A STUDY OF THE INTERRELATIONSHIP BETWEEN THE SOCIAL STRUCTURE AND THE COGNITIVE BELIEF SYSTEM OR “CULTURE" OF A SOCIAL UNIT Thesis TOP ”19 Degree oI M. A. MICHIGAN STATE UNIVERSITY Margaret Brophy I976 IIIIIIIIIIIIIIZIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII 23 10446 3140 3 LIB RA R Y Michigan State University 3mm négw'mflfi’ C sac gauwem Let" A M ABSTRACT A STUDY OF THE INTERRELATIONSHIP BETWEEN THE SOCIAL STRUCTURE AND THE COGNITIVE BELIEF SYSTEM OR "CULTURE" OF A SOCIAL UNIT BY Margaret Brophy This study investigates the interrelationship between the social structure and the cognitive belief system or "culture" of a social unit. While many theorists have discussed this relationship, this thesis has been based largely on the work of Emile Durkheim who specifies that the cognitive belief system of any social unit varies with "the nature and number of channels of communication." Durkheim further stipulates that in order to accurately describe social and cognitive structure we must consider the "aggregate in its totality." In other words, we are not concerned with individuals per se, but rather, the relationships that exist among the members or elements of the social unit and how the aggregate social structure is related to the cultural system. The main assumption of this thesis is that the pattern- ed similarities of people as perceived by the cultural aggregate will correspond to the social structure of that system. In other words, there will be a general corres- pondence between the pattern of similarities perceived Margaret Brophy among members of a social unit and their position in the communication network. From this general reasoning three hypotheses were derived. The first two hypotheses represent a dyadic test of conditions under which networks of interaction vary with the similarity/dissimilarity judgments or "culture" of the system: 1) with frequency of communication; and, 2) the distance or number of links between any pair of individuals. Hypothesis 3 deals with the frequency of interaction of an individual with the group as a whole. The sample consisted of all faculty and funded graduate students in the Department of Communication at Michigan State University. Respondents were provided with the names of all their colleagues in the department and requested to estimate their frequency of communication with each individ- ual during the previous academic term. An eight point rating scale was provided. Additionally, respondents were given 15-16 unique pairs of faculty members and requested to make similarity/dissimilarity judgments by means of direct paired comparisons. Respondents based their esti- mates of dissimilarity as a ratio of a standard distance provided by the investigator. And finally, questionnaire items were used to measure unidimensional attributes as reliability and validity checks of the measurement instru- ments. Data were transformed into indices, and zero order correlations and multiple regression procedures were used to analyze the data. Margaret Brophy The statistical tests of Hypotheses l and 2 fail to indicate any significant relationship between communication frequency, distance or links separating the members of the social unit, and perceived distance in the aggregate (cul- tural) belief system. Although each of these indicators of "distance apart" or "discrepancy of position" in the commun- ication network may well be a reliable and valid indicator of some aspect of overall "distance," there seems reason to suspect, due to low correlations, that these characteristics are not of sufficient sc0pe to accurately describe the totality of the relationship. The results of the statistical tests of Hypothesis 3 clearly indicate a high degree of correspondence between the integrativeness of an individual in the communication network and the perceived similarity of that individual from all other members of the social unit. Additionally, based on the Multiple Correlations of Network Integration and the Galileo Coordinate values, we can conclude that the degree to which an individual is integrated into the com- munication network is clearly an attribute which people recognize when comparing individuals. Given this pattern of findings, and the reliability and validity checks of the measurement instruments, we can not reject the theory based on this analysis. A STUDY OF THE INTERRELATIONSHIP BETWEEN THE SOCIAL STRUCTURE AND THE COGNITIVE BELIEF SYSTEM OR "CULTURE" OF A SOCIAL UNIT BY Margaret Brophy A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Communication 1976 Accepted by the faculty of the Department of Communication, College of Communication Arts, Michigan State University, in partial fulfillment of the requirements for the Master of Arts degree. X12354? Director of Thesis Guidance Committee: Sngz>jxf) , Chairman L17 ['rm9.. (_,./I' pk (if. p.) .' ,LR‘A/h‘ . me If Wm ACKNOWLEDGMENTS Joe, Marylou, Joey and Johana Woelfel whose family life was often interrupted in order that my thesis may be completed. All the BrOphy and Gehrke family members who provided the moral encouragement I needed to finish. Hope Brophy Jr. and Randy Dennis who spent many hours collecting questionnaires. Ruth Langenbacher whose skills and patience made this document presentable. Peter O'Rourke and Senator Fitzgerald for their per— severance and interest in my work. To all those committee members who shared their exper- tise in the field. Especially to JOHN GEHRKE who must be happier than I am to see this document finished. Thank you all. ii TABLE OF CONTENTS Chapter A Page I I. GENERAL THEORY . . . . . . . l A. Social Structure . . . . l B. Culture . . . . . . . 2 C. Interdependence of Culture and Social Structure . . . . 3 II. SPECIAL THEORY . . . . . . . 6 A. Social Structure . . . . 6 B. Culture . . . . . . . 7 C. Interdependence of Culture and Social Structure . . . . 9 Major Premise . . . . . 10 ° Minor Premise . . . . . 10 III. PREVIOUS RESEARCH . . . . . . 10 A. Networks . . . . . . ll B. Person Perception . . . . l6 Formulation of Hypotheses . . . 28 II PROCEDURES AND METHODS . . . . . . 32 III RESULTS . . . . . . . . . . 40 Reliability and Validity of the Instruments . . . . . . . . 53 Summary . . . . . . . . . 62 IV CONCLUSION . . . . . . . . . 63 iii Chapter Counts, APPENDICES . . .- . . . . Appendix 1: Questionnaires l-a. Personal Communication Contact Questionnaire . . . l-b. Galileo Questionnaire . . l-c. Unidimensional Scaling Questions Appendix 2: Statistics 2-a. Galileo Means, Standard Deviations, Variances, Skewness, Kurtoses, Minimum-Maximum Values, Range 2-b. Network Distance Matrix (Link Analysis) . . . . 2-c. Galileo Means Matrix . Z-d. Galileo Coordinate Values . Appendix 3: Galileo Plots 3-a. X - Y Plane . . . . 3-b. Y - Z Plane. 3-c. Plot of First Through Real Dimensions . . . . REFERENCES . . . . . iv Page 68 71 73 74 75 77 78 79 80 81 82 83 Table 1. LIST OF TABLES Descriptive Statistics For The Variables Referred To In HypotheSes 1 and 2 . . . . Descriptive Statistics For The Variables Referred To In Hypothesis 3 . . . . Correlation Coefficients Among The Variables Referred To In Hypotheses 1 and 2 . . . . Correlation Coefficients Among The Variables Referred To In Hypothesis 3 . . . . . Correlation Coefficients For Years With The Department x Network Integration And . Galileo Alienation . . . . . . . . Standardized Partial Regression Coefficients, Multiple Correlation Coefficients, Tests of Significance, and Coefficients of Determina- tion Of Network Integration On Galileo Co- ordinate Dimension Values . . . . . Correlation Coefficients For The Unidimen- sional Scales, Network Integration Scores, Galileo Factor Loadings And The Galileo Alienation Scores . . . . . . . Standardized Partial Regression Coefficients, Multiple Correlation Coefficients, Tests of Significance, and Coefficients of Determina- tion of Ph.D. Committee Membership on Galileo Coordinate Dimension Values . . . . . Standardized Partial Regression Coefficients, Multiple Correlation Coefficients, Tests of Significance, and Coefficients of Determina- tion of M.A. Chairmanship on Galileo Coordinate Dimension Values . . . . . Page 41 43 46 47 51 52 55 59 59 Table . Page 10. Standardized Partial Regression Coefficients, Multiple Correlation Coefficients, Tests of Significance, and Coefficients of Determina- tion of Academic Rank on Galileo Coordinate Dimension Values . . . . . . . . 60 ll. Standardized Partial Regression Coefficients, Multiple Correlation Coefficients, Tests of Significance,.and Coefficients of Determina- tion of Ph.D. Chairmanship on Galileo Coordinate Dimension Values . . . . . 60 12. Standardized Partial Regression Coefficients, Multiple Correlation Coefficients, Tests of Significance, and Coefficients of Determina- tion of M.A. Committee Membership on Galileo Coordinate Dimension Values . . . . . 61 13. Standardized Partial Regression Coefficients, Multiple Correlation Coefficients, Tests of Significance, and Coefficients of Determina- tion of Years with the Department on Galileo Coordinate Dimension Values . . . . . 61 2-a Galileo Means, Standard Deviations, Variances, Skewness, Kurtoses, Counts, Minimum-Maximum Values, Range . . . . . . . . . 75 2-b Network Distance Matrix For Faculty . . . 77 2-c Aggregate Means Matrix of Faculty Members . . 78 2-d Coordinate Dimension Values for Faculty . . 79 vi Figure 3-a. 3-b. LIST OF FIGURES Galileo Plots . . . X - Y Plane . . . . First Through Real Dimensions vii Page 80 81 82 CHAPTER I I. GENERAL THEORY All social units, whether they be dyads, groups, formal organizations, etc., consist of relationships which have be- come interwoven into ongoing patterns of observable regulari- ties. These regularities are manifested in two ways: (1) as a social structure whose elements are organized with roles and activities interrelated; and, (2) as a culture or col- lective cognitive belief system that is associated with patterns or social structure within the process of social organization. Any theory of social organization must invari— ably incorporate the two, and any accurate description of such must be concerned with their relationship. A. SOCIAL STRUCTURE Social structure emerges as interactions and relation— ships evidenced by predictable regularities or arrangements. Social.structure thus consists of patterned recurrent social interactions that maintain their uniformity with some degree of stability over time. This concern is evident, for example, in the following definition of social systems proposed by Parsons and Shils (1951:107): The most general and fundamental property of a system is the interdependence of parts and variables. Interdependence consists in 1 2 the existence of determinant relationships among the parts or variables as contrasted with randomness or variability. In other words, interdependence is order in the relationship among the components of the system . . . . This same notion is also evident in Mead's (Miller, 1973: 43) definition of a system as a: . . . set of entities or concepts so related and connected as to form a unity or a whole . . . and such that there is a sustaining relationship between the entities so related . . . . B. CULTURE Shared cultural ideas in turn emerge from social structure, as the participants communicate with each other about their activities and create common ideas of how social life should and can be organized. Cultural ideas largely reflect and express their underlying patterns of social structure. Although discrepancies often exist between ideas and practices, these differences cannot become too great without imposing severe strains and conflicts on both the social structure and its accompanying culture. As a conse- quence of the emergence of cultural ideas from collective social life, patterns of social order gain unity, stability through time, and functional effectiveness in achieving goals. In addition, each group has a sub-culture of its own, a selected and modified version of some parts of the larger culture. The significance of these sub-cultures lies not so much in what they add to the larger culture as in the fact that without its own culture no group would be more than an assemblage of persons. The common mean- ings, the definitions of the situation, the norms of 3 belief and behavior - all these go to make up the culture of the group. Durkheim (The Division of Labor 1947:14) a half- century ago described this phenomenon and its significance: . . . when a certain number of individuals in the midst of a political society are found to have ideas, interests, sentiments, and occupations not shared by the rest of the population, it is inevitable that they will be attracted toward each other under the in- fluence of these likenesses. They will seek each other out, enter into relations, associ- ate, and thus, little by little, a restricted group, having its special characteristics, will be formed in the midst of the general society. But once the group is formed, a moral life appears naturally carrying the mark of the peculiar conditions in which it was developed . . . Culture, then, is a term applicable not only to the larger society but to its sub-groups as well, and is, indeed, a fundamental feature of the interaction of human beings. C. INTERDEPENDENCE OF CULTURE AND SOCIAL STRUCTURE: The culture of the organization will always influence and shape, as well as reflect, its underlying patterns of social order. Nevertheless, culture and social order will inevitably interpenetrate each other, so that both aspects become fused into an overall process of social organization. Theoretically this notion is not new to the social scientist, (Durkheim 1963, 1938; Mead 1938, 1934, 1973; Parsons 1951) for example, Parsons (l951:5) states that: a social system consists in a plurality of individual actors interacting with each other in a situation . . . whose relation to their situation, including each other, is defined and mediated in terms of a system of culturally structured and shared symbols. . . . Each is 4 indispensable to the other in the sense that without culture there would be no social system and so on around the roster of logical possibilities . . . . Mead (Miller 1973:45) also supports this notion of mutual causality when he states that: . . . Men confer meanings on their environ- ments, which in turn makes their behavior different from what it would have been with— out them. Through meanings the environment adapts to men and men adapt to the environ- ment . . . . The process of social organization necessarily involves at least two interacting persons, but it is not a character- istic of either of them as individual personalities. It arises, rather, through their interaction and communication, as a social order emerges from recurrent social relationships, and as the participants create a shared body of cultural ideas. This is precisely what Durkheim (1963:19) posits when he describes society as collective representations: . . . No doubt each individual contains a part, but the whole is found in no one. In order to understand it as it is one must take the aggre- gate in its totality into consideration . . . . Society does not depend upon the nature of the individual personality. In the fusion from which it results all the individual character- istics, by definition divergent, have neutral- ized each other . . . . Mead (Miller 1973:34) concurs when he states: . . . to share perspectives is to share attitudes, and the attitudes of the community, in their widest meaning, are summarized in terms of the categories with reference to which the experiences of members of the community are interpreted . . . . These common attitudes are not built up out of individual attitudes . . . 5 individual perspectives emerge from within the common perspective, and all are in nature, none in the individual . . . . A phenomenon which exhibits properties that are dis- tinctly its own must have an existence of its own. For example, the substance water cannot accurately be described in terms of its parts, but rather the combination of two parts hydrogen with one part oxygen becomes a unique sub- stance of its own with its own characteristics. This is analogous to describing social units, in that the properties of groups or aggregates may be studied as phenomena in their own right and must be empirically explained as an objective social reality. This thesis will investigate, therefore, the interrela- tionship between culture and social structure. In general, it begins with the assumption that (l) cultural beliefs form an organized system or pattern which must be considered as an aggregate entity rather than a simple collection of indi- vidual representations; and (2) that social structure as well must be considered as a systematic whole entity; and finally (3) that the relationship between culture and social structure must therefore be studied as the relationship between two organized entities. 