Download A Critical Review of Dyads vs. Larger Groups: Unique Characteristics and Implications and more Study Guides, Projects, Research Communication in PDF only on Docsity!
DOCUMENT RESUME
ED 091 780 CS 500 694
AUTHOR Parks, Malcolm R.
TITLE Dyadic Communication from the Perspective of Small
PUB DATE Group^ Research.
NOTE Apr^74
24p.; Paper presented at the Annual Conference of the
Central 4-6, 1974) States Speech Association (Milwaukee, April
EDRS DESCRIPTORS Behavior PRICE MF-$0.75 HC-$1.50 PLUS POSTAGE
Patterns; *Communication (Thought Transfer);
*Educational Research; *Group Dynamics; Higher
Education; Human Relations; Interpersonal
IDENTIFIERS *DyadicRelationship; Communication^ *Literature^ Reviews
ABSTRACT
undertaken in an^ A attemptcritical to^ review identify^ of^ thethose^ small characteristics^ group^ literature^ 3,s
distinguish the dyad, the two-person which
Comparisons of dyads to larger groupsgroup, are made^ from in^ lAxger five categoriesgroups.
major small group variables: influence processes, affiliation of
processes, role differentiation and variability, task factors, and
communication patterns. Ten important empirical and conceptual
distinctions are analyzed, and the major limitations in the research
are suggested: noted. resultsThree implications of research ofbased the onunique largegroups nature of the dyad are
generalized to dyads; results based on dyads may not bemay generalized^ not^ be
to and larger larger groups; groups, and the in distinctive spite of several features similarities of several between dyads
components serve to distinguish the dyad as a unique communicationmajor^ dyadic
system. (Author/RB)
USOILARTMWOFHEALIN EDUCATION&WELFARE trliS Dc)(1,YETIITIMATIONISSINSSFIVIEOi^ tOOCATION MAIDEXACTLYA,WIC(P.tb4.3i0V THEPENSONOQ010C.4%+/A1,0**0.00,4HASBEENRIPII STALEDDONOTNECESSAYIEY"PPE^ NGAtPO,E,1S0fVIE1rO,tOE,01,0%S sus!Off,CtakNAT.EosiALH.,,sylf,j7L0, EDUCATIONP05,11E11,.CRPilt,Cy
Dyadic Communication From The Perspective Of Small Group Research
by Malcolm R. Parks Michigan State University
VATERIAtHASBEEN(iPANTEDBY^ roPEiRODUCE^ tetSnePr. Malcolm R. Parks fu (^) C44.11,,PtAV7OPUAY:41.10,4%1PFPATINGyiros HITHENAT()NAL, SToulE DUCTIO%OurSLE (^) EDUC,AtIUNTHEERICSYSTEMRE FuRIHEHFIEPF 0(PRE; OYS%ER PEPN,S5.,,CN .r THE COF,FUGHT
A Paper Presented at the Anlual Conference of the Central States Speech Association Milwaukee, Wisconsin April 4-6, 1974
Dyadic Communication From the Perspective Of Small Group Research I. (^) Introduction
The dyad, the two-person group, is the most frequent of all social groupings (12, 15, 25, 26). (^) It is also probably the most important of all social groups (52, 63). Ruesch (44) points out that the dyadic relationship between mother and infant is the first relationship experienced in life, at least in our cul- ture, and is essential for individual survival. (^) In his study of boys' gangs, Thrasher (58) found that the two- or three-hoy relationship was often more highly valued by members than the relationship to the larger gang as a whole. It is to the dyadic institution of marriage that most pnons turn to find emotional and physical fulfillment. (^) The dyad is the simplest of all social structures. In Simmel's words (48) it "contains the scheme, germ, .nd materials of innumerable more complex forms." The significance of the dyad, both in terms of numerical frequency and importance, can not be overstated. The present work represents an attempt to critically review the small group literature in a search for those aspects of behavior which are unique to the dyad. The focus of the paper is on those experimental studies in which group size serves as an independent variable and in which some major small group variable functions as the dependent variable. (^) In addition to examining empirical results, the paper will note several basic, conceptual or definitional distinctions between dyads and larger groups
The author would like to thank Dr. W.W. Wimot for his constant encourage- ment. Appreciation is also extended to Dr. Gerald R. Miller, Dr. L. Edna Rogers, John Marlier, and Mark Miller for their patient and helpful comments on the drafts of this paper.
