Research on Interpersonal Relationships

1

Transcript of Research on Interpersonal Relationships

Page 1: Research on Interpersonal Relationships

The Japanese Journal of ExperimentalSocial Psychology. 1991, Vol. 30, No. 3, 259-267

〔Address〕

Research on Interpersonal Relationships

HAROLD H. KELLEY

University of California, Los Angeles

This is primarily an overview of interpersonal

research conducted within social psychlogy. The

study of interpersonal relations is not new in our

field. For example, we can remember the use of

sociometric methods, in the 1930s and 1940s, to

determine the network of attractions and repul-

sions among the individuals in a collection-e. g., a

classroom or a work group. We can also remem-

ber the work on similarity and interpersonal

ttraction. Some of this was correlational, show-

ing that friends tend to hold similar attitudes, and

some of it was conducted in natural •gexperiments•h,

as in Newcomb's famous study of the •gacquaint-

ance process•h, which showed that similarity in

attitudes measured before students came into con-

tact with each other was a good predictor of which

students eventually became friends. Much of this

work in the 1960s had the goal of testing Heider's

balance theory and similar theories. As an exam-

ple of early theoretical attention to interpersonal

relations, we can remember that over half of the

Thibaut and Kelley book on •gThe Social Psychol-

ogy of Groups•h? in 1959, was devoted to an analysis

of the dyad.

However, recently there is a new emphasis

within social psychology. Much of the earlier

work was conducted in the laboratory and studied

first contacts between strangers - their first

impressions of each other and their feelings of

initial attraction. Thibaut's and my theoretical

interest was in dyads in the abstract, rather than in

any specific ones. In contrast, the present empha-

sis is on real life, ongoing relationships. There is

often a reference to •gclose•h or •gpersonal•h relation-

ships, to emphasize that these are persons who are

intimately related and who have personal feelings

about each other. And the term •g-ship•h in •grela-

tionship•h is used implicitly to emphasize that we

are dealing with real, enduring entities and not

simply transient or temporary relations.

In practice, the current work studies young

people in love, people who are making commit-

ments to each other, members of the family, co-

workers in permanent jobs, and stable close friend-

ships. Experimental methods in the laboratory

have been replaced by use of questionnaires and

information about natural interaction. If the labo-

ratory is used at all, its purpose is to enable us to

watch an existing dyad interact under standard

conditions which permit systematic observation.

You will realize that in this shift toward the

study of ongoing relationships, social psychology

has moved into research areas already occupied by

clinical and developmental psychologists and by

family sociologists. Some of us feel that there is

the nucleus of a new discipline here, in the overlap

among social psychology and these other disci-

plines. (I should also mention communication

researchers who, in increasing numbers, have

moved away from their classic focus on mass

communication to the study of face-to-face commu-

nication.) There are some organizational moves

toward the creation of this •gdiscipline•h, with the

formation of a small cross-discipline and interna-

tional society, the publication of a new journal of

personal relationships, and a continuing ad hoc

group of US social psychologists dedicated to this

field of research. (If any of you are interested in

this area of research, we would welcome you as

members of the International Society for the Study

of Personal Relationships. It will hold its biennial

meeting next July in Oxford, England.) The

notion of a new discipline devoted to the study of

personal relationships is based on the idea that the

―259―

Page 2: Research on Interpersonal Relationships

Harold H. Kelley

present research is too dispersed over the existing

disciplines and that it is necessary for clinicians,

developmentalists, sociologists, communication

experts, and social psychologists to bring together

their ideas and methods. From the perspective of

psychology itself, one of the things I find most

appealing about •ginterpersonal relationships•h as a

focus is that it provides a common ground for

interaction among clinical, developmental, person-

ality, and social psychologists - a common around

that has been missing as our large departments of

psychology have become compartmentalized into

•gareas•h.

Some of us feel that social psychology, with its

large armamentarium of concepts and methods, is

likely to form the core of this new discipline of

Interpersonal Relations. Within social psychol-

ogy, the study of ongoing relationships has taken

us into new areas in which it has been possible to

apply our existing concepts. For example, Zick

Rubin's landmark study of love treated it as an

•gattitude•h. That concept, one of our most durable

ones in social psychology, provided a useful way to

conceptualize the complex idea of •glove•h and to

measure it and distinguish it from the related

concept of •gliking•h. Another example : Anne

Peplau's work on loneliness - the feelings associat-

ed with the absence of close relationships - was

interpreted with concepts from Weiner's theory of

attribution and affect. In general, there seems to

be more borrowing of social psychology's concepts

by the other specialities than the reverse. On the

other hand, the clinical and developmental psychol-

ogists seem to have the superior command of

certain methods and research strategies that are

necessary for the interpersonal relations field.