6 II. SPECIAL THEORY Specifically, this thesis will be concerned with the interrelationship between the culture and the social struc- ture of the Department of Communication at Michigan State University. Any attempt to test empirically the theory presented previously must begin by isolating those aspects of social structure and culture which: (1) display observ- able regularities which are intrinsic to that organization under study; and, (2) are so culturally specific as to be identifiable. A. SOCIAL STRUCTURE Social structure shall be described in terms of the patterns (networks) of communication behavior among the members (elements) of the department. Such networks de- scribe how information passes through a system. Structure is not determined by looking at the sum of the communication of an individual, but rather, how individuals fit into the system, that is, how they are related to other elements in the system. Theoretical support for describing structure in these terms is provided by Durkheim (Simpson, 1963:17-19) when he states: . . . Society has for its substratum the mass of associated individuals. The system which they form by uniting together, and which varies according to their geographical disposition and the nature and number of their channels of’communication, is the base from which social life is raised. The representations which form the network of 7 social life arise from the relations be— tween the individuals thus combined or the secondary groups that are between the in- dividuals and the total society . . . . The resultant surpasses the individual as the whole, the part . . . (emphasis added). B. CULTURE The impact of culture and society on cognition has long been a point of convergence for many diverse fields of science; e.g., psychology, sociology, anthr0pology, etc. It is generally believed that shared cultural ideas emerge as individuals communicate with each other and establish norms of belief and behavior and common meanings (defini- tions) of situations that arise. But an assertion of this degree of generality is not very useful and needs to be translated into terms which allow the formulation of ques— tions of an empirical nature. In this thesis culture will be described in terms of person-perception, or rather, how the members define or perceive the other members in the department. Although this procedure does not encompass the totality of the phenomena "culture," it can be considered a crucial aspect which influences and is influenced by inter- action. In all social units (society, formal organizations, dyads, groups, etc.) interactions lead to the formation of common definitions of those objects with which the social unit deals. Just as we create structure in the inanimate world by categorizing stimuli into objects and their at- tributes, so we create order in the world of people by 8 categorizing them and their behavior. It is true that people, as visual stimuli or objects in and of themselves, require the same psychological processes as any other class of physical objects. But, at the same time, this process is far more complex in that we attribute more emotional significance to this class of stimuli. As Tajfel (1964, p. 323 states: "The social world may be less predictable than the physical, but what predictability there is in the social world is achieved by the same processes that are in- volved in the perception of the physical world." Whether we are making judgments based on actual be- havior, or inferential leaps of logic, we attribute to other people psychological traits such as, intentions, purposes, attitudes, motives etc. Additionally, when viewing inter- actions we perceive psychological qualities which we use to describe relationships between people--friendship, love, hate, power, etc. Tagiuri (Handbook of Social Psychology, 1969:396) further adds that: . . . We attribute to a person prOperties of consciousness and self-determination, and the capacity for representation of his environ— ment, which in turn mediates his/her actions. Granted the perceiver may, through his/her own presence and behavior in the phenomenal world of the other, cause changes in the way in which the person whose state he/she is trying to judge presents himself/herself. This of course is much different from the way in which a rock is a source of cues for a perceiver . . . . One other unique characteristic of person perception is worth noting. Because the perceiver and the perceived 9 object (a person) are similar, the perceiver can rely on past experiences in inferring or judging intentions. In practice this excelerates the perception process and allows the perceiver to make judgments with greater confidence. In this thesis, the Communication Department is the social unit to be analyzed, using the Network as a measure of interaction, and the department members as the percep- tual objects with which the social unit deals. It should be the case that persons who interact in the communication de- partment will form common definitions of those persons, and other objects, insofar as they interact about them. Also, people who interact frequently will form similar ideas and be perceived as similar by other members in the department. C. INTERDEPENDENCE OF CULTURE AND SOCIAL STRUCTURE Because when individuals communicate they share infor- mation and experiences, it should be the case that the culture(collective representation) will correspond with the networks of communication behavior. Durkheim (Simpson 1963:17) supports this contention when he states that " . . . systems (collective representations) . . . vary according to . . . the nature and number of channels of communication . . . ." In other words, an accurate description of any social unit must look at the interdependence between the social structure and the cultural structure. It should also be restated that we can talk about structure only in terms of the psychological and behavioral interrelationships 10 within the aggregate. Major Premise - This means that persons are not the key variable understudy, but rather, locations in a common net- work; i.e., persons or groups in the same or equivalent loca- tions in a communication network will develop equivalent conceptions of the collective representation. Minor Premise - Persons who are equivalent (i.e., hold equivalent conceptions of the collective representation) will be seen as similar by others (insofar as those others have contact with them). Following from this, there will be a general corres- pondence between the pattern of similarities perceived among members of an organization and their position in the commun- ication network; i.e., persons close to each other in a communication network will be seen as similar to each other, and persons far apart in the communication network will be perceived as dissimilar. In other words, the patterned similarities of people as perceived by the cultural aggregate will correspond to the pattern of intercommunication found in the network. III. PREVIOUS RESEARCH As stated earlier, the purpose of this thesis is to investigate the interrelationship between the culture (or cognitive belief system) and the social structure of a social organization. In order to be amenable to empirical test a ll theory of this degree of generality must be transposed. Fellowing the work of Durkheim (Simpson 1963:17-19) social structure will be described in terms of networks or "channels of communication." Because Durkheim (The Division of Labor 1947:14) contends that groups display unique sub-cultures of their own (which are modified versions of the larger culture) the literature review shall deal with the group level of analysis. Due to the encompassing nature of the term culture, the more specific relationship of person perception shall be the focus of research and the subject matter of the literature review. It should be noted that we are dealing with the process by which individuals perceive others in slightly different terms than the literature in this field generally focuses. In other words, we are not concerned with person perception as a phenomenon in and of itself, but rather, its interrelationship with the social structure that evolves along with it. A. NETWORKS To say that the parts of a system are mutually dependent is to say that the behavior of each element is predictable in some degree from the behavior of the others, and that the degree of structure in a situation could be measured if the interdependence among the variables could be quantified. Despite the widespread concern with the structure of social behavior, there has not emerged a standardized measurement 12 instrument. This notion of interdependence among parts of a system has received particular attention in the study of small groups. In a review of definitions of "group" Cartwright and Zander (1968:48) suggest that the most crucial determiner of group composition is "the collection of individuals who have relations to one another that make them interdependent to some significant degree." However, studies dealing with small group behavior have differed in their selection of the type of interdependence which is examined. In recent years we have seen an upsurge of studies focusing upon interdependence in terms of communication net- works. (See Glanzer and Glaser 1959, 1961; Collins and Raven 1969 for an extensive review of the literature.) As defined by Farace and Monge (1973:2) "Communication networks arise whenever recurring patterns of interaction occur; these networks make it possible for different types of information to move throughout the system with varying rates of trans- mission, levels of distortion and degree of impact." The significance of such networks can not be underestimated, for such a system coordinates activities, allows for a method of checks and balances, and often is the ingredient which de- termines the success or demise of a group. There has been much criticism of small group research in that the results found in carefully restricted experiment- al settings may not be applicable to on-going groups found in l3 non—artificial settings (Guetzkow 1965). For example, the variable group size has been shown to effect: (1) the number of direct (one step) communication links between members (Rome and Rome 1961; Weick 1969); (2) Requirement of a leader (Weick 1969); (3) Frequency and duration of commun- ication on an individual basis (Weick 1969);(4) composition of the group e.g. as group size increases from three to twelve members it is more likely that members may need to reform into smaller groups. One of the major factors which restricts the utility of small group research in constructing valid, empirically rigorous theory, is the pragmatic question of time. Do groups which work together over a long period of time dis- play different behavior patterns than groups assembled ran— domly for a specific task requiring a limited involvement of time? Both Lorge et_al. (1958) and Burgess (1969) found be- havior differences which were directly related to the length of time groups existed. The types of networks (e.g. circle, wheel etc.) which differentiated effective communication h trial. patterns on a short term-basis disappeared by the 500t Scientists in the field of organizational communication have found contradictory results when replicating small group studies in organizational settings (Cohen, Robinson and Edwards 1969). One reason for this discrepancy could be that artificial groups generally engage in less complex and less meaningful tasks than groups in actual organizations. 14 When we move away from the small group research and deal directly with organizational studies, many new problems arise which hamper theory construction. For example, rather than look at the total organization as a process in which groups interact to accomplish set goals, studies have focused on isolated events employing zero-order correleations (e.g. rate of communication initiated correlated with rank; upward, lateral, downward communiction correlated with rank). As Danowski (1974:14) points out, from the research in this field we have formulated "a large number of unintergrated and perhaps unintegratable two-variable propositions." First attempts to study organizations in terms of an on- going process generally relied upon the formal organizational chart for defining interaction patterns and group formation. Thompson (1956); Berlo gt_al. (1972) and Danowski (1974) found that informal communication patterns (which do not always follow the formal organizational chart) were more accurate in describing how organizations process information and corrdinate activities. In order to advance the study of group processes in formal organizations many sociometric techniques have been developed. (See Farace gt_al, 1973, for an extensive review of these techniques and their limitations.) Two major problems have hindered the utility of these techniques: (1) due to the excessive computer memory required, organizations lander study must be limited in size; and (2) it is difficult 15 to isolate effectively sub-groups in large organizations. Richards (1971), Richards, Farace and Danowski (1973) have developed a computerized program that exceeds any pre- vious techniques by minimizing computer cost and clearly defining what constitutes group formation: (Richards 1974a:21). 1. There must be at least three members 2. Each member must have at least 50% of his/her communication with other members of the same group 3. There must be some path, lying entirely within the group, from each member to each other member 4. There must be no single node (or small set of nodes) which, when removed from the group, causes the rest of the group to fail to meet any of the above criteria 5. There must be no single communication link (or subset of links) which, if cut, causes * the group to fail to meet the above criteria. What makes the NEGOPY program such a useful tool in developing sound theory, is that once a system is accurately described the structural properties that emerge may be re- lated to other non-structural variables. The researcher may also transcend the minimal level of analysis, the indi- vidual, up to the system as a whole. * For a discussion of "Centrality and Leadership see Johnson (1975); Farace, Richards, Monge and Jacobsen (1971); Shaw (1971); Schreiman (1975). 