Group size will be examined in terms of five selected classes of variables:
- Influence processes--including leadership, coalition formation, power, and authority; 2) Affiliation processes--including the issues of cohension and con- formity; 3) Role differentiation and variability; 4) Task factors-- including the topics of performance and problem-solving; and, 5) Communication patterns-- including communication networks, feedback processes, and the distribution of participation. (^) An examination of several general works on small group communica- tion and interaction (11, 22, 23, 39, 45) indicates that these categories of variables are among the most frequently discussed in the small group literature. In examining the relationship of various sized groups (including the dyad) to these variables, the central question will be: "Do Lyads function in any unique ways in terms of these variables?" The paper procedes by: 1) outlining a bas!c perspective for examining small groups in general; 2) examining each of the above classes of variables in turn; and, 3) attempting to evaluate the question of dyadic uniqueness both in terms of the quality of research and the results and implications of that research. II. A Perspective: Group Site and Relationships Small groups are more than aggregations of indlviduals. (^) They represent aggregations of relationships and interrelationships. An essential first-step in the examination of small groups is in terms of the number of possible rela- tionships that exist for any given number of persons (5). Increments in group size imply that more persons are available for acquiring and processing informa- tion and for developing more complex sets of interpersonal relationships (13). Probably the first investigator to formulate this view in mathematical terms was Bossard (9). This formula for determining the number of potential relationships
the group than does the average group member. (11). (^) However, in dyads it may be more difficult to identify the relative contributions made by meAbers. (^) Simmel (48) suggests that the delegation of responsibilities and obligations which character- izes all larger groups is less defined, and perhaps absent, in the dyad. (^) As a result, clear indicators of influence may not be present in the dyad. The exist- ing research appears to support this view. In examining groups of two to twelve members, Bass and Norton (7) report that obsertler-judges were least able to isolate behaviors indicative of leadership in dyads. (^) Hare (23) points out that in groups of less than sir. nembers, observers have great difficulty identifying leaders. It appears that in dyads influence abilities are less attributable to one person than in larger groups. Since the dyad is composed of a single relationship, subsystems or coali- tions are not ipossible by definition (41). (^) An additional member creates a fundamental conceptual and behavioral change in the social unit (8, 46, 47, 48, 60). (^) Triadic relations are typicall'y characterized by a "two-against-one" coalition (10, 17). (^) The addition of new members to a triad, however, does not yield the same kind of alteration in transaction as the addition of one member to a dyad (35, 47, 60). Mills (36) found that over time triads represented a fluid set of power relations which usually resolved itself by the development of a supporting coalition of two stronger individuals against a weaker third. Taylor's (55) research indicates that triads are less stable over time than Ids. In triads and larger groups, subsystems and coalitions become central aspects of the overall group structure. This is a funcamental structural and conceptual distinction between dyads and Larger groups. The fact that dyads lack subsystems or coalitions profoundly influences the use of power and authority. (^) The withdrawal of a single member ends the life of
