INTERACTIONAL BEHAVIOR CHARACTERISTICS OF
SPOUSES RELATED TO EXPERIENCED MARITAL QUALITY

By

Carl Nickeson

A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL
OF THE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE
DEGREE OF DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA

1976

ACKNOWLEDGMENTS

Hugh Davis is first and foremost for mention. Always

available for wild speculation, or more focused discussion, he

"nondirectively" advised and guided me through my last years as

a student and through this dissertation.

Although I have never met Clifford Swensen, he provided

me with much that I needed. In my opinion his published writings

are the best summation of the work in interpersonal relations, and

by way of direct aid he answered all questions I presented through

correspondence.

Franz Epting, co-chairman of my master's committee, also

served on my doctoral committee and was first to help me write

clearly. Nathan Perry, Audrey Schumacher, and Merle Meyer rounded

out the committee, and all contributed to the design of the study

reported here.

Special thanks are due Solon Kimball, as the outside

member of the committee, for his friendly willingness to step in

almost at the last moment, when unforeseen difficulties arose. He

introduced me to anthropological thought and writing--something I

had been interested in for some time but had never studied--which

I regard as a valuable contribution to psychology for the broader

context it provides.

I want to thank the married couples who participated as

subjects in the research and also the three observers who were

essential to the data collection. Without these persons there

would be no dissertation.

To Pat Korb, an exemplar of excellence in many endeavors,

I want to express my appreciation for our friendship of these past

several years. Thanks, too, to transcendental meditation for its

help in making the work represented here a more quality experience.

And to Barbara and Lisa, special thanks for your love and companion-

ship.

TABLE OF CONTENTS

PAGE

ACKNOWLEDGMENTS. . . . . . . . . . . . .ii

LIST OF TABLES .

ABSTRACT . .

CHAPTER I

CHAPTER II

. . . . . . . . . . . . .vi

. . . . . . . . . . . . .vii

INTRODUCTION. . . . . . . . ... .1

REVIEW OF PERTINENT LITERATURE AND DEVELOPMENT
OF THE PRESENT PROBLEM FOR STUDY. . . . 6

Relationship Quality . . . . . . 6
Interpersonal Factors of Personality and
Their Application to Marriage. . . . 9
Theory Suitable for Relating Relationship
Quality to Interpersonal Behavior. . . .14
Role Theory . . . . . . ... .15
Social Exchange Theory. . . . . .19
Hypotheses for the Present Study ... .25

CHAPTER III METHOD ...... . . . . . . . 27

Subjects . .
Instruments. .
Observers. .
Pilot Study. .
Procedure. .
Data Analyses.

CHAPTER IV

RESULTS . . . . . . . . .

Reliability of the Dependent Variable
Measures . . . . . . . .
Characteristics of the Subject Sample
and Comparisons. . . . . ..
Simple Correlational Relationships
Between the Variables . . . .
Findings from Multiple Regression
Analyses . . . . . . . .

. .34

. .34

- .38

. .48

. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .

TABLE OF CONTENTS CONTINUED

CHAPTER V

APPENDICES . .

REFERENCES. . .

BIOGRAPHICAL SKET

DISCUSSION . . . . . . . .

Reliability and Validity of
Present Results . . . . . .
Hypotheses Tests . . . . . .
Alternative Concepts of Interpersonal
Dynamics Relating to the Present Results
Implications of the Present Results. .

APPENDIX I . . . . . .

APPENDIX II- - ...............

APPENDIX III

APPENDIX IV.

APPENDIX V

CH

. .60

. .60
. .64

S.69
. .77

S.81

S.82

S.83

S.85

S.91

S.93

S.95

S.93

LIST OF TABLES

TABLE NO. PAGE

I COMPARISON OF PRESENT SAMPLE WITH LOVE
SCALE NORMS. . . . . . . . .36

II CORRELATIONS BETWEEN QUALITY VARIABLES
AT LEVEL-I AND TIME MARRIED. . . . .39

III CORRELATIONS BETWEEN THE INDEPENDENT
VARIABLE AND THE DEPENDENT VARIABLE
LEVEL-I. ... . . . . . . . .41

IV CORRELATION OF HUSBAND'S AND WIFE'S
DIMENSIONAL BEHAVIOR . . . . .. ... 43

V CORRELATION OF LEVEL-II MEASURES OF
RELATIONSHIP QUALITY WITH LEVEL-II
INDICES OF MATES' INTERACTIONAL
BEHAVIORS. . . . . . . . .44

VI INDIVIDUAL REPORTS OF QUALITY CORRELATED
WITH LEVEL-II INDICES OF MATES'
INTERACTIONAL BEHAVIORS. . . . . .47

VII MULTIPLE REGRESSION EQUATIONS
PREDICTING THE SUM OF MATES' Dl
SCORES FROM THE INDEPENDENT VARIABLES. . .49

VIII MULTIPLE REGRESSION EQUATIONS
PREDICTING THE DIFFERENCE AND SUM
OF MATES' D2 SCORES FROM THE
INDEPENDENT VARIABLES. . . . . . .51

IX MULTIPLE REGRESSION EQUATIONS
PREDICTING THE DIFFERENCE BETWEEN
MATES' D3 SCORES FROM THE
INDEPENDENT VARIABLES. . . . . . .54

X MULTIPLE REGRESSION EQUATIONS
PREDICTING DEPENDENT VARIABLE SCORES,
LEVEL-I, FROM THE INDEPENDENT VARIABLES. .56

Abstract of Dissertation Presented to the Graduate Council
of the University of Florida in Partial Fulfillment of the Requirements
for the Degree of Doctor of Philosophy

INTERACTIONAL BEHAVIOR CHARACTERISTICS OF
SPOUSES RELATED TO EXPERIENCED MARITAL QUALITY

By

Carl Nickeson

March, 1976

Chairman: Hugh Davis
Major Department: Psychology

The purpose of this research was to see if the self-

reports of husbands and wives about the quality of their marital

relationships were related in a consistent fashion to the observed

interaction between the mates of each couple. The literature on

interpersonal relations was reviewed to obtain operational proce-

dures for the study of experienced quality and interactional

behavior and to find theoretical constructs associating these two

realms of variables.

Relationship quality was assessed by a Love Scale and a

Marriage Problems Scale. Interactive behaviors were recorded by

observers using the Bales Interpersonal Rating System, which

included procedures for reducing the data to three dimensions:

power, affection, and attitudes toward social movement or social

convention. These three dimensions have been repeatedly found to

be fundamental factors in interpersonal relations.

A hypothesis was formulated for each interactional dimen-

sion on the basis of social exchange theory and Sullivanian inter-

personal dynamics. It was predicted for the power dimension that

couples reporting the highest quality would relate to one another

in a reciprocal manner; i.e., they would appear more dissimilar

in these behaviors than couples reporting lower quality. An in-

verse function was predicted for the remaining dimensions. Specif-

ically, it was anticipated for the affectional dimension and social

movement or convention dimension that couples reporting the highest

quality would relate to one another in a corresponding manner;

i.e., they would appear more similar in these behaviors than couples

reporting lower quality.

Forty-eight young, recently married (under five years),

childless, college-educated couples participated as subjects in

the research. The most significant finding showed that the great

majority of these couples reported that their marriages were very

high in love and had few problems. In fact, these couples for the

most part scored well above Love Scale norms previously obtained

for both functional and dysfunctional marriages.

The study had been undertaken with the goal of obtaining

a relatively homogeneous sample of subjects so that while the

results would be restricted in generality, any significant findings

would have less chance of being confounded by known or unknown

variables. The sample characteristics noted above show that this

end was well served in the study.

viii

The expected associations between mates' reports of quality,

and a principle of correspondence in their affectional and social

movement behaviors, were found. However, the data were equivocal.

For the social movement dimension the trend was slightly below

statistical significance, and while the results for the affectional

dimension were highly significant, there was also evidence that

simply positive, or friendly, behaviors were associated with the

highest quality.

The data provided no support for the hypothesized rela-

tionship between reported quality and dissimilarity on the power

dimension. Instead, for this dimension there were indications

that the highest quality was associated with a trend toward asser-

tiveness on the part of both spouses.

Additional results indicated some differences between

husbands and wives in the way their reports of quality were asso-

ciated with interactional behaviors. These findings were discussed

in terms of the role theoretical perspective on marriage and as

possible sources of tension in marriage.

Implications were drawn from the findings, reflecting

on the changing nature of marriage in the United States today.

Suggestions appropriate for marriage and premarital counseling were

also made on the basis of the study's results.

CHAPTER I

INTRODUCTION

In our culture, as in most others, marriage is a highly

significant relationship and may perhaps be the most significant

relationship a person may enter. Marriage is regarded as a funda-

mental institution of our society, as a prerequisite for family life

and child-rearing. As with any important social unit there are many

different opinions about the basic nature of marriage, and the sort

of relationship it should be. Some regard a proper marriage as a

union of souls and believe that marriages are made in heaven. This

opinion, of course, is that of the poet, who regards marriage as the

epitome of love and who feels that people should marry only on the

basis of "magnetic attraction." In contrast to idealists with

romantic notions of marriage, others are highly dispassionate about

the relationship. Here are represented individuals who bring mar-

riage under the purview of legal binding and see marriage as pri-

marily a contractual agreement necessary for the protection of

children and women.

Psychologists and sociologists approach marriage in a dif-

ferent way. Rather than intuitions or pragmatic views about the

meaning of marriage, social scientists study the dynamics of the

relationship in terms of observable variables. Looked at in this

way, marriage is seen as an interpersonal relationship that is not

substantially different from many other kinds of relationships

where interpersonal behaviors supply the data of concern.

Researchers have posed the question "Who gets along with

whom?" and in looking for the answer, have studied small groups.

Following the assumption that marriage is similar to other inter-

personal relationships, many findings regarding small groups may

be applicable to the marital dyad. However, a basis for caution in

extrapolating findings on interpersonal relations from small group

research to marriage may be found in the formulation of Secord and

Backman (1964). They noted that there are certain stages through

which relationships pass; namely, sampling, bargaining, commitment,

and institutionalization. Institutionalization is of course the

stage in which most marriages exist, while encounters between mem-

bers of small groups represent sampling or bargaining behaviors.

Because of this distinction as well as others, it is necessary to

specify clearly the nature of the relationship and to qualify

appropriately when making inferences about marriage from general

research on interpersonal relations.

Another related point to take into consideration is that

individuals who are married to each other possess, to a great degree,

a shared history. It is reasonable to assume that spouses have

evolved their interactional patterns on the basis of their past

interactions with each other. A test of this phenomenon may be

seen by comparing mates with interacting persons in most other small

groups. The latter have not jointly developed many behaviors

toward one another and are instead engaging in behaviors of a more

general sort. Spouses deal with one another in many ways on many

levels, from the mundane tasks of maintaining a standard of living

to the most intimate expressions of feeling. There seems to be

no other relationship with as much variety in the types of inter-

actions appropriate, and necessary, to continuance.

Thus, the marriage relationship can be compared and con-

trasted with other interpersonal relationships. However, in con-

sidering research on marriage alone there are found to be still

other differentiating factors within this institutionalized rela-

tionship. Hicks and Platt (1970) reported that family sociologists

have for many years postulated that at least two basic marital types

coexist in the United States: the institutional and the companionship.

According to Hicks and Platt the institutional form is the older,

more firmly established type, but recently, for a variety of social

and personal reasons, there has been a shift toward the companion-

ship marriage. They noted that in the institutional marriage,

adherence to traditional role specifications, customs, and mores

is the factor most significantly related to the happiness and suc-

cess of the relationship. The companionship marriage, on the other

hand, is characterized by the placement of greater emphasis by the

participants on the affective and emotional aspects of the rela-

tionship. Individuals who enter companionship marriages are con-

cerned with equality between the personal rights of each sex, and

in terms of the relationship's quality, greater emphasis seems to

be placed on the interaction of the two personalities involved.

Scanzoni (1972) was very much in agreement with the trend

noted by Hicks and Platt. Scanzoni discussed the changing nature

of the marital relationship as a function of the changing role of

women in our society. He pointed out that while some observers

believe that marriage itself is on the decline, in terms of the

numbers of people who are institutionalizing their relationships,

the data do not support this impression. He indicated that sta-

tistically, Americans are marrying as much as ever. Scanzoni noted

too, however, that the form of marriage is undergoing transformation.

To show the alterations in the nature of marriage, he presented

the historical changes that have taken place in the role relation-

ships of husbands and wives regarding rights and duties. He said

that before 1900 women were almost universally considered the pro-

perty of the husband. Then, at a later time, a woman was frequently

regarded as a complement to a man, and still later she was seen as

a junior partner. Scanzoni stated that most recently in our society,

wives seem to be moving toward achieving status as an equal partner.

He added, however, that the actual societal position held by women

both in and out of the family reamins subordinate to that of men.

The trend toward equality is distinct, nevertheless, at present,

and its effects on marital relations are also apparent.

The concern of the present work is marital interaction.

Specifically, this is a study about the quality of the marriage

relationship as perceived by the mates and the characteristics of

the interpersonal interactions they engage in with each other.

The overall purpose is to find, in marriages distinguished on the

5

basis of the individuals' experience of relationship quality,

differences in the ways the spouses behave with one another on

certain experimental tasks.

CHAPTER II

REVIEW OF PERTINENT LITERATURE AND DEVELOPMENT OF
THE PRESENT PROBLEM FOR STUDY

Relationship Quality

Previous research studies of marital relationships have

used many definitions of relationship quality. The concepts most

typically employed to assess marriage quality have been the norms

of happiness and stability (Hicks & Platt, 1970). Unfortunately,

these are phenomena that are difficult to measure. Hicks and Platt

noted that the words happiness, success, adjustment, and satisfaction

have many nuances of meaning, and behavioral scientists have typi-

cally been unable to formulate precise definitions for them. As

an alternative, researchers have tended to let the subjects provide

their own definitions in reporting happiness or stability.

Many paper and pencil inventories have been developed for

the assessment of over-all quality of the martial relationship.

