maman 0F MM FLQW AS RELATED TO Wm‘ms mmmmn SAWACW 1 m ms SUM/WWW ma: _ _- mmmmam— WSWEW. i:‘ .~ ' 2971 {1‘1 . ~ - o . o .-o .-.q_¢“ .l‘ . PM”) I. '-‘_' "— ‘ (a; r, y I“ 9 "‘ (C13): ‘ In} ' 274 1 . v./h ”‘0 « {xv} 53 A“ FEB 2 11359 ABSTRACT DIRECTION OF MESSAGE FLOW AS RELATED TO SUBORDINATE'S COMMUNICATION SATISFACTION IN THE SUPERVISOR/SUBORDINATE DYAD BY Jerry Buley It is not necessary to emphasize the critical nature of the relationship between the supervisor and the sub- ordinate in organizations. A failure at this important juncture marks a failure in the basic structure of the organization. Previous studies have shown that communication varia- bles such as frequency of contact and initiation of con- tact between subordinate and supervisor are related to the subordinate's satisfaction with his job and with his relationship with his supervisor. A theory which deals with frequency of contact and continuance of behavior in a relationship is social ex- change theory. The chief exponent of this theory is George C. Homans. The purpose of the present study was to seek evidence to support an extension of social ex- change theory to include a new interaction variable called direction of message flow in the supervisor/subordinate dyad. The hypotheses were tested with data derived from a questionnaire completed by 193 clerical employees on all Jerry Buley levels of a department of the local state government. Seventy-five percent of the employees were in non-super- visory positions. The questionnaire was administered on the job at the respondent's desk by the researcher. Four sets of three hypotheses were develOped. Of the twelve hypotheses, three were significant with the shape of the relationship between the variables conform- ing to that predicted by the hypotheses. The subordinate's percentage of solicited informa— tion-givingin the supervisor/subordinate dyad was not related to the subordinate's desired amount of change in percentage or frequency of solicited information-giving, nor was it related to the subordinate's evaluation of his supervisor's communication in the dyad. Subordinates with a moderate percentage of unso- licited information-giving desire less change in that per— centage and evaluate their supervisor's communication more highly than do subordinates with more extreme per- centages. Subordinate's perceived percentage of unso- licited information-giving was not related to his desired amount of change in frequency of unsolicited information— giving. Subordinates who perceived they had a moderate per— centage of total information—giving in the dyad desired less change in that percentage. Subordinate's perceived percentage of total informa- tion-giving was not related to subordinate's desired — — amount nor to communi The dvad st and pax desired ing: 2 quency nation The 0f soc: flow 11 the 31.1} of the ViSOr': ially . Pu. in 0th. look a ShOU1d \ lIn th eqUal the Jerry Buley amount of change in frequency of information-giving, nor to the subordinate's evaluation of his supervisor's communication in the dyad. There was no significant difference between the two dyad styles operationalized in the study (traditional and participation)1 on the three variables: 1) subordinate's desired amount of change in percentage of information-giv- ing; 2) subordinate's desired amount of change in fre- quency of information-giving; and 3) subordinate's eval- uation of the supervisor's communication in the dyad. The findings provide partial support for the extension of social exchange theory to include direction of message flow in the supervisor/subordinate dyad as a predictor of the subordinate's desire to change communication aspects of the relationship and of his evaluation of his super- visor's communication in that relationship. This is espec- ially true for unsolicited information-giving. Further research should operationalize the concept in other ways to further test the relationships, should look at supervisor's satisfaction with communication, and should look at direction of message flow on the group level. The present study indicates that the concept is 1In the traditional dyad, the supervisor has more than or equal to 65% of the message transmission in the dyad. In the participation dyad, the subordinate has greater than or equal to 65% of the solicited information-giving and the supervisor has greater than or equal to 65% of the unso- licited information-giving. Jerry Buley meaningful, is related to important aspects of the super- visor/subordinate relationship, and can be operationalized to provide data which are easily analyzed. DIRECTION OF MESSAGE FLOW AS RELATED TO SUBORDINATE'S COMMUNICATION SATISFACTION IN THE SUPERVISOR/SUBORDINATE DYAD BY Jerry Buley A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF ARTS Department of Communication 1971 ACKNOWLEDGMENTS Early scholars may have been able to say their work was a product of their own endeavors. It is doubtful whether any modern scholar could honestly make that claim. Certainly I cannot do so. However, it is difficult to acknowledge all of the people who have helped me reach this point in my intel— lectual growth. Many have cared and helped in so many different ways. Becoming a scientist is somewhat like designing and building a house. Many people perform many different functions to reach the final product. My wife Jeanie has been my foundation, my reason for being. Dr. Donald K. Darnell has provided direction. Drs. Michael Moore, R. Vincent Farace, and Eugene Jacobson provided struc- ture. Drs. Daniel Wackman and Donald Cushman have helped work out many of the problems which have come along. And, finally, Dr. David K. Berlo drove home the nails of rigor. All of this was made habitable by gentle jibes and the intellectual support from my fellow graduate students. ii TABLE OF CONTENTS LIST OF TABLES . . . . . . . . . . LIST OF FIGURES . . . . . . . . . I. INTRODUCTION . . . . . . . Message Flow in General Message Flow in Dyads. . II. Bidirectionality of Message Flow . . Conclusion . . . . . . . . . . . . . THEORETICAL PERSPECTIVE AND HYPOTHESES Theoretical Perspective . . . . . . Levels of Interaction . . . . . . . Evaluation of Supervisor's Communication . . . . . . . . . . Message Transmission--Two Variables . . . . . . . . . . . . Hypotheses . . . . . . . . . . . . . Solicited Information—Giving . . Unsolicited Information-Giving . Total Information-Giving . . . . Comparison of Traditional and Participation Supervisor/ Subordinate Dyads . . . . . . III.RESEARCH METHODS AND PROCEDURES . . . The Questionnaire . . . . . . . . Description of the Sample . . . . . Preparation for Questionnaire Administration . . . . . . . . . Administration . . . . . . . . . . Statistical Tests of Hypotheses . . iii 26 28 33 35 37 39 41 43 43 49 50 52 54 IV. RESULTS 0 I O O O O O O O O O O O O Solicited Information-Giving . . Unsolicited Information—Giving . Total Information—Giving . . . . Dyad Style: Traditional Versus Participation . . . . . . . . V. SUMMARY AND CONCLUSIONS . . . . . . Summary of Findings . . . . . . . Implications for Future Research LIST OF REFERENCES . . . . . . . . . . APPENDICES A. Pretest Questionnaire B. Letter to Supervisors C. Letter to Employees D. Final Questionnaire iv Table LIST OF TABLES Page % Solicited Information-Giving by Desired Amount of Change of % Solicited Information-Giving . . . . . . . . . . . 62 % Solicited Information-Giving by Desired Amount of Change in Frequency of Solicited Information—Giving . . . . . . 62 % Solicited Information—Giving by Sub— ordinate's Evaluation of Supervisor's Communication in the Dyad . . . . . . . 63 % Unsolicited Information—Giving by Desired Amount of Change of % Unsolicited Information—Giving . . . . . 64 % Unsolicited Information-Giving by Desired Amount of Change in Frequency of Unsolicited Information—Giving . . . 66 % Unsolicited Information—Giving by Subordinate's Evaluation of Super- visor's Communication in the Dyad . . . 67 Amount of Total Information-Giving by Desired Change in Total Information- GiVj-ng O O O O O O O O O O O O O O O O O 71 Amount of Total Information—Giving by Desired Change in Frequency of Total Information-Giving . . . . . . . . 72 Amount of Total Information-Giving by Subordinate's Evaluation of Super— visor's Communication in the Dyad . . . 73 Table 10 ll 12 13 Type of Supervisory Relationship by Mean Amount of Change Desired in Total Message Giving Type of Supervisory Relationship by Mean Desired Amount of Change in Frequency of Interaction Type of Supervisory Relationship by Mean of Subordinate's Evaluation of Supervisor's Communication in the Dyad . . . . . . Significant and Non-Significant Results from the Study . vi Page 74 75 76 80 LIST OF FIGURES Figure Page 1 Rate of Interaction and Enjoyment with Relationship by Authority of Either Member . . . . . . . . . . . . . 6 2 Subordinates' Evaluation of Supervisors' Interest in Subordinates' Personal Problems by Likelihood of Either Member of the Dyad to Initiate Communication . . . . . . . . . . . . . 8 3 Graphic Representation of Frequency of Contact . . . . . . . . . . . . . . . 23 4 Graphic Representation of Percentage of Total Message Flow . . . . . . . . . 24 5 Percentage of Unsolicited Information- Giving by Desired Amount of Change in Percentage Unsolicited Information- Giving . . . . . . . . . . . . . . . . . 65 6 Percentage of Unsolicited Information- Giving by Mean Evaluation of Super- visor's Communication in the Dyad . . . 68 vii CHAPTER ONE INTRODUCTION The purpose of the present chapter is to present a back- ground for the study of the directionality of message flow in the supervisor/subordinate dyad as it is related to the subordinate's desire to change aspects of the communica— tion occurring between them and to the subordinate's eval- uation of his supervisor's communication in their dyad. There are two sections to the chapter. The first provides a very brief overview of the early conceptualizations of message flow. The second section describes the research on message flow in dyads with emphasis on those studies which have been concerned with message flow in the super- visor/subordinate dyad. Message Flow in General Some of the earliest research on message flow was con- cerned with rumor. (Caplow, 1946; Allport and Postman, 1947; Festinger, et al., 1948; and Back, et al., 1950). These and other studies are summarized in Guetzkow (1965). Caplow defined rumor as "an item of information with definite interest connotations transmitted only by informal person-to—person communication within a group." p. 491 (emphasis in original). Caplow and other rumor researchers were interested in the "rapidity of diffusion," number of rumors within 1 2 a given time period, the components of a rumor message, the saturation of the message within a given system, the "veracity" of rumors, and the relationship between rumor channels and formal channels. Caplow noted that there was a bidirectionality in the rumor channels. "These tended to be two-way chan- nels since the communication of rumors is more often than not marked by an exchange. The customary quid pro quo for a rumor is either another rumor or a validity judg— ment upon the one received." (p. 493) An analysis of the rumor research to date led Guetzkow (1965) to say that "these studies seem to re— veal no characteristics which distinguish rumors in any fundamental way from other communications." (p. 562) Next came an emphasis on the flow of messages in small groups contrived in the laboratory. (Bavelas, 1950; Leavitt, 1951; Guetzkow and Simon, 1955). The researchers in the small group area were inter- ested in such message flow concepts as number of mes- sages, duration of communication, and direction of message flow. However, the bulk of the studies deal- ing with communication in small groups was concerned with structure, here defined as a person's unique posi- tion in an ongoing group of people among whom messages are transmitted, where position is defined as the set of people with whom the person communicates. Structure is inextricably tied up with message flow. For example, if a person is at the "center" of 3 a communication network, his potential for transmitting and receiving messages is much higher than persons who are in other positions. Thus while the small group researchers have used the variable "centrality" and found it to be related to satisfaction with job in the group, it may be that the relationship may be the same as that found between frequency of communication and satisfaction with job. Coincidental with the study of message flow in small groups came the study of the diffusion of new ideas, new products and news stories. (Katz and Lazarsfeld, 1955; Deutschman, 1960; Rogers, 1962; and Coleman, Katz, and Menzel, 1966). The diffusion studies were concerned with the pro- cess by which new ideas, products and news stories were learned and adopted by a particular population. Char- acteristics of early learners and adopters, the stages through which a person goes prior to actual adoption, and the roles of the media and the change agent were common areas of research in the diffusion area. Rogers (1962) has summarized more than a decade of the diffu- sion studies. The resources of science have not been marshalled behind the study of the flow of messages in organizations as they have been at one time or another behind the afore- mentioned areas, although it has been of much pragmatic interest to those in the management science and business 4 areas. (Halpin and Winer, 1952; Davis, K., 1953, a&b, and 1968; Turner, 1957; and Rubenstein, 1963 and 1966). Davis (1953, a&b) was interested in the probability of a message being relayed, the proportion of members learning the message given that it was relayed, and the amount of distortion or elimination of information in the message. Rubenstein was interested in the flow of ideas in research and in industrial settings. He was primarily interested in the role of the individual and personality differences in determining whether ideas would be transmitted and accepted. Neither Davis nor Rubenstein related their variables to individuals' de- sire to change their relationship with or their evalua- tion of their superior. Other researchers have asked organization members to compile daily diaries about their communication in the organization to get some indication of the message flow in the organization. (Turk and Wills, 1964; Farace, et al., 1970; and Lawler, 1968). Message Flow in Dyads Frequency of communication has repeatedly been shown to be related to satisfaction with the relationship and with evaluation of performance. Turner (1957) found a positive association between interaction rate and liking of the supervisor. Berlo et a1. (1970) found that the more frequently communication occurs between supervisor 5 and subordinate: l) the more likely either one of them initiates communication in the dyad; 2) the more likely it is that the subordinate will go to the supervisor or the supervisor's boss for work information rather than going elsewhere; and 3) the less likely the subordinate is to say his supervisor is not interested in the sub- ordinate's personal problems. Turk and Wills (1964), in a longitudinal study of student nurse and student physician dyad teams during a period of change in a university hospital, found that rate of interaction is inversely associated with an in- crease in the authority of one member (the student phy— sician) over the other, is related to their expectations of the differences in authority between the two, and is positively associated with enjoyment of the relationship. Authority is defined as one person's direction of another person's activities. The more of person A's activities which can be directed by person B, the more authority person B has over person A. Turk and Wills studied only the relationship of the authority of one person to rate of interaction and to en— joyment with the relationship. Consequently they studied only one half of the possible relationship between their variables. Figure 1a shows this truncation graphically. Figure lb shows the full relationship given that their findings would be mirrored on the other side. H wusmflm m cemumm 4 common m conumm Hm>o 4 sownmm um>o m common Hm>o m GOmumm mo mufluocusd mo hufluosusd mo unflnonuzfi swam o swam swam 30g \gav Amy mflnmcowu ImHmH nuw3 ucmENOncm .HO A: COHHUMHTHGH mo mumm Thu relatic rate of Tu: tionsh. norms 11y re ated i the WC not t‘: VerSE< Their Can 1 mutua T tion n—d n '1‘ on r. t‘ A askeq tien .7 Thus in any