Interpersonal Attraction in Exchange and Communal ... · Interpersonal Attraction in Exchange and...

Journal of Personality and Social Psychology1979, Vol. 37, No. 1, 12-24

Interpersonal Attraction in Exchange andCommunal Relationships

Margaret S. ClarkCarnegie-Mellon University

Judson MillsUniversity of Maryland

Communal relationships, in which the giving of a benefit in response to a needfor the benefit is appropriate, are distinguished from exchange relationships, inwhich the giving of a benefit in response to the receipt of a benefit is appropriate.Based on this distinction, it was hypothesized that the receipt of a benefit afterthe person has been benefited leads to greater attraction when an exchange re-lationship is preferred and decreases attraction when a communal relationship isdesired. These hypotheses were supported in Experiment 1, which used malesubjects. Experiment 2, which used a different manipulation of exchange versuscommunal relationships and female subjects, supported the hypotheses that (a)a request for a benefit after the person is aided by the other leads to greaterattraction when an exchange relationship is expected and decreases attractionwhen a communal relationship is expected, and (b) a request for a benefit in theabsence of prior aid from the other decreases attraction when an exchange rela-tionship is expected.

This research is concerned with how theeffects of receiving a benefit and a requestfor a benefit differ depending on the type ofrelationship one has with the other person.Two kinds of relationships in which personsgive benefits to one another are distinguished,exchange relationships and communal rela-tionships. The stimulus for this distinction wasErving Goffman's (1961, pp. 275-276) dif-ferentiation between social and economicexchange.

In the present theorizing, the term exchangerelationship is used in place of Goffman'sterm economic exchange because many of thebenefits that people give and receive do notinvolve money or things for which a monetaryvalue can be calculated. A benefit can be any-thing a person can choose to give to anotherperson that is of use to the person receivingit.1

In an exchange relationship, members as-

This research was supported by a grant from theNational Science Foundation.

Requests for reprints should be sent to MargaretS. Clark, Department of Psychology, Carnegie-MellonUniversity, Schenley Park, Pittsburgh, Pennsylvania15213.

sume that benefits are given with the expec-tation of receiving a benefit in return. Thereceipt of a benefit incurs a debt or obligationto return a comparable benefit. Each personis concerned with how much he or she re-ceives in exchange for benefiting the otherand how much is owed the other for the bene-fits received.

Since all relationships in which persons giveand receive benefits are social, another termis needed to describe relationships in whicheach person has a concern for the welfare ofthe other. The term communal seems to bethe most appropriate. The typical relationshipbetween family members exemplifies this type

1 Benefits are not the same as rewards, when theterm rewards refers to "the pleasures, satisfactions,and gratifications the person enjoys" (Thibaut &Kelley, 1959, p. 12). The receipt of a benefit usuallyconstitutes a reward, however rewards occur for rea-sons other than the receipt of a benefit. For exam-ple, the rewards that a parent receives from a new-born infant would not fall within the definition ofa benefit, since the infant does not choose to givethem to the parent. The present theorizing is notconcerned with "dependent" relationships in whichone person receives benefits from another but doesnot give benefits to the other.

Copyright 1979 by the American Psychological Association, Inc. 0022-3514/79/3701-0012$00.75

12

INTERPERSONAL ATTRACTION 13

of relationship. Although it might appear toan observer that there is an exchange of bene-fits in communal relationships, the rules con-cerning giving and receiving benefits are dif-ferent than in exchange relationships.

Members of a communal relationship as-sume that each is concerned about the welfareof the other. They have a positive attitudetoward benefiting the other when a need forthe benefit exists. They follow what Pruitt(1972) has labeled "the norm of mutual re-sponsiveness." This rule may create what ap-pears to an observer to be an exchange ofbenefits, but it is distinct from the rule thatgoverns exchange relationships whereby thereceipt of a benefit must be reciprocated bythe giving of a comparable benefit. The rulesconcerning the giving and receiving of bene-fits are what distinguish communal and ex-change relationships, rather than the specificbenefits that are given and received.

From the perspective of the participants ina communal relationship, the benefits givenand received are not part of an exchange.The attribution of motivation for the givingof benefits is different from that in an ex-change relationship. In a communal relation-ship, the receipt of a benefit does not createa specific debt or obligation to return a com-parable benefit, nor does it alter the generalobligation that the members have to aid theother when the other has a need. In a com-munal relationship, the idea that a benefit isgiven in response to a benefit that was re-ceived is compromising, because it calls intoquestion the assumption that each memberresponds to the needs of the other.

Experiment 1

The first study reported here was based onthe assumption, similar to that made byKiesler (1966) in her study of the effect ofperceived role requirements on reactions tofavor-doing, that the giving of a benefit willdecrease attraction if it is inappropriate forthe type of relationship one has with theother. A benefit given in response to a benefitreceived in the past or expected in the futureis appropriate in an exchange relationship butis inappropriate in a communal relationship.

A benefit given specifically because it ful-fills a need is appropriate in a communal rela-tionship but not in an exchange relationship.

If two people have an exchange relationshipand one person benefits the other, it is ap-propriate for the other to give the person acomparable benefit. The receipt of a benefitunder these circumstances should lead togreater attraction. On the other hand, if twopeople have a communal relationship and oneperson benefits the other, it is inappropriatefor the other to give the person a comparablebenefit, since it leaves the impression that thebenefit was given in response to the benefitreceived previously. The other is treating therelationship in terms of exchange, which isinappropriate in a communal relationship.

When a communal relationship does.not yetexist but is desired, the receipt of a benefitshould have the same effect as when a com-munal relationship is assumed to exist. Abenefit from the other after the other hasbeen benefited should reduce attraction ifthere is a desire for a communal relationshipwith the other. If an exchange relationshipis preferred, the receipt of a benefit after theother is benefited should result in greater at-traction. Experiment 1 was conducted to testthese hypotheses.