16 B. PERSON PERCEPTION Person perception refers to the "process by which one comes to know and to think about other persons, their char— acteristics, qualities, and inner states" (Tagiuri 1969:395). The research in this field of inquiry can be categorized into the following content areas: (1) the accuracy of our perceptions of the emotions and personality characteristics of others; and, (2) how infommation about another is processed. The latter is more closely related to the focus of the pre- sent research and thus will be briefly summarized. Extensive reviews of the former may be found in The Handbook of Social Psychology, 1969 (Tagiuri) and Person Perception by Hastorf g£_al., 1970. While the question of how an individual comes to "know" anything at all has been pondered by the great thinkers of all time, the field of person-perception did not develop un— til the later part of the Nineteenth Century. As physical objects a person displays primarily the same characteristics and require the same mental processes as any other physical stimuli. But person perception is also a unique phenomenon in that we additionally perceive or infer psychological prOperties for that person. I shall first briefly describe those characteristics of perception that pertain to all classes of stimuli, and later deal with some working con- ceptualizations about person perception. l7 One's experience of the world is dominated by objects which stand out in space and which have attributes of size, color, depth, weight, shape, etc. Although we are frequent- ly unconscious of our interpretation and inferential leaps of logic about such objects, we do participate in the order— ing or creation of our experiences. In fact, a causal analysis of the process of perception indicates that our per- ceptions are both structured and organized (Leeper 1935), resulting from the organism's active processing of informa- tion (Hastorf 1970). One of the most predominant aspects of a person's par— ticipation in the structuring of his/her empirical world is the process by which objects are placed into categories (Hastorf gt_al., 1970). These categories are the accumula— tion of one's past experiences and present purposes, and are restricted to some extent by the language codes and cultural background of the individual (Dornbusch §E_al., 1965; Passini and Norman 1966). Experience plays a vital role in our process of perception because we rely upon that which is familiar to us, and because our past gives us a sense of orderliness or sequence. Heider (1944; 1958) describes the process by which we develop organized perceptions of others in the following way. First we perceive others as causal agents, that is, capable of controlling or manipulating their behavior to achieve an intended effect. We then make value judgments about the 18 intentions of an individual based on their behavior. Second, because we perceive others as being similar to ourselves, we look for a parallel situation in our life in order to judge the behavior of another. Finally, when we have coded a person's behavior the same way a number of times, we then tend to assume that a person displays stable personality characteristics. Such a process increases the predictability of our interpersonal world so that knowing some things about a person permits us to deduce other characteristics. Without this process, all events would be perceived as unique with no apparent causal relationship. There is a considerable body of literature dealing with the issue of how we form first impressions of another person based on minimal information. Asch (1946), Anderson (1968), and Byrne (1961; 1969) found that most people, given some trait information about another person, generally go on to make inferences about a great variety of other traits not included in the data given, and that within sets of traits, certain ones seem to be more "central" to the set than others. Asch (1968) and Anderson (1968) presented subjects with a hypothetical person, along with a set of personality trait adjectives describing that person, and asked the respondents to rate the person on a scale of likeableness. Byrne (1961; 1969) varied this approach by presenting subjects with a hypothetical person, and the person's attitudes on various issues. Two different models are posited to interpret the variation in attractiveness. 19 Anderson (1962;1965) developed the weighted-average model which "assumes that the subject begins with a neutral impression, which is averaged upward by positive stimuli, and downward by negative stimuli. The greater the number of consistently positive or negative stimuli, the larger the average would be" (Hastorf gt_al., 1970:55). Byrne and Nelson (1965) suggest that attraction is a linear function of the proportion of positive "reinforcement" a view receives. The similarity between these two theories have often been noted (e.g. Anderson 1967; Byrne, Lambreth, Palmer and London 1969, p. 70; Griffit, Byrne and Bond 1971; Kaplan, 1972; Posavac and Pasko, 1971). The two models differ in their theoretical base, particularly in the conceptualiza- tion of the stimuli. Byrne uses reinforcement theory and treats the stimuli as reinforcers. Anderson uses informa- tion integration theory and treats the stimuli as informers. Bruner and Tagiuri (1954) attempted to explain indi- vidual differences in person perception with their Implicit Theory of Personality. They suggested that we all have as— sumptions about the nature of other persons and that these assumptions affect the way we perceive and understand others. Cronbach (1955) had judges rate various stimulus persons and then performed a zero order correlation comparing the traits indicated. He found that judges were indicating different underlying dimensions and thus reflecting their own 20 assumptions rather than objective reality. Similar results were found by Norman (1963), Passini and Norman (1966), and Hakel (1969) whether the perceiver knew the stimulus person or not. Kelley (1955) argues that we are actually revealing personality characteristics of the perceiver when we analyze how they judge others. Walters and Jackson (1966, Messick and Kogan (1966), Shapiro and Taguiri (1959) and Wiggins gt 31. (1969) contend that individual differences support the Implicit Personality Theory by showing a stronger relation- ship to the perceivers' personality than to the stimulus individual. Many investigators have been interested in how per- ceivers interpret inconsistent information in order to arrive at an impression about another individual. Kaplan (1972) found that when stimulus inconsistency exists (e.g. using the adjectives loyal, humorous, unkind, and vulgar to describe a person) individuals will differ in the importance assigned to various stimulus depending upon the predisposition of the subject toward the person being described. Positive or negative disposition subjects rated likableness or persons described by highly likable and dislikable traits under conditions either favoring or discouraging, discounting the inconsistent stimuli. Gallob (1973) found that the greater the degree of change in the overall impression of another that is needed for a perceiver to view as consistent some new information, the lower the probability that the perceiver 21 will believe this new information. Hastorf §t_al., (1970) summarized the ways subjects handle inconsistent information in the following way: (1) Relational Tendency, in which either the inconsistent information is changed in meaning or new traits are inferred to relate the inconsistencies (Asch, 1946; Luchins, 1957) (2) Discounting Tendency, in which part of the stimulus information is either ignored or reduced in importance (Dailey, 1952; Anderson and Jacobsen, 1965; Anderson and Herbert, 1963; Stewart, 1965; Luchin,, 1958; Rosenkrantz and Crockett, 1965; Zazonc, 1960; Cohen, 1961; Leventhal, 1962) (3) Linear Combination, in which the impression is some additive combination of the prOperties of the stimuli (Osgood gt_al., 1957; Warr and Knapper, 1968; Frijda, 1969; Triandis and Fishbein, 1963; Anderson, 1968; Saltiel and Woelfel, 1975). Attribution Theory, developed by Heider (1944, 1958) focused on "how we infer dispositional prOperties of another person from our observation of his/her behavior in social situations" (Hastorf gt_al,, 1970:61). Informational salience (Kanouse, 1972) and perceivers focus of attention (Jones and Nisbett, 1972; Duval and Wickland (1972) are two conditions hypothesized to relate to the process of forming causal attribution. If a perceiver is faced with a number of alternative explanations for an event, he/she will generally adopt the one which is most salient. 22 Duval and Wickland (1972) have Specified that the dif- ferent conclusions a person reaches will depend on whether his/her attention is directed outward to the environment or inward to oneself as a stimulus. Jones and Nisbett (1972) proposed that a person performing a behavior and an observer looking at the behavior will interpret it in dif— ferent terms. Specifically, they suggested that an actor usually attributes his/her own behavior to situational factors, whereas an observer attributes the behavior to qualities or diSpositions of the actor. Jones and Nisbett further suggested that this difference in the interpretation of behavior stems from a basic difference in perceptual orientation. The participant's attention is focused on the situation in which he/she is behaving and does not see the performance or action. The observer's attention, on the other hand, is directed to the actor's behavior, which is figural and dynamic against a situational ground. When asked to explain the reasons for the behavior, each will utilize the information that is salient to him/her while the behavior was being performed. Slightly divergent from the general research in person perception, is the body of literature dealing with the social perception of one's relative standing in the group in which one holds membership. Here the focus is not on how we develop perceptions of others, but rather, how we develop perceptions of ourselves. 23 Cartwright (1951) and Zander (1958) provide evidence that the fundamental part of the individual's perceptual base is to be found in his/her assessment of the perceptions of his/her group. Before an individual acts, one first imagines oneself carrying out the act, then imagines the response of another person to the action. If the imagined response is favorable, one will proceed with the overt act, but if the imagined response is unfavorable, one will modify one's intended action before actually carrying it out (Cooley, 1902; Cottrell, 1942; Mead, 1950; Blumer, 1953; Turner, 1956). In either event, the actual behavior of the other person serves as a check on one's perceptions. The perceptions of the "self" and "other person" are then modi- fied to correspond to the new evidence presented by the actual behavior (Sullivan, 1938). Another group of scientists concerned themselves with the influence of the group on individual judgment (e.g. Asch, 1946; Sherif, 1953; Kelley, 1950; Schachter, 1951). Morris (1956) found that the self concept of men living in a dormi- tory in four-man living units was influenced by others' per- ceptions of them over a period of months spent living toqether. Similar results were found by Miyamoto and Dornbush (1956) in comparing self-perception on four per- sonality traits with actual feelings of others in the group, with their perceptions of others' attitudes, and with their perception of the attitudes of the members of most groups. 24 There is also considerable evidence to support the contention that the group's perception of an individual will have more influence on his/her self—perception when one is highly attracted to the group and when the other group members place a high value on that member's partici- pation (Festinger, 1954; Stotland, 1959). Under these con- ditions the individual will pay more attention to the opinions of the group, and the group members in turn will be more explicit in their evaluation of the individual. Undoubtedly, the most prominent assertion in group research is that people like and interact with those who are most similar to them (e.g., Festinger, 1954; Homans, 1961; Riecken and Homans 1954; Lott and Lott, 1960; Newcomb, 1956). Quite often though we are not dealing with necessar— ily true similarities but perceived similarities. Tagiuri (1958) found that if one subject likes another, he/she tends to think that liking is returned, and if a subject likes two other subjects, he/she will perceive them as liking each other (Kogan and Tagiuri, 1958). Up to this point, research in person perception has generally dealt with the single attribute of one's "liking" of another. There are two major weaknesses in this approach. First, two peOple may be equally liked but for quite differ- ent reasons, one because he/she is friendly, for example, and the other because he/she is responsible. If one 25 assumes evaluations to be a unidimensional continuum and attends to only the level of the manifest response, then these evaluations would be considered equivalent, when in fact they are not. Second, the use of the single attribute "liking” does not encompass all of the criteria used to judge the similarity/dissimilarity of another individual or groups of individuals. For example, I may decide to work with another individual because he/she is dependable, articulate and efficient, whereas, socially I may prefer someone diverse and congenial. we must go beyond unidimen- sional scaling techniques in order to discern the multidi- mensional structure of our perceptual discriminations. A system which enables us to describe the multidimen— sional structure of our person perceptions has recently been develOped. It says: The process of relating objects of thought to each other is the process of definition. The definition of any concept may be taken to be that term's relationship to all other concepts which are used to differ- entiate that referent as a unique object (Fillenbaum and Rapoport, 1971). "Fundamentally this involves taking note of similarities and differences between objects, or identify-- ing the attributes of an object with similar attributes of different objects, and differentiating the attributes of the object from those attributes of the objects which are different." Woelfel (l972:5; Chapter 4), Woelfel (1972:11; Chapter 4) states that: 26 Dissimilarities among objects (whatever those objects may be) may be represented by a continuous numbering system such that two objects considered to be completely identical are assigned a paired dissimilarity score or distance score of zero (0), and objects of increasing dissimilarity are represented by numbers of increasing value. Assuming that the definition of an object or concept is constituted by the pattern of its relationship to other objects, the definition of any object or concept is constituted by the pattern of its relationship to other objects, the definition of any object may be represented by a 1 x n vector, dll' d12, d13, . . . dln’ where d represents the distance or dissimiIErity of object 1 from itself (thus d = 0 by definition), d1 represents the diStance or dissimilarity getween objects 1 and 2, and d represents the distance between the lg; and EBe nth objects. Similarly, the second object may be represented by a second vector, d21, d22, d23, . . . d2n’ and the definition of any set of concepts or objects may therefore be represented in terms of the matrix d d 11' d12' ' ' ' 1n d d 21' d22' ° ° ° 2n d d nl' dn2' ° ' ° nn where any entry d.. represents the dissimilarity or distance betweéa i and i. The distance matrix D describes the static structure of the interrelationships among a set of N objects (persons) at any one point in time for a single individual. This view can be generalized to aggregate or collective meaning quite simply. The collective consciousness, that Iaggregate psychological configuration which constitutes the culture, or a sub-culture, may be represented accurately as 27 the average matrix D, where any entry dij is the arithmetic mean conception of the distances or dissimilarities between object 1 and i as seen by all members of the culture. While the matrix is an accurate representation of a set of cultural definitions, it may be cumbersome due to its size, especially when the number of objects or persons is large. The matrix is of order N, where N equals the number of con- cepts. Matrix 5 describes an implicit vector Space Vk' where k (the dimensionality of the space) :_n-l. K equals the number of independent dimensions which are needed to repro- duce the pattern underlying D} The value k is easily found by factor analysis of the scalar product (5) of D and its transpose, i.e., D D=S. Vk is a spatial coordinate system defined by the dis- tance relation among the c0gnitive objects which are its contents. It has the property that objects defined as similar by any culture will be located close to each other in the space, or more precisely, that the distance between any pair of objects in the space is directly proportional to their perceived dissimilarity. The precise definition of any object, therefore, is given by its location in Vk‘ Matrix 5 may thus be collapsed to vector space Vk' This has the advantage of reducing the data to usable pro- portions and revealing the uniquely shared underlying cultural dimensions. This space Vk' therefore, may be seen to provide a reference coordinate system within which 28 patterns of individual or cultured information may be repre- sented with considerable accuracy. In terms of person perception, this technique which is known as the Galileo Configuration, has two distinct advan- tages. First, it allows us to describe not only individual perceptions or similarity/dissimilarity judgments of one other individual but groups of individuals. Second, we can repre- sent the structure of perception shared by the aggregate. Future references to the aggregate (cultural) belief system will always refer to the results of these operational measures unless otherwise noted. I Gillham (1972) and Gillham and Woelfel (1974) in a similar study in which respondents were asked to rate pairs of faculty members on the attributes of policitcal position and style of professional research found reliable and valid results using the Galileo procedure. Several of the pro- cedures they suggest for assessing the validity of the con- figuration will be used in this research, and will be discussed later. FORMULATION OF HYPOTHESES In conclusion, we now have the methodological and theoretical tools to accurately describe: (1) the organized networks of systematic interaction, or rather, how information passes through a system (NEGOPY Program); and, (2) the organ- ized structure of our person perceptions, or rather, the "culture" of a system (Galileo Configuration). The advances 29 in both of these fields now allows us to progress at levels beyond the minimal unit of analysis and describe the patterns of observable regularities within a system. What is needed now is an analysis of the relationship between communication behavior and the "culture" of a system. In general, the theoretical hypothesis this thesis is designed to test is that persons "close" to each other in terms of social structural location will be perceived as similar to each other by others in the system. We are now in position to make this hypothesis specific in terms of the Operational measures available to us. Given the measure of "closeness" among people in social structure made available by the NEGOPY program and the measures of culturally per- ceived dissimilarities among people made available by the Galileo instruments, three specific hypotheses may be stated: H1: The greater the frequency of interaction between 1nd1V1duals or groups of indi- viduals, the smaller the perceived distance between those individuals in the aggregate (cultural) belief system. Additionally, individuals who have direct communication links with each other should develop similar information, and should be perceived as similar by the aggregate. H2: The greater the number of links necessary to connect any pair of indi- viduals in the communication network, the greater the perceived distance be- tween those individuals in the aggregate (cultural) belief system. 