Both specialities have long traditions of observing

actual behavior where we social psychologists are

too reliant on self-report methods. Both have

traditions of gathering descriptive data for its own

sake, where we social psychologists are too much

devoted to testing hypotheses. Both clinical and

developmental have traditions of longitudinal

methods that enable us to see the development and

decline of relationships, where we social psycholo-

gists are too dependent on one-shot, cross-sectionalstudies. And both other specialities usually study

samples from the broader community where we

social psychologists rely heavily on student sam-

ples. Perhaps the most severe criticism that canbe leveled at the relationship field is that it studies

relationships in the US middle class. There is a

growing body of work in Europe, but badly neededare studies in other nations and cultures. This is

certainly one of the reasons that Japanese research

on relationships can not only expand our database

but also stretch and enrich our theoretical ideas.

I wish now to review two of the major

research themes in current research on interper-

sonal relationships. Then I will briefly outline

one of the major theories in the field - interdepen-

dence theory.

ATTRACTION AND COHESIVENESS

CONCEPTS

I suppose it is obvious that when we think

about a social entity, such as a •ggroup•h or •grela-

tionship•h, a prominent (perhaps the first) question

for research will be •gWhat holds it together ?•h

You will remember that early work on group

dynamics focussed on •gcohesiveness•h : the forces

that draw people to a group and keep them within

it. In the relationship field, there have been sev-

eral analogous concepts, each providing its own

perspective on the question of what draws people

together and keeps them in a relationship.

LOVE emphasizes the positive attractions that

partners have for each other. However, it has

proven to be an elusive concept. Since Rubin's

ground-breaking study, the analysis of love has

gone off in different directions. Numerous types

of love have been distinguished : the short-term,

sexually-based •gpassionate•h love vs. longer-lasting,

more pragmatic •gcompanionate•h love ; playful vs.

selfless love, etc. Similarly, various components

of love have been distinguished, such as its •gneed-

ing•h aspect (the person strongly needs the loved

one) vs. its •gcaring•h aspect (the person is willing

and eager to make sacrifices for the loved one).

Questions have been asked about •glove•h for a

―260―

Page 3: Research on Interpersonal Relationships

Research on Interpersonal Relationships

heterosexual partner, for one's family, and for a

friend. In general, I would say that this area of

research has moved from the simple place where

Rubin left it to a state of disjointed distinctions and

analyses. It seems clear that our field can't do

without the concept of love. Yet, it is not clear

how we are to incorporate it. A conceptual break-

through is due in this area, but I do not see it yet.

Part of the problem is that •glove•h is a widely used

and popular concept, at least in western countries,

and its common usage is extremely varied. There

seems little doubt that a strategy of distinguishing

types or components is the proper one, but so far,

as with the concept of •gintelligence•h, little consen-

sus has emerged from the various empirical

attempts to make such distinctions.

COMMITMENT focuses on factors that hold

people together whether they are attracted to their

relationship or not. It refers to all the factors,

positive and negative, that stably keep people

together. As a consequence, commitment is an

important concept in accounting for the stability or

instability or relationships. Indirectly, the analy-

sis of commitment has been influenced by Lewin's

analysis of the forces acting on group members to

stay within a group, an analysis that identifies the

internal and external positive and negative driving

forces, along with the restraining forces that pro-

vide barriers to leaving. The most influential

work on commitment has been that of Caryl Rus-

bult, a student of John Thibaut. She shows that

commitment is a function not only of satisfaction

with the relationship, but is a (negative) function of

the quality of one's alternatives to that relation-

ship, and a (positive) function of the investments

one has made in the relationship. Thus, her work

incorporates Thibaut and Kelley's concept •gCom-

parison Level for Alternatives•h and adds the con-

cept of •ginvestment•h, which seems to include the

costs of leaving a relationship and of starting up a

new one. Rusbult's work has additional interest,

because she has shown that her analysis applies not

only to personal relationships but also to the com-

mitment a worker may make to his/her company

or organization.