However, the majority of these instruments provide only unidimensional

constructs of good-bad, happy-unhappy, or stable-unstable. The

present study is concerned with phenomena that may be quite subtle,

and for this reason it is necessary to use definitions of quality

that are well differentiated and provide information from a number

of areas relevant to relationship quality. Swensen and Fiore have

have provided two scales that seem suitable to the needs of the

present investigation. Swensen (1973b) has designed a Scale of the

Feelings and Behaviors of Love, and Swensen and Fiore (1975) developed

a Scale of Marriage Problems.

Swensen (1973b) reported that the Love Scale was con-

structed as a step toward the study of love relationships among

normal people. Swensen (1972) discussed how he went about building

this instrument and indicated that beyond defining love as a

relationship phenomenon he had to outline aspects for specific

study. He noted that love is what is inferred from concrete evi-

dence, and since love cannot be studied directly, it is necessary

to study the effects of love. His Love Scale is constructed around

descriptions of the consequences of love and the ways people have

for expressing their love. Unlike a poetic portrayal of magnetic

attraction, Swensen (1972) believed love is not a discrete force.

Instead he viewed it as the extreme positive end of all kinds of

interactive behaviors in relationships. He believed that even eros,

the consummative love between a man and a woman, is represented on

the Love Scale. Eros is different from other kinds of love only in

the breadth and quantity of its expression, not in the introduction

of new qualities missing from other love relationships.

Swensen (1973b) reported that he evolved his Love Scale,

over the course of several factor analytic studies, using an original

pool of items associated with the feelings and behaviors of love.

He stated that seven factors repeatedly appeared in the analyses:

(1) verbal expression of feelings; (2) self-disclosure of personal

facts about oneself; (3) willingness to tolerate the less pleasant

aspects of the loved person; (4) moral support, encouragement, and

interest (non-material evidence of love); (5) feelings that the per-

son felt but had never expressed verbally to the loved person;

(6) giving gifts, doing favors or chores, and providing material

support (material evidence of love); (7) physical expression of

love. In composing the Love Scale from the results of the factor

analytic studies, Swensen eliminated Factor 7 because he found it

to be highly consistent with the type of relationship under consid-

eration (e.g., parent-child, same-sexed friend, etc.) and had little

discriminative value with respect to relationship quality. Thus his

Love Scale contains six subscales, one to represent each of the

remaining factors.

Swensen (1973b) reported that one study (Fiore, 1972) has

shown that married couples who are satisfied with their marriage

score significantly higher than couples with troubled marriages

on all subscales except the unexpressed feelings subscale. Troubled

couples report significantly more unexpressed feelings. In using

this instrument in the present study, high relationship quality is

partially defined as a high score on all subscales except the

unexpressed feelings subscale of the Love Scale.

Swensen and Fiore (1975), in order to construct a scale

of marriage problems, similarly obtained an initial pool of items

which seemed to denote conflict areas in marriages. Using this pool,

they again carried out factor analytic studies and found six main

factors, each reflecting a problem area typical in marriage. The

items grouped as follows: (1) problem-solving, decision making,

and goals of marriage; (2) child-rearing and home labor; (3) relatives

and in-laws; (4) personal care and appearance; (5) money management;

(6) expression of affection, and outside friendship. This problems

scale was then constructed with six subscales, one representing

each of the areas described by a factor. Swensen and Fiore reported

that the problems scale clearly discriminates between functional

and dysfunctional married couples, in terms of total score, and the

mates' degree of agreement on problems. In the present research,

high relationship quality is partially defined in terms of low

scores on the Marriage Problems Scale. The two scales together

provide the total measure of relationship quality used in this study.

Interpersonal Factors of Personality and Their Application to Marriage

Swensen (1973a) reviewed the literature on interpersonal

relations and concluded that a substantial amount of research on dif-

ferent kinds of interactions suggested at least two basic dimensions

to dyadic encounter. He said,

These dimensions are called by different
names, but they have to do with dominance
versus submission, and acceptance versus
rejection. In groups, a third dimension
seems to appear--a dimension that has to
do with whether or not the person helps
the group progress toward its goal.
(Swensen, 1973a; p. 455.)

Foa (1961) had previously come to the same conclusion about the

consistency in the findings of independent investigators who had

shown similarities in the reduced data from.many situations of

interpersonal interaction. Foa presented four separate lines of

research on interpersonal relations, and his review may be regarded

as a major contribution in summarizing this area. A recapitulation

of findings that appeared convergent to Foa will not be included

here; however, Foa considered the work of Leary (1957), Carter

(1954), Borgatta, Cottrell, and Mann (1958), Schaefer (1959), and

Schutz (1958). Foa (1961, 1965, 1966) sought to extend this field

through the application of facet analysis (Guttman, 1958), and he

accomplished a great deal toward validating the conclusion that inter-

personal relations are a function of at least two dimensions.

Several instruments employing self-report methodology have

been developed to assess the position of a subject on each of these

interpersonal dimensions. The labeling of the dimensions has varied

according to the particular focus of an author, and some described

two, while others used three dimensions. Leary (1957) and Schutz

(1958) have provided two of the most prominent instruments for

assessing a person's orientation in terms of the interpersonal

personality factors from his self-report.

Bales (1970) also provided a three-factor system for describ-

ing interpersonal behavior. The Bales system differs, however, from

those just mentioned by virtue of the fact that it is based upon

an observational method for studying the ways people interact in

small groups, including dyads, and not on the self-report of the

subject. Bales' technique for interpersonal ratings is a procedure

in which, from a judge's observations of ongoing interaction, an

individual's interpersonal behaviors may be classified.

Bales (1970) referred to his system as a three-dimensional

spatial model and used his own labels for the interpersonal dimen-

sions. He called his directions down-up (D-U), negative-positive

(N-P), and backward-forward (B-F), and the meaning of each is as

follows: To put something D is to reduce its power, and to make

something U means to give it power. This dimension is equivalent

to the dimension of submission-dominance. To make something N is

to associate it with negative things; to make something P is to

associate it with positive things. This dimension is similar to the

dimension of rejection-acceptance or hostile-friendly. The third

dimension discussed by Bales refers to whether the person regards

things typical of collective acceptance as matters to be disassociated

from legitimacy or whether he places such subjects before the group

for acceptance. Persons tending in the B direction are heretical

and disbelieving toward authority and the rightness of goal orienta-

tions. Fantasies, as opposed to pragmatism and accomplishment,

seem to be prominent in their behavior and attitudes. In contrast

the F person tends to place things he regards as proper pursuits

of the group before them as projects, and he is task oriented.

Persons tending toward F usually strive to get to an objective

and are concerned with problem solving.

The present study employed the Bales system in order to

study the interaction processes of couples in relation to the expe-

rienced quality the mates reported in their marriage. The Bales

observational method is sensitive to the present ongoing behaviors

a person exhibits. Paper and pencil methods, on the other hand,

assess the general attitudes a person holds about his own behavior.

These methods are less suited to the present study because it may

well be the case that a spouse in different situations, with some-

one other than his mate, displays behaviors different from those

that typify his marital interaction. It is believed here that when

spouses' interpersonal behavior patterns are observed in this rela-

tionship-specific way, an investigator improves the likelihood that

a meaningful association may be found between experienced quality

and marital interaction.

While quality has been chosen as the independent variable

for this study, this choice does not mean that the interactional

pattern is assumed to be necessarily a function or derivative of

the individual's experience. It is possible that experienced quality

is a function of the interaction pattern. An investigation could

be made into the question whether the experience or the pattern is

more inherent in an individual's personality, but such answers are

not the purpose of the present study. The theoretical discussion

and formation of hypotheses could be accomplished using either of

the variables studied here as the independent variable, and the

decision in the presentation of this work rested upon the experi-

menter's choice as to how he would develop this theoretical discus-

sion. This statement indicates that quality has been chosen only

by convention as the independent variable. However, it may also

be noted that by making the self-report data the "fixed-point"

about which observers record variations in interaction patterns among

couples, the design allows the individuals for whom the results most

apply to speak first, in a sense.

Swensen (1974) carried out research that is related to

the focus of the present work. Swensen set out to formulate types

of interpersonal marriage relations as functions of lifestyle, and

he attempted to classify marital relationships according to Bales'

system. He hypothesized that in marriage (1) one mate would be

clearly dominant, and the other clearly submissive, versus no one

clearly or consistently dominant; (2) the mates' interaction would

be characterized by warmth and affection, versus hostility and

indifference; and (3) mates would agree on goal conceptualizations

regarding lifestyle and marriage and be moving toward these goals

(forward) or indifferent to such concepts (backward), versus disa-

greement over goal conceptualizations. These hypotheses represent

the bipolar dimensions basic to interpersonal relations, i.e.,

power, affection, and social movement.

Swensen (1973a) was essentially unsuccessful in relating

marital types to the dimensions of Bales. Instead he found that

there appeared to be only one generally applicable dimension in

marriage. This one clear dimension along which he was able to

order the couples' relationships was that of self-actualization.

This dimension at one end included couples who were unhappy and

dysfunctional in their relationship to one another and at the other

end included couples who were committed to one another to such an

extent that the only way they foresaw their marriage ending was by

total non-participation of the spouse. At the two ends of this

dimension he found that the problems the couples reported were

quite different. In sum, the highly actualized couples' conflicts

and stresses were more peripheral to the relationship.

Swensen concluded that on the surface highly self-actualized

couples appeared either conventional or unconventional in lifestyle,

and in each case, at basic levels of observation, each of these

marriages had a uniqueness about it that precluded the assumption

of a particular pattern in terms of interpersonal dimensions or

lifestyle. The people in the actualized relationships appeared

to have arrived at a point with each other where the emotional

needs of both were being met and the multiple components of their

interaction were being handled successfully. The highly self-

actualized couples were secure, distinct from the low-actualized

couples, who were quite insecure about the mate's place in their

lives and generally not satisfied with the marriage.

Theory Suitable for Relating Relationship Quality
to Interpersonal Behavior

So far nothing has been said how the two realms of vari-

ables studied here may be theoretically associated with each other.

Theory appropriate for hypothesis formation will be introduced, but

it is necessary to preface this discussion by acknowledging that

there is no single unified theory from which hypotheses for the

present study can be derived. The field of interpersonal relations

shows a paucity of theoretical backdrop, and it is therefore neces-

sary to take from the literature theoretical conceptions that are

at least partially applicable. Two lines of thinking are presented:

the constructs of (1) role theory, and (2) social exchange theory.

Role Theory

Swensen (1973a) noted that there is no such thing as a

role theory, that it is more appropriate to call the role way of

looking at interpersonal behavior the role perspective. The role

perspective assumes, in line with its theatrical analog, that

performance results from the social prescriptions of others and

that individual differences in role behavior are limited by these

factors (Biddle & Thomas, 1966). Swensen (1973a) continued with

the imagery of the theater, stating that social interaction can be

viewed as taking place upon a stage where parts are played within

the limits of set out words and actions. Normative expectations

about behavior, acquired as part of socialization, are the basic data

of role theory. The study of interpersonal interaction has been

greatly advanced by taking cognizance of the social structure in

which behavior occurs and of the fact that an individual's personality

represents, in some form, the social milieu.

A classic application of role theory concepts to marriage

was the formulation of Parsons and Bales (1955). Swensen (1973a)

pointed out that the Parsons and Bales theory has stimulated more

research than any other role theory, and Tharp (1963) gave even

greater praise when he concluded that this theoretical approach

has been the most promising line to follow in the study of marriage.

An outline of the Parsons and Bales theory begins with their view

of the marriage relationship as a two-member system, with speciali-

zation in separate directions for the wife and for the husband.

In their conceptualization they noted that the husband's role is

specialized in the instrumental direction and the wife's role, in

the expressive direction. It was their conclusion that males are

oriented to meeting the adaptive exigencies and females the integra-

tive. They stressed, however, that these are areas in which one

mate assumes, based upon sex role, the primary responsibility and

that both mates engage to some extent in behaviors most compatible

with the other's role. Parsons and Bales appear to have summarized

the data well in concluding that husbands are expected to be more

technical, executive, and task oriented and wives are expected to

be more supportive, tension-managing, and, in general, the social-

emotional leaders in a marriage.

Research has borne out the impression that couples who

have successfully adapted to social expectations are more stably

married. Tharp's (1964) review of the literature in this area

concluded that marital satisfaction was significantly related to

the achievement of the husband's role by the male and the wife's

role by the female with the husband's adjustment being the more

important. More recent reviews (e.g., Barry, 1970) are less

definite than Tharp; however, it does seem clear that one result

emerging from marriage role research is that husbands and wives

are more likely to be satisfied with their relationship if they

are in agreement on what their roles are to be and if these roles

are congruent with normative expectations (Swensen, 1973a).

The Parsons and Bales (1955) role conceptions for marriage

may be related to the Bales (1970) system of interpersonal behavior

factors in personality. In the latter work Bales noted that there

are sex differences in the distribution of individual types in

his three-dimensional space. He stated:

The members are not typically distri-
buted at random or equally throughout
the space; more are found on the positive
side of the space than on the negative
side, and the tendency to be on the
positive side is very marked for women
as compared to men. More members are
found in the forward part of the space
than in the backward, and this tendency
is more marked for men. On the average,
women are further downward, considerably
more positive, and slightly more back-
ward from the men. Men are about equally
distributed between the upward and down-
ward parts of the space, with a few more
in the downward part, but women are
definitely found more frequently in the
lower part of the space. The differences
are about what might be expected from
the cultural stereotypes of the male
and female adult sex roles. (Bales,
1970; pp. 46-47)

Thus, data from the Bales rating system are well suited

for interpretation by role theory constructs. On the basis of the

role model it could be predicted that the experienced quality of

the marital relationship will be related to the congruence with

which the spouses fulfill their social roles, as represented in

the Bales dimensional space of interpersonal interaction. However,

it is questionable whether the role theory perspective still pro-

vides the best framework for marriage research, as it seemed to at

the time of Parsons' and Bales' (1955) publication, and at the

time of Tharp's (1964) review. Barry (1970), in his review exem-

plifies this trend in that he was not as decisive as Tharp in his

support of the role approach to marriage.