given dyad there would be a curvilinear relationship between authority and the two variables: rate of interaction and enjoyment with the relationship. Turk and Wills found that the actual authority rela- tionship was a result of the member's particularistic norms (those generated in the dyad) and was not necessar- ily related to their universalistic norms (those gener- ated in their peer group). It must be remembered that the working relationship in this study was the dyad, not the group. I would expect this outcome to be re- versed if the working relationship were the group. Their study, they claim, indicated that equality of power can lead to increased interaction, which in turn promotes mutual attractiveness. There is some indication that initiating communica- tion is more satisfying than receiving it. Whyte (1968) draws a strong generalization concerning initiation. Whenever we see a high frequency of initiation down the line and little or no initiation upward, we always find workers expressing dissatisfaction with their superiors and generally with the work situation also. If we find, on the contrary, the subordinates are initiating upward for a significant portion of the time, then we tend to find quite different sentiments expressed toward their super- iors. (p. 56) Lawler (1968), using a diary self-report technique, asked his respondents to rate their satisfaction with the communication which they had initiated and the communica- tion initiated by others with whom they had contact. He 8 found that his respondents were more favorable to their own initiations than to those of their contacts. Berlo et a1. (1970) in a study of lower level employ- ees in a large financial organization found that if the subordinate perceived that each member of the supervisor/ subordinate dyad was about as likely to initiate communi- cation, the subordinate was more likely to say that his supervisor was just as, or more, interested in the sub- ordinate's personal problems as he is in the subordinate's work problems. The same was not true if either member was more likely to initiate than the other. This finding would indicate that there is a curvilinear relationship between initiation and subordinate's perception of his supervisor's interest in the subordinate's personal prob- lems. (Figure 2) High Subordinates' Evaluation of Supervisors' Interest in Subordinates' Personal Problems Low Subordinate Equally Supervisor is more likely likely is more likely to initiate to to initiate initiate Figure 2 9 Another finding from this study was that subordinates who perceived that the supervisor was interested in their personal problems were less likely to have ideas or sugges- tions about how to change things but were more likely to communicate them if they had one. This would seem to indi- cate that a subordinate's desire to change aspects of his job and aspects of his relationship with his supervisor is related to his evaluation of his supervisor's present relation to the subordinate and possibly to the direction of message flow in the dyad. Berkowitz and Bennis (1961) made a comprehensive study across several hierarchical levels in seven outpatient departments in hospitals in Boston. They found that initi- ation is inversely related to rank of the other persons. This compares to the finding by Turk and Wills that rate of initiation is related to authority of the other per- son. (Figure l). Typically, initiation is from super- visor to subordinate, with communication with peers and supervisors generally more satisfying than communication with subordinates. Bidirectionality of Message Flow Blau and Scott (1962) report a study which focused onfifl one aspect of initiation, that of soliciting information. They found that in some dyads either member was as likely to ask the other for information. In other dyads, one member was more likely than the other to ask the other for i fied . ccnta perso city and c sulte frequ empha "this It fa mutua ments MaCDo tion that basis HOWQV messa membt tiOn they 10 for information. These authors found that when they classi- fied groups by the number of reciprocal relationships they contained, there were significant differences between them.1 In low reciprocity groups self-confident and competent persons were more likely to be consulted. In high recipro- city groups the reverse was true. That is, self-confidence and competence were inversely associated with being con— sulted by colleagues. Reciprocal dyads also had a higher frequency of interaction. The concept of reciprocity has received considerable emphasis in the literature. Sherif (1967) has said that "this traffic among men is not haphazard or fortuitous. It falls into a pattern of reciprocities consisting of mutual expectations, role relations, and power arrange— ments." (pp. 19—20). Weiss (1958), Schwartz (1968) and MacDonald (1970) have defined reciprocity as the asser- tion by each of two individuals on separate questionnaires that they have contact with one another on some regular basis. This is the primitive definition of reciprocity; i.e., verification by two people that they have contact. However, knowledge of contact alone tells us only that messages do or can flow between two people. lReciprocal relationships were those in which either member was as likely as the other to solicit informa- tion. If one member was more likely than the other, they were called unilateral relationships. 11 James Davis (1968) defines reciprocity somewhat differ- ently. Given two different sociometric questions, it is the degree to which, if person A chooses person B on one question, person B will choose person A on the other. "We shall call it reciprocity and note that it could be of some interest to research workers interested in 'ex— change' in interpersonal relations. If, for example, it is the case that respect is exchanged for advice, then if one asks persons to name those whom they respect and those to whom they give advice, the two items should show a positive reciprocity coefficient." Davis uses the word symmetry synonymously with the way MacDonald (1970) and others have used reciprocity; i.e., if A lists B, then B lists A for the same sociometric question. Davis' reciprocity notion adds a message flow dimen- sion to contact between individuals similar to that of Blau and Scott. Mayhew, Gray and Richardson (1969) speak of symmetri— cal relationships as those in which power or influence in the relationship moves bidirectionally, as Opposed to unidirectionally as in the assymetrical relationship. Simi— larly, Foa (1958) uses the term balanced dyad which he de- fines as one in which both members receive equal amounts of reward from the interaction. Foa does not, however, mean to imply an equality in the proportion of information flow. Watzlawick, Beavin, and Jackson (1968) have defined symmetrical relationships much as Foa has defined balanced _ __ __ ___ ___—___. — I— _. dyads by e: othe: the' Conc able the ing ing twee COR: tio; 12 dyads. "Symmetrical interaction, then, is characterized by equality and the minimization of differences." On the other hand, the complementary relationship "is based on the maximization of differences." (pp. 68-69) Conclusion This chapter has presented research related to the vari— ables and relationships in the present study. A summary of the major areas of research in and out of organizations deal- ing with message flow was presented. Then, research focus— ing on message flow in the dyad and more specifically be- tween supervisor and subordinate was covered. Finally, conceptualizations and research findings on the bidirec- tional flow of messages in dyads was presented. | __- < — _— T} under: ordine visor' a frar superx ordina uatiox OI bles j write} books, Change this t Change evaIUa from t the dy IQSEEE HO: Variab. CHAPTER TWO THEORETICAL PERSPECTIVE AND HYPOTHESES The goals of the present study are to further our understanding of communication in the supervisor/sub- ordinate dyad and to find ways to improve the super- visor's communication in it. This chapter will provide a framework for the study of communication behavior in supervisor/subordinate dyads as a predictor of the sub- ordinate's desire for the dyad to change and of his eval- uation of the supervisor's communication in the dyad. One theory which has dealt with these and similar varia— bles is social exchange theory. George C. Homans is the writer most commonly associated with this theory. His two books, The Human Group (1950) and Social Behiavior: Its Elementary Forms, (1961) form the backbone of social ex- change theory. The present study will attempt to extend this theory in order to predict a subordinate's desire for change in the supervisor/subordinate relationship and his evaluation of his supervisor's communication in the dyad, from the subordinate's percentage of the message flow in the dyad. Theoretical Perspective Homans analyzes human behavior through the use of two classes of constructs which he calls descriptive terms and variables. The former describe kinds of behavior (i.e., l3 14 activity, sentiment and interaction), and the latter are the quantification of these. Homans defines activity as any voluntary behavior. "The activities that the members of a particular verbal or sym- bolic community say are signs of the attitudes, and feelings a man takes toward another man or other men--these we call sentiments.” (Italics in original; 1961, p. 33). Senti- ments also, thus, are not internal states but "signs of" internal states; and, consequently are also activities. When two persons exchange activities or sentiments, or as Homans puts it, "when an activity (or sentiment) emitted by one man is rewarded (or punished) by an activity emitted by another man, regardless of the kinds of activity each emits, we say that the two have interacted." (Italics in original; 1961, p. 35) Quantity, according to Homans, is a frequency variable. It is "the number of units of activity that the organism in question emits within the given period of time." (Homans, 1961, p. 36). An example of such a quantifica— tion would be the number of times a person has contact with another person over a period of time. Homans states that value has two components, one of which is relatively constant over a period of time, and another which is variable according to the present state of the individual. The first is a rank order of prefer— ences regardless of his state. The second component is dependent upon the degree of satiation or deprivation of 15 the individual. (1961, p. 48-50) Thus, the more deprived a person is of a certain activity and the higher it is on his list of preferred activities, the higher he will value that activity. Homans proposes the following relationships among these constructs and variables. He derived these propositions from "laws" of animal behavior. PROPOSITION ONE: If in the past the occurrence of a particular stimulus situation has been the occa— sion on which a man's activity has been rewarded, then the more similar the present stimulus-situa- tion is to the past one, the more likely he is to emit the activity, or some similar activity, now. (1961, p. 53) PROPOSITION TWO: The more often within a given period of time a man's activity rewards the activ- ity of another, the more often the other will emit the activity. (1961, p. 53) PROPOSITION THREE: The more valuable to a man a unit of activity another gives him, the more often he will emit activity rewarded by the activity of the other. (1961, p. 55) PROPOSITION FOUR: The more often a man has in the recent past received a rewarding activity from ano- ther, the less valuable any further unit of that activity becomes to him. (1961, p. 55) Corollaries l: The frequency of interaction between Person and Other depends upon the frequency with which each rewards the activity of the other and on the value to each of the activity he receives. (1961, p. 55) 2: The more cost Person incurs in emitting an activity, the less often he will emit it. . . . For an activity to incur cost, an alternative and reward— ing activity must be foregone. (1961, p. 59) l6 3: The more often Person has emitted a costly activity, the more costly he finds any further unit of that activity. (1961, p. 60) The Rule of Distributive Justice A man in an exchange relation with another will expect that the rewards of each man be proportional to his costs--the greater the rewards, the greater the costs-- and that net rewards, or profits, of each man be pro- portional to his investments--the greater the invest- ments the greater the profit. (1961, p. 75) When the expected proportionality does not occur, the law of distributive justice is said to have been violated. This leads to Homans' fifth and last proposition. PROPOSITION FIVE: The more to a man's disadvantage the rule of distributive justice fails of realization, the more likely he is to display the emotional behavior we call anger. (1961, p. 75) Homans states that "distributive justice may, of course, fail in the other direction, to the man's advantage rather than to his disadvantage, and then he may feel guilty rather than angry; he has done better for himself than he ought to have done." Frequency of contact is one of Homans' major variables. This is the only quantification of interaction treated in the social exchange theory. Yet frequency of interaction is a gross measure. It is possible to differentiate be- tween several levels of interaction which could provide much finer measures of interaction than frequency. l7 Levels of Interaction The first level is the most primitive. Either there is contact between two people or there is not. Variables assoc— iated with the existence of contact are propinquity, rules, and perceived similarity. An example of the propositions in— corporating these variables is, the closer two people are in distance, the more likely they are to have contact. Propositions incorporating contact as an independent variable are usually stated in the form, "given contact (or interaction, or communication), then . . ." The second level of interaction in dyads looks at con- tact in terms of the duration and/or frequency of occur- rence of contact. These variables come as a response to the question, given that two peOple have contact, how frequently do they interact, and how long do they interact when they do. PrOpositions are usually stated as, "the more frequently two persons interact . . ." It is inter- esting to note that interaction at the first and second levels need not be symbolic. John and Joe may meet every day at the same point on the road, walk down to the lake, fish for two hours, walk back to the road, part, and never interact symbolically the entire time. Most of Homans' prOpositions, from The Human Group, (e.g., the more frequently two peOple interact the greater the amount of liking they will have for one another), are second level prOpositions and apply no less to the inter- action between John and Joe than they do to two people who may spend the same two hours discussing world problems. . _ __— _— _— _— .— __- _— _-——— ._._— — .