The predictions concerning communal rela-tionships might seem contrary to what wouldbe expected from equity theory (Adams,1963). On the basis of equity theory, onemight expect that a benefit from another fol-lowing aid to that other would increase likingin any relationship, because it would reduceinequity. However, the predictions are notinconsistent with a recent discussion of equitytheory (Walster, Walster, & Berscheid, 1978).According to Walster, Walster, and Berscheid:

Another characteristic of intimate relationships,which may add complexity, is that intimates, throughidentification with and empathy for their partners,come to define themselves as a unit; as one couple.They see themselves not merely as individuals in-teracting with others, but also as part of a partner-ship, interacting with other individuals, partnerships,and groups. This characteristic may have a dramaticimpact on intimates' perceptions of what is and isnot equitable, (pp. 152-153)

In Experiment 1 the desire for a communalrelationship was manipulated by using un-

14 MARGARET S. CLARK AND JUDSON MILLS

married males as the subjects and having thepart of the other played by an attractivewoman, who was described as either marriedor unmarried. It was assumed that peopledesire communal relationships with attractiveothers, but only with those available for suchrelationships. It was further assumed that theunmarried woman would be considered avail-able for a communal relationship, whereasthe married woman would not. Thus, it wasassumed that the male subjects would desirea communal relationship with the attractive,unmarried woman but would prefer an ex-change relationship with the attractive, mar-ried woman.

Method

Overview. Under the guise of a study of taskperformance, unmarried male college students workedon a task while a television monitor showed an at-tractive woman working on a similar task in anotherroom. When the subject completed the task, he wasawarded 1 point toward extra credit for finishingon time and given the opportunity to send someof his excess materials to the other, who supposedlyhad a more difficult task. Following receipt of theaid, the other completed her task and was awarded4 points. Some of the subjects received a note fromthe other that thanked them (no-benefit conditions),whereas some received a note from the other thatthanked them and gave them one of her points(benefit conditions). After receiving the note, thesubjects were given information indicating that theother was either married (exchange conditions) orunmarried (communal conditions). Shortly thereafter,liking for the other and expectations concerning afuture discussion with the other were assessed.

Subjects. The subjects were 96 unmarried, malestudents in introductory psychology courses whoparticipated in order to earn extra course credit.They were randomly assigned to one of the fourexperimental conditions: exchange-benefit, exchange-no benefit, communal-benefit, communal-no benefit.

Procedure. Upon arriving for the experiment, thesubject was greeted by the experimenter and toldthat another subject had already arrived and wasin an adjacent room. The subject was seated so thathe could see a video monitor that showed an at-tractive female. The experimenter mentioned thatthe person appearing on the monitor was the othersubject, Tricia. The experimenter told the subjectthat before starting she would explain some thingsabout the studies in which the subject would be par-ticipating, which she had already mentioned to Tricia.

The experimenter said that the first study wasactually one of two she would ask them to partici-pate in that day. Each of the two studies took lessthan half an hour. Although the studies were un-related, she was asking people to participate in both

of them during one session. The first study involvedhaving both subjects work on a vocabulary task.She suspected that people's approaches to solvingthis task varied when certain conditions werechanged. In the condition to which he and Triciahad been randomly assigned, they would be able tosee each other over closed-circuit television but notbe able to talk to each other directly. To enhancecredibility, there was a portable television camerain the room pointing at the subject. Through the useof videotape, what appeared on the monitor was thesame for every subject. When the subject asked whyhe was watching the other person on the monitor,which typically happened, he was told that in thepast it had been found that when people workedseparately on these tasks in the same room theirperformance was often affected by the presence ofthe other person. This might have happened becausepeople could talk to one another or because theycould see one another, and the experiment was de-signed to separate those variables.

The experimenter explained that the second studywould be quite different from the first, since it wouldinvolve much more contact with the other person.She would bring both participants into one room andask them to talk over things that they had in com-,mon. She was interested in the way in which hold-ing common interests helped people to get to knowone another. The experimenter mentioned that inthe past, people who had participated had some-times gotten to know each other quite well.

Vocabulary task. The experimenter next pointedto a batch of letters printed on small cards in frontof the subject and said that his task in the firststudy was to form 10 different four-letter wordsfrom the letters. She went on to say that there weretwo versions of this task and one was more difficultthan the other. She had flipped a coin to determinewhich task each of them would be working on, andthe subject would have the easier one, while Triciawould have the more difficult one. The subject'stask was easier because he had 55 letters rangingfrom A to Z, whereas Tricia had only 45 lettersranging from A to L. The subject was to time him-self on the task with a stopwatch. Although he andTricia were not permitted to speak to each otherdirectly, one type of communication was permitted:They would be allowed to send and request letters.Simple forms were available for this purpose. Theexperimenter would come into the room from timeto time to see if the subject had any messages hewanted to send and would deliver them for him.

The experimenter explained that time was a factorin the first study and that an individual's motiva-tion to do well on a task would obviously affectthe time taken to complete the task. Therefore, shewas taking precautions to insure that motivationto do well was kept at a high level. She wouldaward points toward a possible extra credit forfinishing the task in a minimum amount of time.The subjects were not told exactly what that amountof time would be or how the conversion to creditwould be made, but just to finish as quickly as theycould. The experimenter mentioned that the award-


ing of points to maintain motivation would obviouslynot be necessary in the second study.