30 Both of these hypotheses are correlaries of the main hypothesis of the thesis, i.e., those who are "close" to each other in the social structure will be perceived as similar by those around them. They differ from each other in that "closeness" in social structure is measured in two alternative ways: first, by frequency of communication, and second, by number of links between members of the department. It also follows from the theory presented above, that perceived similarity will vary with interaction, but be- cause respondents are indicating only a small prOportion of their total lifetime interaction, findings should be in the direction anticipated but not necessarily large. As noted above, after determining the patterns of interactions the structural properties that emerge are also of theoretical significance. Of particular interest in this case is the degree to which shared channels of communication contribute to perceptual similarities. It should be the case that the more integrated an individual is in the net- work (the degree to which individual links exist between a particular person and his/her set of contactees) the greater the perceived similarity to other members of the system. H3: The more integrated an individual is in the commun1cation network, the smaller the total perceived distance of that individual from all other persons in the aggregate (cultural) belief system. This is really another variation of the first two hy— potheses, but differs slightly. The first two hypotheses 31 relate pairwise similarities to pairwise interactions. This hypothesis, however, deals with the relationship be- tween any individual and the rest of the group as a whole. Thus the more an individual interacts with the group as an entity, the more similar he/she should become to the group. CHAPTER II PROCEDURES AND METHODS This section will attempt to develop conceptual and operational definitions for the five key variables pre- sented in the previous hypotheses: (1) frequency of com- munication; (2) distance or links between pairs of individuals; (3) network integration; (4) culturally per- ceived distance between pairs of members of the network; and, (5) total culturally perceived difference of any in- dividual in the network from all other persons in the network. The sample consisted of all faculty and funded graduate students in the Department of Communication at Michigan State University yielding an N of sixty-seven (67). Sixty— one questionnaires were returned by approximately 91% of the sample. This sample was itself a subset of a larger study of the College of Communication Arts and Sciences at Michigan State University. Whenever consequences of this subsampling procedure effect the procedure used they will be discussed in that context. Frequency of communication was determined by the "Personal Communication Contact" questionnaire (see Appendix 1 32 33 A), in which the names of the faculty and funded graduate students in the Department of Communication were provided (along with the students in the those responses were included in the analysis). College but not in the Department. names of the faculty and funded graduate Only by and regarding members of the Department Respondents were requested to estimate their frequency of communication with each individual concerning teaching and research. ordinal frequency Once or Several Once or Several Once or Several Once or Several levels were provided: twice times twice times twice times twice times mmmmmmmm coded coded coded coded coded coded coded coded as as as as as as as as (1) (2) (3) (4) (5) (6) (7) (8) The following Responses for the content areas of research and teach- ing were collapsed into a single content category for each respondent. These data were input into the "NEGOPY" network analysis program as described by Richards, Farace, and Danowski (1973). This is a computerized procedure for determining the communication groups in a social system, and the nodes (persons) who link the groups together. Groups are identified according to a standard set criteria as specified in the previous chapter. Various other structural variables are also determined based on the results of group 34 formation and communication frequency. For this analysis, two additional network properties are of interest: (1) the distance matrix or link analysis; and, (2) the network integration scores. Both variables shall be briefly defined. The distance matrix represents the number of links necessary to connect any pair of individuals in the group. In this n by n matrix (n is the number of members for each group) the element in row 1, column j gives the number of steps needed to get from individual i to individual j in the group. If there is some finite number in each element of the matrix, the group will be connected. This means that there will be some path from each individual in the group to every other individual in the group. The longest any path could ever be is n-l steps. The way the distance matrix is constructed is as follows (Richards, l974b:14). A matrix is constructed in which there is a row and a column for each node in the group. All the elements are initial— ized to zero. Whenever there is a link from node i to node j we enter a 1 in row i, column j. If the link is recipro— cated we also enter a l in row j, column i. We then repeatedly perform a Boolean logic Operation which is analogous to raising the matrix to successively higher and higher powers. Instead of entering the cross product of the ith row and jth column as the i, j element in the product matrix, however, we enter the first power on which this value becomes non-zero. 35 (This operation is performed with a series of nested DO-Loops and IF statements in FORTRAN. With careful organization, the process can be optimized to take signif- icantly less time to compute than a standard algebraic multiplication of matrices). We stop raising the matrix to higher powers when one of the two conditions obtains: either (a) all off-diagnol elements be- come non-zero, which implies the group is connected; or (b) when going from any power k to the next power k+l no entries change value, which implies the group is not connected at value k and will never be connected at any level. Network integration scores represent the number of persons with whom an individual is linked and who are also csonnected to each other. The formula for computing the :integration score is as follows: _ (wa) - L i - L (L-l) Where: wf is the ”weighting factor"* for each link which is equal to the number of two-step links connecting individual i and all persons with whom he/she is linked + 1. L = the number of persons with whom person i is linked. Ordinarily, the procedures defined by the Galileo technique for the measurement of the aggregate perceptions of dissimilarities among the 14 faculty would require each \ * I: ‘ FOrmula obtained from a personal communication with <::::1111iam Richards 7/23/76, Stanford University, Stanford, Sallifornia. 36 sample respondent to estimate pairwise dissimilarities for each of the l4(13)/2 = 91 possible pairs of individuals. Since this study is reduced from a larger study of the entire College of Communication, modifications in these procedures were required. First, all 1176 possible pairs of the 49 college faculty were created, then randomly split into five subgroups of 235 pair-comparisons. From these pair comparisons, only those pair comparisons involv- .ing pairs of the 14 departmental faculty were included in ‘this analysis. Exact cell frequencies for each of these :relevant pair comparisons are provided in Appendix 2A. <(Nerall, there were 67 respondents within the communication «department, since funded graduate students were included as :sample respondents (and included in the network analysis) (even though they were not included as objects of scaling in the Galileo questionnaire. (Appendix 13) . Respondents (once again faculty members from outside 't:he department were included in the questionnaire but excluded from analysis) based their estimates of dissimi- lluarity as a ratio of a standard distance provided by the fiLlavestigator in the following way: "If red and white are 50 Galileos apart: How far apart are Professor A and Professor B?" Th is technique has several advantages: First and foremost, :r:.<:> restrictions are placed upon the respondent, who may : eport any positive real value whatever for any pair. n1. -.'..-‘- as 37 Thus the scale is unbounded at the high end and continuous across its entire range. Secondly, because the unit of measure is always the same (i.e., the unit is provided by the investigator in the conditional, "If red and white are g% of the distance from red to white as perceived by each respondent. 50 units apart"), and thus every scale unit is Because the condition of zero distance represents identity between faculty and is hence a true zero, not at all arbi- trary, this scale is a ratio scale. For each distance we then average the estimates across respondents to produce a symmetric matrix of the same order as the number of objects scaled. This matrix tell us the average distance from what is known about any given pro- fessor in the set to what is known about any other. Al- though respondents individually used many different criteria in estimating a distance for each pair, the law of large numbers assures that the scores obtained will be normally distributed about a sample mean score, and that sample mean will converge on the population true score as n_be— comes large. This population true score, that is the true mean dissimilarities estimates for all the members of a culture, is exactly the theoretical definition of culture suggested earlier. Operationally, culture is defined as the matrix D Where any entry dij = kél §%g£-= the distance or dissimi- -laiity between the ith_object and the jth object as esti- InElted by the kth person using the method of direct paired 38 distance estimates, and n = the number of persons making such an estimate. Several qualifications must be made, of course. First, clearly the matrix D, to be exhaustive, would be a g x 3 matrix where c = the number of objects defined by the culture, which is a very large, but finite number. What is at issue, of course, is the measurement of subsets of the matrix D corresponding to segments of the culture under investigation. (In this case this subset includes members of the faculty of the Department of Communication). Secondly, the boundaries of the culture itself need not be so clearly drawn as is implicit in this discussion, and the investigation of subcultures is simply a matter of appro~ priate sampling. Third, the matrix D represents a static picture of the state of a culture at a given point in time. These techniques provide us with measures of all the variables required by the hypotheses. Frequency of com- munication is measured by the ordinal scale ranging from one to eight, corresponding to categories "once or twice a term" through "several times a day." In this 14 x 14 matrix pairwise communication frequency scores are reported. Pairwise distance among people in the social structure is estimated by the NEGOPY distance matrix which is itself based on the number of links required to connect any two faculty members through the total network of 67 faculty 61nd.funded graduate students in the department. Integration 39 into the social structure is estimated by the network integration scores, which again is based on information of the total network of 67 faculty and graduate students. Finally, perceived similarity and dissimilarity of members of the department is measured by the Galileo procedures, which provide both a pairwise dissimilarities score for all pairs of faculty in the department and overall perceived "Alienation" score, or measure of the total distance of any professor from all others. (This measure is therefore reverse-scored; i.e., a low score represents high aliena- tion.) Several additional variables (Appendix 1C) were also measured to establish reliability and validity measures for the principle instruments. These variables will be discussed as they become relevant. CHAPTER III RESULTS In Chapter I three final hypotheses were derived to test the major theoretical pr0positions set forth there. The first two represent a dyadic test of conditions under which networks of interaction vary with the similarity/ dissimilarity judgments or "culture" of the system: (1) with frequency of communication; and (2) the distance or number of links between any pair of individuals. Hypothesis 3 deals with the frequency of interaction of an individual with the group as a whole and the perceived relation of that individual to the group as a whole. In this chapter the findings of statistical tests of the hypotheses are set forth. Hypotheses 1 and 2 shall be discussed conjointly since both are concerned with a dyadic test of the theory rather than the individual focus of Hy- pothesis 3. Reporting an analysis of results are organized within the following format. First, preliminary statistical data (means and standard deviations) are presented in tabular form. Second, each hypothesis is repeated and the statistical procedures (correlation coefficients) and relevant findings are indicated and analyzed. Third, evidence about the 40 41 reliability and validity of the principal measures is pre- sented along with a discussion of the results. Finally, results are summarized at the conclusion of the chapter. TABLE 1 Descriptive Statistics For The Variables Referred To In Hypotheses 1 and 2 (N=6l)* VARIABLES MEANS STANDARD DEVIATIONS FREQUENCY OF INTERACTION 4.3187 4.3815 DISTANCE OR NUMBER OF LINKS 1.2747 .4730 CULTURALLY PERCEIVED' PAIRWISE DISTANCES 62.8897 33.4611 (GALILEO) f The sample size question is complicated with these kinds of measurements. In Table l, the N=61 refers to the total number of faculty and funded graduate students in the depart— ment who responded to the questionnaire, since information from all respondents is utilized to generate these aggregate variables. Particularly for the distance or link analysis, the apprOpriate analysis-and the one performed here-requires an examination of all linkages based on the 67 faculty and students present in the department, rather than the linkages present among only the 14 faculty. Deletion of any members of the department (either faculty or funded graduate students) from this particular analysis may result in an inaccurate description of the overall communication network; e.g. two faculty members who share a common channel of communication via a graduate student, would be considered unlinked. For this reason I utilized the option of the NEGOPY program which forces reciprocation, meaning, if faculty member A reports a link with member B and B did not fill out the questionnaire, I accepted the report of member A as being accurate (resulting in an N of 67). This is not always an advisable practice, but because of my membership in the network, I felt more con- fident in making such a decision. For the individual pairwise measures, however, the sample size is variable. In the case of pairwise frequency of communication scores in the Network 42 analysis, each pairwise frequency is the average of the fre— quency of communication reported by the two people involved. Hence all network pairwise frequencies are based on N=2. Pairwise number of links between any two members of the net- work is based on information drawn from a larger subset of members of the network which is roughly prOportional to the number of links required, but in no case specifically deter- minable without considerable difficulty. Sample size for the pairwise differences in the Galileo Configuration is also variable (see Table 2), depending on how many sample members actually completed estimates of the differences between any two faculty members. These pairwise cell counts, along with other descriptive information, are provided in Appendix 2A. column 6. Table 1 presents the means and standard deviations of variables referred to in Hypotheses 1 and 2. It shows that the average frequency of Interaction is about 4.3 or several times a month and the standard deviation of 4.4 indicates that about two-thirds of the cases lie between several times a week and several times a term. Table 1 also indicates that there are on the average 1.27 steps between any two persons in the department and the standard deviation of .473 indicates that about two-thirds of the cases are linked by either a direct link or a two step indirect link. As Table 1 shows, the average dissimilarity between any two members of the department, as seen by the aggregate, is 62 units. Since these Galileo scores (or culturally perceived pairwise dis- tances) are approximately normally distributed (see Appendix 2C Table 1) these data show that about two-thirds of the possible pairs of faculty are seen as being within 30 or 90 ‘units of each other. Table 2 describes the means and standard deviations of 1‘Jetwork Integration and Galileo Alienation measures of TABLE 2 43 Descriptive Statistics For The Variables Referred To In Hypothesis 3 (N=l4**) VARIABLES MEANS STANDARD DEVIATIONS NETWORK INTEGRATION .6106 .1745 GALILEO ALIENATION 825.090 123.1134 GALILEO FACTOR 1 -.0000 41.1142 GALILEO FACTOR 2 -.0001 39.7374 GALILEO FACTOR 3 -.0000 33.2930 GALILEO FACTOR 4 .0000 20.3123 GALILEO FACTOR 5 2.2714 15.3145 GALILEO FACTOR 6 .