— on t? and g ever persc actic durix orig; tern exis‘ Of a in t] PEri the CO: thir IEVQ Sage Pets Self 18 Homans defines two additional concepts for application on this second level of interaction. They are initiation and origination of interaction. The former occurs when- ever a person emits an action in the presence of another person. Origination, then, is the first initiation of inter- action by one of the members of a dyad after any period during which no interaction occurred in the dyad. The origination concept is the beginning of the idea of pat- tern in the interaction itself (i.e., other than its existence or nonexistence). It is possible for one member of a dyad to be much more likely to originate interaction in the dyad. It is also possible for any origination to be the product of both members' initiation; e.g., when John and Joe met in the road. The third level of interaction incorporates the direc- tign of message flow; that is, who is receiving a message and who is transmitting it in the dyad. Or, over a longer period of interaction, which member is more likely than the other to give the other information. Just as the second level of interaction in a dyad is a combination of the first level and a new concept, the third level is the combination of the first and second levels and another new concept. Thus, proportion of mes- sage flow direction is the third level of interaction. Person A can originate a message flow either toward him— self or away. That is, A can originate a message to B 19 for which B has not asked. Or, A can originate an inter- action requesting information from B. The two patterns of message flow in a dyad will be called solicited and unsolicited information giving. The fourth level of interaction in dyads breaks it into content categories. It represents an answer to the question, given that A and B have contact over time, what are the topics of the interaction. These categories may be broad and inclusive; e.g., Berlo's (1969) "environ- mental," "motivatiOnal," and "instructional" information categories. Or, they may be specific; e.g., the Dewey Decimal System or Library of Congress methods for cate- gorizing library materials. The fifth level of interaction consists of encoding behaviors; e.g., language usage, intonation patterns, channel choice, sequencing and punctuation of messages, etc. Each action-~large or small--of an individual is analyzed in relation to the actions of the other indiv— idual in the dyad. Nonverbal communication (hand motions, gestures, and facial expressions, as well as other body movements) are also included on the fifth level of inter- action. The fifth level of interaction incorporates all previous levels of interaction; and, therefore, it pro- duces a richer form of data. The present study will be concerned primarily with the third level of interaction; i.e., the direction of message flow in a dyad. However, it will focus upon the transfer of symbolic information 20 which occurs in the dyad. Thus, message transmission in a dyad will be defined as the encoding of symbolic infor- mation by one person in the presence of another person who has the potential for and requisite skills to decode that symbolic information.1 If members of a dyad value the reception of messages differently than they value the transmission of messages in the dyad, then prOpositions can be deduced from the propositions contained in the social exchange theory. There is empirical support for the pr0position that transmitting messages is more rewarding to participants than receiving them. Mulder (1966) divided subjects into four conditions in a small group study dealing with task completion. The four conditions were two levels of "self—realization" and two levels of "power." 81: Receives solutions 82: Receives information; forms own solution Pl: Does not transmit information P2: Transmits information After the completion of the task, Mulder had the sub- jects rate their satisfaction with their behavior in the study with the following results: P1 P2 81 4.89 7.11 82 5.37 6.57 1This definition draws upon that presented by David K. Berlo, 1969. 21 Cell entries represent the means of the satisfaction ratings for subjects in that condition. "In an analysis of variance, only the effect of the power (P) variable is significant (p<.OOl); the S variable has no influence at all; the interaction (PxS) is far from significant." (emphasis in original; p. 265) Mulder interpreted his findings as saying that power was more important in terms of satisfying the individual than self-realization. However, as he has Operationalized his variables, there is even a more basic finding here. That is, transmitting messages is more satisfying than receiving messages. .4 U" ‘4- governmental agency explored what may be called the solic- Blau and Scott (1962) in a study of peer dyads in a ited information-giving mode of message transmission in dyads. In general, people tend to interact more frequently with persons they perceive to be their equals in competence. And, while the most competent members had the higher inter— action rates, they were more likely to talk with more dif- ferent people. These authors found two different rela- tionships among these workers. The first they referred to as a reciprocal relationship. In this relationship either member was as likely to ask the other for information. The second type of relationshipthey called a unilateral re- lationship. In this dyad one member was more likely than the other to ask for information. 22 The reciprocal relationship was more likely to occur among participants who perceived themselves to be equiva- lent or nearly equivalent in competence. The members of a unilateral dyad were more likely to perceive that one of them was less competent than the other. BzBlau and Scott theorize that solicited information- giving is rewarding to the giver but is perceived as a cost by the receiver in unilateral dyads. However, the same is not true in reciprocal dyads. In unilateral dyads the person must defer to the more competent mem- ber, and must grant him higher status; thus, the asker perceives it as a cost in terms of the relationship itself. While the asked person perceives being asked as rewarding up to a point, with increasing frequency, he perceives it more and more as a cost. This occurs for at least two reasons. For one thing, satiation sets in; that is, he no longer finds the asker's activity as rewarding as it had been previously. For another, the more often he has to attend to answering requests for information, the less time he has for his own work. Thus, unilateral dyads tend not to interact frequently. On the other hand, in the reciprocal dyad the members are mutually supportive. Even though they may not get answers as good as they might from the more competent source, at the same time they reduce each other's anxi- ety, and serve as objective sources of information for 23 each other. Being asked for information, even though it is from a person equally as competent as oneself, is still rewarding. Thus, message flow seems to have different values de- pending upon whether it is being received or transmitted, and propositions can be developed which relate message flow to valuation of the relationship, desire to change the relationship, and valuation of the other member's activities. It is proposed that the relevant variable on the third level of interaction is not frequency, but rather, an individual's percentage of the transmission in the dyad. The latter is an index of message transmission by one individual relative to the total amount of message transmission in a dyad, while the former is relative to time. We can represent this difference graphically. For example, we can use a circle to represent the total amount of communication in a dyad. The size of the circle will vary with the frequency and/or duration of communication in the dyad, as in Figure 3. . 9 Q TIMES PER DAY 5 Figure 3 24 However, we can divide the circle representing the total amount of communication in the dyad into segments representing the percentage of that total which can be attributed to message transmission of each member of the dyad, as in Figure 4. 75% ’\.v Figure 4 If message transmission in a dyad is more rewarding than message reception, then the latter might be viewed as a cost, because an alternative activity which is more rewarding (i.e., message transmission) is foregone. Then, by Homans' Corollary Three: PROPOSITION ONE: The higher a person's percentage of the message reception in a dyad, the more costly he finds any further unit of that activity in that dyad. We would assume that the following would also be true. PROPOSITION TWO: The lower a person's percentage of the message transmission in a dyad the more re- warding a unit of that activity is in that dyad. Thus, combining the two, we would predict that there is a negative linear relationship between a person's per- centage of the message transmission in a dyad and the re- ward potential of a unit of that activity in that dyad. 25 'Homans incorporates basic economics into his theory and postulates that reward minus cost equals profit. Thus, when a person has a low percentage of the message trans— mission (low reward) and thus a high percentage of the mes- sage reception (high cost) in a dyad, he will have a high loss. Homans argues that the less a man's profit from a par— ticular activity, the more likely he is to change and emit some other activity. There are two ways in which a person can change a relationship. He can change the frequency of interaction and/or he can change his percentage of the message transmission in the dyad. Also, a person with a low percentage of the message transmission in a dyad should perceive that he is receiv- ing less than his "fair share" of the rewards in the re— lationship and thus should desire for his percentage to be closer to that of the other member. Persons having a larger percentage of the message transmission in a dyad, according to prOposition two developed in the present paper, will find message trans- mission less rewarding than those who have a smaller per- centage. In general, the less rewarding an activity is for a person, the less he will tend to emit that activity. Again, if Homans' proposition concerned with the law of distributive justice is true, then persons having more than their "fair share" of the rewards derived from trans- mitting messages in the dyad should feel guilty. This 26 guilt may in turn cause them to want the relationship to change in such a way as to increase the other member's proportion of the message transmission. Persons with a moderate percentage of the message transmission in the dyad are less likely to experience a loss in the relationship. They will view message trans— mission as being more valuable than will those having a higher percentage of the message transmission in their dyads, and they will have a lower message reception cost than will those who have a higher message reception per— centage in their dyads. Thus, those with a moderate percentage of the message transmission in their dyad should be less likely to want the relationship to be changed. That is, they will be less likely to desire for a change in the frequency of interaction or their percentage of message transmission in the dyad than will those having a higher or lower per- centage of the message transmission in their dyads. Evaluation of the Supervisor's Communication by the SfibordInate Research from two different areas has revealed a rela- tionship between direction of message flow and perceived competence. The study reported by Blau and Scott (1962), found that unilateral (in terms of solicited information) dyads were more likely to occur between members who per- ceived that they differed in their levels of competence. 27 The direction of solicited information giving was from the more competent to the less competent. Several researchers have studied this relationship in small groups. They have found that those who are more volu- ble (i.e., more talkative) are more likely to be selected as leaders by members of the group (Gray, Richardson, and Mayhew, 1968; Kirscht, 1959; Riecken, 1948; Shaw, 1959; and Smelser, 1961), are perceived to be more competent and are perceived to have better information and to make better suggestions. (Shaw and Penrod, 1962; and Jaffee and Lucas, 1969). These data lead us to believe that subordinates with a low percentage of the message transmission in the super- visor/subordinate dyad will evaluate their supervisor's communication in the dyad positively and those with higher percentages will tend to evaluate his communication less positively.l However, if a person who has a low percentage of mes- sage transmission in the dyad is incurring a high cost in the relationship, it seems unlikely that he would evaluate his supervisor's communication in the dyad positively. Thus, those with a higher percentage of message flow will evaluate their supervisor's communication less posi- tively. Evaluation of supervisor's communication should 1Compare this to Jain (1971) who found a positive rela- tionship between frequency of interaction and evaluation of supervisor's communication. 28 then increase with a decrease in percentage of message flow until the subordinate's percentage of message flow is low enough that it is viewed as a cost. At that point, further decrease in subordinate's percentage of message flow should lead to decrease in the subordinate's evaluation of his supervisor's communication. Thus, rather than a simple linear relationship between percentage of message transmission in the dyad we would pre- dict a curvilinear relationship. Again, those with a moder- ate percentage of the message transmission in their dyad will be more likely than those with higher or lower per- centages to evaluate their supervisor's communication in the dyad positively. Message Transmission--Two Variables As mentioned earlier, message flow has two dimensions. It is either solicited or not solicited. This increases the complexity of the relationships which have been pre— sented thus far, but not unnecessarily. We have stated that those with moderate percentages of message trans- mission are less likely to desire change in either fre— quency of interaction or percentage of message transmission and are more likely to evaluate their supervisor's communi- cation positively. A person may have a moderate percentage of the message transmission in the dyad as a result of be- ing just as likely as his partner to give solicited infor- mation and just as likely to give unsolicited information 29 in the dyad. However, he can also have a moderate per- centage of message transmission by having a high percent- age of one type of message transmission and a low per- centage of the other type of message transmission. That is, the two counteract each other.1 This "counteracting" possibility has implications for the supervisor/subordinate relationship. Traditionally in organizations messages flow from the top down the chain of command to the bottom. At any given interface between supervisor and subordinate then, messages are more likely 1.7. to flow from supervisor to the subordinate; i.e., the supervisor will have a higher percentage of the message transmission in the dyad. (Berkowitz and Bennis, 1961) If, as has been suggested, a high percentage of the message transmission in one mode can counteract a low percentage of message transmission in the other mode, then supervisors with a high percentage of the unsolic- ited message transmission in the dyad can counteract that by soliciting more information from his subordinate; i.e., increase the subordinate's percentage of solicited message transmission. It would seem that solicited information giving should be more rewarding at any given percentage of 1This is true given that the frequencies of the occur- rence of each type of message transmission are roughly equal. However, this is a weighting problem which will be discussed later in another context. 