Theoretically it is not contrary to the Parsons and Bales

conception to assume that the husband and wife can occupy positions

of equal power, although in contemporary technological culture, the

male role is typically ascribed more social value. Because of this

inequality in the societal valuation of the normative social-sexual

roles, there exists in marital relationships a potential for tension.

In addition to this possible stress, there is another source of dis-

sonance in marriage: it has been easier, at least in the white

middle class and for a number of reasons, for the male to progress

in a vocational field. This aspect of inequality in opportunity

for the sexes and its potential for tension in the relations between

spouses has undoubtedly influenced the interactions of mates.

The socially engendered tension in husband-wife relations

may be providing some of the impetus for change from the institu-

tional to the companionship marriage as the predominant form. The

trend noted in the Hicks and Platt (1970) review, as well as the

cultural change outlined by Scanzoni (1972), implies that the rele-

vance of role theory for understanding the mates' experience of

marital quality is on the decline. As spouses' personal criteria

for measuring the quality of their relationship changes from role

fulfillment to other less traditional standards, so too must

researchers expand their theoretical perspectives so that these

include the current values of the individuals studied.

Social Exchange Theory

Social exchange may provide a more suitable frame within

which the realms of variables studied here may be related and

hypotheses constructed. However, social exchange, like role theory,

must be characterized as a point of view, not regarded as a whole

theory with well-articulated laws (Swensen, 1973a). Basic to the

perspective of social exchange are the concepts of reward and cost.

Reward is anything that meets another person's needs as the person

experiences them; it is a function of the behavior emitted and of

the personality of the receiver. Cost refers to a consequence of

behavior that deters or reduces the probability that a behavior

will be repeated. Homans (1961) has outlined the restrictions and

assumptions of the social exchange perspective on interpersonal

relations: (1) Social behavior is rewarded or punished by the

behavior of another person. (2) When a person acts in a certain

way toward another person, he is punished or rewarded by that person.

(3) The behavior must be actual behavior and not a norm of behavior.

Thus in face-to-face situations people act and react to one another

on the basis of reinforcement principles. Swensen (1973a) put it

simply when he stated that exchange theory sees the interaction between

two people as a function of what each gets out of the relationship:

no payoff in the relationship, no relationship.

Carson (1969) has added to the substance of the social

exchange perspective by integrating it with Sullivanian psychodynamic

theory. Carson regarded personality as the constellation of inter-

personal behaviors a person enacts in order to maintain the security

of his self-picture. This security is seen as dependent on the

extent to which the person is successful in eliciting responsive

behaviors from others in interactions that affirm his self-picture.

Anxiety, a very undesirable organismic state, is experienced when

another's responses are perceived as counter to or incompatible with

the person's expectations. Interpersonal behavior takes place in

sequences related to conscious and unconscious anticipations that

reflect the person's needs. When an anticipated sequence as a function

of the self-picture and the person's security maintenance operations

is interrupted, anxiety results.

He believed that this self-system has evolved developmentally

and chiefly out of the interactions with parents and siblings within

the family. Developmental experiences are the main source of

various interpersonal orientations, described by the interpersonal

personality factors that the individual assumes. He concluded

that by the time adulthood is achieved interpersonal behavior

patterns are relatively fixed and that because the personality is

more or less formed, there are limitations on the ways an adult

can maintain congruency of self in relations. Carson says a person

will tend not to change either self-concept or actual behavior.

He suggested that it may be easier for the person to misperceive

others' behavior in response to his own than to change, but he

also noted that the irrationality of such a course is likely to

make the interaction troublesome. The strategy remaining for the

person, then, is to seek others who, because of their own inter-

personal behavior tendencies, provide congruent responses. In

the process of social bargaining necessary for finding compatible

others, a person perceives the cost of interaction with some people

as too high and terminates the interaction, while with others the

exchange of behaviors is sufficiently rewarding to sustain the

relationship.

Carson has hypothesized that rewarding behaviors are ex-

changed according to a principle of complementarity. He followed

the interpersonal concepts of Leary (1957), and thus addressed only

two of the factors studied here, in Leary's terms submission-dominance

(S-D), and hate-love (H-L). Carson contended that complementarity

occurs on the basis of reciprocity for the S-D axis (i.e., S tends

to induce D, and vice versa) and on the basis of correspondence

for the H-L axis (i.e., H induces H, and L induces L). He regarded

these complementary behaviors as maximally congruent for a person's

self-maintenance operations and therefore rewarding. Based on

similar social exchange reasoning, it is the belief of this writer

that the third dimension of Bales, backward-forward (B-F), also

tends to follow a correspondence principle (i.e., B induces B,

and F induces F).

These hypotheses regarding the complementarity principles

each interpersonal dimension would be expected to follow have been

tested in this research. Carson reported that it has been shown

that complementary responses have a statistically higher probability

of occurrence in interactions than other kinds of responses to the

same evoking circumstances (Leary, 1957; Heller, Myers & Kline,

1963). However, these previous studies have not addressed the

exchange of behavior as a function of experienced quality of the

relationship. Here the assumption is made that mates who report

their relationship as high in quality experience significantly more

rewards than costs in their marital interaction and a higher

reward-cost ratio than those reporting less quality. Thus,

couples may be distinguished on the basis of the quality they

experience, and their interaction patterns may be compared in

terms of the Bales dimensions. The tests of the hypotheses are

made by noting if couples reporting higher quality exhibit inter-

action patterns which follow more closely the outlined principles

of complementarity.

It is important to note the difference between the

principles of complementarity proffered by Carson and tested

here, and the similarity-complementarity issues often associated

with the work of R. F. Winch (e.g., Winch, 1958). Winch and

others have studied personality needs and their relationship to

attraction between mates. Winch (1958) presented data in support

of his proposition that complementary needs are the primary

factor in mate attraction. However, Winch (1963) was quite

explicit in stating that attraction, as he studied it, is a

phenomenon independent from satisfaction, happiness, or some other

criterion of quality experienced by marital partners. He found,

for example, that a nurturant individual is frequently attracted

to a receptive mate, and vice versa, and that submissive persons

and controlling others also typically find each other attractive,

but he made no hypotheses about the satisfaction such mates

experience. Although interesting comparisons may be made between

the scope of the present study and the work of Winch as well

as those who subsequently tested his theory, the foci of the

two approaches to interpersonal relations are clearly different.

Here the concern is experienced quality and interactive behavior,

while Winch studied motivational need and interpersonal attraction.

Bales' system nicely avoids the difficulties involved

in assessment of such constructs as "motivational need" and the

problems Winch's work has encountered. In deriving a method

for measuring interactive behavior, Bales outlined for his

factorially derived dimensions the meaning, or value signifi-

cance, an individual's behavior has for others. He stated

that in his work he attempted to reduce much of the variety of

different persons' behaviors into factors on the basis of the

common consequences that different behaviors have for others.

For example, two entirely different behaviors from a motivational

point of view are talking a great deal and showing anger. However,

by excessive talking, a person tends to reduce the power of

others; likewise, in showing and expressing anger, he affects

others by reducing their power. The traits and behaviors Bales

grouped in the U part of his space are associated with each other,

not necessarily because they have a common personality cause

like an underlying drive or need of the actor, but because they

are evaluated in a similar way by other group members on the

basis of the consequences these behaviors have for them. Others

react in a consistent fashion, regardless of the person's moti-

vation, as they come to know the consequences for them inherent

in relating to a certain person.

Another aspect of the present work also requires clari-

fication. Here Bales' (1970) Interpersonal Rating System is

utilized in order to ascertain the behavior patterns of mates

distinguished on the basis of the self-reported quality of their

marriages. Swensen (1973a) discusses levels of theory as

encompassing the behavior segment, the person, or the dyadic

pattern. There are of course other levels, but the point he

made is that the level at which a formulation addresses itself is

important. Instead of the over-all pattern as the datum, systems

other than Bales' would have enabled the present research to study

behavior at other levels. For example, Longabaugh (1963)

developed a category system at the level of the conversational

unit (sentence or sentence equivalent) and found that interaction

was an exchange process that seemed to follow reward and cost

dynamics.

The Bales system was appropriate for the present study of

marriage in which measures of the mates' experience of relationship

quality were hypothetically related by way of social exchange

dynamics to their pattern of interaction with one another. Social

exchange has been shown to follow reinforcement principles, and

the Bales system, by reflecting overt behavioral consequences,

fits together well with such a theoretical perspective.

Hypotheses for the Present Study

The purpose of this research was to study the nature

of the interpersonal interactions between mates to see if there

were differences in the behavior patterns of married couples who

reported high quality and those who reported lower quality. The

preceding review suggested principles of complementarity for

each dimension of interpersonal behavior assessed as the dependent

variable in the study, and these hypotheses were constructed

on the basis of these principles.

Hypothesis 1: As the mates tend to report higher quality

in their experience of their marriage, observers will note that

they tend to relate to one another in a reciprocal manner with

regard to the D-U dimension in interaction; i.e., mates reporting

higher relationship quality will appear more dissimilar in the

D-U types of behavior they exhibit than those reporting lower

quality.

Hypothesis 2: As mates tend to report higher quality

in their experience of their marriage, observers will note that

they tend to relate to one another in a corresponding manner

with regard to the N-P dimension in interaction; i.e., mates

reporting higher relationship quality will appear more similar

in the N-P types of behavior they exhibit than those reporting

lower quality.

Hypothesis 3: As mates tend to report higher quality in

their experience of their marriage, observers will note that

they tend to relate to one another in a corresponding manner with

26

regard to the B-F dimension in interaction; i.e., mates reporting

higher relationship quality will appear more similar in the B-F

types of behavior they exhibit than those reporting lower quality.

CHAPTER III

METHOD

Subjects

Forty-eight young married couples participated in the

present study. A relatively homogeneous sample of married couples

was sought, and for this reason the following demographic criteria

were selected: (1) both mates between 20 and 30 years of age;

(2) married between one and five years; and (3) with no children.

Couples falling within the demographic limitations were

recruited through the assistance of the Office of Student Affairs

at the University of Florida. This office made available a roll

containing the names of all married students attending the uni-

versity. Couples were contacted at random from this list if

they had a telephone, and their participation was invited. Thus

all subjects were obtained from a college population. Appendix

I contains materials which show the approach taken on the phone

with potential participants.

Instruments

Swensen's ,(1973b) Scale of the Feelings and Behaviors

of Love, and the Scale of Marriage Problems (Swensen & Fiore, 1975)

were used to assess relationship quality. These are self report

instruments, and scoring was carried out according to the authors'

standard instructions. Test-retest reliabilities for the sub-

scales of the Love Scale for subjects of similar age to those

studied here range between 0.77 and 0.96, and for the Problem

Scale 0.85.

Relationship quality is here generally defined as the

higher the score on the Love Scale, the greater the quality;

on the Problems Scale the lower the score, the greater the

quality. Fiore (1972) provided norms for the Love Scale which

he obtained by correlating the instrument with other measures

reflecting the qualitative nature of a marriage. The work

he reported indicated one exception to the above definition which

regards a high Love Scale score as indicative of high quality.

Factor 5, reflecting undisclosed feelings, functions in an oppo-

site manner from the remaining factors of the Scale in that

functional couples show a lower score on this factor, while

dysfunctional couples show a high score. In order to achieve

consistency in scale definition, the numerical value a mate

obtained on this factor was subtracted from the maximum possible,

and this transformed score was then treated in fashion similar to

the other love subscales; i.e., a high transformed score indicated

high quality.

The observational Interpersonal Rating System of Bales

(1970) was employed by observers to assess the position of a

subject on the Bales dimensions as he or she interacted with the

mate. Bales' rating form consists of 26 questions, to which

an observer must give an affirmative, negative, or "I don't

know" response. The observers completed the rating questions

for each mate after watching the interactional session. Appendix II

contains the rating form used in this study.

Bales (1970) determined that each of the questions on

the form relates to one, two, or all three of his interpersonal

behavior dimensions. Bales' scoring procedure, in which a person

may receive a maximum score of 18 for each particular direction

(e.g., 18 D or 18 U, 18 N or 18 P, and 18 B or 18 F) has been

modified here so that the bipolar scores of 18 in either direction

for a dimension are converted into a continuous scale of zero

to 36. On the scale employed in this research a mate located

at the extreme D end of the D-U scale (Dl) received a score of

zero, while a mate in the extreme U direction received a 36; a sub-

ject in the extreme N direction received a zero on the N-P dimen-

sion (D2), and one at the extreme P end received a score of 36;

and, similarly, an extreme B mate received a F-B score (D3) of

zero, and an extreme F person, a score of 36.

Observers

Three women, all doctoral students in various social

science fields, worked as observers in this research. Each ob-

server was paid $2.00 per hour for her assistance.

Training in the use of the Bales rating form was con-

ducted with videotapes of the six couples who participated in the

pilot study for the research. The observers used a training

manual (Appendix III) developed in order to show the sense and

meaning of the rating items. The training procedure consisted

of having the observers first view the training tapes and rate

each mate. Item by item they would then state their responses

to the other observers and justify each on the basis of the

training manual. This author participated in the discussion,

also, and at the end of this training the observers appeared to

have an acceptable and consistent understanding of the Bales form.

Pilot Study

A pilot study with six couples was conducted primarily

for the purpose of determining what interactional tasks could

be given to the couples to fulfill the needs of this research.

The necessity of this pilot work was seen in seeking an interac-

tional situation which would best stimulate the couples in a labor-

atory setting to present behaviors typical of their relationship.

Specifically, tasks were sought which would motivate personal

interest and affective engagement from each partner and which

would not be dismissed in a moment or two. Previous work (e.g.,

Shaw, 1963) on interactional tasks existed, and some of the tasks

suggested by this literature were tried out in the pilot study.

However, none seemed as feasible as tasks devised by the present

author, and three of these were used in the study. The three

tasks and their accompanying instructions may be found in Appendix IV.

Procedure

Couples who agreed to participate on the phone were

given an appointment with the experimenter in the Psychology

Building. When they arrived they were again told the procedures

of the experiment and asked to sign an informed consent document.