ATTACHMENT is a more recent addition to

the set of concepts we can use to explain relation-

ship formation and stability. The research here is

Philip Shaver's creative application to adult rela-

tionships of Bowlby's theory about mother-infant

relations. Bowlby proposed (and his hypothesis

has been confirmed by Ainsworth) that three types

of infants' attachments to their mothers can be

distinguished : (1) secure : children who feel secure

about the relationships and can use it as a safe base

from which to explore the environment (2)

insecure-avoidant : children who are unsure about

the mother's dependability and who avoid or reject

attachment to her, and (3) anxious-ambivalent :

children who, fearful that the mother's love will

not continue, approach her with ambivalence.

Shaver adopts Bowlby's idea that from the early

mother-infant experience, we develop different

•ginner working models•h of love relationships. We

carry those models into adulthood and they shape

the way we approach all potential love partners.

The •gsecure•h attachment person is able to develop

love relationships that provide feelings of security

and confidence. The insecure-avoidant person is

uncomfortable with closeness to a loved person

and doubts that any such relationship will last.

The amxious-ambivalent person falls in love quick-

ly, but feels that their lover does not reciprocate

their desire for stable closeness. Shaver and

Hazen have produced evidence in support of these

ideas, and the study of •gattachment•h is one of the

most active new topics in our field. You will

appreciate the way in which attachment theory

enables us to link early family experiences with

adult attitudes and feelings about close relation-

ships. Shaver's work is an excellent illustration

of my earlier point, that the relationship area

provides a natural setting for cross-fertilization

among sub-areas of psychology - clinical, develop-

mental, personality, and social.

SELF-DISCLOSURE AND INTIMACY

Where love, commitment, and attachment

focus on an individual's attitudes toward a relation-

ship or partner, research on disclosure and inti-

―261―

Page 4: Research on Interpersonal Relationships

Harold H. Kelley

macy focuses on interpersonal processes. The

early studies of self-disclosure grew out of clinical

psychologists' interests in the psychotherapeutic

interview and the personal information their

patients would reveal. The two principal general-

izations from the self-disclosure research of the

1960's and 70's are : (1) Disclosure become more

intimate as partners become better acquainted (not

surprisingly), and (2) the levels of disclosure of the

two partners tend to be mutual (reciprocity of

disclosure - the more one discloses, the more the

other one does). (A major exception to the latter,

of course, is the clinical interview, itself.)

More recently, disclosure and intimacy proces-

ses have become one of the favorite topics among

communication researchers. At the same time, the

concerns have broadened beyond the verbal disclo-

sure of personal facts, and now include the nonver-

bal processes by which feelings are expressed.

The most recent theoretical development is that

researchers have begun to analyze the functions or

purposes served by intimate interaction. The

general notion here, spelled out by Reis and

Shaver, is that true •gintimate•h interaction occurs

when one person is able to reveal important but

private feelings about him/her self, and the other

partner responds in ways that show understanding,

that validate the person's feelings, and that show

affection and caring. You will recognize that this

formulation reflects the influence of the classical

clinical theories of Carl Rogers, Erik Erikson, and

Harry Stack Sullivan. Following these theorists,

in its new emphasis on validation of feelings, inti-

macy research begins to deal with the development

and maintenance of self-indentity and the self-

concept. So here again, the relationship field

makes contact with other areas of research, in this

case, with the social congnition work on self-

concept.

A common assumption in disclosure and inti-

macy research is that the more disclosure, the

better. Indeed, there is growing evidence that

intimate interaction is often beneficial to the par-

ticipant, in warding off loneliness, promoting

psychosocial adjustment, improving one's coping

with illness and stress, etc. However, as Clark

and Reis have noted, there is also evidence to the

contrary. Under some conditions and in some

relationships, intimate disclosure may generate

negative emotions and conflict. There is an

important and subtle point here : Life in close

relationships requires a delicate balance between

controls internal to each person, in which silence is

maintained and coping is managed privately, and

the interpersonal controls which require the shar-

ing of problems and their joint management.

In addition to the two research topics I have

discussed, I will simply mention three other major

areas of work. One concerns power and influence

within relationships, a second concerns conflicts

and quarrels, and a third concerns disordered rela-

tionships and their dissolution. I merely list these

because I want to move on to a consideration of the

theoretical state of the field.