the dyadic group--but not of larger groups. Simmel (48) succinctly observes! "A dyad depends...on each of its two elements...for its life, it needs both, but for its death, only one." This characteristic serves to distinguish the dyad from other groups in several ways. First, in terms of decision-making, it means that there is no "public opinion," audience, or majority pressure in a dyad. (4, 23, 48). There is no majority except unanimity (6). The dyadic group, unlike larger groups, is maintained only by the mutual willingness of all its members. Second, this implies that power, or the ability to exert influence (11), is equally shared within the dyad. Since the continued existence of the dyad relies on the mutual willingness of all its members, each member as the power to destroy the group. (^) The demands of one member can be blunted by the veto or withdrawal of the other (6, 48). While participants may be differentially influential, an attempt to influence the dyadic group by one person will ultimately be only as successful as the other allows it to be. In^ larger^ groups,^ however,^ this^ complete interdependency does not exist. An individual may successfully.influence the group without necessarily successfully influencing all its members. (^) Thus, as group size increases, the consequences of offending, countering, or alienating a given member decrease in severity (49). In larger groups, individuals can exercise power and authority with much less attention to the consequences for other individuals (23). Third, dyads may be unique in terms of their characteristic exchange rela- tions. Because of the intardependency of outcomes, individuals in dyads can not exercise 'their power as freely as they might in larger groups (56). As^ a^ result, members of dyads may more frequently orient their exchanges with greater deference to the other person. In larger groups, however, participants encounter more indirect and ambiguous exchange comparisons. As^ a^ result,^ they^ may^ not^ need^ to
IV. (^) Affiliation Processes
Affiliation processes function to bind individuals more closely together in groups. (^) Cohesion and conformity are the affiliation variables most frequently discussed in small group literature. Most frequently, cohesion is conceptualized as the desire of members to remain in the group (11). (^) Little direct research relating group size to cohesion has been conducted--at least with respect to dyads. Some tangential evidence, however, emerges from researlh on group performance and problem solving. Hackman and Vidmar (21) observed groups of two to seven members in a problem solving sit- uation. In comparfoon to members of the larger groups, members of dyads were found to be: 1) more satisfied with their performance; and, 2) more intensely involved or "invested" in the performance process. (^) In comparison with groups of three, five, and eight, Frank and Anderson (16) report that members of dyads were the most satisfied in terms of liking of the other members. In a study dealing only with dyads, Murdoch and Rosen found a strong tendency toward the formation of normative agreements to protect against disruptions (37). In their review of small group literature relating to size, Thomas and Fink (57) suggest that "group size is an important factor in determining the amount of yielding to conformity pressures." (^) However, at least with small groups, this conclusion is not well supported. (^) In an early study by Asch (3) a negative linear relationship was found between group size and the amount of yielding. The least amount of yielding was reported in dyads. However Goldberg (19) found no dif- ference in the amount of conforming behavior from groups of two and four in an exercise involving the judgment of intelligence from photographs. Similar results were obtained by Kidd (29) in a study of groups of two, four, and six members who made judgments of flicker frequency. On the other hand, Rosenberg (43) examined
conformity pressures and yielding in a length-judging exercise with groups of two to five. (^) His results indicated a curvilinear relationship between group size and conformity with the most yielding occuring in groups of four. Undoubtably the heterogeneity of judgment tasks and the limited range of sizes examined has con- tributed to the inconclusiveness of these results. In any case, the relationship between group size and conformity is not a clear one--especially with respect to comparisons of dyads to larger groups. It must be stressed that the above findings relate primarily to ad hoc experimental groups. (^) In "real" groups, different dynamics might well apply. Simmel (47) and Bales, Hare, and Borgatta (5) suggest that transactions in dyads will tend to conform to an area within which both participants can agree. (^) That is, the conformity variable functions to ensure that transactions remain with bounds of agreement or consensus. This limitation appears to be more clearly evident in dyads than in larger groups (47).