The subject was told to start his stopwatch, beginworking, and stop the watch when he had finished.The experimenter picked up another stopwatch andleft, saying she would give Tricia the watch andstart her on her task. The 55 letters that the sub-ject had to work with allowed him to complete thetask within 10 minutes. When approximately 10minutes had elapsed, the experimenter returned andasked if the subject had any letters he wanted tosend to Tricia. All subjects gave the experimentersome letters for Tricia. At this point, the experimenterlooked at the subject's stopwatch and told him thathe had finished the task in time to get 1 point to-ward extra credit. As she did so, she also filled outa form indicating the time the subject had taken tocomplete the task and that he had earned 1 point.She explained that if Tricia finished her task intime she would earn 4 points, since her task wasmore difficult. The experimenter went on to say thatsince she allowed participants in the study to sendand request letters, she also allowed them to sharepoints they earned. Thus, Tricia could send the sub-ject some of her points if she wanted to do so. Theexperimenter left the room saying she would giveTricia the letters. Tricia continued to work on hertask for about 5 minutes and then finished. The ex-perimenter handed Tricia a form similar to thatgiven the subject earlier. Tricia smiled, wrote a noteon a slip of paper, folded it, and gave it to the ex-perimenter.

Benefit manipulation. Within a few moments,the experimenter returned to the subject's roomand turned off the monitor. She mentioned thatTricia had completed her task within the necessarytime and had received 4 points. She handed thesubject a folded note that she said Tricia hadasked be given to him. In the no benefit condition,the note said, "Thanks for sending the letters." Inthe benefit conditions, the note said, "Thanks forsending the letters. The experimenter said it wouldbe OK to give you one of my points. She said shewould add it onto the points you've already earnedbefore the end of today's session." Which messagethe note contained was unknown to the experi-menter at the time she handed it to the subject.This was accomplished by having the experimenterpick the note out of a container of folded notes ofboth types.

Relationship manipulation. The experimenter toldthe subject that there was one more thing to bedone before getting on to the next study. She saidshe was going to give Tricia some questionnaires tofill out and would then get some more forms forthe subject. In the exchange conditions, the experi-menter said:

Tricia is anxious to get on to the next part of thestudy, since she thinks it will be interesting. Herhusband is coming to pick her up in about halfan hour and she wants to finish before then.

In the communal conditions, she said:

Tricia is anxious to get on to the next part ofthe study, since she thinks it will be interesting.She's new at the university and doesn't knowmany people. She has to be at the administrationbuilding in about half an hour and she wants tofinish before then.

Dependent measures. The experimenter then leftthe room for approximately 5 minutes. When shereappeared she brought two forms, mentioning thatthese were the forms she had told the subject about.She reminded the subject that the second study in-volved having the participants talk over thingsthey had in common with each other. She said thatbefore starting it was necessary to get some idea ofwhat their expectations were in order to controlfor them, since they would vary from person toperson. The subject was asked to fill out a formindicating what he expected the interaction wouldbe like and, in addition, another form indicatingwhat his first impressions of the other person were.The experimenter said that these forms would bekept completely confidential and left the room whilethey were filled out.

The first-impressions form, which was given tothe subject on top of the form concerning expecta-tions about the discussion, asked him to rate howwell 11 traits applied to the other, on a scale from 0(extremely inappropriate) to 20 (extremely appropri-ate). The traits were considerate, friendly, insincere,intelligent, irritating, kind, open-minded, sympa-thetic, understanding, unpleasant, and warm. Thesubject was also asked to indicate his degree of lik-ing for the other, on a scale from 0 (dislike verymuch) to 20 (like very much). The other formasked the subject to indicate how friendly, spon-taneous, relaxed, enjoyable, and smooth he expectedthe discussion to be, on scales from 0 to 20.

Suspicion check. After the subject had completedboth forms, the experimenter casually mentionedthat there was something more to the study andasked whether the subject had any idea what itmight be. The responses of eight persons indicatedsuspicion of the instructions, and they were not in-cluded as subjects. Four persons thought the experi-ment was designed to test reactions to the note,one thought the points might have something to dowith the ratings, two thought Tricia was a part ofthe experiment, and one questioned whether Triciawas actually married. Four of these persons wererun under the exchange-benefit condition, two underthe exchange-no-benefit condition, one under thecommunal-benefit condition, and one under thecommunal-no-benefit condition. In addition, 12other persons were not included as subjects. Sixcould not finish the task within the 10 minutes al-lowed, one did not read the note before filling outthe forms, two failed to follow instructions whenfilling out the forms, one discovered the concealedvideotape recorder beneath the monitor, and twowere married.


Finally, the true purpose of the experiment wasfully explained, and the subject promised not todiscuss it with anyone,

Results

A measure of liking for the other was cal-culated by summing the scores for each ofthe 11 traits and the direct measure of likingon the impressions questionnaire. The scoresfor the favorable traits and for the directrating of liking were the same as the subject'sratings. The scores for the unfavorable traitswere obtained by subtracting the subject'sratings for those characteristics from 20.The means for the experimental conditionsfor the measure of liking are presented inTable 1.

From the hypotheses, it would be expectedthat the scores on the measure of likingwould be greater in the exchange-benefit con-dition than in the exchange-no-benefit con-dition and would be lower in the communal-benefit condition than in the communal-no-benefit condition. From Table 1 it can beseen that the results are in line with the pre-dictions. An analysis of variance of the mea-sure of liking revealed that the interactionbetween type of relationship and benefit wassignificant, /?(!, 92) = 8.35, p < .01. Neitherof the main effects approached significance.A planned comparison indicated that thedifference between the exchange-benefit con-dition and the exchange-no-benefit conditionwas significant, /?(!, 92) = 4.17, p < .05. Asecond planned comparison indicated thatthe difference between the communal-bene-fit condition and the communal-no-benefitcondition was also significant, F(l, 92) =4.37, p< .05.

A measure of anticipated pleasantness ofthe discussion was calculated by summingthe ratings for the questions concerning howfriendly, spontaneous, relaxed, enjoyable,and smooth the subjects expected the discus-sion to be. As expected, the means for thismeasure were greater for the exchange-bene-fit condition than for the exchange-no-bene-fit condition and were lower for the com-munal-benefit condition than for the com-munal-no-benefit condition. However, the

Table 1Means for the Measure of Liking inExperiment /

Benefit from the other

Relationship Benefit No benefit

ExchangeCommunal

193177

176194

Note. The higher the score, the greater the liking.Scores could range from 0 to 240. re = 24 per cell.

interaction between the type of relationshipand benefit was not significant. The main ef-fects were also not significant.

