: McGraw- Hill, 1978. Osgood, Charles; Tannenbaum, Percy; and Suci, George. The Measurement of Meaning. Urbana: University of Illinois Press, 1957. Park, B. An Introduction to Telemedicine: Interactive Tele- vision for Delivery of Health Services. New York: The Alternative Media Center at New York University School of Arts, 1974. Reid, Alex New Direction in Telecommunications Research. Bethesda, MD: ERIC Document Reproduction Service, ED 062 803, 1971. Rice, R. and Case, D. "Electronic Message Systems in the University: A description of Use and Utility.: Journal of Communication, 33 (1983) 131151. Rogers, E. with Shoemaker, F. Communication of Innovations. New York: Free Press, 1971. Rogers, E. with Shoemaker, F. Communication Networks. New York Free Press, 1981. Shoemaker, F. Floyd. Anal sis of Structural Effects 92 Diffusion. Doctoral Dissertation, Department of Communi- cation, Michigan State University, 1971. 97 Schomish, T.P. The Effects of Personal Communication Networks 92 Members Voting in a Merger pf Agricultural Unpublished doctoral dissertation, Cooperatives. University of Wisconsin, 1983. Simon, Herbert A. The Sciences of the Artificial. Cambridge: M.I.T. Press, 1969. Williams, Frederick. The Communications Revolution. Beverly Hills: Sage Publications, 1982. Williams, Frederick . The New Communications. Belmont: Wadsworth Publishing Co., 1984. HICH IonN STATE UNIV. LIBRARIES IWIN“llli"IWIUHllWWI!WIVWIIIIHHII 93106643574 23122