V. Role Differentiation and Variability Simmel (47) and Becker and Useem (8) have hypothesized that dyads will experience a greater variability or individuality than largergroups. (^) The invest- igations of Bales and Borgatta (4) appear to provide some support for this view. In a comparison with groups of three to seven members, dyads showed a greater variability in nearly all categories of the Interaction Process Analysis scheme. Dyads appear to be more varied in terms of the types of transactions they engage in than do larger groups. Becker and Useem (8) contend that the interdependency of the dyad allows the development of idiosyncratic modes of behavior. This variability may result from the development of complementary roles in dyadic transactions. Bales, Hare, and Borgatta (5) report:
type of task. (^) Frank and Anderson (16) draw a distinction between "conjunctive" and "disjunctive" tasks. (^) Conjunctive tasks are those whose solution is a function of the weakest group members; while disjunctive tasksare those whose solution is a function of the best group member. (^) In comparing dyads with groups of three, five, and eight, these investigators discovered that: (^) 1) dyads tended to be the most satisfied with the group's performance when the task was conjunctive; but least satisfied when the task was disjunctive; 2) dyads tended to like disjunc- tive tasks less and conjunctive tasks more than other groups; and, 3) dyads tended to be the most satisfied in terms of liking of other members--regardless of task type. In examining groups of two to seven members, Slater (49) reports that members of dyads were far more likely than members of larger groups to feel that the group was too small to carry out the assigned tasks in an optimum fashion. Hackman and Vidmar (21) report a similar finding. (^) These latter investigators, do note though, that dyads experienced fewer difficulties in organizing and co-ordinating their problem-solving activities than larger groups. While dyads appear to out perPorm individuals both in terms of the number of problems solved and the adequacy of those solutions (24, 30), there is some evi- dence to indicate that dyads take more time per problem than individuals (30). On the other hand, dyads were faster than four person groups in an experiment involving finding the correct response to the "twenty-questions" game (54). The evidence comparing dyads to larger groups has been inconclusive--although the tendency is for dyads to. perform less effectively or efficiently than larger groups. Porter (40) found-the performance of dyads inferior to that of groups of four or eight. (^) In a study of groups ranging from_two to six members, Ziller-(64) found that dyads neither performed particularly well or particularly poorly in
comparison to larger groups in a dot judging exercise. (^) However, there is limited evidence to indicate that dyads generate more ideas in their transactions than do larger groups. (^) Gibb (18) found that "the number of ideas produced was found to increase in a negatively accelerated function of size of group." (^) Lorge E Solomon (31) found that dyc.ds performed somewhat better than groups of five in terms of finding the solution to the Tartaglia transportation problem regardleso of whether members had previous experience with the problem. Lorge and Solomon (33) also have developed two mathematical models describing the probability that a group of a given size will find the solution. (^) These models were derived from data concerning mathematical puzzles involving the transporta- tion of objects or persons. (^) One model applies to the situation in which solution is a function of the abilities of one or a ver few members; while the other applies to the situation in which solution is a function of the pooled abilities of all members. (^) In both models, increases in group size bring an exponential increase in the probabilit:* of solution. (^) This would suggest that dyads would be less likely than other groups to solve such problems. Davis (13) contends that these models probably over-predict group performance. How this would relate to the relative effectiveness of dyads in comparison to larger groups in unknown. In summary; it appears that dyads: 1) are likely to be more satisfied with group members and more involved than larger groups in general; 2) will be either more or less satisfied with their performance than other groups depending on the nature of the task; and, 3) are somewhat inferior to larger groups in terms of the quality and speed of solution. (^) Such conclusions, of course, must be tempered by the generally low degree of comparability among investigations in this area and-by the large number of confusing or contradictory results.
There is evidence that feedback in dyads is characterized by a matching of responses. Messages at one level of intimacy tend to elicit feedback at that same level of intimacy (14, 53, 59). It is not clear, however, that this is unique to dyads. Worthy, Gary and Kahn (62) obtained a similar result in four- person groups. Simmel (48) has observed that the dyad may be unique in toms of forcing individual members to more actively participate. This^ hypothesis^ is^ consistent with the results of Hackman and Vidmar (21) indicating that members of dyads tended to become more intensely involved in a problem-solving exercise than did' members of larger groups. Moreover,^ as^ group^ size^ increases^ the^ amount^ of^ par- ticipation by any given number decreases (