INTERDEPENDENCE CONCEPTS

In view of the variety of topics within the

relationship field, you will not be surprised that the

field has no over-arching theory. Rather, it

includes a number of mini-theories. The closest

thing to an all-inclusive framework is provided by

interdependence concepts, which I now wish to

describe. This kind of theory begins with Kurt

Lewin's insight, that interdependence is the essen-

tial feature of small groups. (In the study of

personal relationships, we study the smallest of

groups - the pair of lovers or close friends, the

small family group).

By interdependence, we mean simply that the

various members influence and are influenced by

one another. Several years ago, eight other rela-

tionship researchers and myself provided the

sketch in Figure 1 as a means of representing

•ginterdependence•h. The figure shows a dyad, con-

sisting of P and O. Associated with each person is

a chain of events, largely generated by that person

and connected by causal arrows within each chain.

Their interaction is represented by a pattern of

connections between their chains, these also being

represented by causal arrows. They are interde-

―262―

Page 5: Research on Interpersonal Relationships

Research on Interpersonal Relationships

Figure 1

The basic data of a dyadic relationship. Each person has a chain of events, eachchain including affect, thought, and action. The events are causally connected

within each chain (shown by arrows from one P to another or from one O toanother) and the two chains are causally interconnected (shown by arrows from aP to an O or from an O to a P). The interchain connections constitute the essentialfeature of interpersonal relationships.

―263―

Page 6: Research on Interpersonal Relationships

Harold H. Kelley

pendent in the sense that the arrows go in both

directions : Elements in P's chain of events affect

O's chain, and similarly, elements in O's chain of

events affect P's chain. In short, the interdepen-

dence within the dyad simply means that there are

mutual effects between the two persons. Their

behaviors, emotions, and thoughts are mutually

and causally interconnected.

The preceding formulation is quite general

and atheoretical. It represents a chain of events

associated with each person and the possible lines

of causality within and between those events.

However, there is no specification of what the

events or the causal arrows mean. Consistent

with this abstract model, a number of different

specific •ginterdependence theories•h can be devel-

oped, depending on what types of events and what

kinds of causal connections are defined as impor-

tant.

We used this schematic model as a guide in

defining what we meant by a •gclose relationship•h-

the title of our project and eventual book.

Accordingly, we defined a •gclose relationship•h as

•gone of strong, frequent, and diverse interdepen-

dence that lasts over a considerable peried of

time.•h Recently, Berscheid has used this defini-

tion to construct a scale of relationship closeness.

From detailed questionnaires, she obtains self-

report measures of (1) •gfrequency•h : number of

minutes per day the pair spends together, by them-

selves, (2) •gdiversity•h : the number of different

activities they do together, and (3) •gstrength•h :

amount of influence they have on each other.

These are combined into a single Relationship

Closeness Index, and it is used to predict stability

and, in the event of breakup, the degree of emo-

tional distress. Berscheid's results show that

closeness is a good predictor of how long a rela-

tionship will last. This is consistent with much

other evidence, that the more interdependent are

two people, the more stable is their relationship.

Her results are particularly interesting from a

methodological perspective. The •gcloseness•h

index, derived from reports of specific details of

the pair's life, is only slightly correlated with direct

subjective ratings of •gcloseness•h. In other words,

persons' subjective sense of how close they are is

not at all a good reflection of how close they are

behaviorally. As Berscheid notes and as various

parts of her data suggest, this has a very interest-

ing consequence. Because of a discrepancy

between feelings of closeness and actual behavior-

al interdependence, people who, subjectively, are

only moderately •gclose•h, are often in for an un-

pleasant surprise when their relationship breaks

up. Their subjective feelings of moderate close-

ness leave them unprepared for the great distress

they have to go through in order to disentangle

themselves from their great behavioral dependence

on the partner.

In basing her index of relationship closeness

on our general and atheoretical definition of close-

ness, Berscheid's scale is extremely behavioristic

in purpose, even though reliant, in its execution, on

subjective reports. Ideally, she would like to fol-

low a couple around and count the frequency,

diversity, and strength of their interactions. And

she is atheoretical in her willingness to count

everything and give it all pretty much the same

weight in calculating the aggregate score. In

contrast, other versions of interdependence analy-

sis become theoretical in the sense that they desig-

nate certain types of events as the important ones

and focus their attention, at least initially, on those

particular events. The most common focus is on

the rewardcosts aspects of events - their reinforc-

ing or need-reducing properties. These rewards

and costs are referred to as •goutcomes•h, so these

theories define interdependence in terms of how

people affect each others' outcomes.