30 solicited information giving than is unsolicited informa- tion giving at that given percentage of unsolicited infor— mation giving in a dyad. Both should be rewarding, be- cause they are both forms of information giving. However, solicited information giving has a prerequisite that the other person asks for the information. Asking for information is an acknowledgement that the other person knows more than the asker. It is rewarding to know that other peOple believe we know more than they do. Thus solicited information should potentially be more rewarding than is unsolicited information. Homans felt that his theory of social behavior was limited to the explanation of non-normative non-rule governed behavior. He says that his theory cannot explain "why the more or less explicit rules of a society or some organization within it are what they are." More specific- ally, he says, I shall not be concerned with the behavior of a man so far as he holds authority over others by reasons of appointment by a public or private corporation. Thus I shall not be concerned with the behavior of a supervisor so long as he acts as the representa- tive of a firm, employed to get his subordinates to carry out rules. I shall on the other hand, be much interested in the behavior of a man who through his own exertions as an individual in his face-to- face dealings with other individuals, acquires in- fluence over them not de 'ure but de facto. But it is not impossible, though it may be d1fficult, for a supervisor to exercise both sorts of authority, and then, if I am to explain his actual behavior, I shall certainly have to take account of the power he exercises by virtue of his official position. Others following Homans have disagreed as to the applic- ability of his theory to the supervisor/subordinate dyad. 31 Abrahamson (1968) differentiates between institutional and subinstitutional behavior as does Homans. The latter, he says, is interaction which is not predominantly the result of clearly defined roles and role expectations. As such, each party is largely able to determine his own way of acting; he is not occupying a role which carries with it clear—cut behavioral expectations. (p. 4) 0n the other hand, Caplow (1964) argues that the ob- servation that there are informal and formal statuses "obscures the more important fact that the correlation be- tween prescribed and observed status orders is very close. Accidents of personality and circumstances introduce status differences among people in nominally equal positions or reduce status differences between adjacent unequal positions, but major inequalities are seldom removed--let alone re- versed--in this way." (p. 102) Another argument in support of the applicability of the theory to the supervisor/subordinate dyad is that the relationships between the variables which Homans has pre- dicted with his theory have been supported by studies in organizations and more Specifically by studies of the supervisor/subordinate relationship. (Turk and Wills, 1964; Berlo, 1969; Jain, 1971.) It can also be argued that with the addition of direction of message flow to the basic theory, the theory becomes even more appropriate for the description of com— munication in the supervisor/subordinate dyad. The 32 participatory models of organizational management developed by McGregor (1957), Likert (1961), and others assume that when a subordinate person is given an opportunity to deter- mine how goals are achieved, he will be more satisfied than if he is told exactly what to do and when to do it. (Vroom and Deci, 1970). This assumption is based upon a philosophy of man which expresses a belief that man must strike a balance between being inner or other directed. Maslow (1943) would suggest that the expanded theory would thus find differential application depending upon the level to which each person's needs have been fulfilled in the organization. Those who have their physiological, safety, and love needs satisfied are more likely to be con— cerned with their self-concept or self-esteem than are those who do not have these more basic needs fulfilled. The self-esteem needs as described by Maslow are, First, the desire for strength, for achievement, for adequacy, for confidence in the face of the world, and for independence and freedom. Second, we have what we may call the desire for reputation or pres- tige (defining it as respect or esteem from other people), recognition, attention, importance or appreciation. (p. 32) If self-esteem is garnered through interaction with others, and direction of message flow is associated with value, then self-esteem would seem to vary with the direction of message flow in a dyad. Litwak (1961) would argue that the extended theory would be more applicable to some jobs than to others. He says that participation in making decisions is crucial 33 where the job requires people "to identify themselves with organizational goals, to cooperate in their social relations, and to communicate." He continues by saying that, "Since these involve social skills, participation in making deci— sions is important where jobs are chiefly defined by those abilities. Put differently, jobs characterized by social skills might be carried out most efficiently under a hori- zontal structure of authority, that permits all individuals to participate in decisions." Litwak further argues that organizations should be designed around a complex model of bureaucracy which permits differentiation of managerial style by job characteristics. Short cycle high redundant jobs and few social skills require less participation. On the other hand, low redundancy in a job which requires social skills requires more participation. Thus, the ex- panded theory would seem to find better application for those supervisor/subordinate dyads which occur in low re- dundancy jobs which require social skills than in those which do not conform to this description. Hypotheses Several sets of hypotheses will be tested in the pre— sent study. One set is concerned with the subordinate's percentage of solicited information giving. Another will be concerned.with the subordinate's percentage of unso— licited information giving. A third set will combine the subordinate's percentage of solicited and his percentage 34 of unsolicited information giving to form a total informa- tion giving index. One last set of hypotheses will com— pare the traditional supervisor/subordinate relationship in which the supervisor has a higher percentage of infor- mation giving in both modes than does the subordinate with the supervisor/subordinate relationship in which the super- visor is more likely to give unsolicited information and the subordinate is more likely to give solicited informa- tion. Each set of hypotheses will have the following depen- dent variables: Subordinate's desire to change the fre- quency of interaction, subordinate's desire to change his percentage of message transmission, and subordinate's evaluation of his supervisor's communication in the dyad. Argyle (1967) and Turk and Wills (1964) would argue that we should expect most people to be satisfied with their present relationship with their supervisor no matter what the direction of message flow in the dyad might be. They would probably state that most people will not wish to change their relationship because it would disturb the present steady state, thus decreasing the predicta- bility of the partner's behavior. Also, since the pre— sent relationship is a result of the particularistic norms of the dyad (that is, the mode of interaction for each participant has been established by the members of 'that dyad) it represents a compromise that each can work with. 35 Ikmever, I believe that these arguments apply best to less flnmally prescribed relationships than those found in Huesupervisor/subordinate dyad. Informal relation- ships are much more flexible and the members can usually readhaasteady state or compromise position which is some- what acceptable to both. The latter may not be as true in Um:supervisor/subordinate dyad which is in part gov- erned by the rules and policies of the organization, in part by universalistic norms for all supervisor/subordinate dyads, and less in part by the particularistic norms devel- Oped by the members of the dyad. Thus, while the counter argument has strength relevant to the relationship being studied, we expect that the follow- ing hypotheses will find more support in the supervisor/ subordinate dyad and less support in the friend or lover dyads, for example. Solicited Information Giving Hla: There is a curvilinear relationship between subordinates' percentage of solicited infor— mation giving in a dyad and the amount of change in that percentage they will desire; such that, those with a moderate percentage will desire less change than those with more extreme percentages. 36 Hlb: There will be a curvilinear relationship between subordinates' percentage of solicited informa- tion giving in a supervisor/subordinate dyad and the desired change in frequency of solicited information giving in the dyad; such that, those with a moderate percentage will desire less change in frequency than will those with more extreme percentages. ch: There will be a curvilinear relationship between subordinates' percentage of solicited informa- tion giving in a dyad and the subordinates' evaluation of their supervisor's communication in the dyad; such that, those with a moderate percentage will evaluate their supervisor's communication more positively than will those with more extreme percentages. Although not contained in the hypotheses, it was sug— gested earlier that solicited information giving should potentially be more rewarding than unsolicited informa- tion. Thus, while I predict the relationships suggested in these hypotheses, I would not be surprised to find that the distribution will be skewed such that subordinates with somewhat less than moderate percentages of solicited information giving will still desire little change in -m-z. u “‘4" (Sn- 37 frequency of interaction or in their percentage of solicited information, and will still evaluate their supervisor more positively than will those with more extreme scores. The reason for suggesting that the distributions will be skewed is that both members of the supervisor/subordinate dyad are acting within a set of role expectations which pre- scribe that the relationship operate such that the super— T; visor is more likely to give the subordinate unsolicited E?“ information and the subordinate is more likely to ask him for information. This ensures that the supervisor is more likely to give solicited information in the dyad. u Given the normative aspect of these expectations, the subordinate may be less likely to desire a change when the relationship conforms to his expectations, and may be more likely to evaluate his supervisor positively when the supervisor conforms to the subordinate's expec- tations. Unsolicited Information Giving H2a: There is a curvilinear relationship between subordinates' percentage of unsolicited infor- mation giving and the amount of change they will desire in that percentage; such that, those with moderate percentages will desire less change than those with more extreme percentages. 38 H2b: There is a curvilinear relationship between subordinates' percentage of unsolicited infor- mation giving and frequency of occurrence of unsolicited information giving they will desire in the dyad; such that, those with moderate percentages will desire less change in fre- quency of unsolicited information giving in the dyad than will those with more extreme percentages. H20: There will be a curvilinear relationship between subordinates' percentage of unsolicited informa- tion giving and subordinates' evaluation of their supervisor's communication in the dyad; such that, those with a moderate percentage will evaluate their supervisor's communication more positively than will those with more ex— treme percentages. Again, although not contained in the hypotheses, it is expected that the relationship between the subordinates' percentage of unsolicited information giving and the de- pendent variables will be somewhat skewed. It is expected that the subordinate who is somewhat less likely than the supervisor to give unsolicited information in the dyad will be nmme likely than subordinates with more extreme scores to desire less change in his percentage of giving unsolicited information; will be more likely to desire 39 for the frequency of unsolicited information giving to remain the same or increase; and will be more likely to evaluate his supervisor's communication in the dyad posi— tively. A comment should be made at this point. As stated near the beginning of the present chapter, there are two components to value. One of these components is fairly . constant over time; the other varies with the individual's “7 degree of deprivation and satiation. : However, the first component will vary from individual 5. to individual for any particular value object. Thus, B value for message transmission will vary across indiv- iduals. This means that members of dyads will find given percentages of message transmission differentially reward- ing. Thus, while overall the theoretic base used in the pre- sent study should predict the general relationships among the variables presented, further specification of indiv- idual differences could increase the precision of the pre- dictions and the results. Total Information Givigg As mentioned earlier, the reason for treating this as a different set of hypotheses is that in some dyads sub- ordinates will not have moderate percentages of message transmission in either mode, but still may have a mod- erate percentage overall. This happens as a consequence of one member having a high percentage in one mode and the 40 other member having a high percentage in the other mode. H3a: H3b: H3c: There is a curvilinear relationship between subordinates' percentage of total information giving and whether they will desire a change in that percentage; such that, subordinates with moderate percentages will desire less change than will subordinates with more ex— treme percentages. There is a curvilinear relationship between subordinates' percentage of total information giving and the amount of change they will de- sire in frequency of interaction in the super- visor/subordinate dyad; such that, those with moderate percentages will desire less change in the frequency of interaction than will those subordinates with more extreme per- centages. There is a curvilinear relationship between subordinates' percentages of total information giving and the subordinates' evaluation of the supervisor's communication in the dyad; such that, subordinates with moderate percentages evaluate the supervisor more positively than. subordinates with more extreme percentages. 1", .' ' ”J": 1.- 41 Comparison of Traditional and Participation Supervisor/ Sfibordinate Dyads In the traditional supervisor/subordinate dyad, the supanfisor has a higher percentage of message transmission in.both modes than the subordinate. In the participation supervisor/subordinate dyad, the supervisor has a higher percentage of unsolicited information giving and the sub- ordinate has a higher percentage of solicited information giving. That is, the supervisor in the latter dyad is equal or more likely to ask the subordinate for informa- tion than the subordinate is to ask the supervisor for information. H4a: Subordinates in participation supervisor/ subordinate dyads will desire less change in total message giving than will those in tra- ditional dyads. H4b: Subordinates in participation supervisor/ subordinate dyads will be significantly less likely than those in traditional dyads to desire for their frequency of interaction to change. H4C: 42 Subordinates in a participation supervisor/ subordinate dyad will evaluate their super— visor's communication in the dyad more posi— tively than will subordinates in traditional supervisor/subordinate dyads. -!-n. I I— '. _— ‘4 c CHAPTER THREE RESEARCH METHODS AND PROCEDURES This chapter will present the operations for testing the hypotheses. First, the design and construction of the questionnaire is covered. This is followed by a description of the sample. Then, the administration of the questionnaire along with other procedures is pre- sented. Finally, the statistics used to test the signif- icance of the results end the chapter. The Questionnaire Pretest An initial questionnaire was designed and administered to seventeen individuals. All of these respondents held information processing jobs. Seven of them worked as tellers and clerks in a loan company. Five of them worked in a branch of the local state judiciary as lawyers and clerks. The remaining five consisted of one manager of public relations of a large local industry and four secre- taries. The key questions on this questionnaire dealt with perceived and desired likelihood of solicited and unso— licited information giving. The sequence of questions measuring perceived and desired likelihood of solicited information giving are presented below. The actual questionnaire is presented in the appendix. 43 44 Think of the times when you talk with your supervisor. Some of these conversations are about work. Sometimes you ask him for work information and sometimes he asks you for work information. (Percentages of respondents using each category are shown in the space before the foil.) Which of you is more likely to ask the other for work information? 12% He is more likely 12 We are about equal 76 I am more likely How do you feel about this? 0% Prefer he ask me more 60 I'm satisfied 22 Prefer I ask him more (NOTE: 6% wrote in "dissatisfied") 12 DK, NR Which of you is more likely to give the other work information which the other had not previously asked about? 