-.1805 10.8505 GALILEO FACTOR 7 .1098 8.9950 GALILEO FACTOR 8 -.1467 .5844 GALILEO FACTOR 9 1.8929 -5.3817 GALILEO FACTOR 10 -.2051 -3.9497 GALILEO FACTOR 11 .0001 -9.0445 GALILEO FACTOR 12 .0001 -18.3275 GALILEO FACTOR 13 0.0000 -32.5860 14 -33.6010 GALILEO FACTOR 5.3508 * See footnote below Table 1 regarding N for Galileo Factors. * * Each individual or faculty Network Integration score is computed from the total N of 67. Reported here is the mean for the 14 faculty members based on the judgments of 67 sample members. 44 Hypothesis 3 along with the means and standard deviations of the Galileo Factors (dimensions). Since a Network Integra- tion score of 1.0 represents complete Integration, and 0.0 complete isolation, the mean .61 and standard deviation of .17 indicates, with small exception, most members are rela- tively integrated into the department. The Galileo Aliena- tion mean of 825.0 represents, of course, the total distance between a person and all others and so correSponds to (N-l) x the average Galileo distance within rounding error (63.47 x 13 = 825). More importantly however, the large standard deviation of 123.11 indicates substantial variance in how similar the faculty members are perceived. Since the Galileo Configuration is centered at the cen- troid of all concepts, the mean of all Galileo factors is normalized to zero. The actual numbers in Table 2 therefore represent a useful check of rounding error in the interme- diate step of key punching (these errors are small enough to be considered insignificant). The standard deviations of the factors, however, represent the relative contribution of each of the factors to the overall Galileo Configuration. Factor 8, for example, represents the NULL vector since a complete factorization of the 14 x 14 distance matrix should yield n-l or 13 factors. Again this is a useful check of the rounding error which is again very small. Factors 9-14 are negative and represent non-euclidianism in the Galileo space. These negative roots have been interpreted as uncertainties on the 45 part of the respondents. In this case these uncertainties are fairly small relative to the positive dimensions of the space. Additionally, Appendix 2 contains the Distance Matrix or link analysis computed from the NEGOPY program along with the Galileo Configuration, coordinate dimensions, and the Aggregate Means matrix. Several tentative conclusions can be drawn from these data. First, there is little variance in the number of links between any two faculty members. This hinders any statistical relationships. Second, the Galileo Factor Matrix and Configuration is multidimensional indicating that respondents are discriminating across a number of factors. Though factors vary in their contribution to the total variance, recent evidence supports the use of all factors in analysis (Barnett 1976; Barnett and Woelfel, 1976). In order to test the linear relationship between inter— action and perceived similarity, zero order correlation co- efficients were computed. These correlation coefficients will provide an estimate of the deviation of scores from a straight regression line. Table 3 reports the intercorrela— tions among the variables referred to in Hypotheses l and 2, and Table 4 reports the intercorrelations among the variables in Hypothesis 3. 46 TABLE 3 Correlation Coefficients Among The Variables Referred To In Hypotheses 1 and 2 VARIABLES 1 2 3 l. FREQUENCY OF INTERACTION 1.000 2. DISTANCE -.6861 1.000 3. CULTURALLY PER- CEIVED PAIRWISE DISTANCES (GALILEO CONFIGURATION) -.0614 .0482 1.000 HYPOTHESES TESTS Hypothesis 1 The greater the frequency of interaction between individuals or groups of individu- als, the smaller the perceived distance between those individuals in the aggregate (cultural) belief system. HYPOtheSIS 2The greater the number of links necessary to connect any pair of individuals in the communication network, the greater the per- ceived distance between those individuals in the aggregate (cultural) belief system. As Table 3 indicates, the correlation coefficients for frequency of interaction and distance or links between indi- viduals with perceived similarity are -.0614 and .0482 re- spectively. Based on the conceptualization developed in Chapter I we would expect the correlations to be low since respondents are reporting only a small proportion of their total lifetime communication, but such correlations fail to provide strong enough evidence to accept the prOposed 47 SOB... m~m~of Sad. 38%. ~88. $5. SSS. ~28... 88.8w? 3 . 8:88 x m3-r 2.25... 880. 88s. Sea. 838... Res. 383. 3 8.8% max 28a... 889- 8.48. 88?. 88°. 88o. 808. $48.- 2 58% 3x 21.8.- R30. 838- 888. 88o... 88o. 825. Sam. ~H 8.8% 2x 32o. 258. mH-E 88o... ~88. 88o. 88o... 28m. 2 8.8% fix 888.- 088.- 482... :84. 8an 35%. 8~8. 339- 2 58% 3x SST- Seam. SSH. 88°. meomm... Rex: ~88. 885. a “88% 2x 3:5. 348.. ~33... v.28. SRO. 88m. 320. H-8r m 55% ax 8084 «.88. ~82. 38o... 220.- «~20... 88o... 938... A 58% mx 88o; BER. ~88. «.83.- 488.- -m8. ARR. e 88% ex 88o; R~m~. ~23. :38. ~38. 882. m 88% ex 8°84 88°. 880. 88°. «#2.. s 88% me 8084 088... 825.. «.88.- m 58% ex 8804 88c... ~88... ~ 58% mx 88o; 834.- H 8.8% ~x 8804 .szH Roswmz ax ax w” ox mx ex mx ~x ax 83281, m 3858s: 5 8. Shana SEEMS 9:. 88.2 358888 883880 «ugh. 48 88¢; ~8-. ~88. ~88. 88%- ~88... ~88... $-~. 88882 S 8888 x 1884 8-o... SSS. ~88. 88o. ~88... :38. 3 58% 8x 8804 88c. ~88. -~8. --o... ~88. - 58% 3x 88o; ~88.u ~88. ~38. ~88. - 58% 2x 8804 82:. ~88. ~82..- S 58% -x 8°84 ~83. 820... 3 58% :x 88c; 88%. m 58% 3x 8°84 m 56% max max «ax max max flax QHx ax mmaomwhm> vuflfiflflv 49 hypotheses. Hypothes1s 3The more integrated an individual is in the communication network, the smaller the total perceived distance of that individual from all other persons in the aggregate (cultural) belief system. The two variables referred to in Hypothesis 3 are (1) Network Integration (computed directly from the NEGOPY program) which indicates the degree to which individual links exist between a particular person and his/her set of contactees; and (2) the perceived similarity/dissimilarity of an individ— ual from all other individuals in the cultural aggregate. The latter variable, referred to as Galileo Alienation, is determined by taking the sum of all the entities in that indi~ vidual's row in the Galileo distance matrix. Since any cell d.j of the distance matrix represents the distance of the 1th 1 th row will th person from the jth person, then the sum of the 1 represent the total distance, dissimilarity, of the 1 person from all other persons. As Table 4 indicates, the correlation coefficient between Network Integration and Galileo Alienation is —.29l7, sig. .156.* While apparently small this means that nearly 10% of These significance levels, as all others derived from the Galileo Configuration, are substantial underestimates of the true statistical significance. This is so since the number of persons rated in Galileo (n=l4) is considered by the program to be the number of observations (see footnote page 41). The Galileo Configuration however, is itself an aggregate derived in this case from (n (n-l)/2') paired com- parisons times 61 (sample size) observations. In any event, 50 the individual differences perceived among members of the department can be accounted for by their integration in the communication network alone. This correlation does not take into account the length of time that individuals have been in the network and should therefore be an underestimate of the effect of intra-departmental communication on individual dif— ferences. This is born out by the correlation coefficient of -.3009 between Galileo Alienation and the number of years in the Department (Table 7). Consequently, we might suspect an interaction effect between Network Integration and Years with the Department. Table 5 indicates that when Years with the Department is multiplied by Network Integration its correla- tion with Galileo Alienation is -.37957. This is a very sub— stantial correlation since it indicates that nearly 15% of the individual differences perceived among department members can be accounted for by their communication history in the department. Finally, some additional evidence germane to this hy- pothesis is available in the form of the correlation co- efficients of Network Integration with individual Galileo Factors. This information is summarized in Table 4. Inasmuch as the correlation between Network Integration and Galileo Alienation is lower than the correlations between significance tests in this analysis are not appropriate since the members of the department represent the complete pOpula- tion of interest and no inferences to a more general popula- tion are required. 51 TABLE 5 Correlation Coefficients For Years With The Department x Network Integration And Galileo Alienation VARIABLES X l 2 X1 Years x Network Integration 1.00000 X Galileo Alienation -.37957 1.00000 2 Network Integration and several of the Galileo Factors, there is reason to believe that there is more to the relationship than a simple correlation. It may be the case that integra— tiveness in the communication network is an attribute which people recognize when comparing individuals. The test for this is a straightforward regression analysis. The regression equation is as follows: Yi ‘ bo + bilxil + bizxiz + bi3xi3 + bi4xi4 + biSXiS + biGXiG + e where: Yi = The Network Integration score of the ith individual bi“ = the set of unstandardized regression coefficients 3 or partial slopes of the six unidimensional scales Xi. = is known as a constant, namely the value of the 3 independent variable, or rather, the six highest correlated factor loadings e = stochastic (error) terms As Table 6 indicates, the Multiple Correlation between Network Integration and the six Galileo factor loadings is 52 TABLE 6 Standardized Partial Regression Coefficients, Multiple Correlation Coefficients, Tests of Significance, and Coefficients of Determination 0f Network Integration On Galileo Coordinate Dimension.Va1ues Multiple R .95330 R Square .90879 Degrees of Freedom 6,7 F 11.62393 Significance .002 Accumulated F R . Significance Beta Factor 1 11.761512 .40483 .011 -.3916862 Factor 12 ' 11.522152 .50669 .012 .4155880 Factor 11 10.974002' .63764 .013 .3940481 Factor 6 3.1442338 .81920‘ .119 .2473755 Factor 9 3.0672969 .81923 .123 -.2580013 Factor 8 18.238496 .95330 .004 -.6548335 .95330 and is significant at the .002 level. Additionally, the Beta weights of Factors 1, 8, ll, 12 are below the .05 level of significance. Not only does this coefficient indicate a high degree of correspondence between the variables, but it also signifies that the measurement in instruments for the aggregate cultural system (Galileo) and Network Integration 53 have no more than a small unreliability component. Further- more, the notion that department members take into account the degree to which an individual is integrated into the network when making evaluations provides additional evidence of the validity of the measurement instruments. RELIABILITY AND VALIDITY OF THE INSTRUMENTS The substantial multiple correlations between Network Integration and the Galileo factors, along with the theo- retically sensible relationship between the interaction of years with the department and Network Integration correlated with Galileo Alienation provide fairly strong evidence that both instruments (i.e., the NEGOPY program and the Galileo) are reliably and validly measuring the variables of interest. Nevertheless, the near-zero correlations among pairwise dis- tances among the faculty, number of links, and Network dis- tance require a more careful analysis before we are willing to accept the results at their face value. Reliability and validity are best assessed by looking at the pattern of correlations among the Network and Galileo measures and additional variables whose relations to these measures are predicted or known. Accordingly, we made measures of several other variables whose reliability and validity are not suspect, and whose relationship to the other Network and Galileo measures could reasonably be predicted. The variables are: (1) Years with the department; (2) Academic rank (e.g., Associate professor); (3)-the number of 54 Ph.D. committees one chairs; (4) the number of memberships one holds on Ph.D. committees; (5) the number of M.A. com- mittees one chairs; and, (6) the number of memberships one holds on M.A. committees. It is reasonable to assume that the members of the department consider the preceding attri- butes when making evaluations of the other members,* and thus should reaSonably be expected to be represented in the aggre— gate cultural perceptions measured by the Galileo procedure. Table 7 summarizes the correlation coefficients for the unidimensional scales, Network Integration scores, Galileo Factor loadings, and the Galileo Alienation scores. As indi- cated, both Network Integration and Galileo Alienation fail to correlate highly with any of the unidimensional scales, and thus provide little evidence of measurement reliability or validity. However, Network Integration and a number of Galileo Factor loadings do correlate significantly with each other (Factor 8, r = -.6622 sig. = .005; Factor 6, r = .5373 sig. = .025). Moreover, there are additional correlations which are large but not significant (Factor 1, r = -.4048; Factor 9,.r = .3702; Factor 11, r = .3871; Factor 12, r = .3043). The lack of significance may be due to the small sample size, but since we are dealing with the whole popula- tion and not generalizing to another, the significance levels * Other ethical and technical reasons prohibit measuring all possible attributes. 55 ~e-~. ~m~4~.u eemmo.u --o. -o-. --v.n -~m~.n omm~m.u --o. e--. o~¢m~.u mmess. «Hoes ooooo.n moooo. ooooo. Hoooo. ~mo~¢. ~H-~. ¢--.a -~m~. oeo-.n -o~o.n --o. mevo~.u -uam ~oooo. Hoooo.u Hoooo. ooooo. mmee~. m-~o.n moeso.u --~.u oe~mo.u o--. s~4o~.u -vo~. -8es ooooo.u ~oooo. Hoooo. ~oooo.u o~mo~.u ~m~oe.u --o.u --~.n ~vaao. vasom. -~e~. o--. -uam -~oe. ~o~e~.r e-~o.u ~o-o. -o-.a oe-o.a --o.u oo-~.u ~e~s~. -~ee. v~¢-. o~eeo.a oases Homoo. m~o-.r ~m~e~.u -o~o. -m-.n ~e-e. -~o~.u emamo. -m-.u --o.u --o. o~o-. a can m--. ~v-o. o-o~. --o. se-~. -v~o.u --o._ --~. -~o~. ~e-~. ~o~eo. -~oe.u m can odoeo.n --o.u e-~o.u o-oo.u e~o~o.n --e.u m--.u e~m-.u -v~o.u momma. o~moo.n o-~o.u ~ use -o~o. 44040.