OUTCOME INTERDEPENDENCE

•gOutcome interdependence•h is the kind of the-

ory John Thibaut and I have developed over the

years, beginning with our 1959 book. This kind of

theory is often described as •gexchange theory•h,

inasmuch as it uses a loose economic metaphor, of

people trading or exchanging rewarding goods or

services. Thibaut and I have tried to divorce

ourselves from the •gexchange•h label, and instead

―264―

Page 7: Research on Interpersonal Relationships

Research on Interpersonal Relationships

Table 1 Six levels of analysis in interdependence theory

have referred to our theory as •ginterdependence

theory•h. One purpose is to emphasize that inter-

dependence involves not only exchange phenonema

but also coordination phenomena.

In general, Thibaut and Kelley interdepen-

dence theory is about how people deal with the

problems posed by their control over each other's

outcomes. Table 1 shows that the theory begins

with a characterization of situations of interdepen-

dence. Then, the theory makes various assump-

tions about the persons who interact in those situa-

tions. These are assumptions about the goals or

motives that people bring to their interaction.

Finally, the theory examines interaction process,

which is viewed as the joint product of the situa-

tions and of the people who act in reasonable ways

to pursue their various goals in those situations.

I should emphasize that at its core, interdepen-

dence theory is a THEORY OF SITUATIONS,

very much is the tradition of Kurt Lewin's •glife-

space•h. And process is treated as a joint function

of •gpersons•h and •gsituation•h, analogous to Lewin's

famous formula that B=f (P,E).

I. Simple Exchange. (Interdependence in

forming relationships.)

I have labelled this •gsimple exchange•h because

it is what people usually have in mind when they

think of •gexchange theory•h. Table 1 shows that

the situation is defined in terms of a field of poten-

tial partners who afford the person possible out-

comes. The persons are assumed to be motivated

by self interest They try to maximize their own

outcomes (maximize their rewards and minimize

their costs). The prototypic process phenomenon

is that of pair formation. A major finding here is

what is known as the •gmatching hypothesis•h, that

the persons who form pairs tend to be similar or

•gmatched•h in their general attractiveness.

―265―

Page 8: Research on Interpersonal Relationships

Harold H. Kelley

II. Simple Interdependence. (Interdependence

in sustaining membership.)

The second level of analysis begins with per-

sons who are paired off but who may voluntarily

dissolve their relationships. Table 1 shows that

the salient feature of each person's situation is the

alternative partners available should the person

leave the current relationship. Again, the person

is assumed to be oriented primarily toward maxim-

izing his/her own outcomes, and the prototypic

process phenomena are dependence, dissolution,

and power.

Following Thibaut and Kelley (1958) we may

define each person's dependence on the pair as the

degree to which the outcomes there exceed those

available from the person's best alternative part-

ners. A major consequence of dependence is that

it determines the likelihood the person will remain

in the relationship. The more dependent a self-

interested person is, the more that person will want

the relationship to continue and, therefore, the

more that person will do to keep the relationship

intact. The major findings at this level have to do

with the consequences of unequal dependence : the

more dependent partner is less likely to leave, is

more likely to worry about the partner's leaving

and, therefore, has less power in the relationship,

and does more to anticipate the partner's needs.

III. Outcome Matrices : (Interdependence in

outcomes obtained within the relationship.)

The third level of analysis examines in greater

detail how variations in outcomes are controlled

within a relationship. Thus, this analysis is rele-

vant not only to voluntary pairings but also to

nonvoluntary ones. It helps us understand the

ways people exercise control even in relationships

they cannot readily leave. Table 1 shows that we

describe situations in terms of patterns of outcome

control, using the outcome matrix as our descrip-

tive tool. Again, we assume persons are self-

interested, and we study, for example, processes

related to exchange vs. coordination. The major

results here have to do with the contrasting prob-

lems, norms and rules, and process scenarios that

result from the exchange vs. coordination aspects

of interdependence.

IV. Outcome Decision Rules. (interdependence

in outcome rule and dispositions).