Discussion

The results of Experiment 1 provide sup-port for the hypothesis that when a com-munal relationship is desired, a benefit fol-lowing prior aid decreases attraction. Whenthe attractive woman they had aided wasunmarried (communal conditions), the un-married male subjects liked her less whenshe gave them a benefit than when she didnot. When the attractive woman who wasaided turned out to be married (exchangeconditions), she was liked more when shegave the subject a benefit than when shedid not. The results for the married womanare consistent with the hypothesis that thereceipt of a benefit after the other is bene-fited leads to greater attraction when anexchange relationship is preferred.

It might be thought that the lower likingof the unmarried woman in the benefit condi-tion than in the no-benefit condition couldbe explained by a difference in the anticipa-tion of future interaction, a variable whichhas been shown to influence liking in previ-ous research (Barley & Berscheid, 1967;Mirels & Mills, 1964). Such an interpreta-tion would make the assumption that therepayment by the unmarried woman withwhom the subject wished to interact sug-gested that she did not wish to interact withhim. However, this assumption is reasonableonly if the future interaction is of the kindreferred to here as a communal relationship.


If the future interaction involved an explicitexchange of benefits, there is no reason therepayment would have suggested that the un-married women did not wish to interact withthe subject. Since an interpretation in termsof the anticipation of future interaction re-quires a distinction between different typesof interaction similar to the distinction be-tween communal and exchange relationships,it is not an alternative explanation but es-sentially the same interpretation in some-what different language.

A different interpretation might be sug-gested that has to do with the role relation-ships of males and females. It might be ar-gued that the male subjects subscribed to a"traditional" rule that males should givegifts to females, who should gracefully ac-cept them and not attempt to repay theirbenefactor, and the unmarried woman whogave them a point violated this rule. Such aninterpretation would assume that the malesdid not apply the same rule to their rela-tionships with a woman who is married,which does not seem consistent with tradi-tional values concerning the role of menvis-a-vis women. If the interpretation is re-stricted to the relation of men and womenin romantic relationships, then it is not analternative explanation, since romantic re-lationships are communal relationships.

That the desire for a communal relation-ship was induced by creating a situation inwhich there was a possibility of a romanticrelationship with an attractive member ofthe opposite sex was not fortuitous. A ro-mantic relationship that might lead to the de-velopment of a family relationship throughmarriage is a particularly appropriate situa-tion for the study of communal relation-ships, since relationships between familymembers are the most typical kind of com-munal relationships. However, the distinc-tion between communal and exchange rela-tionships is not restricted to romantic rela-tionships with members of the opposite sex.The same effect should occur in situationsin which a communal relationship, such asfriendship, is desired or expected with amember of the same sex.

It was assumed not only that the other

was perceived as available for a communalrelationship in the communal conditions butalso that she was regarded as an attractivepartner for such a relationship. If the other isunattractive, a communal relationship withher should not be desired even if she is avail-able for such a relationship. People do notdesire communal relationships with peoplethey dislike. An exchange relationship shouldbe preferred with an unattractive other, andthus a benefit from such a person after heor she has been aided should lead to greaterattraction.

Since the effect found in the first studyinvolves the assumption that the benefit thatthe person received from the other is per-ceived as a response to the previous benefitthat the other received, it should not occurif the other had not received something ofvalue from the person. The receipt of a bene-fit when the other has not been aided previ-ously should lead to greater liking when aGommunal relationship is expected or de-sired. The rule in communal relationships isto respond to a need rather than to recipro-cate benefits. The giving of a benefit whenno prior help has been received is appropri-ate for a communal relationship if there is ormight be a need for the benefit.

Some evidence that the receipt of a bene-fit when the other is not aided previouslyproduces greater liking when there is a com-munal relationship than when there is not isprovided by the study by Kiesler (1966).She found that a partner on a cooperativetask was liked more when he shared his win-nings with the subject than when he did notshare, whereas an opponent on a competitivetask was liked about the same when heshared and did not share. Partnership on acooperative task should create the expecta-tion of a communal relationship, and inKiesler's study the subject always lost, sothe partner did not receive any aid fromthe subject prior to benefiting the subject.However, the possibility that the subject'slosing could have been construed as a kindof aid for his opponent in the competitiveconditions complicates the comparison ofthe cooperative and competitive conditionsof Kiesler's study.


Experiment 2

The distinction between communal andexchange relationships also has implicationsfor reactions to a request for a benefit. If itis true that in an exchange relationship anybenefit given by one member to the othercreates a debt or obligation to return a com-parable benefit, a request for a benefit fromanother after one has been given aid by thatother creates an opportunity to repay thedebt. Thus, such a request following aidshould be appropriate in an exchange rela-tionship. Since it provides an opportunity toeliminate any tension caused by the presenceof the debt, it should increase liking for theother.

The idea that the recipient of a benefitwill like his or her benefactor more if he orshe can return the benefit has been expressedbefore (Mauss, 1954). Several studies haveshown that recipients of benefits like thedonor more if they are able to repay thebenefit than if they are not able to repay,whether the opportunity is provided by thedonor's specifically requesting that the otherrepay the benefit (Gergen, Ellsworth, Mas-lach, & Seipel, 197S) or whether the op-portunity to repay is provided but repaymentis not specifically asked for (Castro, 1974;Gross & Latane, 1974).

In a communal relationship, to request abenefit after having given another person aidis inappropriate. It may imply that theoriginal aid was not given with the intent ofsatisfying a need but rather with the expec-tation of receiving something in return, whichmay be taken as an indication that the otherdoes not desire involvement in a communalrelationship. Assuming that beginning ormaintaining a communal relationship withanother is desirable, such an implicationshould be frustrating and therefore result indecreased liking.