- «oeao.u -~oo. --o.n oemeo. -o~o.u -~e~. ~e¢e~.u -4-.u --v.u --~. 8 use --~. --H. Hmomo. --o. ammHH. ~mmm~.g m-oe.n mmevo. mmoom.u o--.n mmm-. ~GG-. m 84% ooooo. ooooo.u Hoooo. -~o~.a o-oo. v--.n -e-. movw~. ~o-~. o~o-. --~.n 4 use ooooo.s Hoooo.u -~me. --o.u omm-. --~. --o.u --~.u --~.: ¢~e~o.u ~ cam ~oooo.u mm~mv. «Hams. --v. ~o~ee. -~oe. m-s~.u m--.u ~He-.n ~ 8am mee-. me-o. ~v-~.u ~v-~.u --o.u mm~m~. -ee~. -voe.u H use ~G-o. --~. ~m-~. --o.u oo-~.n smoo~.u --~.u H oeuHAeu ~eo-. --m. o-~o. ~mvo~.n ~m~mo. --~.n 2a: q-~m. --m. -~o~.u s-~v.u mm-o. Us: emamo. --m.u --o.u ~o-~.n same HHvoe. -~o~.n -~e~.u comm ~m~e~. --o.n Hmom s-~e.n was 4 one ~ one ~ one H can H om-~mu .gmz “use case. norm smom mus uaHumz mmuoom ao~nmam-a.oo-~mo was was macaques schema oo-~mo .mmuoom coaumummuaH xuozuoz .mmsmom anaesmamsacaas are non muawsoammmpo cosuasmuuoo 5. EH. 56 webmo.l manna. mmoma. mambo. mmvmo.l HmHmH. MOHNN.I vvvoo.l moamo. ooooo. Noooo. mamoo. mamwo.l hmmvo. mmmoo.l vowoo.l mvmmo. Noooo.l mNmNm. mevqm. mmmmo. mqhao.l hhvvo. movem.l omvoa. mmomm. bvaho.l mmhao. memNo. mHNNv. hovva. mohao.l mmNoo.I mmmmv.l woomm.l mmmom.l Hmoqm.l oammm. moaoa. mmhoo. mmvmm.l mmmmH.I mmmmo. hmvma. mmmhm. vaumm maQflm NHUdm HHUdm caumm m 0mm m 0mm 5 Uflm w Uflm ma 0mm NH 0mm 3 one 3 0mm m 0mm m own home m use m can h mflflfla 57 may be eliminated as a criteria for acceptance. Since all Galileo Factors are orthogonal by definition,* the correlations are not redundant and indicate a substantial interrelatedness between Network Integration and perceived similarity. Thus there would seem to be ample evidence to rule out the possibility of unreliability for these two measurement instruments. Additional evidence of the reliability and validity of the cultural aggregate as measured by the Galileo instruments can be generated by regressing the unidimensional measures on Galileo Coordinate values. This follows from the "equivalence hypothesis" proposed by Gillham and Woelfel (l975:4) which states: . . . the scaling theory on which Galileo rests must assume that the Matrix D will represent the pattern of differences among the stimuli across whatever attributes the respondent perceives them to differ at the time and under the circumstances that the measurements are made. . . . If the "equivalence hypothesis" is correct, the space D should be the space within which the attribute vectors used by the respondents to differentiate the faculty are arrayed. Gillham and Woelfel (1975:14) further state that finding the projection of a vector or a set of vectors on a unidimensional regression model. So an optimal test of the hypothesis in * Deviations from othogonality due to rounding errors are negligible except for the smallest of the Galileo factors. 58 this instance consists in the goodness of fit of the six regression equations to the data: Y1 = bilxl + bi2X2 + bi3x3 + bi4x4 + biSXS + bi6x6 + e where: Y = 14 x 1 vector of scores on the ith unidimensional scale i' = The set of unstandardized regression coefficients 3 or partial slopes of the six unidimensional scales xi = is a known constant, namely the value of the independent variable, or rather, the six highest correlated factor loadings e = stochastic (error) terms Based on the correlation coefficients summarized in Table 7, each unidimensional scale was regressed on these six factors which correlated most highly.* The results of this analysis (Tables 8-13) show that the six attributes are clearly represented in the matrix of factor loadings. The Multiple Correlations among the six factor loadings and, (1) Years with the Department (Multiple R .88207 sig. .044); (2) Academic rank in the Department (Mul- tiple R .88922 519. .036); (3) Ph.D. Chairmanship (Multiple R .81858 sig. .142); (4) Ph.D. Committee Membership (Multiple R .81858 sig, ,041; (5) M.A. Chairmanship (Multiple R .94032 Only six factors were chosen since regression of the 14 x 1 vector on a 14 x 13 matrix of coefficients will always result in 100% of the variance explained. Since the factors are constrained to be othogonal, no redundacy results from this procedure. 59 TABLE 8 Standardized Partial Regression Coefficients, Multiple Correlation Coefficients, Tests of Significance, and Coefficients of Determination of Ph.D. Committee Membership on Galileo Coordinate Dimension Values MULTIPLE R .88504 R SQUARE .78329 DEGREES OF FREEDOM 6,7 F 4.21692 SIGNIFICANCE .041 Signifi- F R cance Beta FACTOR 2 4.5515931 .44203 .070 .3781628 FACTOR 3 1.4181148 .58761 .273 .2218817 FACTOR 12 .19810053 .64919 .670 -.0869670 FACTOR 7 7.0606974 .74586 .033 -.4820060 FACTOR 10 .80959214 .75593 .398 -.1708147 FACTOR 14 6.8433890 .88504 .035 -.5150953 TABLE 9 Standardized Partial Regression Coefficients, Multiple Correlation Coefficients, Tests of Significance, and Coefficients of Determination of M.A. Chairmanship on Galileo Coordinate Dimension Values MULTIPLE R .94032 R SQUARE .88420 DEGREES OF FREEDOM 6,7 P 8.90801 SIGNIFICANCE .005 Signifi- F R cance Beta FACTOR 1 8.4424202 .37841 .023 -.3737772 FACTOR 4 .63438366 .41564 .452 -.1061842 FACTOR 3 5.1442290 .47484 .058 .2982521 FACTOR 2 16.318769 .68519 .005 .5223654 FACTOR 7 11.324239 .86572 .012 -.4451895 FACTOR 5 8.1439816 .94032 .025 —.3970637 60 TABLE 10 Standardized Partial Regression Coefficients, Multiple Correlation Coefficients, Tests of Significance, and Coefficients of Determination of Academic Rank on Galileo Coordinate Dimension Values MULTIPLE R .88922 R SQUARE .79071 DEGREES OF FREEDOM 6,7 F 4.40783 SIGNIFICANCE .036 Signifi- F R cance Beta FACTOR 3.0173980 .34735 .126 -.3024111 FACTOR 1.0331866 .45467 .343 -.1838125 FACTOR ll 12.682675 .68037 .009 .6999421 FACTOR 6 .71765942 .77314 .425 -.l703576 FACTOR 10 .48040372 .79475 .510 .1406098 FACTOR 5 5.3207440 .88922 .054 -.4820076 TABLE 11 Standardized Partial Regression Coefficients, Multiple Correlation Coefficients, Tests of Significance, and Coefficients of Determination of Ph.D. Chairmanship on Galileo Coordinate Dimension Values MULTIPLE R .81858 R SQUARE .67008 DEGREES OF FREEDOM 6,7 F 2.36954 SIGNIFICANCE .142 Signifi- F R cance Beta FACTOR 4.2928558 .40877 .077 .4865606 FACTOR 2.3279503 .48664 .171 .4392694 FACTOR .93263074 .54901 .366 -.2494383 FACTOR 10 .71556501 .80545 .797 .0718694 FACTOR 8 .45264891 .81858 .523 -.1971641 61 TABLE 12 Standardized Partial Regression Coefficients, Multiple Correlation Coefficients, Tests of Significance, and Coefficients of Determination of M.A. Committee Membership on Galileo Coordinate Dimension Values MULTIPLE R .90766 R SQUARE .82384 DEGREES OF FREEDOM 6,7 F 5.45623 SIGNIFICANCE .021 Signifi- F R cance Beta FACTOR 5 .97565776 .36563 .356 -.l787767 FACTOR l3 2.2561258 .43399 .177 .2398993 FACTOR l4 5.6225845 .58097 .050 -.4008541 FACTOR 9 6.4993874 .75669 .038 .4451928 FACTOR 7 5.5382313 .84933 .051 -.413l719 FACTOR ll 4.0721593 .90766 .083 -.367314 TABLE 13 Standardized Partial Regression Coefficients, Multiple Correlation Coefficients, Tests of Significance, and Coefficients of Determination of Years with the Department on Galileo Coordinate Dimension Values MULTIPLE R .88207 R SQUARE .77804 DEGREES OF FREEDOM 6,7 F 4.08956 SIGNIFICANCE .044 Signifi- F R cance Beta FACTOR 1 4.1983922 .36488 .080 .3649795 FACTOR 12 2.5274387 .62250 .156 -.3094874 FACTOR 4 1.0055487 .66053 .349 .1866364 FACTOR 8.1396905 .77738 .025 -.5340847 FACTOR l4 1.6946022 .88207 .234 -.2482798 FACTOR 5 4.4033730 .074 .4195786 62 sig. .005); (6) M.A. Membership (Multiple R .90766 sig. .021) provide support for the "equivalence hypothesis." ‘In other words, respondents are clearly considering these attributes when differentiating among faculty members. Given the pattern of findings presented above, it seems reasonable to conclude, under the conditions of the present research, that the Galileo procedures produce a stable and precise measurement system which is equivalent to very ex- tensive applications of the best of conventional measurement systems. SUMMARY The results indicate that Hypotheses l and 2 are not supported. Evidence of support for Hypothesis 3 and the reliability and validity tests indicate that the Galileo measures of the cultural aggregate and Network Integration scores are reliable. Validity checks of network distance, or number of links between pairs of individuals, and frequency of communication are not amenable to straightforward tests with the data at hand. We must at least entertain the sus- picion that the failure to find support for Hypotheses 1 and 2 may be attributable to certain characteristics of these two variables. In the next chapter we will discuss this possi— bility and provide a more generalized interpretation of the results along with suggestions for future research. CHAPTER IV CONCLUSION In Chapter I theoretical support was presented concern- ing the interrelationship between the social structure and the cognitive belief system or "culture" of a social unit. Out of the ongoing pattern of intercommunication there arises a cultural belief system or rather, common defini- tions of those objects with which the social unit deals. The general acceptance of this premise can be found through- out many fields of inquiry, such as sociology, psychology, and anthropology. While many theorists have discussed this relationship, this thesis has been based largely on the work of Emile Durkheim who specifies that the cognitive belief system of any social unit varies with "the nature and number of channels of communication." He further states that we can only accurately describe social and cognitive structure by considering "the aggregate in its totality." In other words, we are concerned with how individuals fit into the system, how they are related to the other elements or members, and also, how the aggregate social structure is related to the cultural system. Following from this general reasoning we should find that: 63 64 Persons or groups in the same or equivalent locations in a communication network will develop equivalent conceptions of the collective representation. Although the major premise is of theoretical signif- icance in that it applies to all social units, this degree of generality is not amenable to empirical test. Any empirical test would require the specification of some con- crete subset of cultural objects which may be observed in some setting. Thus the following minor premise was derived: Persons who are equivalent (i.e. hold equivalent conceptions of the collective representation) will be seen as similar by others (insofar as those others have contact with them). It should be the case that the patterned similarities of people as perceived by the cultural aggregate will cor- respond to the social structure of that system. In other words, there will be a general correspondence between the pattern of similarities perceived among members of a social unit and their position in the communication network. From this general reasoning three hypotheses were derived: H1: The greater the frequency of inter- action between individuals or groups of individuals, the smaller the per- ceived distance between those individ- uals in the aggregate (cultural) belief system. The greater the number of links neces- sary to connect any pair of individuals in the communication network, the greater the perceived distance between those individuals in the aggregate (cultural) belief system. 65 H3: The more integrated an individual is in the communication network, the smaller the total perceived distance of that individial from all other persons in the aggregate (cultural) belief system. The results of the statistical tests presented in Chapter III clearly indicate a high degree of correspond- ence between the integrativeness of an individual in the communication network and the perceived similarity of that individual from all other members of the social unit. In fact, 10% of the perceived differences among the members can be accounted for solely on the basis of this relation- ship. Additionally, when the interaction effect of years in the department with Network Integration is considered, 15% of the variance in the Galileo Alienation Index can be accounted for. And finally, based on the Multiple Correla- tion of Network Integration and the Galileo Coordinate values we can conclude that the degree to which an individ- ual is integrated into the communication network is an at- tribute which people recognize when comparing individuals. Given this pattern of findings, it is reasonable to conclude that the more integrated an individual is in the communica- tion network the smaller the perceived distance of that in- dividual from all other members of the social unit. Thus, Hypothesis 3 is supported. The statistical tests of Hypotheses 1 and 2 fail to indicate any significant relationship between communication frequency, distance or links separating the members of the 66 social unit, and perceived distance in the aggregate (cul- tural) belief system. These two hypotheses, in a theoret- ical sense, are concerned with differences in location in the social structure as related to differences in percep- tion among the members of the system. The extent to which persons are perceived as different from each other is measured by the Galileo Means matrix, whose validity in this case seems well supported by its interrelationship with the six unidimensional attributes and with the Network Integration scores. Hypotheses l and 2 differ from each other in that each utilizes a different measure of the theoretical concept "distance apart" or "discrepancy of position" in the com- munication network. Although each of these may well be a reliable and valid indicator of some aspect of the overall "distance" in the network, there seems reason to suspect, due to the low correlations, that these characteristics-- Network distance and frequency of communication--are not themselves a complete measure of overall communication "distance" between persons. In fact, a more appropriate concept of "network dis— tance" might well be the inverse of the exact "volume of flow" of information between pairs of individuals. Such a variable would include not only number of links between persons, but also frequency of interaction, duration of average interaction, and "average information content" of 67 an interaction. Such a measure, across time, might be con- strued as a "communication history" for every dyad in the network, and it is probably to this variable that perceived differences among persons is related. CONCLUSION Though we failed to find a significant relationship be— tween "distance apart" or "discrepancy of position" in the communication network and perceived similarity, there is reason to believe that the operationalization of this vari- able was not of sufficient scope to accurately describe the complexity of this interrelationship. Considering the more substantial findings in terms of Hypothesis 3, that the more integrated an individual is in the communication net- work the smaller the perceived distance of that individual from all other persons in the aggregate (cultural) belief system, we can not reject the theory based on this analysis. APPENDIX I: QUESTIONNAIRES l-a Personal Communication Contact Questionnaire 1-b Galileo Questionnaire l-c unidimensional Scaling Questions 68 MICHIGAN STATE UNIVERS ITY College of Communication Arts EAST LANSING ' MICHIGAN ' 48824 Department of Communication June 24, 1974 Dear College miter: With the College's Ccmrunication Arts Building becauing more of a real- ity , it seems appropriate to include information about the oamnmication networks in the College as the architectural design proceeds. The pro— jectI amocnductinghasbeendiscussedwithDeanOyer, withtheDepart- ment maiman, with sate faculty groups, and, I believe, in departmental faculty meetings. In the enclosed questionnaire, you will find a list of the faculty and selected graduate students in the College, from Spring Term. The first few questions ask for some brief background information about yourself . Then, you are asked to indicate how often you ccmnmicated with other nenbers of the College about teaching and research topics during Spring Term. I am asking for approximations, because I recognize that this is a difficult question to answer precisely. You are also asked to indicate sure other features of your ocummication patterns in the College. In particular, you are asked to indicate the frequency with which you would refer to talk with other College mariners. While the first set of questions gives information on oomunication networks as they presently exist, the additional information cn preference is of importance for