Each mate was then administered the Love Scale and the Marriage

Problems Scale. Although they completed the inventories in the

same room with each other, the experimenter impressed upon them

the necessity of foregoing collaboration on the scales and the

need to withhold their comments to each other until the end of

the experimental procedures.

When a couple had completed the scales, they were taken

into a videotape studio. An effort was made by the experimenter

to give this taping room an informal appearance, although the

mates generally reported that a homey effect had not been achieved

and that the bright lights were disturbing. The camera was con-

tained out of sight behind a one-way mirror, and the couple

sat in chairs side by side at a coffee table upon which a micro-

phone rested. They were left alone in this room with a task

card and instructions. The experimenter sat in the camera room,

and operated the recording equipment, leaving it only to provide

additional task cards as the mates requested them. A total of

15 minutes of interaction was recorded for observation and rating

by the observers. The couple was allowed to spend all 15 min-

utes on the first task or to request up to two additional tasks.

The majority of the couples continued to discuss the first topic

until stopped by the experimenter, and only two couples finished

with all three in less than the allowed 15 minutes. These latter

two were told to remain in the room and to discuss something of

significance to them.

At the conclusion of the interactional portion the

experimenter discussed the experiment with couples who indicated

an interest. All couples were again informed that feedback would

be available upon compilation of the results.

Data Analyses

Sample means were calculated for each variable studied

in this research. The present sample's scores on the Love Scale

were compared by between group t-tests with norms (Fiore, 1972)

previously compiled for each subscale. Comparisons by t-statis-

tics were also made for sex differences on the dependent vari-

ables.

More extensive analyses were carried out by a Statis-

tical Analysis System (SAS) computer program (Barr & Goodnight,

1972; Service, 1972) in order to find the correlational rela-

tionships between the independent and dependent variables of the

study. A univariate correlation matrix containing all possible

Pearson product-moment correlation coefficients was derived.

The variables were treated at two levels (Swensen, 1973a) in

these computations: the level of the person or individual (Level-I)

and the level of the dyad or couple (Level-II). The Level-I

33

variables were simply the scores as taken from each individual

mate, while the Level-II variables required derivation. They

were calculated as the difference between the mates' scores on

a variable or the summation of their scores.

Using the results of the simple correlational analyses

as guidelines indicating associational trends, selected multiple

regression analyses were computed by the SAS program for multi-

variate relationships that showed promise of significance. These

latter analyses also treated the data at both Level-I and Level-II.

CHAPTER IV

RESULTS

Reliability of the Dependent Variable Measures

The observers in the study showed the following over-all

percentages of agreement (Katz, 1966) in their responses to the

items of the rating form: 01 with 02 = 0.80, 01 with 03 = 0.81,

and 02 with 03 = 0.83. These data indicated that the observers

achieved a high degree of consistency with each other and an

acceptable level of reliability.

Characteristics of the Subject Sample and Comparisons

The mean age of the subjects was 24.1 years (SD = 2.8)

for men, and 23.1 years (SD = 2.9) for women. The mean time

married was 33.0 months (SD = 17.2), and only four couples with

any children (one very young child each) were included. Thus,

the demographic criteria for subjects were essentially fulfilled

by the sample.

The data indicated that the procedure for the recruit-

ment of couples encountered some fairly strong self-selection

processes in the original pool of potential subjects, which may

have affected the sample's composition. Only 20% of the couples

contacted eventually participated in the study. Of the remain-

ing 80%, 23% were not acceptable in terms of the study's

demographic criteria, leaving 57% who declined for various

reasons. The largest portion of these latter couples stated

that they were simply not interested in participation, while

others gave excuses that appeared to make participation impos-

sible (e.g., spouse out of town for extended period, separation,

etc.) or failed to appear at the appointed time.

Data reflecting characteristics of the present sample

on the variables relevant to the hypotheses of this study

--relationship quality and interactional behavior--are next

presented. With regard to the independent variable, the spouses'

experienced quality of their marriage, comparisons can be made

between the means of the present couples' scores on the Love

Scale and of norms Fiore (1972) acquired for two groups of

couples, one having functional marriages (FM) and one dysfunc-

tional marriages (DM). The mean for each Love Scale factor

for the present sample is presented in Table I along with the

means of Fiore's groups.

Table I shows that the present sample of couples was

a highly FM group. Another way to present this result is by

comparing the over-all Love Scale score with Fiore's over-all

norm. This measure shows that 79% of the couples in the present

sample were above Fiore's FM norm (mean = 245), only 2% were

below the DM norm (mean = 217), and 19% of the couples were between

the two norms.

Norms for the Problem Scale are not available at the

present time. The Problem Scale results for the sample studied

TABLE I

COMPARISON OF PRESENT SAMPLE WITH LOVE SCALE NORMS

Love Scale
Factor Present Sample Fiore's Groups Comparisons

t-Tests: t-Tests:
Present Present
mean SD FM mean DM mean with FM with DM

1 43.7 7.72 38.8 31.8 4.4 10.7

2 49.5 4.37 45.2 41.1 6.8 13.0

3 43.3 5.11 37.5 36.0 8.0 11.8

4 62.5 4.59 61.2 53.3 2.0 10.0

5 25.1+ 7.94 26.4+ 31.6+ 1.2 5.6

6 37.0 4.43 34.8 32.6 3.5 7.3

p = 0.1, all other t-values: p C .025 (DF = 95)
+ Factor 5, undisclosed feelings, reflects numerically
the reverse of the other Love Scale factors, as noted in Chapter
III, with FM couples expected to score lower than DM couples.

here were as follows: Factor 1 = 5.67 (SD = 3.83), Factor 2 =

2.60 (SD = 2.61), Factor 3 = 2.23 (SD = 1.87), Factor 4 = 2.61

(SD = 1.88), Factor 5 = 2.34 (SD = 2.12), Factor 6 = 1.96

(SD = 1.69). The maximum possible on the scale is 84, and these

results are quite low relative to the scale's ceiling, as would

be expected in a highly FM sample.

The dependent variable of this study, the Bales dimen-

sions, showed that the vast majority of the mates of these highly

FM couples appeared in the UPF octant of the Bales interactional

space. The mates were distributed in the Bales space in terms

of the Bales personality types as follows: UPF = 69 (72%);

UPB = 6 (6%); DPF = 6 (6%); UNF = 4 (4%); PF = 3 (3%); DNF = 2

(2%); DNB = 2 (2%); UNB, UP, UF, and DP = 1 each (1%).

Treating the dependent variable data as dimensional

scores along the continuous scales described in Chapter III,

the results for husbands were: mean on Dl = 23.6 (SD = 3.0);

mean on D2 = 28.6 (SD = 7.5); and mean on D3 = 22.7 (SD = 3.2).

For wives: D1 = 22.2 (SD = 4.4); D2 = 29.7 (SD = 6.5); D3 = 21.6

(SD = 4.1). Tests by t-statistics for differences between the

sexes on their scores on the Bales dimensions were not statis-

tically significant.

Simple Correlational Relationships Between the Variables

The quality of relationship variables, Level-I, corre-

lated with each other and with time married are presented in

Table II. These results reflect on the validity of the two

scales used to assess mates' experienced quality and the effect

of time on experienced quality.

Table II shows that the Love Scale scores and Problem

Scale scores are negatively correlated with each other. This

negative relationship was found in the scores of both husbands

and wives for the two instruments and indicates support for the

definition of quality used in the study; i.e., quality is taken

as a high score on the Love Scale and a low score on the

Problems Scale. It is also apparent that within a couple, the

husband's measures of quality are significantly correlated

with the wife's, indicating that the husbands' and wives'

experiences of quality in the marriages studied here are not

markedly disparate.

Another result of interest shown in Table II is the

significant relationship between the length of the marriage and

the husband's experience of quality. A negative correlation

was found for both love and total quality with time married for

Appendix V contains an explanation of the abbreviations for
terms used in Table II and all subsequent tables. It is suggested
that this Appendix be consulted before an attempt is made to
read the tables.

HL(

MOSMARR -0.
(.0

HLOVET

HPROBT

HTOTQUAL

WLOVET

WPROBT

TABLE II

CORRELATIONS BETWEEN QUALITY VARIABLES
AT LEVEL-I AND TIME MARRIED

VET HPROBT HTOTQUAL WLOVET WPR(

.27, 0.30 -0.34 0.00 0.
)5) (0.04) (0.01) (0.99) (0.

-0.33 0.70 0.49 -0.
(0.02) (0.00) (0.00) (0.

-0.90 -0.39 0.
(0.00) (0.01) (0.

0.51 -0.
(0.00) (0.

-0.
(0.

OBT

15
29)

30
04)

52
00)

53
00)

34
02)

Significance level shown in parenthesis.

WTOTQUAL

-0.12
(0.58)

0.43
(0.00)

-0.57
(0.00)

0.63
(0.00)

0.67
(0.00)

-0.92
(0.00)

the males, and a positive correlation may be noted between

time and their reports of problems. These associations, however,

were not found for the wives' reports.

Table III shows results in which quality variables,

Level-I, have been correlated with the Bales dimensions, also

Level-I. These findings were pertinent to the theoretical

questions of the study and were part of the results used for

determining possible multivariate relationships, examined by

multiple regression analyses, between the two types of data

studied here.

The strongest findings in Table III show a relation-

ship between each mate's quality scores and their own D2 scores

as well as their mate's D2 score. Only the correlations for

wives are statistically significant, with the husband's coef-

ficients indicating a trend. These results show that the ex-

perience of high quality (much reported love and few problems)

is associated with P behavior for both mates, particularly for

wives.

Two other findings in Table III, to be examined later

in detail, may be noted. First, the husband's D2 score approaches

a significant (P = 0.11) positive correlation with time married.

Thus, while Table II showed that over time the husband's exper-

ience of quality declines, his observed P behavior tends to

increase. The second result is an association between the

wife's report of quality (primarily the problem component)

and the husband's Dl behavior score. This finding indicates

TABLE III

CORRELATIONS BETWEEN THE INDEPENDENT VARIABLE
AND THE DEPENDENT VARIABLE LEVEL-I

HD1 HD2 HD3 WD1 WD2 WD3

MOSMARR -0.04 0.22 0.15 0.00 0.14 0.11
(.80) (.11) (.31) (.99) (.66) (.53)

HLOVET -0.20 0.26 -0.05 0.06 0.15 0.07
(.16) (.07) (.73) (.67) (.29) (.64)

HPROBT -0.16 -0.19 -0.08 0.04 -0.27 -0.04
(.27) (.17) (.60) (.78) (.06) (.79)

HTOTQUAL 0.03 0.27 0.04 0.00 0.27 0.06
(.83) (.06) (.80) (.98) (.06) (.69)

WLOVET 0.19 0.34 0.10 0.06 0.31 -0.02
(.19) (.02) (.52) (.71) (.03) (.86)

WPROBT -0.40 -0.30 -0.11 -0.15 -0.35 0.15
(.00) (.04) (.55) (.31) (.02) (.28)

WTOTQUAL 0.39 0.37 0.13 0.13 0.39 -0.13
(.00) (.01) (.64) (.65) (.00) (.63)

Significance level shown in parenthesis.

that wives tended to report higher quality as their husbands'

behavior showed a trend toward U.

A final Level-I correlational analysis was performed

in order to find the relationships between the husband's and

wife's observed behavior. Table IV shows these results.

These results indicate significant positive correlations

between husbands and wives on D2 and D3. Thus, with regard to

N-P behaviors and B-F behaviors husbands and wives in the present

sample tended to be similar.

The data were also analyzed to find the simple correla-

tional relations between the two categories of variables, reported

quality and observed interactional behaviors, at Level-II.

These findings were relevant to testing the hypotheses of the

study and were used to delineate trends for multivariate ex-

ploration. Table V presents these Level-II results.

The observed differences in mate's interpersonal behavior

scores represent a numerical method for determining the degree

of similarity or dissimilarity of mates' dimensional behaviors,

and these dyadic units have been correlated with their reports

of quality as preliminary tests of the study's hypotheses.

None of the correlation coefficients in the first three rows

of Table V, indicating the difference in husbands' and wives'

behaviors, are significant at the 0.05 level of confidence.

This finding indicates that mutual reports of experienced quality

(composed of either the sum of the husband's and wife's indi-

vidual reports or defined on the basis of the disparity of

TABLE IV

CORRELATION OF HUSBAND'S AND WIFE'S DIMENSIONAL BEHAVIOR

HDI HD2 HD3

WD 1 -0.07 0.26 -0.22
(.62) (.07) (.13)

WD 2 -0.00 0.51 0.05
(.98) (.00) (.73)

WD 3 -0.20 0.27 0.29
(.17) (0.06) (0.04)

Significance level shown
in parenthesis.

TABLE V

CORRELATION OF LEVEL-II MEASURES OF RELATIONSHIP QUALITY
WITH LEVEL-II INDICES OF MATES' INTERACTIONAL BEHAVIORS

COUPLE COUPLE COUPLE COUPLE COUPLE
LOVESUM PROBSUM TOTQUAL LOVEDIFF PROBDIFF
DIFFD 1 -0.07 0.12 -0.12 0.06 0.07
(.65) (.59) (.57) (.68) (.61)

DIFFD 2 -0.18 0.23 -0.24 0.16 0.15
(.22) (.11) (.09) (.26) (.29)

DIFFD 3 -0.09 0.22 -0.21 0.20 0.13
(.53) (.12) (.16) (.16) (.64)

SUMD 1 0.05 -0.25 0.20 0.02 -0.38
(.73) (.08) (.16) (.88) (.01)

SUMD 2 0.36 -0.36 0.42 0.01 -0.18
(.01) (.01) (.00) (.92) (.23)

SUMD 3 0.03 -0.01 0.02 -0.18 0.04
(.81) (.94) (.88) (.23) (.77)

Significance level shown in parenthesis.

their reports) are not strongly related to the similarity or

dissimilarity in their observed behaviors. However, in the

cases of D2 and D3, portions of the data approached significance.