If one has not been previously aided byanother and there is no opportunity to aidthe other in the future, a request from thatother is inappropriate in an exchange rela-tionship. In an exchange relationship, a per-son who has not been aided by another shouldlike the other more when he or she does not

ask for a benefit than when he or she doesask for a benefit.

However, requesting a benefit in the ab-sence of prior aid is appropriate in a com-munal relationship. Such a request impliesthat the other desires a communal relation-ship and, assuming that beginning or main-taining such a relationship is desirable, itshould result in increased liking. Jones andWortman (1973) suggest that asking anotherfor a benefit is a way of conveying that wethink highly of them. They say, "This tacticis likely to convey that we feel good aboutour relationship with the target person, sinceit is not customary to ask people to do favorsfor us unless our relationship is a relativelygood one" (Jones & Wortman, 1973, p. 13).

The implications of the distinction betweencommunal and exchange relationships for re-actions to requests for a benefit in the con-text of prior aid or lack of prior aid wereinvestigated in Experiment 2. A situationwas arranged in which aid could be given tothe subject and, later, a different benefit re-quested from the subject. The anticipationof a communal relationship was created byleading the subject to believe that the otherwanted to meet people and that she wouldbe discussing common interests with theother. The anticipation of an exchange rela-tionship was created by not mentioning any-thing about the other wanting to meet people,implying that the other was very busy andleading the subject to believe that she wouldbe discussing differences in interests withthe other.

The following hypotheses were tested inthe second experiment:

1. A request for a benefit from anotherafter the person is aided by the other leadsto greater attraction when an exchange rela-tionship is expected.

2. When a communal relationship is ex-pected, a request for a benefit after the per-son is aided by the other decreases attraction.

3. A request for a benefit in the absenceof aid from the other decreases attractionwhen an exchange relationship is expected.

4. When a communal relationship is ex-pected, a request for a benefit in the absenceof aid from the other increases attraction.


Method

Overview. Under the guise of a study of taskperformance, female college students worked on atask while a television monitor showed anotherfemale working on a similar task in another room.Some of the subjects were told that the other wasmarried, had a child, and lived far from the uni-versity, and that she and the subject would be dis-cussing differences in interests in a second study(exchange conditions). Other subjects were toldthat the other was new at the university and didnot know many people, and that she and the sub-ject would be discussing common interests in asecond study (communal conditions). The otherfemale finished the task, received 1 point, and gavethe subject aid on her supposedly more difficult taskor did not give the subject aid. The other femalethen requested a point from the subject or did notrequest a point. Finally, the subject's liking forthe other and expectations concerning the futurediscussion with the other were assessed.

Subjects. The subjects were 80 female, intro-ductory psychology students who received extracredit toward their course grade for their participa-tion. They were randomly assigned to one of theeight experimental conditions: exchange-aid-request,exchange-aid-no request, exchange-no aid-request,exchange-no aid-no request, communal-aid-request,communal-aid-no request, communal-no aid-request,and communal-no aid-no request.

Procedure. Upon arriving for the study, the sub-ject was greeted by the experimenter and told thatthe other subject scheduled to participate at thesame time had already arrived and was waiting inanother room for the experiment to begin. The ex-perimenter explained that it would take a little timefor the equipment to warm up before the experi-ment could begin.

Relationship manipulation. In the communal con-ditions, the experimenter casually stated that theother person was anxious to begin because:

She thinks it will be interesting. She's new at theuniversity, doesn't know many people, and she'sinterested in getting to know people.

In the exchange conditions, the experimenter casuallysaid:

She wants to finish soon. Her husband is comingby to pick her up, then they have to pick up herchild and go home to Columbia (a city some dis-tance from the university).

The experimenter said that the first study wasactually one of two short, unrelated studies theywould be asked to participate in that day. The ra-tionale for the first study was the same as in Ex-periment 1. The experimenter went on to say thatthe second study would be quite different from thefirst. In the communal conditions she continued:

What we're going to do is bring you both intoone room. We want you to talk over common in-terests. We're interested in finding out how peo-ple get to know one another. We try to create arelaxed atmosphere, and actually, in the past we'vefound that some of the people have gotten toknow one another quite well.

In the exchange conditions she continued:

What we're going to do is bring you both intoone room. We want you to talk over differences ininterests. We're doing this because most peopleavoid talking about differences in interests andwe're interested in getting people's reactions todoing so.

Vocabulary task. After the subject had signedan experimental consent form, the procedure forthe vocabulary task was explained in the samemanner as in Experiment 1, except that the subjectwas told that she would be performing the moredifficult task while the other person would be per-forming the easier task. The experimenter pointedout that since the subject's task was the more dif-ficult one, she would have a chance to earn 4 pointstoward the extra credit, whereas the other person,Tricia, only had a chance to earn 1 point, since shehad the easier task. As in Experiment 1, the ex-perimenter mentioned that the awarding of pointsto maintain motivation obviously would not benecessary in the second study.

After the same instructions concerning the stop-watches as in Experiment 1, the subject was leftalone to work on the task for a short time. Withthe 45 letters the subject had it was impossible forher to finish the task in that time. Subjects typicallyfinished between five and seven words. During thetime the subject was working, the experimenter,who always wore a lab coat so that changes inclothing over days could not be detected, could beseen on the monitor starting Tricia on her task andthen leaving the room. Tricia finished her taskeasily. After a short time the experimenter reenteredthe other room, and Tricia could be seen pushingsome extra letters to the front of the table. At thispoint the experimenter stepped in front of thecamera, blocking the subject's view of the otherso that the subject could not see whether the otherhanded the letters to the experimenter. Finally, theexperimenter could be seen leaving the room, andTricia sat back in her chair.