59% He is more likely 35 We are equally likely 6 I am more likely DK, NR How do you feel about this? 12% Prefer he give me more 60 I'm satisfied Prefer I give him more (NOTE: 14% wrote in "dissatisfied") 14 DK, NR One of the problems with the questions used in the pre— test was that one foil in each question usually collected a very large number of respondents. Another was that the foils to the question used to index a person's desire for change were not perceived as exhaustive by some respondents. It was also felt that in order to relate the variables in the manner suggested by the hypotheses that no point 45 along the dimension being quantified have an evaluative dimension. The "we are about equal" foil seemed to have that failing. The same may be true for the "I'm satis- fied" foil to the desire for change question. Consequently, the items were altered drastically. The final form of the solicited information-giving sequence of questions is presented below. Questions concerned with unsolicited information giving followed the same format. The actual questionnaire is attached as an appendix to the present study. Think about the times when one of you is giving the other work information which the other has not re- quested. Sometimes you give your boss work infor- mation which he or she has not requested and some- times your boss gives you work information which you have not requested. Of all the times when one of you is giving the other work information which the other has not requested, what percentage of those times are you the one doing the giving compared to the percentage of times that your boss is the one doing the giving? It's about . . . 95% me and 5% my boss 80% me and 20% my boss 65% me and 35% my boss 50% me and 50% my boss 35% me and 65% my boss 20% me and 80% my boss 5% me and 95% my boss IIHHI 95% me and 5% my boss 80% me and 20% my boss 65% me and 35% my boss 50% me and 50% my boss 35% me and 65% my boss 20% me and 80% my boss 5% me and 95% my boss How often during the day does one of you give the other work information which the other has not re- quested? It's about . . . 20 or more times a day 15 19 times a day 10 14 times a day 5 - 9 times a day l — 4 times a day less than once a day How often would you prefer that one of you be giving the other work information which the other has not 20 or more times a day 19 times a day 10 - 14 times a day 5 - 9 times a day l - 4 times a day less than once a day [.1 U1 I The scales used to index subordinates' evaluation of supervisors' communication in the dyad were not pretested “.8. W. _O' I 47 for the present study. However, they have been used in a prior study, also of a bureaucratic organization. 1 The following items loaded on the same factor in a fac- tor analysis of all questions in the study. sis consisted of an orthogonal rotation of the inter-item correlation matrix for a varimax solution. The Kiel- Wrigley criterion was used for terminating rotation. 2 How would you rate your supervisor's communication on the following points? riate space.) (Place an X in the approp- Scale No. Good Average Poor 2 Being easy to talk to about problems or complaints 7 Taking prompt action on problems or com- plaints 3 Listening to what I say 4 Giving me the infor- mation I need to do a good job 6 Explaining clearly what he wants 5 Letting me know how well I am doing 1 Letting me know what he thinks of my prOposals or ideas Code 3 2 l l Berlo et a1. (1970) 2Kiel, (1966) The factor analy- Scale No. 48 Second Item Highest Item Standard Loading Loading» Mean Deviation l .71 -.26 2.45 .71 2 .69 -.08 2.77 .54 3 .68 -.18 2.68 .56 4 .66 -.29 2.44 .69 5 .64 -.27 2.29 .73 6 .64 —.25 2.42 .65 7 .61 -.22 2.45 .70 Average factor loading is .66. The number of people responding to this questionnaire was 289. The complete questionnaire consisted of the following items in the order in which they are listed below. 1. 2. A practice question for those questions concerned with percentage of message flow. A question asking for general frequency of com- munication. A set of questions asking for . . . a. Perceived percentage of solicited informa- tion—giving. b. Desired percentage of solicited information- giving. c. Perceived frequency of occurrence of solicited information-giving. d. Desired frequency of solicited information- giving. A set of questions asking for . . . a. Perceived percentage of unsolicited infor— mation-giving. b. Desired percentage of unsolicited informa- tion-giving. c. Perceived frequency of occurrence of un- solicited information-giving. d. Desired frequency of unsolicited infor— mation-giving. 49 5. The set of items used to index evaluation of supervisor's communication in the dyad. (7 items) 6. Demographic questions which asked for . . . a. Age of respondent b. Sex of respondent c. Age of supervisor d. Sex of supervisor e. Length of time under present supervisor f. Respondent's rank in the organization. Description of the Sample The sample consisted of persons who perform clerical tasks, such as filing, bookkeeping, and accounting in an office of the state government of Michigan. Ninety percent of the respondents were female. Sixty-five percent of the respondents worked for male supervisors. The average age of the respondents was 35 years. Aver- age age of the supervisor was around 45. (Supervisor's age was provided by the subordinate and so may be some- what in error.) Seventy-five percent of the respondents were first- level employees. Seventeen percent were first—level supervisors; and seven percent were second-level mana- gers, or higher. 50 Preparation for Questionnaire Administration Because the questionnaire contained what could be anxi- ety producing questions for the respondents, it was decided that the normal channels of distribution should not be util— ized. If the supervisor distributed questionnaires, asking his subordinates to evaluate him, it is likely that the responses would be biased. Using normal distribution channels also has another weakness; it would allow the respondents to discuss the questionnaire among themselves while they were completing it. This would tend to bias their answers to the questions. It was decided that the questionnaire would be dis- tributed and collected by the researcher in the various work areas instead of using normal channels. This pro- cedure had the advantage of linking the questionnaire directly with the researcher, rather than with the organ- ization or the supervisor. Also, it still allowed for minimal interruption of ongoing activities. In order to accomplish this, a letter was prepared by the researcher and sent out to all supervisors under the signature of the director of the organization. The letter (contained in the appendix of the present study) stated: 1) the researcher's interest in the supervisor/ subordinate dyad; 2) that the researcher was performing the research for a thesis; and 3) the method for admin- istering the questionnaire in the work areas. The letter was sent three working days prior to the day of “_ ‘l..‘“_" . 51 administration. The supervisors were told to inform their people of the administration so that little time would be wasted when the researcher came to the work area with the questionnaire. Another letter (also contained in the appendix of the present study) was written as the cover sheet to the ques- tionnaire. This letter also describes the researcher's interest in the supervisor/subordinate dyad, and states that the research is being performed as part of a thesis requirement. Additionally, the respondents were asked to be as accurate as possible in their responses. This letter had many purposes. First, it eliminated the need to give the same information orally to each work group. This prevented any bias which might be due to giv- ing different information to the different work groups. Second, since this letter was signed by the researcher, it further identified the questionnaire with the researcher rather than the organization. This was done to decrease any anxiety the respondents might have about the ques- tionnaire; and as a result, to increase the accuracy of their responses. Thirdly, the letter attempted to provide a rationale for the need to collect the data so that the respondents would perceive that they were performing a useful function. 52 Administration On the day prescribed by the letter to the supervisors, the researcher and the secretary to the director of the organization went to each work area to administer the questionnaires to the respondents at their desks. In most cases, the questionnaires were placed in front of the respondent by the secretary or by the researcher. F“ In the larger departments, either the supervisor or a a subordinate helped with the distribution and collection of the questionnaires. Most of the respondents completed the questionnaire ‘ET; in five to seven minutes. When the respondents in a work area had finished the questionnaires, they were collected; and the researcher and the secretary went on to another work area. The entire data collection took little more than an hour. Consequently, there was little opportunity for respondents who had completed the questionnaire in one area to talk with respondents who had yet to complete the questionnaire in another area. Also, there was little Opportunity for the respondents to discuss the questionnaire among themselves during the administration, particularly in the smaller work areas. It was possible to observe everyone in the smaller areas and verify that there was little interaction. In the larger work areas, it was impossible to view all respon- dents at the same time. Consequently, it was possible for respondents to interact without'the researcher being aware of it. 53 Since the researcher collected the major proportion of the questionnaires himself, he also received the respon- dents' volunteered evaluations of the questionnaire. None of these was related to the questions which operationalized the direction of message flow. The total number of questionnaires collected by the researcher was 203. Three of these were blank, and two contained unuseable data. There were officially 230 working in the organization at the time of the administration of the questionnaire. The disparity between the number of completed questionnaires and the number of employees is explained by absences and by the mobility of the staff. Some of the workers were away from their desks during the data collection period. These people were checking files, etc., in another part of the building. For two reasons, it was decided not to return for those who had not completed the questionnaires. First, it would be difficult to find these people, since names were not requested on the questionnaire. And, secondly, the results obtained from these questionnaires might be biased by the later respondents' having discussed the question— naire with earlier respondents. The total number of questionnaires transcribed to computer cards was 198. The computer program used to calculate Eta could not handle missing data; i.e., non- responses. Consequently, those hypotheses requiring 54 the correlation coefficient for their test were tested on the 193 respondents who had completed all responses. Statistical Tests of Hypotheses Several statistics will be computed for each of the curvilinearity hypotheses. These are: Pearson's Product Moment Correlation (r); the correlation ratio (n); the F-test of the significance of each of these; and an F- test of the difference between the correlation ratio and the correlation coefficient. The formulae used for the F-tests are as follows. r2 1. Frxy = 2 (l-r )/(N-2) d.f. = 1 n-2 (p. 308)1 nz/G-l 2. Fnyx = 2 (l-n )/(N-G) d.f. = G-l, N-G (p. 308)1 (nz-r2)/(G-2) 3. Fer-nyX = 2 (l—n )/(N-G) d.f. = G-2, N-G (p. 314)1 The correlation coefficient (r) is an index of degree of linear association between two variables. Eta (n) is an index of the total association including any curvi- linearity component, of one variable to the other. lFrom McNemar (1969) 55 Formula Number 3 above is a test of the difference between r2 and n2. When the difference between the two is signi- ficantly large, the linear hypotheses must be abandoned in favor of a more complex function. A significant dif- ference indicates that there is a Significant curvilinear component in the relationship between the two variables. It does not mean, however, that the curvilinear component is as predicted in the hypotheses. Therefore, for each test the means of the arrays of the independent variable will be plotted to see if they conform to the curve specified in the hypotheses. The l3 Riv level of significance for the tests will be p<.05. Hla, b, c The solicited information scale will be reflected in the analysis so that the percentage of information ask- ing becomes the percentage of solicited information giv- ing. For Hla the absolute difference between perceived and desired percentage of solicited information will be correlated with perceived percentage of solicited infor— mation. For Hlb the perceived percentage of solicited infor- mation giving will be correlated with the absolute dif- ference between perceived and desired frequency of solic- ited information giving. For ch the evaluation of supervisors' communication in the dyad scales will be summed to form an index. This 56 index will then be correlated with the subordinates' per- ceived percentage of solicited information giving in the dyad. H2a, b, c These tests will be performed exactly as the previous three tests; with the exception that the perceived unsolic— ited information giving scale need not be reflected. '1 H3a, b, c In order to combine the perceived percentages from (4 both kinds of information giving, it will be necessary to weight these modes according to the frequency with which they occur in the dyad. This will be accomplished according to the formula: Subordinate's total percentage of message transmission = P_(F_ +P.(F.) 51g 819 uig uig F +F. Slg uig Where P . = subordinate's perceived percentage of 519 solicited information giving P . = subordinate's perceived percentage of U19 unsolicited information giving F = frequency of occurrence solicited infor— sig mation giving in the dyad F , = frequency of occurrence unsolicited “19 information giving in the dyad 57 For actual computation the coded value of each response will be used rather than the percentage represented in the foil. The values resulting from the formula will vary from 0 to 6, corresponding to the following percentages: 0 = 95% subordinate, 5% supervisor l = 80% subordinate, 20% supervisor 2 = 65% subordinate, 35% supervisor 3 = 50% subordinate, 50% supervisor 4 = 35% subordinate, 65% supervisor .«A‘u’ie W 3'?" 5 = 20% subordinate, 80% supervisor 6 = 5% subordinate, 95% supervisor ‘vua v.1“. “4:4! '. For H3a a chi-square test will be performed which will use the absolute difference between scores derived from the coded values of perceived and desired percentages of subordinates' message transmission in both modes. Re- spondents will then be divided into two categories: Desire change and desire no change. These will then be compared to total percentage of message transmission. The level of significance used will be p<.05. For H3b the total percentage of message transmission will be compared, in a fashion similar to that used in testing H3a, to the absolute difference between perceived frequency of occurrence of each mode and desired fre— quency of occurrence of each mode. A chi-square test of significance will be used with p<.05. For H3c the total percentage of message transmission will be correlated with the index of the subordinates' 58 evaluation of the supervisors' communication in the dyad. Again Eta will be used to test the significance of the curvilinearity. H4a, b, c The traditional and participation dyads will be drawn from the rest of the sample in the following manner: Fr... Traditional dyads: All subordinates who report that 2 their supervisors have a percentage of 65% or over in both modes of message transmission. ‘ Participation dyads: All subordinates who report that Ej‘ their supervisors have a percentage of 65% or over for unsolicited information giving, but a percentage of 35% or lower for solicited information giving.l A two-tailed t—test of the differences between means will be used to test the significance of the relationships, p<.05. For H4a, the means on total change in percentage of those desiring no change will be compared between tradi- tional and participation dyads. 1Note that participation here does not refer to the par- ticipatory forms of management suggested by such writers as Argyris (1964) and McGregor (1970). It merely implies that both members have a high level of information-giv— ing but in different modes. The subordinate has a higher proportion of the solicited information-giving and the supervisor has a higher prOportion of unsolicited infor- mation giving. 