All of the trends were in the direction expected on the basis

of the study's hypotheses, as shown by the positive or negative

sign of the association. The data showed that the sum of the

mates' quality scores approached a negative correlation with

the D2 and D3 differences in their observed behavior, indicating

that as reported quality increases, the amount of dissimilarity

in the mates' behaviors decreases. The measures of their disa-

greement on love and problems, found to be associated with lower

reported quality (Fiore, 1972), also tend to be associated with

greater dissimilarity in their observed behaviors. These latter

results, however, have even less statistical confidence than

the trends of the love and problem sums. These trends, indica-

ting that high quality may be associated with smaller differences

between spouses on D2 and D3, were given further consideration

in the multiple regression analyses.

The last three rows of Table V present the correlations

of the quality variables with the totals of the spouses on the

observed behavior variables. These behavioral sums indicate

the strength of the dyad's tendency to behave in the U, P, or

F directions, and the theoretical significance of these sum

terms is taken up in Chapter V. Only the coefficients for

D2 are statistically significant, and these indicate that

increases in total reported quality for the dyad are associated

with a trend toward increases in the total amount of P behavior

observed in their interaction.

Additional correlational analyses were carried out

relating reported quality variables, Level-I, to the interactional

behavior variables, Level-II. The data were analyzed in this way

because it was believed that dyadic measures of reported quality

which did not discriminate between couples where the mates were

in close agreement in their experience of quality and couples

where the mates experienced the marriage differently might

tend to confound some of the relationships in the results.

These results appear in Table VI.

The results presented in the first three rows of Table

VI, as in Table V, are tests of the research hypotheses of this

study. Only two of the correlations related to the differences

in mates' scores on the dimensions are of statistical signifi-

cance, and these show a negative association between the wife's

report of quality (primarily her Problem Scale score) and dis-

similarity in their D2 behavior. However, the positive and

negative correlational trends of the data were, in every case,

in the hypothesized direction and these trends merited further

analyses by multiple regression analyses.

The last three rows of Table VI present the associations

between the individual mates' quality reports and the sum of their

behavioral dimensions. It is apparent that for Dl only the

wives' reports are of significance. The results show that as

TABLE VI

INDIVIDUAL REPORTS OF QUALITY CORRELATED WITH
LEVEL-II INDICES OF MATES' INTERACTIONAL BEHAVIORS

HLOVET HPROBT HTOTQUAL WLOVET WPROBT WTOTQUAL

DIFFD 1 -0.04 0.06 -0.07 -0.06 0.14 -0.14
(.74) (.64) (.62) (.65) (.32) (.66)

DIFFD 2 -0.20 0.03 -0.12 -0.10 0.34 -0.31
(.16) (.79) (.57) (.50) (.01) (.03)

DIFFD 3 -0.16 0.23 -0.24 0.00 0.16 -0.13
(.29) (.11) (.09) (.99) (.25) (.62)

SUMD 1 -0.06 -0.06 0.02 0.16 -0.36 0.35
(.67) (.69) (.91) (.29) (.01) (.01)

SUMD 2 0.24 -0.26 0.31 0.38 -0.37 0.44
(.09) (.06) (.02) (.01) (.01) (.00)

SUMD 3 0.02 -0.06 0.06 0.04 0.05 -0.02
(.88) (.65) (.68) (.78) (.75) (.88)

Significance level shown in parenthesis.

wives report greater quality, the total observed behavior in the

dyad tends toward the U direction. With one exception, every

correlation for D2 is significant. These results indicate

that as both mates individually report greater quality, the

couples' behavior tends to increase in the P direction.

Findings from Multiple Regression Analyses

On the basis of the results and trends found in the

simple correlational analyses, multiple regression analyses were

computed for associations between variables that showed promise

of significance. These analyses were carried out using data

from each of the two levels of interest to the present work.

The remainder of this chapter is a presentation of the statis-

tically significant multiple regression equations relating the

independent and dependent variables.

With regard to data relevant to hypothesis I, there were

no significant correlation coefficients associating mate's dis-

similarity on Dl with the independent variables. On the basis

of this finding, no significant multivariate functions for pre-

dicting Dl differences between mates were expected, and although

several potential equations were attempted, none were found.

However, consistencies were noted in the correlational

results for the couple's behavioral sum on D1 as it was related to

quality scores. Table VII presents equations predicting the

couple's sum on Dl from the independent variables, and the import of

these findings is addressed in Chapter V.

49

TABLE VII

MULTIPLE REGRESSION EQUATIONS PREDICTING THE SUM OF
MATES' D1 SCORES FROM THE INDEPENDENT VARIABLES

1. E(SUMD 1) + 49.16 + (-0.02)COUPLEPROBSUM + (-0.30)COUPLEPROBDIFF;

R = 0.15, p = 0.02.

2. E(SUMD 1) = 46.91 + (0.009)WLOVET + (-0.18)WPROBT;

R2 = 0.13, p = 0.04.

3. E(SUMD 1) = 50.24 + (-0.05)HLOVET + (O.10)HPROBT + (0.04)WLOVET
+ (-0.24)WPROBT;

R2 = 0.20, p = 0.04.

Equation 1, predicting from Level-II independent var-

iables, shows that as the reported quality increases, the

totality of the mates' observed U behavior also increases.

Equations 2 and 3 predict the D1 sum from the individuals'

reports of love and problems. Equation 2 shows a trend similar

to equation 1 in that for the wife, as her reports of quality

increase, the amount of observed U behavior increases. Equation

3, however, shows that the opposite trend is true for husbands;

i.e., as his experience of love decreases and problems increase,

the total U behavior increases. Thus the trend in equation 1

is determined by the wives' scores, and in fact it is apparent

in equation 3 that her problem report carries significantly

greater predictive weight.

Many consistencies were noted above in the results

relating the independent variables to D2, Level-II. Multiple

regression equations of significance, or approaching signifi-

cance, predicting D2, Level-II, are presented in Table VIII.

The first seven equations in Table VIII represent tests

of hypothesis 2. The remaining equations in the table present

a dyadic index of interactional behavior similar to the dependent

variable term employed in Table VII. As noted above, the theore-

tical significance of the sum scores used in these equations is

discussed in the chapter following.

Equations 1 through 5 in Table VIII predict the dif-

ference between mates' D2 scores from independent variables

TABLE VIII

MULTIPLE REGRESSION EQUATIONS PREDICTING THE DIFFERENCE AND
SUM OF MATES' D2 SCORES FROM THE INDEPENDENT VARIABLES

1. E(DIFFD 2) + 18.97 + (9.43)HTOTQUAL + (-28.30)WTOTQUAL;

R2 = 0.11, p = 0.07.

2. E(DIFFD 2) + 10.09 + (-0.04)HLOVET + (-0.13)HPROBT + (O.O1)WLOVET
+ (0.23)WPROBT;

R2 = 0.07, p = 0.08.

3. E(DIFFD 2) = 2.02 + (-0.12)HPROBT + (0.24)WPROBT;

R2 = 0.15, p = 0.02.

4. E(DIFFD 2) = 7.81 + (-0.02)HLOVET + (0.17)WPROBT;

R2 = 0.13, p = 0.04.

5. E(DIFFD 2) = 11.73 + (-0.03)HLOVET + (-0.14)HPROBT + (0.23)WPROBT;

R2 = 0.17, p 0.04.

6. E(DIFFD 2) = 9.41 + (-0.01)COUPLELOVESUM + (0.06)COUPLEPROBSUM;

R2 = 0.06, p = 0.25.

7. E(DIFFD 2) = 1.97 + (0.07)COUPLELOVEDIFF + (0.16)COUPLEPROBDIFF;

R2 = 0.06, p = 0.26.

8. E(SUMD 2) = 20.70 + (0.02)HLOVET + (-0.02)HPROBT + (0.15)WLOVET
+ (-0.32)WPROBT;
2
R = 0.21, p = 0.04.

9. E(SUMD 2) = 39.45 + (0.09)HLOVET + (-0.28)HPROBT;
2
R = 0.10, p = 0.10.

10. E(SUMD 2) = 22.37 + (0.15)WLOVET + (-0.33)WPROBT;

R2 = 0.20, p = 0.006.

11. E(SUMD 2) = 25.68 + (0.07)COUPLELOVESUM + (-0.19)COUPLEPROBDIFF;

2 0.18, p 0.01.
R = 0.18, p = 0.01.

at Level-I (individuals' scores), while equations 6 and 7 treat

both variables at Level-II dyadicc scores). The latter two

only approach significance and are shown because of the trends

they indicate. Equation 1 predicts the mates' dependent vari-

able differences from their over-all quality scores, and equa-

tions 2 through 5 predict from the various components of the

husband's and wife's report of quality, i.e., love and problems.

It is obvious from the first seven equations that a

strong predictive relationship exists in which the quality vari-

ables account for up to 17% of the variance in the difference

between the mates' D2 scores. Hypothesis 2 (for D2) predicts

that as less quality is reported, the mates should appear to

observers as more dissimilar in their D2 behaviors. The hypo-

thesized trend is followed by the husbands' love scores and

the wives' problem scores, but not by the trend of his problem

scores. (The one significant equation including the wives'

Love Scale scores indicated a trend for this data opposite to

that hypothesized; however, the weight this factor carried was

negligible, and there seemed to be no consistent relationship

for this variable.) One result of these opposing trends for

the husband and wife is that in equations 1, 6, and 7, where

the components are combined, the strength of prediction is greatly

reduced.

Equations 8 through 11 predict the sum of the spouses'

D2 scores from the independent variables. The D2 sum reflects

the total amount of P behavior observed in their interaction.

These equations account for up to 21% of the variance in depen-

dent variable values, a slightly better prediction than the best

for the mates' differences on D2 in equations 1 through 7. In

fact, by using the behavioral sum as the dependent variable

term, the results do not show the sex differences found in the

first seven equations, which indicated opposing trends for hus-

bands and wives. These last equations show unequivocally that

as the mates reported greater quality, the observers noted a

larger total of P behavior.

Tests of hypothesis 3 were made by correlation and regres-

sion analyses in which the mates' differences on D3 constituted

the dependent variable term. Although no equations relating

the independent variables to the mates' D3 differences were

statistically significant, Table IX presents four which approach

significance.

Hypothesis 3 predicted that as less quality is reported

by the mates, they should appear to observers as more dissimilar

in their D3 behaviors. All four of the equations in Table IX are

in accord with the hypothesis; i.e., as the mates reported less

quality in their marriage, observers noted greater differences

in their D3 behavior. The associations noted in these equations

are rather weak, however, with the best predictors accounting

for only 9% of the dependent variable variance.

With regard to the mates' D3 total, i.e., the total

amount of F behavior observed, it was shown that no simple cor-

relations relating the independent variables to the D3 sum even

TABLE IX

MULTIPLE REGRESSION EQUATIONS PREDICTING THE DIFFERENCE
BETWEEN MATES' D3 SCORES FROM THE INDEPENDENT VARIABLES

1. E(DIFFD 3) = -0.22 + (0.05)COUPLELOVEDIFF + (0.05)COUPLEPROBSUM;

R = 0.09, p = 0.10.

2. E(DIFFD 3) = 0.71 + (0.06)COUPLELOVEDIFF + (0.10)COUPLEPROBDIFF;

R2 = 0.07, p = 0.20.

3. E(DIFFD 3) = 4.91 + (-0.01)HLOVET + (0.08)HPROBT;

R2 = 0.06, p = 0.25.

4. E(DIFFD 3) = 0.83 + (0.08)HPROBT + (0.02)WPROBT;

R2 = 0.06, p = 0.27.

approached significance. As would be expected from this finding,

no multivariate statistics were found to be of significance for

this dependent variable term, and for this reason, none are

reported.

While the Level-II dependent variable terms providing

data on the dyadic pattern of a couple's interaction are of

greatest theoretical interest to this study, it was noted in the

simple correlational results that predictions for HDI, HD2, and

WD2 might be made from the independent variables. No other

Level-I dependent variables showed significant associations

with the mates' reports of quality, and multiple regression

analyses were made using only the above terms as dependent vari-

ables. These results are presented in Table X. Although the

significance of these data is discussed in the next chapter, it

may be said here that these equations make it possible to note

some trends which contribute to the patterns found in the Level-II

results.

It is apparent from the equations presented in Table X

for E(HD1) that the values of this variable may be predicted

from the independent variables. Equations 1 and 2 show that

the husband's Dl score can be predicted from both the husband's

and wife's individual experience of quality. The wife's report

of problems appears to be the most salient factor for prediction,

although both mates' reports of love are strong contributors

in the equations. The wife's report of love functions in the

opposite fashion from the husband's, in that the wife's love

TABLE X

MULTIPLE REGRESSION EQUATIONS PREDICTING DEPENDENT VARIABLE
SCORES, LEVEL-I, FROM THE INDEPENDENT VARIABLES

Husbands' Dl behavior:

1. E(HD 1) = 31.40 + (-0.06)HLOVET + (0.01)HPROBT + (0.04)WLOVET
+ (-0.14)WPROBT;

R = 0.32, p 0.002.

2. E(HD 1) = 34.54 + (-0.04)HLOVET + (-0.08)HPROBT;

R2 = 0.10, p = 0.10.

Husbands' D2 behavior:

3. E(HD 2) = -1.58 + (6.28)HTOTQUAL + (33.23)WTOTQUAL;

R = 0.14, p = 0.03

4. E(HD 2) = -2.71 + (0.15)MOSMARR + (0.06)HLOVET + (-0.05)HPROBT
+ (0.05)WLOVET + (-0.15)WPROBT;
2
R = 0.25, p = 0.02.

5. E(HD 2) = 1.03 + (0.03)HLOVET + (0.03)HPROBT + (0.08)WLOVET
+ (-0.15)WPROBT;

R2 = 0.16, p = 0.10.

6. E(HD 2) = 12.10 + (0.07)HLOVET + (-0.10)HPROBT;

2 0.0 0.14.
R = 0.08, p = 0.14.