planning the new facilities. The second half of the questionnaire asks for you to make some ocupari- sons of College neubers in terns of the similarity or distance between them. This information provides another formof network that is also representative of the types of interaction taking place in the College. I ask you to place your natre on the questionnaire only for the purpose of enabling ne to construct the overall ccnmmication networks . Your nanewillbeinnediatelytransferredtoaoodemmberandngogeinthe College will have access to the list of names and numbers. Also, the questionnaires therrselves will not be made available to anyone else in the College in any way that allows the identification of specific indi- viduals. Nevertheless, initially your nane is needed so that the com- munication network can be constructed. Thank you very much for your assistance in this project. It is inpor- tant to me, and I believe the results (even in a general fonn) will be of help in planning for the future College facilities. Please retum the questionnaires to your departnental secretary for pickup. A manila envelope is provided for your convenience. Thank you again for your help. Margaret Brophy 517 S. Kedzie Hall Communication Department 69 In this section we are concerned with the frequency of your commun- ication about TEACHING and REEAKl-I. On the following pages we are pro- viding you with a list of names of your colleagues in the College. We would like you to indicate for TEACHING and separately for RESEARCH: l) the average frequency of camunication (if any) with each person, and 2) the mode or means by which you and the person communicate, and, 3) which of you initiates the commmication, and 4) the frequency you would have preferra to communicate with the person. Please use the following scale to indicate the approximate number of time; in the past term, you communicated with each individual on Efie Iist. First, describe your communication about RESEARCH (including your own work, the research of others in the college, and research being con— ducted outside the college, etc.). Second, use the same scale to indicate the frequency of your communication about Teaching (i.e., discussing course content, course materials , teaching styles and strategies , problems and successes, etc.). IFYOUCDDWNIC‘AIED: Onceortwice during the term ..... Writeal Several times during the term... ..Write a 2 (hoe or twice during the month. .. .Write a 3 Several times during the monE. .. .Write a 4 (hoe or twice during the week ..... Write a 5 Several times during the week.....Write a 6 Once or twice during the Q1 ...... Write a 7 Several times during the gay......Write a 8 When you and the other person communicate , who generally began or INITIATED the contact? IF: You did, almost always ................. Write a 1 You did, usually .............. . ........ Write a 2 Both of you did about equally ........ . .Write a 3 The other person did, usually .......... Write a 4 The other person did, almost always... .Write a 5 Please indicate the usual mode of communication: either ORAL (face to face, telephone, in meetings, etc.), or WRITTEN (proposals, rqaorts, distribution of papers, memos, letters, etc.) . IF IT WAS: Always oral ...... . . .......... . . ..Write a 1 Usually oral ..................... Write a 2 Equally oral and written. . . . . . . . .Write a 3 Usually written .................. Write a 4 Always written ................... Write a 5 70 Nextwewouldliketo findout sure things aboutthe communication. contacts you would like to have had during the term. IIhis information will become part of an abstract "network" (without specific names) that may be useful in designing the future College building layout. Suppose You and all other college personnel were free to do as much commmica- —tI'n_g about RESEARZH and TEACHING as each persm wanted. How often would you PREEER to talk with each person? IF YOU WCIILD HAVE PREFERIED 'IO WICATE: (hoe or twice during the term ......... Write a 1 Several times during the term ......... Write a 2 (hoe or twice during the month ........ Write a 3 Several times during the month ........ Write a 4 Qiceor twice during theweek ......... WriteaS Several times during the week ......... Write a 6 mos or twice during the day. ......... Write a 7 Several times during the day .......... Write a 8 Note: After you have made these four decisions about a person, move on to the next name and make the decisions again. FREDUEDCY OF CDNMZJNICATICN & PREFERRED WICATICN Onceortwiceaterm ..... l Severaltimesaterm ..... 2 Once or twice amonth....3 Several times amonth....4 Chceortwiceaweek ..... 5 Severaltimesaweek ..... 6 Once or twice ad_ay ...... 7 Several times a_d_ay_......8 EDIE OF MNICATIGNI: Always oral. . . ........... 1 Usually oral. . ........... 2 Equally oral & written. ..3 Usually written .......... 4 Always written ........... 5 71 M10 INITIATFS the WICATION? You do, almost always ........... 1 You do usually ................ . .2 Both of you do about equally....3 The other person does usually. ..4 'Ihe other person, almost always.5 i E KOZWCOMZ‘JW mmewHt-JHZH MUOK‘ omwwmusmwvu «nzmcomww mtdtatt’Hr-JHZH [11003 UMWNMMMW'U Individual A* Individual B Individual C Individual D Individual E IndividualF' Individual G Individual H Individual I Individual J Individual K Individual L Individual M Individual N * Actual names were provided in the distributed questionnaire. 72 Just as we can measure the difference between a pair of physical objects (in terms of inches, feet, miles, or other mnits) we can also measure the distance between pairs of concepts, ideas, or peOple. We will call this unit a "GALIIED." In the following questions we want you to estimate the differences or distances between the members of each pair. As you will see at the top of each page we have provided you with an example of the differences or distances between two concepts: RED and MIITF. are 50 Galileos apart. Now, consider each pair carefully, and then estimate the number of Galileos you feel separate the pair. Ifyaifeeluneyare_<fl_o_s§£toget1erflianthepairofconceptsinflne guide, writeammberless than50. Iftlneyare furtherapartthanthe pair of concepts in the guide, write a number more than 50. 73 Panther: Red and White are 50 Galileos apart. HCW FAR APART ARE: Galileos Professor" A and B Professor A and C Professor A and D Professor A and E Professor A and F Professor A and G Professor A and H Professor A and I Professor A and J Professor A and K Professor A and L Professor A and M Professor A and N Professor A and 0 Professor A and P Professor A and Q Professor A and R Professor A and S Professor A and T Professor A and U Professor A and V Professor A and W Professor A and X Professor A and Y Professor A and Z 1”Actual names were provided in the distributed questionnaire. 74 Name 2. Age 3. Years With the Department 4. Position: a. Full Professor b. Associate Professor c. Assistant Professor d. Instructor e. Assistant Instructor f. Lecturer 9. Other (please specify) If Graduate Student : h. lst year Graduate Student 1 . 2nd year Graduate Student j. 3rd year Graduate Student k. 4th year Graduate Student 1. more than 4 years as a Graduate Student m. Other (please specify) Please describe your main academic and/or research areas of interest: Approximately how many M.A. and Ph.D. DEGREE committees in the College do you serve on, as: Ph. D. Chairman 93 Member only M.A. Chairman QR Matber only How many UNIVEIBITY committees do you serve on? W) How many COIIEGE committees do you serve on? - W) How many PROFESSIONAL ASSOCIATIONS do you belong to? (Number) How many PmFESSIONAL CINVENI‘IONS did you attend since June 1 of last year? Number) Approximately how many PUBLICATIONS did you have since June 1 of last year (including books, jounnal articles, convention papers , reports/monographs) ? (Number) 2-b 2-c APPENDIX 2: STATISTICS Galileo Means, Standard Deviations, Variances, Skewness, Kurtoses, Counts, Minimum-Maximum Values, Range Network Distance Matrix For Faculty Aggregate Means Matrix of Faculty Members Coordinate Dimension Values For Faculty 75 .omcom .moch> Eseflxmz IESEHCHE .mucsoo .momounsm .mmmczmxm .moo:maum> .mcoflum~>mo ouwosmum .mccmz ooaaamo fl :wom.¢ same «we mzcuuqszumac ucaau>¢ o.oo a.aa~ e.~ ~ m~:.~ an». mo~.o:_« .a9.nn ~ms.ow n~ :— aoosu aoamq e.~ a “do.“ «be. ea—.e~cs nmo.n: m~c.n~ m— :« o.o: a.- o.m~ o -e.« -a. am~.~c~ am~.au w~a.~: .s :— o.on a.am a.o~ m e~o. -~. aoo.eiq c:s.~« occ.mm cu :~ e.mm cows e.gw o -a.— “no. ~oc.«:m :c:.c¢ =ee.a: c as c.8eu e.a~— a.:« o ~«m.~ moo. oau.o-~ ¢.u.~: m~o.oc o :. e.mo e.eed a.m ~ ocm.~ cam. cc~.:«e~ o:c.~n aoo.m: ~ :. e.oam a.eom a. o ~me.n emu.“ o«m.n:~m~ doe.cma occ.n~_ c :L e.m= e.mo o.o~ o -n.. ~c~.- oe«.~e~ mma.o~ m~..~m m :. e.m~u e.oo~ a.m~ ~ -:.n ~o:.« cu¢.:ccm m-.am ¢~m.a~ : :— e.m~ e.ae~ e.m~ o o~o. -~.o occ.u.c ~o:.o~ -o.me ~ :~ e.m:« e.cm« ¢.m c ~90.“ cacoo em~.ame~ ecesm: ecc.cc m :. e.mc~ a.oc~ a.m« o ~mo.~ emm. ~cc.c-~ ano.cm em~.~e a :. e.m~ coon o.a~ n cam. o~m.u -~.a~ ce:.c ~co..~ - nu o.mo a.oau a.mn n new. n:c. ~¢c.c~o -m.a~ cac.ac «a n— o.ao a.oa~ o.c: o as..~ ~oc. -:.~uo oem.:~ n~o.oc as - a.am c.9o 9.9“ e e-.. -~.u oc¢.c- ~co.o~ ~co.o. o a. e.ao ooe~ coon m amaou mac. eaa.:c: «cm.d~ oao.ew c n. a.e~ coco a.a~ J om~o coo. aoa.wm aoo.m coo.w~ ~ mu o.ao: o.ee¢ e. m mm~.~ ~e~.~ aco.nse- -~.ew« aoa.c=~ c ~. e.mo a.ae~ a.m c o-.~ ewe. owc.meo ~ee.en w~c.o. m m. a.m: comm coca m n~e.« men. aea.con no:.~— eoa.c~ : «a c.9m c.9o a.en m ~oc. n«~.u e¢n.¢~n -:.e— eea.~w n n. aooau 9.9m“ e.v: m oma.~ rma.« amc.ex:_ n¢:.cn ea~.c~ N nu o.c«n e.ga; 0.50 4 oen.— amm. ~c~.:cem~ arc.e~. em~.c- . n- e.a¢ cocoa e.ao : eco.a one. ae~.¢- oc~osu cam.- a. N. a.ao coca“ a.a: c :~m.« we“. cwa.a~n cm~.c— ~c..ce cu ~¢ a.om coco e.a~ a «do. as“. occ.c~r cc:.¢n om~.—~ : m. e.mm e.m~ o.e~ e ooa.~ ~La. ccm.a~: ~cc.- . ~co.a: c ~. o.mm e.ea a.m~ e awn." :Ne. oee.e¢~ cam.o. ~co.¢: s - o.eo~ aooaw a. m -d.~ :m..~ n:m.::nm :o«.n~ eat..c a Na e.oa~ o.oe~ =.~ c c~c.: eec.s oec.ec- :e~.~m e~c.:m m ~d a.o~ e.am e.e: : ~mm.~ ::_._- =c~.r— am~.¢ aoo.~: : ~. a.a~ c.am o.sn a ~sc.. c-. ~c~.:: -a.a oaa.o~ ~ ~. swim xHQmemd '76 9.99. 9.99. 9.99 9.99~ 9.9a. e.gom 9.9a 9.99~ 9.9m 9.99. 9.9:. 9.9» 9.99 9.9~ 9.99. 9.9- 9.m~ 9.m~ 9.99. 9.a~ 9.m9 9.99 9.9~ 9.9. 9.sm 9.94 9.m9 9.m9 9.~9: 9.99 9.9~ 9.~9 9.mm 9.9m 9.9¢~ 9.99. 9.99. 9.90 9... 9.99 9.e- 9.99 9.:~ 9.99 9.99 9.m4~ 9.99. 9.m~ 9.9: 9.eoe 9.9. 9.mm 9.99: 9.9m 9.~c. 9.99 9.9. 992.9 9.9u~ 9.99~ 9.99. 9.9- 9.99~ 9.99m 9.99. 9.99: 9.99. 9.99M 9.9m“ 9.9: 9.99 9.999 9.99~ 9.9m~ 9.99 9.999 9.m~n 9.99 9.99 9.99. 9.99 9.999 9.9~ 9.9~ 9.m~ 9.99. 9.999 9.99u 9.ms 9.999 9.mo 9.99 9.999 9.999 9.99m 9.m~ 9.99 9.999 9.99m 9.999 9.m~ 9.9m 9.9~ 9.9m~ 9.99~ 9.99 9.99 9.999 9.9K 9.mm 9.999 9.99 9.n- 9.999 9.9~ 44> .x¢z 9.9m 9.9N 9.9— 9.9~ 9.9: 9.99 9.9: 9.9N 9.9: 9.9~ 9.99 9.— 9.9m 9.9a 9.9« 9.9m 9.99 9.99 9.9N 9.9 9.99 9.99 9.99 9.9M 9.9N 9.99 9.99 9.99 9.9 9.9m 9.9 9.n 9.99 a. 9.99 a. 9.9 9.99 9.9: 9.9a 9.9» 9.99 9.9 a. 9.9 9.m 9.9 9.9 9.9: 9.9— 9.m~ a. 9.9~ 9.9» 9.99 9.9: 9.99 499 d N00WNNU‘NU‘NU‘NNONNOU‘U‘NONCOWOQOOU‘QCC‘NQO‘O‘NRAQmOONO‘ONCONCCOC ~9..~ ~.~.. ~.~.~ ....~ ....~ so... .-.~ ~.... 9.... ....~ ....~ ..9.. ..a. a... 9.... o...~ as... .9... cm... .~... ~ma. ~.~.. ~...~ o-.~ -~.. -~.. .4... ~.~.~ .em.~ ~.e.~ ..~.~ 9...~ m.... ....~ ~.:.. 9.... ..~.. ~m... .~... ~.~.. so..~ so... a...~ :.~.. ..... ~.o.~ m.~.~ so... .~... :9... c.... -c.. mm... ~9... maa.~ c... a... m.m:~a:y :mztDZ by“ @:£o ms;.« «.9... Mao.u «meow 99.. cu $09.9: «.9.~ ch. 9.9.9 9:9.“ ore. 99—. ~99. o9s. ewe.“ rc~.o :m~.o 929.- ~09. ~99.- ~99. 999.: 099. eds.» 959. @¢~.o 999.9 :~9.« 9~:.u uNJ. :99. escoo 00:. -~. 4N9.: can. m-.o ::n. 999. ~0:.u aso. 999.9 9:9. 9:9. c9~.u mom. @:¢.o swc. c-.« 999.0 o~m.c an.4 cc~.o one. man.: 999.: mmmzzmzm er. ohm-Jr. GUN .6...m..N Jr..— .t..~ mmwowem: ~3ro~vtm flaw-Jud: arsoffir ~9¢.w99:u e9c.c.¢ 9.A.9:~n 99:.:~9. a~c.-. 9mm..¢: 999.99~ 9-.9:~: 9mm.~m99 mc~.n.~ 9.c.ma~ cor..... ow~.cwc 999.9t9. 999.99: 990.com 999.999 cru.hu~ 9c~..c. -~.~c. 999.:mm nrr.cc:9~ 99¢.cce 99~.~9: :»~.ecc 999.9cn ~oo.e- on—.m99un -u.9mnu ncm.:—9nn 99~.~¢m 999.com 9a9.9.9 999.crwc n9:.:~c :.~.e:: -4.~9. 9.5.999 o¢~.:.~w ~:m.c~.n are... 999.:u~ mc~.~¢nc~ 39 3.4.9 9o:.:~m mrk.npc:w ccnoumr ~s¢.nmcw 5.9.9 outm 999.4“ wcrdu 24> a home trcu 965.cm 99m.9m ~:m.- ~4~.~o @59.nm 904.99 n9~.¢u -:.¢- cm9.4~ ~90.cw m¢~.ne -n.nn ~m¢.u~ 9m:.o~ 099.90 m9c.9~ ~9~.o~ ~un.- 3:9.nn $99.m~ 49:.Nn Nn:.un :sc.c« 90m.¢~ sns.:u no:.nu 9@~.- 99¢.MN mc~.oou ~9~.o~ 9c~.u~ ooa.o~ anu.o~ -u.m« m9¢.:9n n99.0n 609.999 nn~.n~ 999.99 9cm.9~ ~n9.«c J-.9~ nno.«~ -u.9u N9~.:N www.mh Osn.om mmo.O ~m9.mu N~:.Ncu som.mu c¢m.0« som.omu suc.~u sur.um 99n.o~ 940.; .>mo .zcbm «cu mu—hmuheum ONfloaT ccm.a~ mmfiou: scooor :9~.9? m-.:~ 999.99 7N:.¢t 5.9— on»... €99.99 @€~.:: m~—.mu 96w.~w 999.mc wsn.$¢ soc.~r nla.sw OSN.J~ Mfi¢.wt 999.~m 999.0w cow-G: hwc.s¢ 9m~.oJ amsonr 999.m: nnn.c: 999.0m 999.0Ma ~0u.:: 999.:m msn.an eca.cn 9¢~.9m .nou.cdu n:~awm 999.90. um9.~: 999.~m 999.n9 ~¢—.49« oc~.:m 999.cN 509.:— 999.99 999.me 999.09 .. z .9... 999.:m ~ms.mu- 99~.nm ££N.Ce ~$c.~:u 99~.~w ~¢«.- 999.99 9—som9 ram: CDC-N a. Add“!“NnflNanNW-fm"N”JmflflNHJWCNdNMJJ‘cNQ«N‘HJU‘ONSC‘U‘NWJU‘ONC: N. m« a. a. 9— a. a. u— «a «a «— an 9— 99 9d 99 9— COO coo-c- INMM:JJU‘U\WJ‘~°~G~D~D\D~NNNNNc&mCDCQO‘U‘C‘O‘C‘QU‘O‘ ca 77 TABLE 2-b* Network Distance Matrix For Faculty oh 5. \IO‘ 10. 11. 12. 13. 14. PERSON PERSON PERSON PERSON PERSON PERSON PERSON PERSON PERSON PERSON PERSON PERSON PERSON PERSON A 0 1 1 1 2 2 1 1 1 2 2 2 2 2 B 1 0 l 1 2 2 1 1 l l 1 2 2 2 C 1 1 0 2 1 1 1 l 1 2 2 2 1 2 D 1 1 2 0 2 l 1 1 1 l 2 2 l 2 E 2 2 1 2 0 2 2 2 1 1 1 1 2 1 F 2 2 1 1 2 0 2 2 2 2 3 3 1 2 G 1 1 1 1 2 2 0 1 1 1 2 1 2 2 H 1 1 1 1 2 2 1 0 1 1 2 2 2 2 I 1 1 1 1 1 2 1 1 0 1 1 1 2 1 J 2 1 2 1 l 2 1 1 l 0 1 2 2 1 K 2 1 2 2 1 3 2 2 1 1 0 1 3 1 L 2 2 2 2 1 3 1 2 1 2 1 0 2 2 M 2 2 1 1 2 1 2 2 2 2 3 2 0 2 N 2 2 2 2 1 2 2 2 l l 1 2 2 0 * Actual distance matrix would be 67 x 67, only the upper left section is represented here, giving the scores of faculty members. total 67 x 67 matrix. Scores referred to in the text are based on the 78 ooo. 9mm.om 3.9.2. 3.9.99 ooo.mm 999.99 3.9.99 2 Ewan 3 909. hbbéN 999.99 mmmdw bmwém ooo.mm 2 7699mm ma 999. oomfin 93.99 omNAm 999.99 A Bmmmm NH 999. mNm.Hv 999.99 39.99 & gm AH 999. oom.~.m 99960 9 599mm 9H coo. oomdv H 6mm?“ m 999. m gm 9 3” ma NH Ha 9H m 9 999.3. www.mma mNH.mm H9999 mmm.mm 999.99 933mm 2 59mm 3 999.3. 909.9: www.mm 999.9N ~5me CONdh omN.mm.n 2 ammmfi ma 990.99 99923 £9.99 oomfiv 999.9m mdem oom.o~. A 7599mm NH mNHéh oomNm vaom 933mb 999.99 mmN.vm mNH.mH M Ema HA mhmdm hmmfim mvafim mwNéh mmm.mm 999.Nm ooo.mm .9 75mm?” 9H hmwfiw owed". omh.mm 999.9. mmm.9v ooodm 999.9NH H Emma.” m 99.73. oooém mhmfim 999.9m omN.oN 93.9.: mvafim m Bmmmm 9 999. oomdma hmmfiv 999.Nm 99969 hoaéoa mmN.vm U zommmm N. 999. 999.9N hwaéa ooo.mm 99969 999.99 .m 5mm m 999. SKUON oooém emanaa OONAH m 7699mm m 999. omN.mN hmvaH ooN.~.m 0 gm 9 999. 99.72. 999.99 0 ammumm m 999. 9.2..ma m ammmm N 999. N 8959 H N. o m 9 m N H DIN Emma 8.98: 3.8% no ..ng 882 38683 79 N979999HI 9N9.9999HI 999.9999... 999.999HI 9N9.N99I 99H.HHN... 99N. WHOOmzmuHm 999.N 9N9.HH 999.9H H99.I 9N9.9H 999.9... 9N9. 2 7599mm 9H 999.99 N99.99I 999.9H 9H7 99N.HI 999.9 99N.! E 29959 9H 9HN.9H... H99.9I H99.9I 999.N 99H.HH 9NH.NI 999.I .H 699mm NH H99.9NI 999.9! 999.9N: 999.NHI 999.I 9N9. 999. M Emmmm HH 999.9I 999.9 N99.HNI N99.9H H99.9I 999.NI 999. 9 209.999 9H 9N9.9 H99.99.. 99N.9HI 9N9.9! HN9.9I 999.9I 99H.I H Ema—d 9 999.9NI 999.9... 999.N 999.9HI 999.N... 999.NI 999.I a Zommmm 9 999.NN 9H9.99 HNN.9I 999.Hl NNN.9I 999.H 9H7 0 Zummmm 9 N99.99 999.99 999.H 999.9! 999. H99.I NNH.I .9 59999 9 N9H.99.. 99H.9H... 999.9I 999.9I 999.N 9979H 999.I m gm 9 999.99.. N99.9... 999.9N 999.9 9N9.9... 999.NI 999.... O zommmm 9 9979.. 999.9H! 9N9.NNI HNH.9H 999. H99.H 999.... U gum 9 999.N9I 999.9 9N9.9N 9H9.I 999.NI 9H9.N 9NH. 9 299999 N 999.99 999.99.. 999.9 H99... 99N. 997 999. .9 76959 H 9H 9H NH HH 9H 9 9 99N . 9NHH 999 .999H 999 . N9N9 NH9 .9999 999 . 9999H 999 .9N99N 999 . 999HN 998979.me 999.9 H99.9H... 9N9.9HI 999.9H H9H.9H 999.9 9H9.99I 2 7699mm 9H 99N.9 9H9.9 9N9.HH H99.9 999.H9 999.99... 999.99.. 2 Emmmm 9H H99.9I H99.NN 9NH.9H H9N.HNI 999.9H! 9N9.HHI 999.N: H Zommmm NH 999.NI H99.H.. 999.9N HH9.99 999.9N 999.9N 999.N M 59.9mm HH 9H9.9 999.9H NH9.HNI 9NN.NN 999.99 999.9H 999.9 9 $99 9H 999.9HI 99N.9I N99.9HI 99N.HNI 999.9H 99N.N9I 999.9I H 7599mm 9 9H9.NN 99N.9 999.9! 999.9... 9N9.99I 999.9HI 999.9 m 59%.: 9 999.N! 999.N 9H9.9 999.9H... 999.9N: 999.99 H99.99! 0 Zommmm 9 99N.HI 999.HI 999... 9N9.H 999. 