VIII. Evaluation Obviously, the importance of the differences between dyads and larger groups depends upon the nature and quality of the research upon which such distinctions are based. Unfortunately,^ the^ small^ grollip^ literature^ suffers^ from^ a^ number^ of conceptual and methodological faults. McGrath^ and^ Altman^ (34)^ describe^ it^ as^ an "...unabated race toward more empirical knowledge and the comparative absence of integrating theory..." Golembiewski^ (20)^ cites^ "substantial^ limitations^ on^ the reliability and validity of existing results" as a stumbling, block to the develop- ment of theory. Rogers^ (41)^ notes^ the^ lack^ of^ a^ "unifying^ descriptive^ frame." Since these weaknesses detract from the significance of the distinctions drawn above, it is necessary we explore them in somewhat greater detail. Three^ general weaknesses of the research seem important to note. First, small group research has been characterized by a lack of constnaus among investigators at to the conceptual or operationalmeanings^ of^ major
constructs. (^) Investigators have developed new terminology rather than clarifying existing constructs. As a result, a large number of conceptually similar con- structs remain inadequately related in the research (20, 23, 41). An example of this would be the previously noted difficulties in relating role differentiation and leadership. (^) Concepts and operationalizations are frequently vague and ill- specified (5, 50). (^) The heterogeneity of approaches has precluded the development
of cumulative results which would ultimately lead to theoretical advances (20, 57). The relative ease with which small group research may often be conducted (50) and the limited use of multivariate techniques (S7) have only served to exacerbate this general difficulty. Second, the heavy reliance on ad hoc experimental groups has raised ques- tions as to: (^) 1) the extent to which such aggregations can be considered real groups at all; and, 2) the extent to which one can generalize from experimental results to existing, real, or free-forming groups. In many experiments it is doubtful that the collectivities ever really function as a "group" (20). Kelley and Thibaut (27), for example, have pointed to the generally low levels of involve- ment and activity observed in many experimental problem-solving groups. (^) The fact that most experimental groups have no prior history of interaction as a group and no expectation of any future interaction together may curtail the development of important features in the dynamics of interaction (32, 41). As a result, the validity of generalizations of experimental findings to groups outside the labor- atory is frequently questionable (11, 20). Since most experiments have the "double artificiality" of the laboratory setting and the use of temporary groups, our knowledge of "real" groups is severely restricted (1, 2). Third, the variable of group size has not received systematic treatment (20). Studies comparing different group sites often cite little or no theoretical or
- (^) Variability of behavior is greatest in dyads.
- (^) Communication networks are not possible in dyads.
- (^) Feedback processes appear to involve greater self- disclosure and intimacy in dyads.
- (^) Amount of participation by any given member is greatest in dyads. In terms of several variables which appear central to small group analysis, then, the dyad seems to emerge as a distinct and unique social unit. (^) As Zimit and Schneider (65) conclude: (^) "the dyad is the group size which shows the most discrepancy from the other group sizes with respect to the behavior of the members." (^) InTheir discussion of size, Cartwright and Zander (11) comment: ...on the basis of available findings one would have more reasons to draw a boundary between groups of two and groups of three than at any other critical size. The investigator or theorist must interpret the implications of such a boundary with care. (^) There are two important limitations for theory construction on small groups in general. (^) First, these distinctions imply that the investi- gator must account for the potential uniqueness of the dyad. (^) Many generaliza- tions based on larger groups might not apply to dyads. Second, these distinctions between dyads and larger groups imply that the investigator must be wary of generalizing results based on dyads to larger groups. (^) Frequently, investigators assume that a knowledge of dyads will provide a basis for understanding other social groupings. Thibaut and Kelley's (56) viewpoint is typical in this regard: ...we assume that if we can achieve a clear understanding of the dyad we can subsequently extend our understanding to encompass the problem of larger and more complex social relationships. While an understanding of the dyad may be, and,probably is, essential to the explanation of larger and more complex social phenomena, one must not confuse
an understandirg of the dyad with an understanding of larger groups. They are not the same. (^) As a result, the generslization of dyadic findings to larger groups is frequently inappropriate (51). This is an important point since large amounts of social psychological research are conducted with dyads and then gener- alized to larger social contexts. In a broader view, the dyad emerges as a unique system. Certainly there are similarities between several of the componenta of dyadic sy.5.tems and larger group systems. (^) However (^) taken as a whole the dyad represents a unique system as a result of the variation introduced by its distinctive components or functions. At this level, the dyadic system is quite different from the systems characterizing larger groups. The dyad is not simply another small group. The frequency with which dyadic systems occur and their individual and cultural importance under- score the necessity for further specification of the structure, functions, and operating characteristics of dyadic systems.