Aid manipulation. Shortly thereafter, the ex-perimenter reentered the subject's room and saidthat Tricia had finished her task and received 1point. The experimenter turned off the monitor,commenting that it wouldn't be needed any more.In the aid conditions she said, "Tricia asked me togive you these letters," and handed the subject someletters. In the no-aid conditions there was nomention of the letters. The experimenter then leftthe room, telling the subject she would be back


Table 2Means for the Measure of Liking inExperiment 2

Aid from & request for benefit fromthe other

Relation- Aid- Aid- No aid- No aid-ship request no request request no request

Exchange 173Communal 156

149191

149179

173177

Note. The higher the score, the greater the liking.Scores could range from 0 to 240. n = 10 per cell.

shortly. In approximately 3 minutes, she returned,and regardless of whether the subject had finished(none in the no-aid conditions did, whereas mostin the aid conditions did), she told the subject shehad done well enough to receive the 4 points towardextra credit. She filled out a form indicating thisand handed it to the subject. The experimenter saidthat was all there was to the first study, except forfilling out a form if the subject wanted to requestany points from Tricia. Before the subject had anopportunity to look at the form, the experimentersaid she would check with Tricia and see if shewanted to fill out a form and left the room for ap-proximately 2 minutes.

Request manipulation. When she appeared, theexperimenter handed the subject a folded note, sup-posedly from Tricia, in which a box was checkedindicating either that she wished to request 1 pointfrom the subject (request conditions) or that shedid not wish to request any points (no-request con-ditions). The experimenter was unaware of howthe form was checked, having drawn it from a con-tainer of folded forms checked in both ways. If thesubject wished to fill out a form to request pointsfrom the other, the experimenter took it.

Dependent measures. Next, the experimenter re-minded the subject that the second study would in-volve having both subjects talk over common in-terests (communal conditions) or differences in in-terests (exchange conditions). Before starting thestudy it was necessary to get some idea of whattheir expectations were about the forthcoming in-teraction in order to control for those expectations,since they might vary from person to person. There-fore, she was asking the subject to fill out two formsindicating what her first impressions of the otherperson were and also what she expected the discus-sion to be like. The subject was told that theseforms would be kept completely confidential.

The first-impressions form, which was given thesubject on top of the form concerning expectationsabout this discussion, was the same as the first-im-pression form used hi Experiment 1. The other formasked the subject to indicate how friendly, spon-taneous, strained, enjoyable, and awkward she ex-pected the discussion to be, on scales from 1 to 20.

Suspicion check. The experimenter left andwatched through a one-way mirror until she couldsee that the subject had finished the forms. She thenwaited approximately 30 additional sec and reenteredthe room. As she picked up the forms, she casuallysaid that there was something more to the studythan she had mentioned before and asked the sub-ject if she had any idea of what it might be. Theresponses of eight persons indicated suspicion ofthe instructions, and they were not included as sub-jects. Five persons suspected that the other personwas not actually in the other room, one thoughtthat the other's task had been intentionally madeeasy, one thought that the request did not actuallycome from the other person, and one thought thatneither the aid nor the request came from the otherperson. All five people who suspected that the otherperson was not actually in the other room had beenin another study by a different experimenter, inwhich they were told that another person partici-pated but later found out that there was no suchother person. Three of the eight persons were rununder the exchange-aid-request condition, two underthe exchange-aid-no-request condition, one underthe exchange-no-aid-request condition, one underthe communal-aid-request condition, and one underthe communal-aid-no-request condition. In addition,four other people were not included as subjects. Tworequested points from the other person, one did notread the request for a point before filling out thequestionnaires, and one asked so many questions thatthe experimenter could not leave the room beforeappearing on the monitor.

Finally, the true purpose of the experiment wasfully explained, and the subject promised not to dis-cuss it with anyone.

Results

A measure of liking for the other was cal-culated in the same way as in Experiment 1.The means for the experimental conditionsfor the measure of liking are presented inTable 2. An analysis of variance of the mea-sure of liking revealed that the main effectof type of relationship was significant, F(l,72) = 6.21, p < .05. Neither of the othermain effects nor any of the two-way interac-tions were significant. The three-way interac-tion between type of relationship, aid, andrequest was significant, F(l, 72) = 12.73, p <.001.

From the first hypothesis, it would be ex-pected that scores on the measure of likingwould be greater in the exchange-aid-requestcondition than in the exchange-aid-^no-requestcondition. As can be seen in Table 2, thedifference was in the predicted direction. Aplanned comparison indicated that this dif-


ference was significant, F(l, 72) =4.03, p< .05.

From the second hypothesis, it would beexpected that liking would be less in the com-munal-aid-request condition than in thecommunal-aid-no-request condition. As canbe seen in Table 2, the difference was aspredicted. A planned comparison indicatedthat this difference was significant, F(l, 72)= 8.60, p< .01.

From the third hypothesis, it would be ex-pected that liking would be less in the ex-change-no-aid-request condition than in theexchange-no-aid-no-request condition. As canbe seen in Table 2, the difference was as pre-dicted. A planned comparison indicated thatthis difference was significant, F(l, 72) =4.07, p < .05.

From the fourth hypothesis, it would beexpected that liking would be greater in thecommunal-no-aid-request condition than inthe communal-no-aid-no-request condition.As can be seen in Table 2, the means for thesetwo conditions were very similar. A plannedcomparison indicated that the difference be-tween these two means was not significant.