59 For H4b, the means on total change in frequency of interaction will be compared between traditional and participation dyads. For H4c, the means of the evaluations of supervisor's communication will be compared between traditional and participation dyads. CHAPTER FOUR RESULTS There will be four sets of hypotheses tested in this chapter. Each set will consist of the relationship of one variable to desired amount of change in percentage of message flow, desired amount of change in frequency of contact, and evaluation of supervisor's communication. The first set of hypotheses is concerned with the relationship between solicited information-giving and: l) desired amount of change in percentage of solicited information-giving; 2) desired amount of change in fre- quency of solicited information-giving; and 3) evalua- tion of supervisor's communication. The second set of hypotheses is concerned with the relationship between unsolicited information-giving and: l) desired amount of change in percentage of unsolicited information giving; 2) desired amount of change in fre— quency of unsolicited information-giving; and 3) evalua- tion of supervisor's communication. The third set of hypotheses is concerned with the relationship between perceived percentage of total in— formation flow and: l) desired amount of change in per- centage of total information flow; 2) desired amount of change in frequency of total information flow; and 3) evaluation of supervisor's communication. 60 61 The fourth and final set of hypotheses is concerned with the differences between participation and balanced supervisor/subordinate dyads on three variables: 1) de- sired amount of change in total percentage of message flow; 2) desired amount of change in total frequency of message flow; and 3) evaluation of supervisor's com- munication. Solicited Information Giving Hla: There is a curvilinear relationship between subordinates' percentage of solicited infor- mation giving in a dyad and the amount of change in that percentage they will desire; such that, those with a moderate percentage will desire less change than those with more extreme percentages. The correlation coefficient and Eta between the two variables was not significant. With these data, the null hypothesis cannot be rejected. (Table 1) 62 Table l % Solicited Information Giving by Desired Amount of Change of % Solicited Information Giving F d.f. N = 193 r2 1.753 1,191 r = .095 n2 1.770 6,186 n = .232 n2-r2 1.768 5,186 '7 Hlb: There will be a curvilinear relationship between . subordinates' percentage of solicited informa- tion giving in a supervisor/subordinate dyad if and the desired change in frequency of solic- E9 ited information giving in the dyad; such that, those with a moderate percentage will desire less change in frequency than will those with more extreme percentages. Table 2 % Solicited Information Giving by Desired Amount of Change in Frequency of Solicited Information Giving F d.g. N = 193 r2 .172 1,191 r = .03 n2 .923 6,186 n = .17 n2-r2 1.073 5,186 There is not a significant linear or curvilinear re- lationship between the two variables. Consequently the null hypothesis cannot be rejected. (Table 2) ch: % S 63 There will be a curvilinear relationship between subordinates' percentage of solicited informa- tion giving in a dyad and the subordinates' evaluation of their supervisor's communica— tion in the dyad; such that, those with a moderate percentage will evaluate their super- visor's communication more positively than 3 will those with more extreme percentages. Table 3 , .,-. - "‘ 4.1. A- ,' pK'. .' . I olicited Information Giving by Subordinate's Evaluation of Supervisor's Communication in the Dyad F d.f. N = 193 r2 .557 1,191 r = -.05 n2 1.127 6,186 n = .04 n2-r2 1.240 5,186 The correlation coefficient and the Eta were not significant. Thus, the null hypothesis cannot be rejected. The three hypotheses in the first set did not receive support from the data. Unsolicited Information Giving H2a: There is a curvilinear relationship between subordinates' percentage of unsolicited infor- mation giving and the amount of change they will desire in that percentage; such that, 64 those with moderate percentages will desire less change than those with more extreme per- centages. Table 4 % Unsolicited Information Giving by Desired Amount of Change of % Unsolicited Information Giving F d.f. F‘ N = 193 r2 1.722 1,191 r = -.09 n2 3.897* 6,186 n = .33 n2-r2 4.302* 5,186 :1 *Significant past the .05 level. 1? The correlation coefficient was not significant. However, Eta and the difference between Eta and r2 were significant past the .05 level. Thus, the linear hypo- thesis can be rejected (Table 4) Figure 5 presents the plot of the means of the arrays for unsolicited information giving. The curve for the relationship is complex with those desiring least change occurring at "50% me" and "35% me." There is also a minor mode occurring again at "85% me." Thus, hypothesis 2a finds support in the data. m onsmflm mmH u z on mm mm mm ma ma HH 0 c .r mom .r wmm mcH>HO .. wow coaumEnOmcH UmuHOHHOmCD .. wma mmmgcwoumm CH mmcmno .. woa mo USSOE< . pmufimma .. wm - - . wo A A _ _ _ . wm wow wmm wom wmo wow wmm mcH>lecoHumEuOmcH pmuHOAHOmCD mo ommucoonom 65 H2b: 66 There is a curvilinear relationship between subordinates' percentage of unsolicited informa- tion giving and frequency of occurrence of unso- licited information giving they will desire in the dyad; such that, those with moderate per- centages will desire less change in frequency of unsolicited information giving in the dyad than will those with more extreme percentages. Table 5 % Unsolicited Information Giving by Desired Amount of Change in Frequency of Unsolicited Information Giving F d.f. N = 193 r2 .052 1,191 r = .02 n2 .611 6,186 n = .14 nZ—rz .723 5,186 There is no curvilinear or linear relationship between unsolicited information giving and amount of desired change in frequency of unsolicited information giving. (Table 5) H2c: There will be a curvilinear relationship be- tween subordinates' percentage of unsolicited information giving and subordinates' evalua— tion of their supervisor's communication in the dyad; such that, those with a moderate 67 percentage will evaluate their supervisor's communication more positively than will those with more extreme percentages. The correlation coefficient was not significant. Eta was significant as was the difference between Eta and r2. Therefore, the linear hypothesis may be rejected and the curvilinear hypothesis finds support in the data. (Table 6) Table 6 % of Unsolicited Information Giving by Subordinate's Evaluation of Supervisor's Communication in the Dyad F _élil N = 193 r2 .385 1,191 r = .045 n2 3.000* 6,186 n = .297 nZ-rz 6.980* 5,186 *Significant past the .05 level. The plot of the means of the arrays gives general sup- port to the shape of the curve prescribed by the hypothesis. There is, however, a minor mode occurring at 80% of the message flow. (Figure 6) mmH m musmflm N 2 mm mm mm mm ma H H II ma 1 I m 00 I" \O .u HH 1. ma 1 ea wm wow wmm wom wmm mcH>H0 GOHDOEMOmcH UODAOHHOmCD mo mmoucmoumm 30_.H 66mm 028 ch COHDMOHGDEEOU m.HOmH>uomsm mo coflpmsam>m cmmz swam 68 69 Thus in the second set of hypotheses, perceived per- centage of unsolicited information-giving is curvilinearly related in the manner predicted by the hypotheses to the amount of change in percentage of unsolicited information- giving the subordinate will desire and to subordinate's evaluation of his supervisor's communication in the dyad. It was not found to be related either linearly or curvi- linearly to the amount of change in frequency desired by the subordinate. Total Information-Giving H3a: There is a curvilinear relationship between subordinates' percentage of £2331 informa- tion giving and whether they will desire a change in that percentage; such that, sub- ordinates with moderate percentages will de— sire less change than will subordinates with more extreme percentages. Amount of total information flow is a transformed vari- able ranging from zero to six. It is the percentage of each mode times its frequency of occurrence summed and divided by the sum of their frequencies of occurrence. In the actual transformations the recoded values of per- centages and frequencies were used. The recoded values conform to the following percentages: 70 O = 95% subordinate, 5% supervisor l = 80% subordinate, 20% supervisor 2 = 65% subordinate, 35% supervisor 3 = 50% subordinate, 50% supervisor 4 = 35% subordinate, 65% supervisor 5 = 20% subordinate, 80% supervisor 6 - 5% subordinate, 95% supervisor A respondent with a value of 4 transmits thirty-five percent of the total information in the dyad. The super- visor in the same dyad transmits 65% of the total infor- mation in the dyad. Thus direction of message flow, both solicited and unsolicited is from supervisor to subord- inate. The lower the number, the higher the respondent's pro- portion of total information flow, with the value of 3 corresponding to the supervisor/subordinate dyad in which either member is equally likely to transmit messages. The number of respondents who perceive that they have a very large proportion of the total information flow in the dyad was very small. Consequently, their responses were collapsed in order to be analyzed. The category 2.5 to 3.4 represents the "moderate" percentage category. 71 As can be seen in Table 7, X2 is significant past the .05 level. Thus, hypothesis 3a receives support from the data. Table 7 Amount of Total Information Giving by Desired Change in Total Information Giving Amount of Total Information Giving Desired Change in Message Flow 0-2.4 2.5-3.4 3.5-4.4 4.5-5.4 5.5-6.0 Desire no Change 18% 58% 49% 41% 42% n = 16 48 45 41 45 X2=9.648*; d.f. = 4 *Significant past the .05 level H3b: There is a curvilinear relationship between subordinates' percentage of total informa- tion giving and the amount of change they will desire in frequency of interaction in the super— visor/subordinate dyad; such that, those with moderate percentages will desire less change in the frequency of interaction than will those subordinates with more extreme percent- ages. 72 Table 8 shows that X2 is not significant past the .05 level, although the 2.5-3.4 category again had a larger percentage than did any of the other categories. Hypo- thesis 3b is not supported by the data. Table 8 Amount of Total Information Giving by Desired Change in Frequency of Total Information Giving Amount of Total Information Giving Desired Change in Frequency 0-2.4 2.5-3.4 3.5-2.4 4.5-5.4 5.5-6.0 Desire no Change 31% 48% 40% 46% 42% n = 16 48 45 41 45 X2=.8507; d.f. = 4 H3c: There is a curvilinear relationship between sub- ordinates' percentage of total message giving and the subordinates' evaluation of the super- visor's communication in the dyad; such that, subordinates with moderate percentages evalu- ate the supervisor more positively than sub- ordinates with more extreme percentages. The correlation coefficient, Eta and the difference between the two were not significant. (Table 9) 73 Table 9 Amount of Total Information Giving by Subordinate's Evaluation of Supervisor's Communication in the Dyad F d.f. N = 193 r2 1.377 1,191 r = -.083 n2 .838 6,186 n = .148 nZ-rz .721 5,186 The hypothesis was not supported by the data. In the third set of hypotheses, one of the three was significant. Subordinates' perceived percentage of total information-giving is curvilinearly related in the manner predicted by the hypotheses to the amount of change in that percentage he will desire. Perceived per- centage of total information-giving is not related to amount of change desired in frequency of total informa- tion giving, nor is it related to subordinates' evaluation of the supervisor's communication in the dyad. Dyad Style: Traditional versus Participative A two-alternative test of significance of (t) was used for testing the following hypotheses with p<.05. H4a: Subordinates in participation supervisor/ subordinate dyads will desire less change in total message giving than will those in tra- ditional dyads. 74 Table 10 presents the results of the analysis. Table 10 Type of Supervisory Relationship by Mean Amount of Change Desired in Total Message Giving Supervisor/Subordinate Dyads Traditional Participation Mean amount of Desired change .839 .885 Standard Deviation .882 .891 Number in each group 87 26 t = .2291; d.f. = 111 The null hypothesis cannot be rejected, given the present data. H4b: Subordinates in participation supervisor/ subordinate dyads will be significantly less likely than those in traditional dyads to desire for their frequency of interaction to change. Table 11 presents the results of the analysis. 75 Table 11 Type of Supervisory Relationship by Mean Desired Amount of Change in Frequency of Interaction Supervisor/Subordinate Dyads Traditional Participation Mean amount of desired change in frequency 1.080 1.192 Standard Deviation 1.096 .828 Number in each group 87 26 t = .0525; d.f. = 111 The difference between the means is not significant. Thus, the null hypothesis cannot be rejected. H4c: Subordinates in a participation supervisor/ subordinate dyad will evaluate their supervisor's communication in the dyad more positively than will subordinates in traditional supervisor/ subordinate dyads. Table 12 presents the results of the analysis. 76 Table 12 Type of Supervisory Relationship by Mean of Subordinate's Evaluation of Supervisor's Communication in the Dyad Supervisor/Subordinate Dyads Traditional Participation Mean evaluation of super- visor's communication in the dyad 9.942 10.000 Standard Deviation 3.767 2.948 Number in each group 87 26 t = .0716;d.f. - 111 The difference between the means is not significant past the .05 level. The null hypothesis cannot be rejected, given the present data. In the fourth set of hypotheses there were no signifi- cant differences between traditional and participative dyads on the three variables: 1) desired amount of change in per- centage of total information—giving; 2) desired amount of change in frequency of total information-giving; and 3) eval— uation of supervisor's communication in the dyad. CHAPTER FIVE SUMMARY AND CONCLUSIONS Chapter Five summarizes the results of the present study. In addition, it presents some implications as to how the communication-oriented manager may vary his behavior so as to change the subordinate's level of satisfaction with the communication in the supervisor/subordinate dyad. Also suggested in the present chapter are other variables which may be related to direction of message flow in the hier- archically differentiated dyad. Lines of future research are also suggested. Summary of Findings Hypotheses la, b, and c failed to receive support from the data. These hypotheses were concerned with the rela— tionship of perceived percentage of solicited information- giving to three other variables: 1) desired change in per- centage of solicited information-giving; 2) desired change in frequency of solicited information-giving; and 3) evalu- ation of the supervisor's communication in the dyad. Hypotheses 2a, b, and c received some support from the data. These hypotheses predicted a curvilinear relationship between perceived percentage of unsolicited information- giving and each of three variables: 1) desired change in percentage of unsolicited information-giving; 2) desired change in frequency of unsolicited information—giving; and 77 78 3) evaluation of the supervisor's communication in the dyad. A curvilinear relationship of the form predicted by the hypotheses was found between the criterion vari- able and the desired change in percentage of unsolicited information-giving, and between the criterion variable and evaluation of the supervisor's communication in the dyad. The relationship did not hold for the criterion variable and desired change in frequency of unsolicited information-giving. Thus, those subordinates who perceive that they are as likely as their supervisor to give the other unsolic- ited information are less likely to desire to change that percentage and will evaluate their supervisor's communica- tion more highly than will subordinates with either a smaller or larger percentage of unsolicited information- giving in the dyad. The third set of hypotheses (3a, b, and c) received less support. These hypotheses were concerned with the relationship between total information-giving and three variables: 1) desired change in percentage of total in- formation-giving; 2) total desired change in frequency; and 3) evaluation of supervisor's communication in the dyad. Only the first hypothesis received support. Those subordinates who perceived that they were equally likely to give the other information (solicited or unso- licited) were less likely to desire to change that aspect of their relationship. 