57

Wives' D2 behavior:

7. E(WD 2) = 2.41 + (3.30)HTOTQUAL + (32.46)WTOTQUAL;
2
R = 0.16, p = 0.02.

8. E(WD 2) = 19.68 + (-0.01)HLOVET + (-0.05)HPROBT + (0.06)WLOVET
+ (-0.16)WPROBT;

R2 = 0.16, p = 0.10.

9. E(WD 2) = 17.61 + (-0.04)HPROBT + (0.05)WLOVET + (-0.16)WPROBT;

R2 = 0.16, p = 0.05.

10. E(WD 2) = 16.08 + (0.06)WLOVET + (-0.18)WPROBT;

R = 0.16, p = 0.02.

score is positively related to the husband's D1 score, while

the husband's love score carries a negative sign. This finding

indicates that in interactions where the husband is observed

as tending toward U, the wives have reported greater quality

(greater love, fewer problems) and in contrast, the husbands

have reported less love. (The husband's report of problems is

associated in a manner similar to the wife's; however, in

equation 1 it has a negligible weight and appears to have little

predictive significance.)

Equations 3 through 6, Table X, show a strong predic-

tive relationship between the independent variables and the

husband's D2 behavior. Equation 3 relates the individual mates'

total experience of quality to the husband's D2 behavior. The

wife's report of quality has greater predictive weight than the

husband's, although both factors have a positive sign, indicating

that as the mates' experience of quality increases, the husband's

behavior is observed as tending in the P direction. Equations

4, 5, and 6 present the husband's and wife's experience of

quality in terms of the love and problem components and predict

HD2 from these. The trends observed in these equations are con-

sistent with those noted for the over-all quality factors. Similar

to the equations for HD1 the wife's report of problems is the

most salient factor here also. In addition, equation 4 includes

the factor MOSMARR and indicates that the length of the marriage

accounts for a sizable portion of variance in the prediction of

the husband's D2 behavior. This equation indicates that the

longer the couple has been married, the greater the husband's

observed tendency toward P behavior.

Equations 7 through 10 show that the wife's observed

D2 behavior can be predicted from the independent variables.

The trends here for WD2 are similar to the trends noted for HD2.

Equation 7 shows that the wife's global report of quality is

much more significant in predicting her D2 behavior than the

husband's, although both the mates' reports are positively

associated with an increase in the wife's movement toward P

behavior. Equations 8, 9, and 10 present the love and problem

components of the mates' reported quality. The husband's report

of love is practically negligible in accounting for the wife's

D2 behavior, while her report of problems has the greatest

weight. These results show that as mates' reports of problems

grow in magnitude, observers scored the wife's behavior as tending

away from the P direction. The opposite relationship was found

for the wife's report of love; i.e., as she reported more love,

she tended toward P behavior. The time-married factor showed

no predictive significance for wives' D2 behavior, in contrast

to the relationship reported for the husbands' D2 behavior and

this factor.

CHAPTER V

DISCUSSION

Reliability and Validity of Present Results

The observers of this study agreed at a reasonably

high level in their assessments of interactional behavior, and

this result indicates that the dependent variable data are

reliable. In addition, with respect to the demographic limita-

tions outlined for the study, the subjects were well within these

guides and constituted a relatively homogeneous population of

young, college-educated, childless, recently married couples.

In these two respects, in terms of sampling procedures, the re-

quirements of the study appear to have been fulfilled adequately.

The data also showed, however, that the rate of parti-

cipation among couples contacted as potential subjects was fairly

meager. Self-selection, then, was a prominent factor in the

assemblage of the sample. There is no way of knowing the effects

of self-selection factors on the data, and the couples contacted

surely had a variety of reasons for declining to participate,

but it is apparent that DM couples were not represented in the

sample here. The strongest and clearest finding of this study

was that the great majority of the couples who participated

reported their experience of marriage was very high in love, with

few problems. The comparisons with Fiore's (1972) samples

showed that the present sample included exclusively FM couples.

Along with the self-selection factors operative in the

procedures for obtaining couples, other confounding processes

may have influenced this study. It is possible that the persons

who did participate responded to the independent variable mea-

sures with a social desirability set. The items of the Love

and Problem Scales are quite direct in the way they ask about

feelings toward and experiences with the mate, and it is easy

for a person to respond to these inventories from the perspec-

tive of how they would like the relationship to be or how they

would like others to see it.

It would have been possible to assess the social desir-

ability of the items, but such measures were not taken in the

present research and no previous data are available that would

allow social desirability to be studied as a variable in this

work. In any case the point that seems to override such possible

flaws in the study's design is that while social desirability

and other factors may have diluted the applicability of the

findings, it can also be said that any significant results

presented here were found almost in spite of possible confounding

factors. Thus, while the trends noted in this report consequently

lose some of their generality, they still possess a certain

clarity because the persons for whom they apply are well defined

and the possible confounding factors are recognized.

Along with this consistency among subjects on the inde-

pendent variables, it was found in this sample of FM couples

that the large majority also appeared to the observers to tend

in their interactions toward the UPF octant of the Bales (1970)

three-dimensional space; most of the remaining subjects appeared

in the parts of this space close to the UPF octant. In Chapter I

the study of Swensen (1973a) was reported in which he found no

consistent patterns of life style for FM persons. Although

Swensen had attempted to categorize the couples according to

the Bales dimensions, his data were not obtained in so direct

a way as by observers rating interactions they witnessed. His

data were instead derived from the mates' self reports about

their lifestyle. In contrast, the present sample of FM subjects

tended strongly in their ongoing behaviors in the UPF direction

and also showed the more subtle results discussed throughout

this chapter.

It is possible that the observed behaviors were sig-

nificantly biased by the experimental procedure rather than

being a representative sample of the mates' interaction with

each other. Several of the subjects remarked that they had an

"on-stage" feeling and believed that this anxiety strongly

influenced their behavior, thus implying that their behavior

was public behavior and their actions were directed to third

persons, rather than the mate, and that their dyadic behavior

would have been significantly different from what the observers

were able to see. However, other participants reported that

they were at ease and that their interaction in the laboratory

was quite typical.

The problems presented by the implications of unnatural-

ness are of course inherent in any laboratory research. While

they lead to justified criticism, the questions arising may also

be regarded as limited in scope. Even though the experiment

created tension, it is possible to assume that the mates handled

this tension in ways typical of their interaction. The situation

may have been of a particular kind, to their perceptions, but

it should not have been unique or the primary determiner of

their interaction. It seems reasonable to accept the dependent

variable data as valid and representative and to proceed from

the finding that in the FM sample, consistent patterns of inter-

active behavior in terms of the Bales dimensions were observed.

Data pertaining to the validity of the independent var-

iables were presented in Chapter IV. It was noted that the

Love and Problem Scales were related to each other in a fashion

consistent with the study's definition of quality and that there

was considerable agreement among the husband's and wife's reports

of quality.

These constraints thus define the context in which all

subsequent findings must be viewed. Within this context the

present study seems to show meaningfully how marriage partners

integrate with each other the three basic dimensions of inter-

personal interaction: power, affection, and social movement.

Hypotheses Tests

The hypotheses formulated for this research were tested

by the data through correlational and regression analyses. It

has been noted that the hypotheses were being tested here, not by

findings from a broad sample, but in a closely homogeneous group,

and therefore, questions about interaction dynamics across a

large range of marital situations could not be answered by the

present results.

Hypothesis 1 stated:

As the mates tend to report higher quality
in their experience of their marriage,
observers will note that they tend to
relate to one another in a reciprocal
manner with regard to the D-U dimension
in interaction; i.e., mates reporting
higher relationship quality will appear
more dissimilar in the D-U types of behavior
they exhibit than those reporting lower
quality.

The results of analyses which treated the difference

between the mates' D1 behaviors as the dependent variable serve

as this hypothesis' test. The analyses indicated that there were

no significant relationships between the independent variables

and the mates' D1 differences. Therefore, hypothesis 1 was

necessarily rejected.

However, the data implied that an alternative hypothesis

might be appropriate for accounting for the dynamics by which

the mates exchanged their Dl behavior. A significant trend

appeared in the data which showed that as the wife experienced

greater quality, the sum of the couple's behavior on Dl tended

in the U direction. The husband's report of quality, though,

was associated in the opposite manner; i.e., as the husbands

reported less quality, the couple's total Dl behavior tended

toward U. The wives' reports carried significantly greater

weights, which accounted for the combined husband and wife quality

score being similar to the trend for wives.

These results clearly showed that the hypothesized

social exchange dynamic was not applicable in any straightforward

manner to the data of the present sample. The results showed

that experienced quality was related to Dl behavior in an unex-

pected fashion and that the husband's report of quality was

associated with D1 behavior in a way different from the wife's.

Alternative explanations of these results are discussed in the

next section of this chapter.

Hypothesis 2 stated:

As mates tend to report higher quality in
their experience of their marriage, obser-
vers will note that they tend to relate
to one another in a corresponding manner
with regard to the N-P dimension in inter-
action; i.e., mates reporting higher rela-
tionship quality will appear more similar
in the N-P types of behavior they exhibit
than those reporting lower quality.

The results of the present study gave strong support to

hypothesis 2, and there were several forms which this support

took. First, the results indicated a strong positive correlation

between the husband's and wife's D2 scores. Since most of the

entire subject pool of the study reported high quality, there

appeared to be a strong association between experienced quality

and mates' D2 similarity. Second, the simple correlational

findings indicated a trend in which the difference quantity

between mates on D2 was negatively associated with reported

quality, although these results for the most part only approached

statistical significance. Finally, the multiple regression

findings showed that larger differences in the mates' D2 scores

could be predicted for couples who reported lower experienced

quality than those who reported high quality.

The multiple regression analyses indicated an additional

trend, however. It appeared that the prediction of greater

differences in mates' D2 scores from diminished reports of quality

depended primarily on the wives' Love and Problem Scale scores,

while the husbands' data actually showed some indications (though

with significantly less weight) of a counter trend. When the

mates' D2 scores were summed, showing the total tendency of the

spouses toward P behavior, the results for the husband and wife

were consistent with each other, and a larger portion of the

dependent variable's variance was accounted for by the indepen-

dent variable. In the latter equations also, the wives' quality

reports carried greater weight. These results regarding the

sum of the mates D2 behavior thus indicated that the data supported

a modification of hypothesis 2 to the effect that mates who

reported higher experienced quality showed greater tendencies

toward P behaviors than those who reported lower quality. This

modification implied that persons who may be similar but tend

toward N behaviors in their interaction would not experience

as much quality as P persons. This matter is also discussed

in the next section.

Hypothesis 3 stated:

As mates tend to report higher quality in
their experience of their marriage, obser-
vers will note that they tend to relate to
one another in a corresponding manner with
regard to the B-F dimension in interaction;
i.e., mates reporting higher relationship
quality will appear more similar in the B-F
types of behavior they exhibit than those
reporting lower quality.

The findings of this study lent support to hypothesis 3.

There were no indications in the results that alternative hypo-

theses would give an equal or better account of the data. The

support for hypothesis 3, as for hypothesis 2, came in several

forms. First, there was a significant correlation between the

husband's and wife's D3 scores for the entire sample, and since

the sample as a whole consisted of FM couples, a trend of

association between reported quality and D3 similarity was

inferred. Second, a correlational trend was found, although it

only approached significance, in which increases in reported

quality were associated with decreases in the mates' D3 difference.

The third piece of evidence came from the multiple regression

equations developed for predicting the mates' D3 difference.

These equations only approached significance, but they indicated

a clear trend in which the D3 difference was negatively associated

with the various terms reflecting reported quality.

Over-all then, only hypothesis 3 was supported without

the implication that an alternative hypothesis would be more

appropriate, although at only a moderate degree of confidence.

For the remaining hypotheses, hypothesis 2 received support, but

an alternative hypothesis was suggested by the data, and the

other, hypothesis 1, was unsupported, with strong indications

that an alternative hypothesis was required. These implications

are discussed in the next section. Several comments can be made

here, however, about D3, the hypothesis about which there was

the least-confounded support in the study.

This last dimension of Bales, B-F, refers to the attitudes

a person holds toward tasks and projects. The two other dimen-

sions of Bales may be regarded as related to value directions

that are relevant primarily to the interaction between the spouses,

and in fact these two dimensions are basically the areas of con-

cern to the role theory view of marriage: power, or dominance,

and affection. The B-F dimension, however, may be seen as refer-

ring to factors beyond the immediate interaction of the mates, and

this dimension seems to reflect the person's orientation to mat-

ters of the world external to a particular interaction. The

B-F dimension seems most closely a measure of the feelings and

attitudes one has toward things in general, rather than toward

the mate. The data here showed that a person who is oriented

to the world in terms of attitudes consistent with conventional

goals found greatest satisfaction with a mate who is similar and,

likewise, a person who is more concerned with fantasies, ideals,

or theoretical possibilities than accomplishment found a mate

of similar orientation most compatible.

Alternative Concepts of Interpersonal Dynamics
Relating to the Present Results

The results have shown that theoretical constructions

other than those used in formulating the original hypotheses

may be appropriate for associating the mates' D1 and D2 behavior

patterns with their reports of marital quality. The hypotheses

of the study were tests of the dynamics Carson (1969) proposed

for social exchange. Usually, for any set of empirical data,

numerous explanatory hypotheses are potentially available, but

the discussion here is limited to the two perspectives introduced

in Chapter I: role theory concepts and social exchange.

The results showed a difference between husbands and

wives in the way their reports of experienced quality were

associated with D1 and D2 behavioral trends. Sex differences

between mates' behaviors are the bases of role theory, and if

the sex differences found here are compatible with those hypo-

thesized by the role perspective, these data might best be

interpreted in this way. Role theory suggests that the husband

leads the couple in areas requiring instrumental behavior and the

wife leads when affective behaviors are primarily involved.

The behavior observed and rated in this study was

generated by tasks given in an experimental context. It was

noted earlier that the topics for discussion constituting the

experimental tasks were selected because they seemed to stimulate

personal involvement and affective responses, but it seems more

realistic to assume the prominence of the task's instrumental

components, e.g., holding a 15-minute conversation at the request

of the experimenter. On this basis, in couples reporting higher

quality the husband might, in fulfilling his role, be expected

to lead in the task and show the observers more U behavior than

his wife.