Another way of looking at the results is tocompare the aid and the no-aid conditions.It would be expected that liking would begreater in the exchange-aid-request conditionthan in the exchange-no-aid-request condi-tion. As can be seen in Table 2, this expecteddifference was obtained. A planned compari-son indicated that the difference was signifi-cant, F(l, 72) = 3.99, p < .05. It would beexpected that liking would be less in the com-munal-aid-request condition than in thecommunal-no-aid-request condition. As canbe seen in Table 2, this expected differencewas obtained. A planned comparison indicatedthat the difference was marginally significant,F(l, 72) =3.93, p< .06. It would be ex-pected that liking would be less in the ex-change-aid-no-request condition than in theexchange-no-aid-no-request condition. As canbe seen in Table 2, this expected differencewas obtained. A planned comparison indi-cated that it was significant, F(l, 72) = 4.10,p < .05. Finally, it would be expected thatliking would be greater in the communal-aid—no-request condition than in the communal-

Table 3Means for the Measure of AnticipatedPleasantness of the Discussion inExperiment 2

Aid from & request for benefit fromthe other

Relation- Aid- Aid- No aid- No aid-ship request no request request no request

ExchangeCommunal

5663

5972

5870

5952

Note. The higher the score, the more positive thesubjects' expectations for the discussion. Scorescould range from 5 to 100. n = 10 per cell.

no-aid-no-request condition. Although themeans were in the expected direction, theplanned comparison indicated that this dif-ference was not significant.

A measure of anticipated pleasantness ofthe discussion was calculated by summing thescores on the questions concerning howfriendly, spontaneous, strained, enjoyable, andawkward the subjects expected the discus-sion to be. The scores for the favorable char-acteristics were the same as the subjects'ratings. The scores for the unfavorable charac-teristics were obtained by subtracting thesubjects' ratings for those characteristics from21. The means for the experimental condi-tions for the measure of anticipated pleasant-ness of the discussion are presented in Table 3.

An analysis of variance of the measure ofanticipated pleasantness of the discussion re-vealed that the main effect of type of rela-tionship was significant, F(\, 72) =5.01, p< .05. The main effects of aid and of requestand the interactions between type of relation-ship and aid and between type of relationshipand request were not significant. The inter-action between aid and request was significant,F(l, 72) = 7.31, p < .01, and the three-wayinteraction between type of relationship, aid,and request was also significant, F(l, 72) =5.61, p < .05.

As can be seen in Table 3, anticipatedpleasantness was approximately the same inall four of the exchange conditions, but therewere differences within the communal condi-tions. Anticipated pleasantness was lower inthe communal-aid-request condition than in


the communal^aid-no-request condition. Asimple comparison indicated that this differ-ence approached significance, F(l, 72) =2.99, p < .10. Anticipated pleasantness wasgreater in the communal-no aid-request con-dition than in the communal-no-aid-no-re-quest condition. A simple comparison of thisdifference was significant, F(l, 72) = 11.19,p < .001.

Discussion

In general, the results supported the hy-potheses concerning reactions to a requestfor a benefit based on the distinction betweencommunal and exchange relationships. As ex-pected from the hypothesis that a requestfor a benefit after the person is aided by theother leads to greater attraction when an ex-change relationship is expected, it was foundthat liking for the other was higher in theexchange-aid-request condition than in theexchange-aid-no-request condition. As pre-dicted from the hypothesis that a request fora benefit after the person is aided decreasesattraction when a communal relationship isexpected, liking was lower in the communal-aid-request condition than in the communal-aid-no-request condition. In line with thehypothesis that a request for a 'benefit in theabsence of aid from the other decreases at-traction when an exchange relationship is ex-pected, liking for the other was lower in theexchange-no-aid-request condition than in theexchange-no-aid-no-request condition.

The hypothesis that a request for a benefitin the absence of aid from the other increasesattraction when a communal relationship isexpected was not supported; there was nodifference in liking between the communal-no-aid-request condition and the communal-no-aid-no-request condition. The subjects inthe communal-no-aid-request condition mayhave been somewhat uncertain about the in-tentions of the other. Although the requestmay have indicated to the subject that theother wanted a communal relationship withher and consequently led the subject to ex-pect such a relationship, it also may havereminded the subject that the other had notgiven her aid earlier. This reminder may haveraised doubt about whether the other would

behave in an appropriate way for a communalrelationship. This could explain why the re-quest did not result in increased liking for theother in the communal-no-aid-request con-dition.

As would be expected from the distinctionbetween communal and exchange relation-ships, liking was greater in the exchange-^aid-request condition than in the exchange-no-aid-request condition, marginally less in thecommunal-aid-request condition than in thecommunal-no-aid-request condition, and lessin the exchange-aid-no-request condition thanin the exchange-no aid-no-request condition.The greater liking in the exchange-aid-re-quest condition than in the exchange-no-aid-request condition could be due to a generaltendency for aid to increase liking, as well asto the appropriateness of the request. How-ever, the fact that liking was less in the ex-change-aid-no-request condition than in theexchange-no-aid-no-request condition is op-posite to what would be expected from ageneral tendency for aid to increase liking,but follows from the idea that differences inliking are due to the appropriateness of theother's behavior for the type of relationship.That liking was less in the communal-aid-request condition than in the communal-no-aid-request condition is also opposite to thetendency for aid to increase liking and con-sistent with the effect on liking of the appro-priateness of the other's behavior for the typeof relationship.

Since the focus was on the interactive ef-fects of type of relationship, a main effect oftype of relationship was not specifically pre-dicted. However, the significant main effectthat was obtained for liking is understandablein view of the operations used to manipulatetype of relationship. Among other things,subjects in the communal conditions may haveexpected the other to be more similar thandid subjects in the exchange conditions.

The results for the measure of anticipatedpleasantness of the interaction were also con-sistent with the distinction between communaland exchange relationships. Anticipated pleas-antness was similar in the four exchangeconditions, marginally less in the communal-aid-request condition than in the communal-


aid-no-request condition, and greater in thecommunal-no-aid-request condition than inthe communal-no-aid-no-request condition.