79 The fourth set of hypotheses (4a, b, and c) received no support from the data. It was hypothesized that par- 1 would differ on three ticipative and traditional dyads variables: I) desired change in percentage of total information-giving; 2) desired change in frequency of information—giving; and 3) evaluation of supervisor's communication in the dyad. There was no significant difference between the two types of relationships. Table 13 indicates that the supervisor cannot manipu- late the subordinate's satisfaction with the communication in the dyad by asking the subordinate more questions (i.e., increasing the subordinate's percentage of solicited infor- mation-giving).2 However, he can increase the subordinate's satisfaction with communication in the dyad by attempting to equalize the subordinate's percentage of unsolicited information—giving. This is more difficult than is equal- izing solicited information giving between them. The lat- ter is equalized by asking the subordinate for more or 1In the participative dyad the subordinate has 65% or more of the solicited information-giving while the supervisor has 65% or more of the unsolicited information-giving. In the traditional dyad, the supervisor has 65% or more of the solicited and unsolicited information-giving in the dyad. Thus iH—Ehe participation dyad each has a high percentage of message transmission but in different modes. In the traditional dyad, the supervisor has a high percentage in both modes; and, thus the subordinate has a lower percentage in both modes. 2Unless he increases the frequency of contact at the same time. Frequency of contact is linearly related to the satisfaction variables with correlations ranging from .27 to .29. uCMOAMHcmHm uoz uGMUAMHcmHm uoz ucmofluflcmflm ucMOHchmHm uoz ammo map 28 coflumoflcsafioo m.u0mfl>nmmsm mo coflpm5H0>m “cavemacmflm poz uCMUHMHcmHm uoz Damnamflcmflm uoz ucmoHMHcmHm uoz mochqum CH mmcmno pmnflmmo pcmoHMHcmHm uoz mamum pmmo pcm0flmwcmflm mcH>HU coflumEH0mcH mo mmmucmoumm Hmuoa pcMOHmwcmHm aca>wo coaumEHOMcH pouHoHHOmcD mo mmmvcmonmm pm>flmouom ucmoHMHcmHm poz mafi>flo coflpmfihomcH pmufloflaom m0 mmmucmoumm pm>flmonmm mmmucmoumm CH mmcmsu pmnflmwo mpsvm mcu Scum muHSmmm uGMOHMHcmHmIcOZ cam ucmofimflcmflm ma GHQME 80 81 less information than previously. In order to equalize unsolicited information-giving, the supervisor must create a climate in which the subordinate feels free to bring unsolicited information to him if the supervisor has a larger percentage than the subordinate. Or, the supervisor should attempt to increase his own percentage of unsolic- ited information-giving if his percentage is lower than the subordinate's.1 It is interesting to note that the desired amount of change in frequency of information—giving in either mode was not related to any of the criterion variables in the manner predicted by the hypotheses. Frequency of contact, as well as frequency of information-giving in either mode were related linearly with evaluation of supervisor's communication. Evaluation of Supervisor's Communication Frequency of contact r = .29 Frequency of solicited information-giving r = .26 Frequency of unsolicited information-giving r = .24 1It would be difficult to believe that this situation would occur often. 82 This would indicate that frequency of contact, as suggested by several of the studies mentioned earlier in this study, is related linearly to satisfaction with communication, such that the higher the frequency of con- tact the higher the level of satisfaction. However, direction of message flow is not related to frequency of contact or to desired change in frequency of contact. The fact that both direction of unsolicited informa- tion-giving and frequency of contact explain a portion of the variance in the evaluation of supervisors on communication and that the two are independent, both linearly and curvilinearly, indicates that a theory con- taining both will be stronger than a theory containing only one. In this case, social exchange theory can be extended to include direction of message flow. Implications for Future Research Variables which have been related to change and inno- vativeness in the diffusion research (such as age, mobility, and education) may have an effect on the relationship be— tween direction of message flow and desire to change the direction of message flow. That is, for younger, more mobile, and more educated pe0ple the relationship will be stronger than for older, less mobile, less educated peOple. 83 It is also suspected that the relationship between message flow direction and desire to change the direction of message flow will vary with the duration of the rela- tionship between the two members of the dyad. If, as suggested by Turk and Wills (1964), the behaviors of mem- bers of a dyad and their expectations concerning the re- lationship are at first determined by previous experience and other external factors, then they should be more desirous of change if the direction of message flow is not as they would want it. However, as the relationship con- tinues over time the expectancies and behaviors of the members are more and more determined by previous experi- ence in that dyad. The members will be less desirous of change in the relationship, no matter what the direction of message flow may be. While no direct analysis of this was performed in the present study, there was indeed a linear correspondence between what the subordinate perceived his relationship with his supervisor to be and what he desired it to be. Fifty—five percent of the respondents, for example, said that they desired no change in their present percentage of solicited information-giving, no matter what percent- age they perceived they had. The correlations between perceived percentage and frequency of information-giving 84 and desired percentage and frequency of information-giv- ing ranged from .45 for percentage of solicited informa- tion-giving to .53 for percentage of unsolicited infor- mation-giving. This leads to another theoretic point. The rela- tionship between direction of message flow and desire to change the direction of message flow is stronger for those who are dissatisfied (that is, desire some change) than for those who are not. While seemingly a simple point, this says that while the existing relationship is the one most subordinates desire, those who are dis- satisfied with their present relationship can be predicted by the direction of message flow in the dyad. It is sus- pected, then, that if all persons desiring no change in their present relationship are removed from the analy- sis, the resulting relationship between direction of mes- sage flow (both solicited and unsolicited) would be much stronger, and the subordinate's evaluation of his super— visor's communication would be more strongly related to the direction of message flow in the manner set out by the hypotheses in the present study. A question for further research is, what is the relationship between frequency of contact, direction of message flow, and dissatisfaction with communication in the supervisor/ subordinate dyad. A question of immediate concern is why solicited information-giving is not related to subordinate's 85 communication satisfaction in the dyad. We have evi- dence that it is related in non—hierarchically differ— entiated dyads from the study reported by Blau and Scott. (1962) The supervisor/subordinate dyad is based upon norms and prescriptions set out by the organization. Status in that relationship is non-negotiable. On the other hand in non-hierarchically differentiated dyads, the norms and prescriptions for the particular relationship are not set out by the organization but are determined by the partic- ipants. Thus, status is a negotiable item. Blau and Scott argue that in this relationship, the act of asking for information is associated with the giving up of status to the person being asked. Solicited information-giving in the non-hierarchically differentiated dyad, because it is related to the negotiability of status in the relationship, is also related to the desire to change aspects of the communication in the relationship. In the supervisor/ subordinate dyad, status is not negotiable and, there- fore, solicited information-giving is not related to the desire to change the communication in the relationship. However, the giving of unsolicited information in the non-hierarchically differentiated dyad may be related to a member's desire to change the communication in the dyad. The Blau and Scott study only focused on the solicited information-giving mode. Therefore, another direction of future research should be testing the relationship 86 of direction of message flow in the supervisor/subord- inate dyad to satisfaction with the communication in the relationship. There may be a methodological ambiguity in the instru— ment used. It did not allow the respondent to make a dis- tinction between his idealized and realistic expectations for desired percentage of message transmission. A respon— dent may have interpreted the question to be asking for a "best of all worlds" answer; or, he may have interpreted it to be asking for a "given present constraints" answer. This ambiguity of interpretation calls the precision of the data into question. Further research would seek to eliminate this ambiguity by pointing out the distinction to the respondent and then asking him for his realistic desired percentage of message transmission. There is one final theoretic point that must be made. Homans specifies that he does not feel his theory predicts for a dyad in which the rights of the members are deter- mined by a role in the organization or institution. It would seem that the theory also would not apply to those relationships in which the rights of the members are determined by the organization itself or by another organ- ization. For example, a supervisor in a profit-making organization generally has the right to fire a subordinate for incompetence. This is one of the reward/punishment alternatives Open to him. The same is not true for the 87 supervisor in a bureaucracy. This right is withheld from him by another organization called civil service. Communication would seem to take on more importance as a reward/punishment mechanism as reward/punishment alter- natives available to the supervisor in the hierarchically differentiated dyad become fewer. The thrust of this argument is toward the need for specifying the level of rights of (or the range of reward/punishment alternatives available to) the supervisor in the particular organiza- tion being studied. Two areas of importance which have not been touched upon in the present study are supervisor's satisfaction with communication and group satisfaction with communica— tion. Since the present study found support for the pro- position that a relationship exists between direction of message flow and communication satisfaction of the sub- ordinate, consideration should be given to further studies with both supervisor and subordinate satisfaction, as well as with communication satisfaction across group members. LI ST OF REFERENCES LI ST OF REFERENCES Abrahamson, Mark. Interpersonal Accommodation. N.Y.: D. Van Nostrand Co., Inc., 1966. Adams, J. Stacy and Romney, A. Kimball. A functional analysis of authority. Psychol. Rev. 66, (4), 1959. 234-251. Allport, G. W. and Postman, L. The Psychology of Rumor. N.Y.: Holt, 1947. Argyle, M. The Psychology of Interpersonal Behavior. ‘_ London: Penguin Books, Ltd., 1967. afyh 15Q~ 7:? I” l3 /" Argyris, C. Integrating the Individual and the Organi- zation. N. Y.: Wiley, 1964. ,, Back, K., Festinger, L., Hymovitch, 8., Kelly, H., Schacter, S., and Thibaut, J. The methodology of studying rumor transmission. Hum. Relat., 1950, 306-312. I") f“ /-l- ‘4' Bavelas, A. Communication patterns in task oriented 3 groups. J. Acoust. Soc. Amer., 22, 725-730, 1950. ' Berkowitz, Norman H., and Bennis, Warren. Interaction patterns in formal service organizations. Admin. Sci. Q.6, 1961/62, 26-50. Berlo, D. K. Essays on Communication. Mimeographed. Mighigan State University, 1969. ~/.Berlo, D. K., Farace, R. V., MacDonald, D. S., Buley, Jerry L., and Connelly, Richard. Organizational Communication: A Firstline ManageriaICCommunIca- tion System. Dept. Comm., MichIgan State Univers- ity, April, 1970. Blau, Peter M. Patterns of interaction among a group of officials in a government agency. Hum. Relat., 7 (3), 1954. Blau, P. M. and Scott, W. R. Formal Organizations. San Francisco, Calif.: Chandler Pub. Co., 1962. Caplow, T. Rumors in war. Social Forces. 25, 1946— 1947., also in Some Theories of Organization. Rubenstein, A. H. and Haberstroh) C. J. (Ed§.) Homewood, 111.: Dorsey Press, 1966. 88 89 Coleman, J. S., Katz, E., and Menzel, H. Medical Inno- vation. N.Y.: Bobbs-Merrill Co., Inc., 1966. Davis, J. A. Statistical analysis of pair relationships: Symmetry, subjective consistency and reciprocity. Sociometry, 1968, (31), 102-119. ‘57Davis, K. A method of studying communication patterns in organizations. Personnel Psych., 1953A, 6, 301- 302a. Davis, K. Management communication and the grapevine. Harvard Bus., Rev., 1953b, 31, 5, 43-49. Davis, K. Success of chain-of-command oral communication in a manufacturing management group. Academy of Mgmt. J., Dec., 1968. Deutschmann, Paul J. and Danielson, Wayne A. Diffusion of the major news story. J. Q. Summer, 1960, 345- 355. Festinger, L., Cartwright, D., Barber, Kathleen, Fleisch, Juliet, Gottsdanker, Josephine, Keysen, Annette, and Leavitt, Gloria. A study of rumor: Its origin and spread. Hum. Relat., 1948, I, 464-486. #iFoa, U. G. Behavior norms and social rewards in a dyad. Behav. Sci., 1958, (3), 323-334. Gray, L. N., Richardson, J. T., and Mayhew, B. H., Jr. Influence attempts and effective power: A re-exam- ination of an unsubstantiated hypothesis. Sociometry, 1968, 31, 245-258. : Guetzkow, H. Communication in Organizations, in Handbook of Organizations. James G. March (Ed.) Chicago: Rand McNally 8 Co., 1965, 534-573. , Hm r3! -" H 0 73110437.? P") ’12» 5133‘" xGuetzkow, H., and Simon, H. A. The impact of certain com- munication nets upon organization and performance in task oriented groups. Mgmt. Sci., 1, 233-250, 1955. Halpin, A. W. and Winer, B. J. The leadership behavior of the airplane commander. Columbus: Ohio State University Research Foundation, 1952. Homans, G. C. The Human Group. N.Y.: Harcourt, Brace, 1950. xaaaga_ ,4, Homans, G. C. Social Behavior: Its Elementary Forms. N.Y.: Harcourt, Brace, 1961. ‘ 90 Jaffee, ., and Lucas. Effects of talking and correct- ness of decision in small groups. J. of Soc. Psychol., 1969, 247-254. Jain, Harish C. Communication effectiveness and super- visory performance in two urban hospitals. Paper presented to International Communication Assoc., Phoenix, Ariz., 1971. u(Katz, Elihu and Lazersfeld, Paul F. Personal Influence.'fi ”j N.Y.: Free Press of Glencoe, 1955. solution of rotating varying numbers of factors with differing initial commonality estimates. Unpublished M.A. Thesis. Dept. Comm., Michigan State University, ' 1966. \ Kiel, Donald Franklin. The effects upon the factorial F7? Kirscht, J. P., Lodahl, T. M., and Haire, M. Some fac- tors in the selection of leaders by members of small groups. J. Abnorm. Soc. Psych., 1959, 58, 406-408. ii Lawler, E. Manager's attitudes toward interaction epi- sodes. J. App. Psychol., 1968, 52 (6, Pt. 1), 432- 439. Leavitt, H. J. Some effects of certain communication patterns on group performance. J. Abnorm. Soc. Psychol., 1951, 46, 38-50. Likert, R. New Patterns of Management. N.Y.: McGraw— Hill Book Co., 1961. "(Litwak, E. Models of bureaucracy which permit conflict. Am. J. Soc., 67, (2), 1961, 177-184. ‘-MacDona1d, D. S. Communication roles and communication content in a bureaucratic setting. PhD. Thesis, Michigan State University, 1970. VrMaslow, T. A theory of human motivation. Psychol. Rev., J“ 50, 1943, 370-396. Mayhew, B. H., Gray, H. L. and Richardson, J. T. Behav— ioral measurement of operating power structures: characterizations of asymmetrical interaction. Sociometry, 1969, 32, (4), 474-489. !;McGregor, D. The human side of enterprise, in Mana e- ment and Motivation. Vroom, V., and Deci, E. (Eds.) Baltimore, Md: Penguin Books, Inc., 1970. 