These considerations imply that the mates' perception

of the task is an important variable. Unfortunately, the sex

role emphasis inherent in.the task used here is not known, and

further research would have to be conducted in which task was

formally studied. In any case, no association was found between

reported quality and differences between mates on Dl, and thus

hypothesis 1 was not supported. This result indicated of course

that neither sex consistently showed more U behavior than the

mates.

Although no association was found relating reported

quality to mates' D1 differences, the findings did indicate

that as the wives' experience of quality increased, the totality

of the mates' U behavior increased. However, the husbands'

quality reports related to the sum of the observed Dl behavior

in an opposite fashion, but with significantly less weight than

the wives' reports. Data analyses were performed in which the

components of the couples' Dl sums, i.e., the individual mates'

Dl behaviors, were separated and predictive equations for these

Level-I dependent variable units were sought. It was found

that no relationship existed between the independent variables

and the wives' Dl scores. However, for the husbands' D1 scores

it was shown that as they tended toward U behaviors, their wives

reported greater quality, while the husbands themselves reported

less quality, although this latter association was statistically

weaker than the positive association between U behavior and

quality for wives. This finding indicated for males' Dl behaviors

a relationship between variables similar to that found for the

Dl sum. Thus it appears that husbands' Dl behaviors alone

change in values in a consistent fashion and account for the

relationship found between the D1 sum and the independent vari-

ables.

These dissected results showed that as the wife reported

greater quality in the marriage, the husband's behaviors on the

task tended toward U, and caused the total amount of U behavior

observed for the couple to also increase. While the husband's

reports of quality were statistically less significant, never-

theless, as he reported greater quality, his behavior appeared

less U to the observers. In summary, in marriages experienced

by the wives as high in quality, the husband appeared U, and

a weaker trend showed that in marriages the husbands found

highest in quality, they moved away from U in their behavior.

In interpreting these results it appears that the

husband and the wife have different expectations about the male's

D-U behavior. Wives like those studied here may want their

husbands to tend toward U on tasks such as the experimenter

provided, and such a tendency on his part may be associated with

the wife's greatest marital satisfaction. In contrast, the

husband may be content to behave more moderately in terms of

this dimension. These results suggested opposition between hus-

bands and wives in the behavioral expectations they hold for

the male role with differential effects on their experiences

of quality in their marriage.

It is necessary to remember that the entire sample of

subjects was composed of FM couples. With this sample character-

istic in mind, this trend as a possible source of conflict in

the mates' relationships may also be regarded in the light of

a statement made by Bales (1970). Bales said,

In a two person group there is no literal
third person, hence there is no impartial
or impersonal judge. If a two person group
cannot run on the assumption of "love" it
cannot run, since neither norms nor coali-
tions have their usual representation in
actual third persons. Individuals in a
two person group tend to adjust to their
problem of having no third person, or
judge, to appeal to, by being very care-
ful to maintain the appearance, at least,
of solidarity. They tend to have high
rates of agreeing, and low rates of disa-
greeing, but they also have high rates of
showing tension--all indications that they
tend to suppress disagreement and negative
feeling. (1970, p. 79)

It would be interesting to see if these same trends

exist and appear with greater strength in a sample of couples

reporting low marital quality, on the assumption that they are

less skillful at suppressing their negative feelings, with

negative consequences resulting for the mates' experience of

the marriage. Apparently the present sample of marriage partners

do not make into big issues the role disagreements they may have.

The divergent expectations they may hold appeared as only subtle

indications of tension around the D-U behavioral dimension,

the dimension reflecting power, and the mates' experience of

quality did not seem severely affected.

The N-P dimension relates to the affectional behavior

of the spouses. By inferring from role theory it would be ex-

pected that the wife would be more sensitive to the interactional

factors this dimension reflects. The results here showed that

the wife's report of her experience of marital quality carried

weights generally three times as great as the husband's for

predicting the total D2 score, a measure of the total P behavior

observed in the interaction. The wives' independent variable

scores were also more strongly associated with the individual

husband and wife D2 behaviors and showed that as she reported

greater quality, both she and her husband tended toward P

behavior. Thus, wives in this sample follow the role theory

predictions regarding their behavior by reflecting stronger

associations between their experience of the marriage and the

affective interactional dimension.

Recapitulating, this study has shown, on the basis of

its data reflecting the power dimension in interpersonal rela-

tions and the affective dimension, results that are appropriate

for role theoretical considerations. A possbile conflict was

noted between males and females over their expectations about

the husband's role in a marriage, although in the present sample

the conflict was adequately suppressed. It was also concluded

that the wives studied here appeared to be fulfilling the model

for their behavior prescribed by role theory.

A third result from the study is also relevant to this

discussion. It was found that the husbands' reports of marital

quality were negatively associated with the length of the marriage;

i.e., as the time married increased, their experience of quality

declined. This trend indicates that as the couple's shared

history increased and an interaction pattern unique to that

relationship had time to evolve, the husband's experience of

quality became reduced. Wive's data did not show this trend.

These three findings all imply that the wife's experience

of quality is more closely related to the interaction she engages

in with her mate. The absence of time-effect results for her

indicated that over time she does not tend to grow dissatisfied

with the relationship, as her preconceptions and images of her

husband are replaced by interactional knowledge of him from a

variety of interpersonal situations. The male, on the other

hand, may experience, along with the dissolution of his ideali-

zations about his mate, increasing demands from the relationship.

These exigencies were especially strong in terms of the behaviors

subsumed by the power dimension, although the wife also appeared

quite sensitive to the husband's affectional behaviors.

A speculative conclusion that may be drawn from these

findings is that males may find marriage a more difficult and

formidable relationship than women. This tentative assertion

from the findings implies that male socialization, which emphasizes

the instrumental role, leaves men less prepared by comparison

to women for engagement in a highly complex and demanding inter-

personal relationship like marriage.

Questions for additional research abound in this specu-

lation. The nature of marriage is a field for a great deal of

future work, and it would be interesting to study in detail how

socialization and interpersonal dynamics are related to changes

in the institutionalization of love between persons. One may

conjecture as to how husbands' and wives' attitudes about role

behavior and divisions of labor are related to interpersonal

dynamics in marriage. Which sex's expectations for their pro-

spective mates are the most dissonant with the others' self-

pictures? How do changing life-styles relate to changes in

interpersonal interaction in relationships such as marriage?

A final theoretical note is pertinent to the findings

regarding Bales' N-P dimension. Although evidence was found

which supported the hypothesis derived from social exchange

theory for D2, additional trends in the results indicated that

hypothesis 2 might be modified. To review these results, it

was found that the positive correlation between quality and

mate similarity on D2 was possibly attributable to an even

stronger positive correlation between quality and the mates'

tendency toward P behavior. The sample, as noted, was composed

of FM couples, and these couples were seen as predominately P

in their behaviors. Because of this sample limitation, the

evidence for modification of the hypothesis may be a function

of the sample primarily, and the general compatibility of mates

who are similar on this dimension would be found in a larger

sample. It would be necessary to study additional couples

whose N behaviors were more prevalent as a more definitive test

of the hypothesis.

Hypothesis 2 cannot be rejected on the basis of the

present results, and an alternative hypothesis relating reported

quality to P behavior must be regarded as untested here. However,

there are theoretical reasons for believing that such an alterna-

tive formulation might be better than hypothesis 2.

From the social exchange perspective Carson (1969)

discussed persons whose development included what he called a

"negative transformation." He stated that these persons may

carry a great deal of guilt and, for them, some reward is received

when they prompt negative interactions, because they are then

able to alleviate their own feelings of self-negativity. In

this way they gain support of their self-image from another.

However, this form of relating imposes severe limitations on their

relationships. Even though the security of the self-picture is

maintained by negative interaction, it seems inevitable that the

experienced quality of their relationships is compromised. In

short, while similarity is the principle by which compatibility

dynamics may be best understood on this dimension, for the N

person the quality of the relationships is bound to be low.

Carson outlined some interpersonal dynamics regarding

the degree to which a person is dependent on others for a par-

ticular kind of response. He viewed great dependence in relation-

ships as a state in which there is a strong need for particular

types of interpersonal responses and the person's tolerance for

variability in the kinds of others' responses is constricted.

It follows that the N person has a very narrow range in his

ability to accept responses from others, because the N person is

always close to experiencing anxiety and insecurity. His self-

image is rigid and brittle, and so are his expectations about

other's behavior. All his relationships carry a high cost, and

consequently alternatives, in the form of new relationships

replacing his on-going ones, are not feasible. Thus, even though

a N person may find compatibility with a mate who is similar

to him, he still is not likely to experience high quality. It

may be reasonable to suppose that for N persons a sustained mar-

riage of high experienced quality is an impossibility.

Implications of the Present Results

Most of the subjects who participated in this study

tended toward the UPF part of the Bales space, and Bales (1970)

has provided descriptions of the personality characteristics

of such persons: they show primarily an interest in social solidarity

and social success, loyalty, friendliness, and a spirit of cooperation.

These personal characteristics manifested in interper-

sonal behaviors may be a determiner of the quality experienced

by a person in his relationships. It has already been noted

that the person tending toward N in his behavior may be severely

limited in the relationship quality he can experience. However,

it is also plausible to suppose that the way a person experiences

relationships in general, going back to infancy and prior to

marriage, in terms of love and other aspects of quality, deter-

mines his interactive style. Experiences in intimacy and the

ability to love may thus be prerequisite to quality in inter-

action.

In Chapter I this same dilemma was raised with regard

to the choice of independent variable, and this issue was skirted

in designing the study. There, as for this discussion, no data

are available to settle the question about which set of variables

is more inherent in personality. In any case the present results

have shown that these two aspects--interactional personality

characteristics and high experienced quality--are associated

with each other to a highly significant degree. Moreover, person-

ality was only a tangential interest of this work; the nature of

marriage was selected as the main focus of study.

This discussion of experienced quality and interpersonal

behavior may be linked to the subject of marital form. It may

be assumed that the criteria for satisfaction in a companionship

marriage are different from such criteria for an institutional

marriage. It was noted in the beginning of this report that in

a companionship marriage the mates have choose one another on

the basis of what each is as a person, rather than on the basis of

parental wishes or the values of some other societal authority.

The characteristics of the present sample of persons, listed above,

would seem to be the sort of traits essential to a companionship

marriage and less important to an institutional marriage. On the

basis of the findings and the inferences drawn from these results

regarding interactive styles, it appears that the couples studied

here enjoy companionship marriages. Also, on the basis of their

demographic characteristics some support is given to such a claim;

i.e., it is the young and recently married who would be expected

to be in the vanguard of the contemporary trend toward companion-

ship marriages.

These considerations about companionship marriage and the

interactional aspects of such relationships may be relevant to

marriage and to pre-marital counselors. The findings here provide

clues to ways in which a married couple can ameliorate a deteriora-

ting relationship or to the possibilities of success a proposed

marriage may have. In counseling a couple seeking advice, it

would seem, discussing with them the interactional behaviors they

engage in might be highly fruitful. It is well established that

power, affection, and social movement are fundamental concerns in

interpersonal relations, and for the highly functional couples

studied here, different principles of compatibility were found

for each interpersonal dimension.

Based on the present findings, a counselor might tell a

couple that power may be a most difficult area to balance in their

relationship. He could say that they may find that they hold

opposing expectations, especially for the male's behavior, and

give suggestions as to how they may best handle their difficulties.

He could inform them that the woman is usually the most sensitive

partner with respect to behavior in the affective area, that men

have a greater tendency than women to simply grow tired of the

relationship, independent of the woman's feelings about it or the

interactional behaviors they engage in with each other. A couple

may also be counseled about the values they possess with regard

to matters beyond their marital interaction, which may nevertheless

influence their experience of the relationship. With respect to

this subject it has been shown here that similarity on the social

movement dimension is quite important for marital satisfaction.

The implications from the results of this study may simply

provide another way for counselors to say what they already say.

However, the findings here have been obtained from couples pos-

sessing highly functional relationships, and since a marriage of

high quality is of course the primary goal of counseling, sug-

gestions made from such a data base may be particularly appropriate.

It is hoped that the information provided here may contribute to

the field of marriage.

APPENDICES

APPENDIX I

APPROACH TAKEN ON PHONE IN RECRUITMENT OF SUBJECTS

Mr./Mrs. ? My name is Carl Nickeson. I am a

graduate student in psychology, and right now I am doing my

dissertation research on marriage. I am contacting you to see if

you are a couple without children who have been married between

one and five years and possibly interested in participating in

a study that takes one hour. If you are interested let me give

you some more details.

I am specifically studying interaction between husbands

and wives. What I ask you to do in the hour is to first take a

questionnaire, and then to engage in what I call an interaction

sequence. This interaction sequence consists of my giving you

and your spouse a card with a topic on it and asking you to dis-

cuss the subject between yourselves. The interaction will be

studied by having three fellow graduate students observe your

discussion and rate it on some scales.

The meeting will be at your convenience--evening, day, or

weekend. As a consequence of this flexibility I cannot have the

observers actually present for many of the interactions, and so

I ask you to allow me to videotape your discussion for them. The

tape is erased after they rate it, and all materials are strictly

confidential.

After the results are compiled I will be glad to meet again

and tell you your scores, and what they mean. Will you participate?

APPENDIX II

RATING FORM USED BY OBSERVERS IN THE STUDY AND SCORING KEY

RATING FORM FOR OBSERVERS (from Bales, 1970)

Respond "yes" ( ), "no" ( ), or uncertain (?) to each question for
each spouse.