The results for the measure of pleasantnesscan be understood in terms of the assumptionthat one will anticipate interaction with theother to be more pleasant if one expects acommunal relationship than if one expects anexchange relationship. In the exchange condi-tion, subjects were led to expect an exchangerelationship by the experimental instructionsgiven in those conditions. Subjects in the com-munal-aid-request condition should have ex-pected an exchange relationship, because therules of a communal relationship were vio-lated by the request following the prior aid.Subjects in the communal-no-aid-no-requestcondition should also have expected an ex-change relationship, because the rules of acommunal relationship were violated by theother's failure to respond to their needs orto request something from the subject forwhich the other presumably had a need.

The interpretation of the results for themeasure of anticipated pleasantness mightappear inconsistent with the fact that likingwas not significantly higher in the communal-no-aid-request condition than in the com-munal-no-aid-no-request condition. However,it is possible that the request for a benefit inthe absence of prior aid was sufficient tocreate an expectation of a communal rela-tionship yet insufficient to increase liking. Asmentioned earlier, the request for aid in thecommunal-no-aid-request condition may havebeen taken as an indication that the otherfelt positively toward the subject and thusled the subject to expect a communal rela-tionship with that other, but it may alsohave reminded the subject that the other hadnot fulfilled her needs earlier, resulting inambivalent feelings toward the other.

General Discussion

While it is assumed that the distinction be-tween communal and exchange relationshipsis made implicitly by most people in theirinteractions with others, it is not assumedthat they are explicitly aware of the distinc-tion or are able to describe how it affectstheir reactions. Certainly they do not use the

terms communal and exchange relationships.It is also not assumed that everyone makesthe distinction in the same way. Some peoplerestrict their communal relationships to onlya very few persons, whereas others have com-munal relationships with a wide circle ofothers. There are some people who do notmake the distinction at all. Some people treatevery relationship, even relationships withmembers of their own immediate family, interms of exchange.

It is possible for a person to have both acommunal relationship and an exchange re-lationship with the same other, for example,when a person sells something to a friend orhires a family member as an employee. Insuch instances, a distinction is typically madebetween what is appropriate for the business(exchange) relationship and what is appro-priate for the family or friendship (com-munal) relationship. Exchange relationshipssometimes can develop into communal rela-tionships, such as when a merchant and acustomer become close friends or when anemployer and an employee marry.

The lack of attention paid to communalrelationships in previous research on inter-personal attraction may be accounted for bythe fact that almost all of the past researchhas involved attraction toward persons whoare not only previously unknown to the sub-ject but who are not expected by the subjectever to be known in the future. Communalrelationships involve an expectation of a long-term relationship, whereas exchange relation-ships need not be long-term. However, thevariables of communal versus exchange rela-tionship and expected length of the relation-ship are conceptually independent. Exchangerelationships may be expected to continueover a long period.

If it is true that treating a communal rela-tionship in terms of exchange compromisesthe relationship, then exchange theories ofinterpersonal attraction (e.g., Secord & Back-man, 1974, chapter 7) may create a mislead-ing impression about the development andbreakup of intimate relationships. The ideathat exchange is the basis of intimate relation-ships may actually have the effect of impair-ing such relationships. For example, the rec-


ommendation, which seems to be growing inpopularity, that prior to marriage a marriagecontract be drawn up that specifies in detailwhat each partner expects from the other,should, if followed, tend to undermine therelationship.

If the theoretical viewpoint of this researchis correct, a communal relationship will bestrained by dickering about what each of thepartners will do for the other. Of course, ifone of the partners in a communal relation-ship is convinced that he or she is being ex-ploited by the other because that person isconcerned about the other's welfare while theother is not concerned about his or her wel-fare, the communal relationship has disin-tegrated. If this happens in a marriage, theremay be attempts to preserve the marriage bychanging it into an exchange relationshipthrough dickering.

References

Adams, J. S. Toward an understanding of inequity.Journal of Abnormal and Social Psychology, 1963,67, 422-436.

Castro, M. A. C. Reactions to receiving aid as afunction of cost to donor and opportunity to aid.Journal of Applied Social Psychology, 1974, 4,194-209.

Barley, J. M., & Berscheid, E. Increased liking as aresult of the anticipation of personal contact. Hu-man Relations, 1967, 20, 29-40.

Gergen, K. J., Ellsworth, P., Maslach, C., & Seipel,M. Obligation, donor resources, and reactions toaid in three cultures. Journal of Personality andSocial Psychology, 1975, 31, 390-400.

Goffman, E. Asylums. Garden City, N.Y.: AnchorBooks, 1961.

Gross, A. E., & Latane, J. G. Receiving help, re-ciprocation, and interpersonal attraction. Journalof Applied Social Psychology, 1974, 4, 210-223.

Jones, E. E., & Wortman, C. Ingratiation: An at-tributional approach. Morristown, N.J.: GeneralLearning Press, 1973.

Keisler, S. B. The effect of perceived role require-ments on reactions to favor doing. Journal ofExperimental Social Psychology, 1966, 2, 198-210.

Mauss, M. The gift: Forms and functions of ex-change in archaic societies. Glencoe, III.: Free Press,1954.

Mirels, H., & Mills, J. Perception of the pleasantnessand competence of a partner. Journal of Abnor-mal and Social Psychology, 1964, 68, 456-459.

Pruitt, D. G. Methods for resolving differences ofinterest: A theoretical analysis. Journal of SocialIssues, 1972, 28, 133-154.

Secord, P. F., & Backman, C. W. Social psychology(2nd ed.). New York: McGraw-Hill, 1974.

Thibaut, J. W., & Kelley, H. H. The social psychol-ogy of groups. N.Y.: Wiley, 1959.

Walster, E., Walster, G. W., & Berscheid, E. Equity:Theory and research. Boston: Allyn & Bacon, 1978.

Received November 7, 1977 •

Interpersonal Attraction in Exchange and Communal ... · Interpersonal Attraction in Exchange and...

Documents

Transcript of Interpersonal Attraction in Exchange and Communal ... · Interpersonal Attraction in Exchange and...