91 McNemar, Q. Psychological Statistics (IV Edition). N.Y.: John Wiley and Sons, Inc., 1969. Mulder, M. The power variable in communication experi— ments, in Communication and Culture. Smith, A. G. (Ed.) N.Y.: Holt, Rinehart, and Winston, 1966. Riecken, H. W. The effect of talkativeness on the ability to influence group solutions to problems. Sociometry, 1948, 21, 309-321. Rogers, Everett M. Diffusion of Innovations. N.Y.: Free Press, 1962. Rubenstein, Albert H. Problems in the measurement of interpersonal communication in an ongoing situation. Sociometry, 1952. Rubenstein, Albert H. Studies of idea flow in research and develOpment, in Some Theories of Organization. Rubenstein, A. and HaBerstroh, C. (Eds.i Rev. Ed., Homewood, Ill: Dorsey Press, 1966. Rubenstein, Albert H. Field studies in idea flow and pro- ject selection in industry, in Operations Research in Research and Deve10pment, Dean, B. V. TEd.) N.Y.: John Wiley, 1963. Schwartz, D. F. Liaison-Communication roles in a formal organization. Fargo, N.D.: Communimetrics Report No. 1, Dept. Comm., North Dakota State University, 1968. Shaw, M. E. Some effects of unequal distribution of information upon group performance in various com— munication nets. J. Abnorm. Soc. Psychol., 49, 1954, 547-553. V/,"Shaw, M. E. Some effects of individually prominent V/ behavior upon group effectiveness and member satis- faction. J. Abnorm. Soc. Psychol., 1959, 59, 382— 386. Shaw, M. E., and Penrod, W. T., Jr. Validity of infor— mation, attempted influence, and quality of group decisions. Psych. Rep;, 1962, 10, 19-23. Sherif, M. Social Interaction: Process and Products. Chicago: Aldine, 1967. V(Smelser, W. T. Dominance as a factor in achievement and ~’ perception in cooperative problem solving interac- tions. J. Abnorm. Soc. Psychol., 62, 535-542, 1961. 92 Turk, Herman, and Wills, G. Robert. Authority and Inter- action. Sociometry, 1964, 1-18. Turner, A. N. Foreman, job and company. Hum. Relat., 10, 100-112, 1957. :Vroom, V., and Deci, E. (Eds.) Management and Motiva— J tion. Baltimore, Md.: Penguin Books, Inc., 1970. Watzlawick, P., Beavin, J. H., and Jackson, D. J. Pragmatics of Human Communication. N.Y.: W. W. Norton, 1967. Weiss, R. S. Processes of Organization. Ann Arbor: Institute for Soc. Res., University of Michigan, 1956. Whyte, W. F. An interaction approach to the theory of organization, in Organizational Behavior and the Practice of Management. Hampton, D. R., Summer, C. E., and Webber, R. A. (Eds.) Glenview, Ill: Scott, Forsman and Co., 1968, 52-71. APPENDICES APPENDIX A SUPERVISOR/SUBORDINATE Message Flow Questionnaire We are interested in how messages tend to flow between supervisors and the peeple Who work for them. Consequently, we are asking many people in many different org» inations to tell us something about their communication with their supervisors. We do‘gg£_want your name on this questionnaire. However, we would like for you to answer each of the questions as objectively as you can. 1. First, we would like to know how long you have worked for your present supervisor? years w? months 2. How frequently you talk with him? several times a day--(about how many times a day . ) once a day less than once a day, but more than once a week once a week less than once a week l Think about the times when you talk with your supervisor. Some of these con- versations are about work. Sometimes you ask him for work information and some- times he asks you for work information. '3. Which of you is more likely to ask the other for work information? ’he is more-likely we are about equal I am more likely ! 4. How do you feel about this? ____lvd prefer that he ask me for information more frequently-.. ___1 'm satisfied I'd prefer that I could ask him for information more frequently Other times when you and your immediate supervisor are talking about work, some- times be gives you information which you had not previously asked for and some- times you give him information which he had not previously asked for. 5. Which of you is more likely to give the other work information which the other had not previously asked about? he is more likely we are equally likely .__} am more likely 6. How do you feel about this? I'd prefer that he give me information more frequently I'm satisfied “_m_l'd prefer that I give him information more frequently r»: .4- .3 1.: e ' --v :3 .4. - >111} n—J nu .- ~..... ”w"; LL; . -‘.. '.. w...» a'..'.-._,-. ‘: -... .- c. . ... .. .: -...... - .... .. -.,. z . ..- cation. leen he st.ar£.s a convex 3ntian with vznu show; work: shouting» n3 sans yer .— for information; sometimes he gie es you informatio :1 which you had regs is3zd prom viously; and sometimes he gives you information which you had not previously requested. 7. 8. Of the times when your supervisor starts a conversation with you about work in which he either asks for information or gives unrequested information, what percentage of those times does he 0 . . ask you for information give you information which you had not requested How satisfied are you with this? I'd prefer he'd ask for information more I'm satisfied I'd prefer he'd give unrequested information more This time we want you to think about the times when you start a conversation wi .h your supervisor about work. Sometimes you ask him for information; other times y: i give him work information which he requested; and other times you give him work information which he had not requested. 9. 0f the times when you start a conversation with your supervisor about wo1;k in which you ask him for information or give him no equested information, what percentage of those times do you . . . ask him for information give him information which he had not requested 10. How satisfied are you with this? I' d prefer to ask him for information more frequently 1' m satisfied I'd prefer to give him unrequested information more frequently 11. In general, how satisfied are you with your relationship with your boss? gvery satisfied satisfied netiher satisfied nor dissatisfied dissatisfied very dissatisfied 12. If your relationship with your boss were different could you work more efficiently? ‘;Yes I don't know 310 Please complete the following questions. Thank you. 13. 14. 16. 17. Your occupatioq_u__fl_ _._-1-_1.1 -1“. 1-11-- __11 Your age ears 15. Your sex ( )Female ( )Male M Your bossT; sex )Female ( )Male Your boss' 8 age (approximately) _years '1‘. .: 11 a ‘13 » APPENDIX B _T-I 101 (HIV. S-CI) STATE OF MICHIGAN DEPARTMENT OF TREASURY LANSlNG INTER OFFICE To All Supervisors - Income Tax Section FROM Chester Zawislak, Director DATE Igndl.2l, 1971 SUBJECT Questionnaire on Communications FRESNO. Early next week on Monday afternoon, April 26, 1971, Mr. Jerry Buley will be in your area to administer a questionnaire to you and your people. Mr. Buley is a graduate student in the Department of Communication at Michigan State University. His area of interest is communication in organizations. Mr. Buley's specific interest is in the communication which occurs between supervisor and subordinate and his questionnaire has been designed to collect data on the communication patterns which occur in that relationship. He is performing this research as part of his Master's degree requirements. The procedure for administering the questionnaire will be as follows: At some point in the day Mr. Buley and someone from Personnel will distribute the questionnaires in your area and then in a few minutes come back to collect them. The questionnaire is short and should take no more than five to seven minutes to complete. Thank you very much for your cooperation. .D_ APPENDIX C HICH 16AM STATE UNIVERSITY em mute - m .3 April. 25, 1971 Dear Employee: For as such tine as people have spent and will continue to spend talking with their supervisors. science really can tell us very little about what happens in that particular commication situation. The following questionnaire in an attempt to fill the gap in our knowledge about mi- cation between subordinates and supervisors. My one is Jerry Buley. l as a. graduate student in the Deputsent of Communication at hichigsn Stats University. fly area of interest is munication in organizations; specifically. «munication between subordinate and supervisor. The date tron this questionnaire will be part of a Master's thesis that l as writing which deals with this leper- tant topic. 1 want to emphasize that there are no "right" answers to the questions. just ygur answers. The questionnaire is going to provide as with a ”pictu'e" of munication between subordinates and their supervisors. I do not have any idea what that picture §____hould look like. Also. I went to esphasise ey need for you to be as aecurate as you can be when answering the questions. To the extent that your answers are accurate, 593; m. we can be sure to learn useful information about commieation. Thank you very such for your ties and for completing the questionnaire. Iwill be mixing a summary sheet of the results available to you “more in about a nonth for those who may be interested in the'a. _. Y) ,Tn:::e.~" 'It": 3‘q!h‘ '.," :0". 11", q‘ ’., 0 I C ' --- ‘ _ '~ w-uz: t 37.“ '..'-f. . - a :... .. (3.: ‘~ I; e .- o :Jm; n ' ' . . f1" -7-' . t . ' -, :2 :. uh: ‘ . .! ~1«.7~*'.'-.20. ‘11." 3'. ma S .' S‘w r~'= -' . 7;; ‘L a. ' e, :13 . .'. ‘1' l a... ' .e,- . : ‘|.!’ "l. 0" H!) I; ’ I: :: n}: "- ‘ ; . '. 753'.” ' Jag-1’. . '2': ' :1} . . mos ,Qi.’ . J. * .‘1‘3‘ ' .\ .-.r : - ' J. 2: ‘ :9 2:1!” mm":- ' " w ‘19! .1: 1. . quire: 'L'vm' 351311».- .si .‘mi -.5‘-- :-_. ‘3. e'. ‘ ’ e g ‘ -a ‘ ' . "‘ s ' . I e ' e . .61-51 25!." d3 4.? ’H .111 0G .a‘ w- . J . r ‘. 1‘.. 4 1 . 3 ‘ , ' . J e‘ ~ f 9'1:’.: 31,(i I ' it"J‘. -‘Te "e: I ; L ' I. I ,. fii' . .. ‘. 'I- ' ' K. mused-30:: D I ‘ .- ' 3.; .3 .A ‘ ' z- .. tn.“- ‘. ! ‘\'- ; . a .u . a ' . . , ‘e ‘13 ,s418.1. n‘h‘t’ 43 ' '25". 3 . . ‘ '._ ‘fifutjsic : - . ‘ ' “175:9”, I. 4 1.. ' ~19 a". n 9."; . ..~'; ‘. . 3' :. 1.51335 2,". nu. :3 "'15: :5 .£\’- 1w am V‘ :1 1' z. 9423.75 mm} Hf" . F:- . ~ =-.4: ~2:;1~‘ ‘ ' ... 133901372931}: . I” 3012",".‘31131. '1. ~. 'iUO“ 7.91 4.: ' .".' w? M’. UO‘.’ 03- ‘._.L"-\'i:. . .'!.-".‘ fiai.‘ . ié‘. ‘vi‘f:"-’""3'. ' i" "t g": n.:...-'e‘;i 3'. --fi, . “‘ ; . . ‘n \ J ‘U ' ' . ' ,0; ' ”fl...’ .-" '\.'.' ‘ .— ‘1' I}. ' t :xie 5 i APPENDIX D MESSAGE FLOW QUESTIONNAIRE We are interested in how messages tend to flow between supervisors and subordinates. 'Fherefore, we are asking many people here to answer several questions about their communication with their boss. We do not want your name on this questionnaire, but ‘we do want to emphasize that we need accurate information from you. Some of the questions in this questionnaire are concerned with the percentage of ‘time that you are giving or asking for information from your boss compared to the percentage of time that your boss is giving or asking for information from you. IFor example: Suppose you talk with your boss about ten times a week about television. Some of those times you might ask your boss for information about television shows and some of those times your boss might ask you for such information. Let's say 'that two times a week you ask your boss for information about television shows and ‘eight times a week your boss asks you about television shows. We could say then, ‘that of all the times when one of you is asking the other for information about ‘television shows, 20% of those times you are asking your boss and 80% of those 'times your boss is asking you. 'Thus, you would answer a typical question in this questionnaire as in the example below. EXAMPLE Of all the times when one of you is asking the other for information about television shows, what percentage of those times are you doing the asking compared to the percentage of times your boss is doing the asking? It's about. . . 95% me and 5% my‘boss 80% me and 20% my boss 65% me and 35% my boss 50% me and 50% my boss 35% me and 65% my boss 23 2096 me and 8096 my boss 5% me and 95% my boss OKAY, LET'S GET ON WITH THE QUESTIONNAIRE. 1. First, we want to know approximately how frequently you and your boss communicate. 20 or more times a day 15 - 19 times a day 10 in times a day 5 - 9 times a day l n times a day less than once a day I. c. I“ .1 . ' n I. .l. . ..o. ... .. ..J .... T . u. \ . 1 v . .- t . u . u _ .vI\ u «a .s .s a II .- l .i . ..." 4‘ .... ... l . y. .. .. I. a . A. a. C! ... \u . .l .. .l. .¢. «I . . . .l L . ..A Z ., . . s . .. . . .. . or... .. l 4 . . . E . . . ... . ... .-. I t .. . p 1.1 p... d . ~ “ . . . ..1 .n .. l>~ 4 . . .— .. .- . . 1!. ...! . ..tt, _ w .3 .. I m pl; ... . .. J .i _ ._ ... . .. ...) ..u . 0" t s . . . .v .. -|.. I... .‘I w I. .. . . ... .. . . 5 ... .. a . l W. . .l .u c A.. ,9. . . a. In.” r . ._ fl . ,. I\‘ . ... .1. ...; .0. .L .. . (a . . .. P, ,r. .... .. . ... ; . .... .. u. ,l u . L \ .. M. . . . e n. . ”k - - .1 L ., “I a 4.4 . ..ll .9 IL 1 . .... .. .... ..l a 1‘ L . ‘ ~. . .- ... . - . ... . . . .I “... ..-—...... . .fi... .3 " . .73"\.' .n l Now, think about the times when you and your boss talk about work. -2- Sometimes you ask your boss for work information and sometimes your boss asks you for work information. 2. Of all the times when one of you is asking the other for information about work, what percentage of those times are you the one doing the asking compared to the percentage of times that your boss is the one doing the asking? It's about. and and and and and and and % my 20% my 35% my 50% my 65% my 80% my 95% my boss boss boss boss boss boss boss What percentage of the times that one of you is asking for work information would you prefer that you were the one doing the asking compared to the percentage of time that he is the one doing the asking? me me me and boss boss boss boss boss boss boss or more times a day — 19 times a day lO - in times a day 5 - 9 times a day l - n times a day less than once a day How often would you prefer that one of you be asking the other for work information? I'd prefer about. 20 15 10 or more times a day - 19 times a day in times a day 9 times a day n times a day than once a day -3- iNcw think about the times when one of you is giving the other work information 'which the other has not requested. Sometimes you give your boss work information 'which he or she has not requested and sometimes your boss gives you work informa- tion which you have not requested. 6. 7. Of all the times when one of you is giving the other work information which the other has not requested, what percentage of those times are you the one doing the giving compared to the percentage of times that your boss is the one doing the giving? It's about. . . 95% me and 5% my boss 80% me and 20% my boss 65% me and 35% my boss 50% me and 50% my boss 35% me and 65% my boss 20% me and 80% my boss 5% me and 95% my boss What percentage of the times that one of you is giving the other work infor- mation which the other has not requested would you prefer to be the one doing the giving compared to the percentage of the time that your boss is the one doing the giving? I'd prefer about. 95% me and 5% my boss 80% me and 20% my boss 65% me and 35% my boss 50% me and 50% my boss 35% me and 65% my boss 20% me and 80% my boss 5% me and 95% my boss How often during the day does one of you give the other work information which the other has not requested? It's about. . 20 or more times a day 15 - 19 times a day 10 14 times a day 5 - 9 times a day l - 4 times a day less than once a day How often would you prefer that one of you be giving the other work information which the other has not requested? I'd prefer about. . . 20 or more times a day 15 - 19 times a day 10 In times a day 5 - 9 times a day l - n times a day less than once a day n Us 13. I.” ...4- " fit. L. L What LL, What \s" tnxat 10. 13. 1%. 16. .4. How would you rate your boss's communication on the following points? (Place an X in the apprOpriate space.) AVER« GOOD AGE 133R being easy to talk to about problems or complaints taking prompt action on problems or complaints listening to what I say giving me information I need to do a good job explaining clearly what he or she wants letting me know how well I am doing letting me know what he or she thinks of a proposal or suggestion I have made What is your :gc? years What is your sex? female male What is your boss's age (approximately)? years What is your boss's sex? female male How long have you worked for your present boss? years months —-4 Do you have anyone working under you? no yes -» (Do you have any supervisors working under you?) no ‘798 ___..J Because the analysis which will be performed requires that every question be completed or the questionnaire must be thrown out, would you please check back over the questionnaire to be sure that you have answered every question? Thank you very much for your time. "Immmm