H W

1. Is S's rate of participation generally high?
2. Does S seem to rate himself/herself highly on all
good and/or socially popular traits?
3. Does S seem valuable for a logical task?
4. Is S's rate of giving suggestions on group tasks high?
5. Does S make inhibitory demands and want to enforce
discipline?
6. Does S seem dominating?
7. Does S seem to demand pleasure and gratification?
8. Does S make many jokes or show many fantasies?
9. Does S seem able to give a lot of affection?
10. Does S seem friendly in his behavior?
11. Is S's rate of giving agreement generally high?
12. Is S generally very strongly work oriented?
13. Does S seem to emphasize moderation, value-determined
restraint?
14. Does S seem unfriendly in his behavior?
15. Does S seem to feel that others are generally too
conforming to conventional social expectations?
16. Does S seem preoccupied with wishful fantasies?
17. Do you feel liking for S?
18. Does S seem calm, understanding?
19. Does S seem to believe that equality and humanitarian
concern for others are important?
20. Does S seem very introverted, serious, shy, introspective?
21. Does S seem to plow persistently ahead with great inertia?
22. Does S seem resentful?
23. Does S seem to accept failure and withdrawal for himself/
herself?
24. Does S seem to withhold cooperation passively?
25. Does S seem to be appealing for understanding?
26. Does S seem to devaluate himself/herself?

84

Key to Dimensional Loadings

1. No = D Yes = U 14. No = P Yes = N
2. No = DN Yes = UP 15. No = PF Yes = NB
3. No = DNB Yes = UPF 16. No = F Yes = B
4. No = DB Yes = UF 17. No = NF Yes = PB
5. No = DPB Yes = UNF 18. No = UN Yes = DP
6. No = DP Yes = UN 19. No = UNB Yes = DPF
7. No = DPF Yes = UNB 20. No = UB Yes = DF
8. No = DF Yes = UB 21. No = UPB Yes = DNF
9. No = DNF Yes = UPB 22. No = UP Yes = DN
10. No = N Yes = P 23. No = UPF Yes = DNB
11. No = NB Yes = PF 24. No = UF Yes = DB
12. No = B Yes = F 25. No = UNF Yes = DPB
13. No = PB Yes = NF 26. No = U Yes = D

APPENDIX III

TRAINING MANUAL AND INSTRUCTIONS FOR USE OF THE
INTERPERSONAL BEHAVIOR RATING FORM OF BALES

The questions on the rating form are interdependent. Dif-
ferent questions may relate to similar behaviors a S exhibits. As
an observer you are to answer each question by itself, without re-
gard as to whether it may be partially redundant with another. For
example, you may answer #3 "yes" on the basis of S's giving of sug-
gestions. Number 4 also may be answered affirmatively on the basis
of S's suggestions, and, because you have already "used" his sug-
gestion behavior in #3, you must not hesitate to respond "yes"
to #4, also.
In considering your ratings of the mates in this study
please make your ratings of the husband independent from those of
the wife, and vice versa. Of course, their behaviors are inter-
actional, and so a particular style on the part of one has an effect
on the behavior of the other. However, in responding to the questions
on the rating form you must disregard the rating you gave to one
person when you rate the spouse.
I would like to suggest some rules to insure that the rating
given one does not effect the rating given the spouse. 1) Go down
the form two times, once for the husband, and once for the wife;
2) Conceal the mate's rating when going down the second time; and,
3) Alternate across couples which spouse you rate first, i.e., H
then W, and W then H.

Criteria and behaviors for answering affirmatively to a rating question

1. Is S's rate of participation generally high?
This category refers to the amount of interaction initiated
or acknowledged, and "generally high" refers to S's participation
in at least 42% of the interactional acts, e.g., S must do no less
than 42% of the talking.

2. Does S seem to rate self highly on all good and/or socially
popular traits?
Behaviors indicating social extroversion, friendliness, and
self-assurance, e.g., taking the initiative in friendly acts, show-
ing comradery, expressing empathy and agreement, etc. Also,
making statements advocating social success.

3. Does S seem valuable for a logical task?
Leadership behaviors, S promotes solidarity and interest
in the experimental task, e.g., conciliatory, mediating behaviors,
tact, diplomacy, urging cooperation, paternalistic suggestions,
etc.

4. Is S's rate of giving suggestions on group tasks high?

This category includes routine control of communication
and the direction of the couple's attention to task problems
as well as substantive suggestions. Thus, mentioning a pro-
blem to be discussed, calling attention to what one is going to
say, or pointing out the relevance of what one is saying or
doing are instances. This includes even signals that are meant
to control attention, e.g., clearing the throat, etc.

5. Does S make inhibitory demands and want to enforce discipline?
This refers to the fusion of an ascendant attitude,
moralistic orientation, and negative feeling toward the other,
e.g., expressions of prejudice, advocacy of autocratic authority,
attempts to control, regulate, govern, direct, or supervise in
a manner which seems arbitrary, and in which the freedom of
choice or consent of the other person is either greatly limited
or non-existent, with the implication that he has no right
to protest or modify the demand, but is expected to follow
the directions immediately without argument, e.g., "Hurry up",
"Stop that". Also included are indications that one is indig-
nant, offended, or insulted, such as grim expressions, etc.

6. Does S seem dominating?
This includes aggression, such as conspicuous attempts
to override the other in conversation, interrupting, inter-
fering with his speaking, gratuitiously finishing his sen-
tence when he does not want help, insisting on finishing one's
own sentences or complaining, finding fault. Strong assertion
of one's claims, trying to outdo, argumentation. Also, glaring,
frowning, moving in a threatening manner; indications of
envy, and jealousy, or attempts to take something away from the other.

7. Does S seem to demand pleasure and gratification?
Include entertaining, loud, guffawing behavior; shows behavior
indicating non-submissiveness to authority, excessive non-con-
formity or independence, acts which, from the point of view
of the person in authority, are seen as disobedient, rebellious,
irresponsible. S is disrespectful, discourteous, impudent,
bold, saucy, flippant. Also, attempts to attract attention by
mannerisms, expressive gestures, emphatic or extravagant
speech, posturing and posing for effect, bragging, praise of
self, exhibitionistic behavior, or posing as unique and mysterious
with a negative element present in this behavior.

8. Does S make many jokes or show many fantasies?
Include dramatizing, where S presents images or poten-
tial emotional symbols to the listener. A joke is a story or
an anecdote about persons or personified beings in which
diverse acts and diverse feelings are portrayed in a non-conven-
tional manner. Look for stories or tale-like expressions
dramatized by the color of their action, giving non-serious
or non-literal suggestion, e.g., clowning, bantering, kidding, etc.

NOTE: #8 may sometimes be similar to #7, but may be dis-
tinguished from #7 by its absence of a negative element.

9. Does S seem able to give a lot of affection?
Behaviors that put the other at ease, show spontaneous
warmth; reassuring, takes the initiative in praising, rewarding,
boosting the other, and giving approval or encouragement when
he is having difficulty in performing adequately. Giving
strokes or interpersonal rewards that are not conditional upon
conformity to some expected pattern. Any manifestations which
O interprets as nurturant, gentle, maternal, paternal, benevolent,
humanitarian, merciful, or charitable is included.

10. Does S seem friendly in his behavior?
Appears relaxed, well-adjusted to the situation. S
assumes equality between himself and the other. He is congenial,
sociable, affiliative, cordial, or informal. He compliments
the other, gives him credit, shows enthusiasm for his views,
applauds him, and even may give him approval, provided the
status implications are those of equality.

11. Is S's rate of giving agreement generally high?
Less overt forms include giving signs of recognition as
the other speaks; showing interest, receptiveness, readiness,
responsiveness, such as looking at the speaker, etc. Giving
signs of attention as a means of encouraging him to say what
he wishes, e.g., by nodding the head, saying "M-hmm", etc.,
completing what he says, or otherwise aiding and facilitating
communication.
More substantial forms seem to commit the agreeing member
to the content of what has been said, and as if they might be
relied upon later. A person may express confirmation, con-
viction, accord, concurrence, assent about facts, belief,
inferences, etc., e.g., "That is the way I see it too,"
"I hope so too," "I also feel that way," etc.

12. Is S generally very strongly work oriented?
Participation consists of task oriented opinion or analysis,
e.g., "Maybe it could be . .," "If we add two and two . .,"
"If that's true then we can guess that ... ." S proceeds
instrumentally toward goals given by E; goals of the task are
not thought of as self-generated, but as received, and accepted
from a source of authority. S advocates conservative group
belief, e.g., statements of moral obligation, statements of
policy, intention, law.

13. Does S seem to emphasize moderation, value-determined restraint?
Strives to be "objective" in tone; tends to arouse guilt
in others; blandly ammends the other's opinions about the

situation in a distant impersonal way. Appears mainly task-
or value-oriented, and persists conscientiously in spite of
resistance from others.

14. Does S seem unfriendly in his behavior?
Shows disagreement with the other, suspiciousness. More
generally, seems isolated psychically detatched, isolated,
indifferent, impersonal, formal, distant, unsocial, reserved,
secluded, unapproachable, exclusive, or forbidding especially
in responding to an approach of the other. Passive refusals
to act which frustrate the other may be included, e.g., acts
in which one thwarts, balks, blocks, or obstructs the way of
the other. Acts of "defensiveness" often fit in the category.

15. Does S seem to feel that others are generally too conforming
to conventional social expectations?
S indicates a negative attitude both toward the other,
and toward the values which he expresses. S is non-compliant,
and is negativistic, stubborn, resistant, obstinate, refractory,
contrary, sulky, or sullen toward the efforts or imagined
efforts of the other toward restraint.

16. Does S seem preoccupied with wishful fantasies?
S expresses improbable ambitions, seems unable to decide
things. Expresses rejection of traditional belief. Does not
depend upon logical analysis as the means of making his points,
but instead uses pictures, and feeling to form and guide his
statements. He dramatizes, jokes, and laughs to a relatively
high extent. His statements may be cryptic and vague.

17. Do you feel liking for S?
Look for acts that seem friendly, which set aside or have
no relevance to the group task. S gives permissive grins or
shrugs, knowing looks, or wags of the head which indicate to
the other that he is accepted and liked. S shows warmth.

18. Does S seem calm, understanding?
Look for acts in which there are signs of positive feeling
or emotion, and which at the same time are somewhat submissive.
S shows tolerance toward the other's behavior, stable, seems
to trust in the goodness of others. S expresses gratitude
or appreciation, shows admiration, esteem, respect, wonder, or
reverence toward the-other. Sometimes S's positive feeling
may appear primarily in contrast to the other, as, when in
response to aggression directed toward him, the person is sub-
missive, acquiescent. Allows self to be talked down, surren-
dering, giving in, acknowledging defeat; standing aside and
letting the other aggressively push.

19. Does S seem to believe that equality and humanitarian concern
for others is important?

This question relates to displays of submissive friendli-
ness; showing respect and concern for both the other person
and conventional norms, e.g., by admitting errors and over-
sights, admitting that some objection or disapproval of the
other is valid, conceding a point to the other on the basis
of a logical proof or a value-based argument, asking the other's
pardon when shown wrong. Include confessions of ignorance or
incapacity, acts of apology, etc., unless they are so marked
or extreme as to indicate underlying negativity. (Submissive
friendliness overdone begins to seem unfriendly.) Look for
acts of genuine altruism, modesty, humbleness, respect.

20. Does S seem very introverted, serious, shy, introspective?
Low rate of dramatizing, and joking. Engages in acts
which seem to be over-careful, over-cautious, over-prudent,
vigilant, tense, and inhibited because of fear of possible
blame. Indications that the person is over-scrupulous, conform-
ing, conscientious, conventional, or dutiful because of fear
of breaking group norms, and thus experiencing disapproval
and guilt.

21. Does S seem to plow persistently ahead with great inertia?
Acts of self-sacrifice for higher values, or apparent
self-sacrifice which seem unfriendly, because although sub-
missive and value-oriented, they imply an excessively harsh
picture of the other, e.g., attempts to shame the other into
some kind of desired behavior by acting as if injured, hurt,
put-upon; acts which attempt to place the responsibility for
the solution of one's own problems on the other which imply
the other has not given the aid or support he should have
given; acts of servility or fawning with ulterior purposes;
acts of self-condemnation, shame, etc.

22. Does S seem resentful?
Acts which are at the same time relatively passive and
yet expressive of negative feeling. Indications of attitudes
which seem over-cool, frigid, inexpressive, or unsmiling may
be included. Passive forms of rejection, such as remaining
rigid, silent, inexpressive, impassive, or responseless in
the face of approaches of the other, or any passive with-
holding of love and friendship. Also, and manifestation of a
partially repressed negative reaction to the other which seems
cranky, touchy, testy, irritable, or sulky.

23. Does S seem to accept failure and withdrawal for himself/herself?
Indications that one feels his efforts have failed, that
some problems confronting him in his earlier or previous ef-
forts to conform still remain; expressions of feeling frustrated,
thwarted, or deprived; expressions of discontent, disappoint-
ment, discouragement, resignation, may be included, if the per-
son conveys the feeling of rejection for both the affection

of other persons and the demands of norms. Actions or the
display of attitudes which indicate that the person is inat-
tentive, bored, or psychically withdrawn from the other and
the problems at hand, e.g., slouching, yawning, looking away,
and letting the eyes wander, refusing to talk loud enough to
be heard.

24. Does S seem to withhold cooperation passively?
Submissive and non-conforming, yet not necessarily clearly
negative in feeling about the other as a person. Several
varieties of acts are scored in this category, not all of
which seem similar on a superficial level. Laughter, in par-
ticular, may seem quite different from signs of anxiety and
tension, yet both these behaviors may result in or be signs of
passive resistance.
Minor outbreaks of reactive anxiety may be included, such
as appearing startled, disconcerted, alarmed, dismayed, per-
turbed. Hesitation, fluster, flushing, blocking-up, etc. may
also be included. More passive forms of hanging back are also
included, such as evading the actual content of requests,
shrinking from what is felt to be threatening, etc.

25. Does S seem to be appealing for understanding?
Submissive friendly responses, which at the same time
indicate deviance from the value or task orientations of the
other. S is unlikely to arouse dislike or negativity in the
other, and in fact seems filled with good will, however, his
behavior is somewhat tangential to movement of the conversation,
progress on the task, etc., e.g., smiling, giggling, grinning
in response to a compliment.

26. Does S seem to devaluate self?
Submissive, low rate of participation, puts statements
in informational rather than opinion form.