9H9.99I H99.99 .9 gm 9 999.9I 999.HI 999.9NI 999.9 999.9N! H99.9HI 999.N m 5955 9 999.9I 999.HI 9H9.NH 999.NN 99N.99I 999.NNI 999.9 Q gm 9 N99.9 999.9N... 999.9H N9H.9NI 999.HH.. 9NN.9HI N99.9 0 7699mm 9 999... 9HH... N99H.H 999.9N... N99.H9 N99.99 N99.9H m 59.9mm N 999.I NN9.H 99H.H.. 999.HI H9H.99I 999.99 H99.99 fl 5955 H 9 9 9 9 9 N H Each Mom mmDHM> 835.59 mumfiouooo UIN 159a. APPENDIX 3: GALILEO PLOTS 3-a X - Y Plane 3-b Y - Z Plane 3-c Plot of First Through Real Dimensions 80 APPENDIX 3 9A X-Y Plane 06 .B 0K 0 ON J E LJ .C ‘H ‘D I F0 'M Figure 3-a Galileo Plots. 81 Y-Z Plane B ‘M Jo O OI N. K F. ..L C E o H: G A' ° D Figure 3-b X - Y Plane. 82 X-Y Plane ml. F Y-Z Plane Figure 3-c First Through Real Dimensions. 'REFERENCES 83 REFERENCES Anderson, N. H. "Application of an Additive Model to Impression Formation," Science, 138 (1962), 817-818. Anderson, N. H., and S. Herbert. "Effects of Concomitant Verbal Recall on Order Effects in Personality Forma- tion," Journal of Verbal Learning and Verbal Behavior, (February 1963), 379L391. Anderson, N. H. ”Averaging Versus Adding as a Stimulus - Combination Rule in Impression Formation," Journal of Experimental Psychology, 70 (1965), 394-400. Anderson, N. H., and A. Jacobsen. "Effects of Stimulus Inconcsistency and Discounting Instructions in Person- ality Impression Formation," Journal of Personality and Social Psychology, 2 (1965), 531-539. Anderson, N. H. "Averaging Model Analysis of Set-size Effect in Impression Formation," Journal of Experi- mental Psychology, 75 (1967), 158-165. Anderson, N. H. "Likableness Ratings of 555 Personality Trait Words," Journal of Personality and Social Asch, S. E. "Forming Impressions of Personality," Journal of Abnormal and Social Psychology, 41 (1946), 558-290. Asch, S. E. Review of "A Theory of Cognitive Dissonance by Festinger," Contempora£y_Psychology, 3 (1968), 194-195. Barnett, G. and J. Woelfel. "On the Dimensionality of Psychological Processes," paper presented at the Annual Meeting of the Psychometric Society, Bell Laboratories, Murray Hill, New Jersey, (April 1976). Barnett, G. Bilingual Information Processing: The Effects of Communication on Semantic Structure. Ph.D. disser- tation, East Lansing: Department of Communication, Michigan State University, (1976). 84 Berlo, D. K., R. V. Farace, P. R. Monge, S. A. Betty, and J. A. Danowski. "An Analysis of Communication Struc- ture of the Office of Civil Defense," East Lansing, Michigan: Department of Communication, Michigan State University, (1972). (Report to the Office of Civil Defense.) Blumer, H. "Collective Behavior," New Outline of the Principles of Sociology, ed. A. M. Lee. New York: Barnes and Noble, (1946). Bruner, J. S., and R. Taguiri. "Person Perception," The Handbook of Social Psychology, ed. G. Lindzey. Vol 2, Reading, Mass.:7 Addison Wesley, (1954). Burgess, R. L. "Communication Networks and Behavioral Consequences," Human Relations, 22 (1969), 137-159. Brunswick, E. Perception and the Representative Design of Psychological Egperiments, Berkley and Los Angeles: University of California Press, (1956). Byrne, D. "Interpersonal Attraction and Attitude Similarity," Journal of Abnormal and Social Psychology, 62 (1961), 713-715. Byrne, D. and D. Nelson. "Attraction as a Linear Function of Proportion of Positive Reinforcement," Journal of Social Psychology, 1 (1965), 659-663. Byrne, D., J. Lambreth, J. Palmer, and O. London. "Sequen- tial Effects as a Function of Explicit and Implicit Interpolated Attraction Responses," Journal of Person- ality and Social Psychology, 13 (1969), 70-78. Cartwright, D. "Achieving Change in PeOple: Some Applica- tions of Group Dynamics Theory," Human Relations, 4 (1951), 381-392. Cartwright, D. and H. Zander. Group Dynamics, New York: Harper and Row, (1968). Cohen, A. "Cognitive Tuning as a Factor Affecting Impression Formation," Journal of Personality, 29 (1961), 235-245. Cohen, A., E. L. Robinson, and J. L. Edwards. "Experiments in Organizational Embeddedness," Administrative Science Quarterly, 14 (1969), 208-221. 85 Collins, B. E., and B. H. Raven. "Group Structure: Attraction, Coalition, Communication and Power," Th3 Handbook of Social Psychology, ed. G. Lindzey. Vol. 4, Reading, Mass.: Addison-Wesley, (1969). Cooley, C. H. Human Nature and the Social Order, New York: Scribner's, (1902). Cottrell, L. S. "The Analysis of Situational Fields in Social Pscyhology," American Sociological Review, 7 (1942), 370-382. Cronbach, L. J. "Processes Affecting Scores on 'Understand- ing of Others' and 'Assumed Similarity,'" Psychological Bulletin, 52 (1955), 177-193. Dailey, C. A. "The Effects of Premature Conclusion Upon the Acquisition of Understanding of a Person," Journal of Psychology, 33 (1952), 133-152. Danowski, J. A. Environmental Uncertainty, Group Communica- tion Structure and Stress, M.A. thesis, East Lansing: Department of Communication, Michigan State University, (1974). Dornbusch, S. M., A. H. Hastorf, S. A. Richardson, R. E. Muzzy, and R. S. Vreeland. "The Perceiver and Perceived: Their Relative Influence on Categories of Interpersonal Perception," Journal of Personality and Social Psychology, 1 (1965), 434-440. Durkheim, E. The Rules of Socialogical Method, trans. G. E. Catlin. Chicago: The University of Chicago Press, (1938). Durkheim, E. The Division of Labor in Society, trans. G. Simpson. Glencoe: The Free Press, (1947). Durkheim, E. Selectigns From His Work. Introduction and Commentaries by G. Simpson, Thomas Y. Crowell Company, University of New York,(l963). Farace, R. V., W. D. Richards, P. R. Monge, and E. Jacobson. "Analysis of Human Communication Networks in Large Social Systems," (unpublished paper). East Lansing: Michigan State University, Department of Communication, (1973). Farace, R. V. and P. R. Monge. "Analysis of Human Communica- tion Networks in Large Social Systems," research pro- posal submitted to National Science Foundation, Depart- ment of Communication, Michigan State University, February, (1973). 86 Festinger, L. "A Theory of Social Comparison Processes," Human Relations, 7 (1954), 117-140. Fillenbaum, S. and A. Rapaport. Structures in the Sub- jective Lexicon, New York: Academic Press, (1971). Frijda, D. H. "Recognition of Emotion," Advances in Experimental Social Ps cholo , ed. L.‘§erkowitz, New York: Academic Press, (1969). Gallob, H. F., and G. W. Fischer. "Some Relationships Between Social Inference, Cognitive Balance and Change in Impression," Journal of Personality and Social Psychology, 26 (1973), 16-22. Gillham, J. R. "The Aggregation of Shared Information in a Socialogy Department," unpublished Ph.D. disserta- tion, Urbana: University of Illinois, (1962). Gillham, J. R. and J. Woelfel. "A Sociological Illustra- tion of Construct Explication Using Multi-dimensional Scaling," Central States Sociological Association Annual Meeting, Windsor-Ontario, (1974). Glanzer, M. and R. Glaser. "Techniques for the Study of Group Structure and Behavior: I Analysis of Struc- ture," Psychological Bulletin, 56 (1959), 317-332. Glanzer, M. and R. Glaser. "Techniques for the Study of Group Structure and Behavior: II Empirical Studies of the Effects of Structure in Small Groups," Psy- chological Bulletin, 58 (1961), 1-27. Griffitt, W. B., D. Byrne, and B. H. Bond. "Proportion of positive Adjectives and Personal Relevance of Adjectival Descriptions as Determinants of Attraction," Journal of Experimental Social Psychology, 7 (1971), 111:1—1. Guetzhow, H. "Communication in Organizations," Studies on Behavigr in Organizations: A Research Symposium, ed. R. W. Bowers, Athens, Georgia: University of Georgia Press, (1965). Hakel, M. "Significance of Implicit Personality Theories for Personality Research and Theory," Proceedings of the Psychological Association, (1969). Hastorf, A. H., D. J. Schneider, and J. Polefka. Person Perception. Massachusetts: Addison-Wesley, (1970). Heider, F. "Social Perception and Phenomenal Causality," Psychological Review, 51 (1944), 358-374. 87 Heider F. The Psychology of Interpersonal Relations, New York: Wiley, (1958). Homans, G. C. Social Behavior: Its Elementary Forms, New York: Harcourt, Brace, and World, (1961). Johnson, J. D. "The Analysis of Organizational Communica- tion Phenomena Over Time: Some Selected Methods Applied to Communication Networks," paper contributed to International Communication Association Convention, Chicago, (1975). Jones, E. E. and R. E. Nisbett. The Actor and the Observer: Divergent Perceptions of the Causes of Behavior," Attribution: Perceivigg the Causes of Behavior, ed. Jones et al., Morristown, New Jersey: General Learning Press,—Tl§72). Kanouse, D. E. "Language, Labeling and Attribution," Attribstion: Perceiving the Causes sf Behavior, ed. E. E. Jones 2E al., Morristown, New Jersey: General Learning Press,"Tl972). Kaplan, M. F. "Interpersonal Attraction as a Function of Relatedness of Similar and Dissimilar Attitudes," Journal of Experimental Research in Personality, 6 Kelley, G. A. The Psychology of Personal Constructs, New York: Norton, (1955). Kelley, H. "The Warm-Cold Variable in First Impressions," Journal of Personality, 18 (1950), 431-439. Kogan, N. and R. Tagiuri. "Interpersonal Preference and Cognitive Organization," Journal of Abnormal and Social Psychology, 56 (1958), 113-116. Leeper, R. "The Role of Motivation in Learning: A Study of the Phenomenon of Differential Motivation Control of the Utilization of Habits," Journal of Genetic Psy- chology, 46 (1935), 3-40. Leventhal, H. "The Effects of Set and Discrepancy on Impression Change," Journal of Personality, 30 (1962), 1-15. Large, I, D. Fox, J. Davitz, and M. Brenner. "A Survey of Studies Contrasting the Quality of Group Performance and Individual Performance," Psychological Bulletin 55 (1958), 337-372. 88 Lott, B. E., and A. J. Lott. "The Formation of Positive Attitudes Toward Group Members," Journal of Abnormal and Social Psychology, 61 (1960), 297-300. Luchin, A. S. "Primacy-Recency in Impression Formation," The Order of Presentation in Persuasion, ed. C. Hovland, New Haven, Conn.: Yale University Press, (1957). Luchin, A. S. "Definiteness of Impression and Primacy- Recency in Communication," Journal of Social Psychology, 48 (1958), 275-290. Luchin, A. S. "Definiteness of Impression and Primacy- Recency Effects in Impression Formation," Jogrnal of Abnormal and Social Psychology, 68 (1964), 335-388. Manis, M. S., S. C. Cornell, and J. C. Moore. "The Trans- mission of Attitude-Relevant Information Through a Communication Chain," Jggrnal of Personality and Social Psychology. 30 (1974), 81-94. Mead, G. H. Mind, Self and Society, (posthumous) ed. C. M. Morris, Chicago: University of Chicago Press, (1934). Mead, G. H. The Philosophy of the Act, edited, with Intro- duction by, Charles W. Morris, in collaboration with J. Brewster, A. Dunham, D. Miller, Chicago, Ill: The University of Chicago Press, (1938), (1950). Mead, G. H. The Social Psychology of George Herbert Meed, ed. Anselm Strauss, Chicago: UniVersity of Chicago Press, (1956). Messick, S. and N. Kogan. "Personality Consistencies in Judgment: Dimensions of Role Constructs," Multi- variate Behavioral Research, 1 (1966), 165-175. Miller, D. L., and G. H. Mead. Self, Language and the World, Austin: University of Texas Press, (1973). Miyamoto, S. F., and S. M. Dornbusch. "A Test of Inter- actionist Hypotheses of Self-Conception," American Journal of Sociology. 61 (1956), 399-403. Monge, P. R., K. Kirste, and J. Edwards. "A Preliminary Causal Model of the Formation of Communication Structure in Large Organizations," paper presented to the International Communication Association, New Orleans, (1974). 89 Morris, C. W. {ysrieties of Human Value, Chicago: University of Chicago Press, (1956). Newcomb, T. M. "The Prediction of Interpersonal Attraction," American Psychologist, 11 (1956), 575-586. Norman, W. T. "Toward an Adequate Taxonomy of Personality Attributes: Replicated Factor Structure in Peer Nomination Personality Ratings," Journal of Abnormal and Social Psychology, 66 (1963), 574-583. Osgood, C., G. J. Suci, and P. H. Tannenbaum. The Measure- ment of Meaning, Urbana, Illinois: University of Illinois Press, (1957). Parsons, T. The Social System, New York: The Free Press of Glencoe, (I951). Parsons, T. and E. A. Shils. Toward a General Theory of Action, Cambridge: Harvard University Press, (1951). Passini, F. T. and W. T. Norman. "A Universal Conception of Personality Structure?," Journal of Personality and Social Psychology, 4 (1966), 44-49. Posavac, E. J. and S. J. Pasko. "Interpersonal Attraction and Confidence of Attitudes and Attitude Similarity," Psychometric Science, 23 (1971), 433-435. Richards, W. D. "A Conceptually Based Method for the Analy- sis of Communication Networks in Large Complex Organ- izations," paper presented to the International Communication Association Conference, Phoenix, (1971). Richards, W. D. "Network Analysis in Large Complex Systems: Theoretical Base,” paper presented to the International Communication Association Conference, New Orleans, (1974a). Richards, W. D. "Network Analysis in Large Complex Systems: Techniques and Methods: Tools," paper presented to the International Communication Association Convention, New Orleans, (1974b). Richards, W. D., R. V. Farace, and J. A. Danowski. NEGOPY: Program Description, unpublished paper, East Lansing, Mich.: Department of Communication, Michigan State University Press, (1973). 90 Riecken, H. W. and G. C. Homans. "Psychological Aspects of Social Structure," Handbook of Social Psycholqu. ed. G. Lindzey, Cambridge Mass: Addison-Wesley, (1954), 786-832. Rome, 8. and B. Rome. "The Leviathan Technique for Large- Group Analysis," Behavior Science, 6 (1961), 148-152. Rosenkrantz, P. S. and W. Crockett. "Some Factors Influ- encing the Assimilation of Disparate Information in Impression Formation," Journal of Personality and ’ Social Psychology, 9 (1965), 283-294. Saltiel, J. and J. Woelfel. "Accumulated Information and Attitude Change," Human Communication Research, 4 (1975), 333-334. Schachter, 8. "Deviation, Rejection, and Communication," Journal of Abnormal and Social Psychology, 46 (1951), 190-207. Schreiman, D. B. and J. D. Johnson. "A Model of Cognitive Complexity and Network Role," paper contributed to International Communication Association, Chicago, (1975). Shapiro, D. and R. Taguiri. "Sex Differences in Inferring Personality Traits," Journal of Psychology, 47 (1959), 127-136. Shaw, M. E. Grogp Dynamics: The Psychology of Small Group Behavior, New York: McGraw-Hill Book Company, (1971). Sherif, M. and C. Sherif. Groups in Harmony and Tension, New York: Harper, (1953). Simpson, G. EmilssDurkheim, Selections From His Work, New York: Thomas Y. Crowell Company, University of New York, (1963). Stewart, R. "Effects of Continuous Responding on the Order Effects in Personality Impression Formation," Journal of Personality and Social Psychology, 1 (1965), 161-165. Stotland, E. "Determinants of Attraction to Groups," Journal of Social Psychology, 49_(l959), 71-80. Stotland, E., N. B. Cattrell, and G. Laing. "Group Inter- action and Perceived Similarity of Members," Journal of Abnormal and Social Psychology, 61 (1960), 335-340. 91 Tagiuri, R. and L. Petrulla. Person Percsption and Inter- personal Behavior, Stanford: Stanford UniVersity Press, (1958?? Tagiuri, R. "Person Perception," The Handbook of Social Psychology, ed. G. Lindzey and E. Aronson, Reading, Mass.: Addison-Wesley, V01. 3’(l969), 395-449. Tajfel, H. and S. D. Cawasjee. "Value and the Accentuation of Judged Differences: A Confirmation," Journal of Abnormal and Social Psychology, 59 (1959), 436-439. Thompson, J. D. "Authority and Power in 'Identical' Organizations," American Journal of SociologY: 62 (1956), 290-301. Triandis, H. C. and M. Fishbein. "Cognitive Interaction in Person Perception," Journal of Obnormal and Social Psychology, 67 (1963), 446-453. Turner, R. H. "Role Taking, Role Standpoint, and Reference- Group Behavior," American Journal of Sociolqup 61 (1956), 316-328. Walters, H. A. and D. N. Jackson. "Group and Individual Regularities in Trait Inference: A Multidimensional Scaling Analysis," Multidimensional Behavioral Research, 1 (1966), 145-163. Warr, D. B. and C. Knapper. The Perception of People and Events, New York: Wiley, (1968). Weick, K. E. The Social Psychology of Organizing, Reading, Mass.: Addison-Wesley, (1969). Wiggins, N., P. Hoffman, and T. Taber. "Types of Judges and Cue Utilization in Judgments of Intelligence," Journal of Personality and Social Psychology, 12 (1969), 52-59} Woelfel, J. Sociology and Science, (forthcoming) (1972). Zazonc, R. "The Process of Cognitive Turning in Communica- tion," Journal of Abnormal and Social Psychology, 61 (1960), 159-167. "7'11? 11111119111111) is