At the three preceding levels of analysis, we

have assumed that people's behavior is governed

exclusively by self interest. At various choice

points each person follows the decision rule •gMaxi-

mize my own outcomes•h. At level IV, we expand

the set of •gdecision rules•h that people are assumed

to follow. Specifically, we examine how interde-

pendent people apply rules that take account of

their partner's outcomes. In doing so, they exhibit

various interpersonal dispositions, such as consid-

erateness, altruism, competitiveness, and fairness.

The major idea at this level is that people are

interdependent in these outcome rules and disposi-

tions. So we now must describe their situations in

terms of the patterns of outcomes they obtain both

from their behaviors and from the rules they follow.

The major results here show that people infer

underlying dispositions for each other's behavior

and that those inferences give rise to strong af-

fective reactions.

V. Transition Lists (Interdependence in tran-

sitions between situations.)

The 2•~2 outcome matrix is a powerful

conceptual tool in the study of interpersonal rela-

tions. However, it has clear limitations. These

stem primarily from the fact that it describes the

situation as it exists at a given moment. These

limitations are apparent in laboratory work that

compares a given game matrix as it is played

under different sequential or temporal constraints.

It becomes clear that any complete characteriza-

tion of interdependence requires our having a

method to represent these sequential-temporal con-

straints. That method is provided by transition

lists. Table 1 shows that at Level V we use transi-

tion lists to describe situations in terms of patterns

of transition control. We return to the assump-

tion that persons are self-interested, and we study

such processes as interaction and control

―266―

Page 9: Research on Interpersonal Relationships

Research on Interpersonal Relationships

sequences. Some of the interesting results have to

do with conflict, not about what should occur but

about when it should occur, i. e., conflict about the

tempo and rhythm of interaction.

VI. Transition Decision Rules : (Interdepen-

dence in transition rules and dispositions).

At the VIth and final level of analysis, we

examine transition decision rules and the corre-

sponding transition or temporal dispositions. We

use here the same analytic strategy that we used at

Level IV. First, we analyze the kinds of interde-

pendence problems people face, as identified by

transition lists. Then, we ask what decision rules

would be adaptive in coping with them. We con-

sider how self-interest should be expanded through

responsiveness to the sequential and temporal

structure of situations. The earlier analysis was

based on the domain of outcome matrices. Here it

is based on the much more extensive (and less well

understood) domain of transition lists.

Some of the most interesting results here have

to do with sustaining a common line of activity

over time. Solutions to this problem are described

under the general rubric of •gcommitment•h. The

adaptive decision rules are those suggested by

research in individual psychology on cognitive

dissonance reduction, delay of gratification, and

impulse control. Some of these rules involve spe-

cial information processing mechanisms, such as

emphasizing the benefits of a chosen line of action,

or suppressing thoughts about tempting diversions

from that line of action. Like all rules, their

adaptive value depends on their being exercised

contingently. One's commitment to a relationship

should depend on the partner's commitment.

With interdependence that stretches over time, the

shared interest in one another's commitment will

usually lead to a negotiation of shared commit-

ment. This illustrates the important point, that

the time dimension of relationships encourages

adaptation through shared rules (norms), and

reduces the reliance on rules that each person

applies more or less independently.

REMARKS

I hope these brief comments will explain this

emerging research field to you (if you are not

already familiar with it). It is one I consider to be

at the core of social psychology, with its focus on

interpersonal influence and perception, and its

concern about real interaction processes. Histori-

cally, the field is interesting in what it draws from

earlier work and in the new uses to which old data

and ideas are being put. It draws much motiva-

tion from people's interests in heterosexual rela-

tions. And not surprisingly, both genders are re-

presented about equally among the workers in the

field. Certainly, many of the major contributions

have been made by women, among whom I would

especially mention Ellen Berscheid and Elaine

Hatfield (once, Elaine Walster). The field also

draws much impetus with practical concerns about

such problems as the absence of relationships (lone-

liness) and the pathologies of distressed relation-

ships. As I have noted, the database is presently

based for the most part on relationships in the

United States, and undoubtedly reflects some pecu-

liar historical and cultural factors existing there.

I wonder how much our ideas and results will apply

to Japanese relationships, and I am confident that

research on these relationships will contribute to a

sound conceptualization of close relationships.

―267―