Date

0-7639

ml\\\\\\\\\\\l\l\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\l

3 1293 10382 5844

This is to certify that the

thesis entitled

MOTHERS' AND DAUGHTERS' PERCEPTIONS
OF THEIR RELATIONSHIP AND SEPARATION

IN LATE ADOLJEgCENCE
presente y

DEVORAH RUTH SMITH 3H,;u

has been accepted towards fulfillment
of the requirements for

Psychology

 

M.A. flame in

 

Major professor

2-27-81

  
  
 

LIBRAR I

Adulthgan Sum
nivcm'cy

 
    

OVERDUE FINES:
25¢ per day per item

RETURNING LIBRARY MATERIALS:

Place in book return to remove
charge from circulation records

       
 

1/ Al\\

tat! .n{\\‘ in

‘ ' any

, 4»

"I a ..
W3 E NOV @3 96M) 0
FEB agnigi; 1}

      
 

 

 

MOTHERS' AND DAUGHTERS' PERCEPTIONS
OF THEIR RELATIONSHIP AND SEPARATION

IN LATE ADOLESCENCE

BY

Devorah Ruth Smith

A THESIS

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF ARTS

Department of Psychology

1981

ABSTRACT

MOTHERS' AND DAUGHTERS' PERCEPTIONS
OF THEIR RELATIONSHIP AND SEPARATION
IN LATE ADOLESCENCE

BY

Devorah R. Smith

This study explored the interactive nature of mothers' and their
late adolescent daughters' approaches to separation. It examined the
linkages between their interactional patterns and the daughters' percep-
tions of home-leaving. Twenty-six volunteer, middle—class mother-
daughter pairs provided the data.

Overall, mothers' perceptions of own and daughters' behaviors
correlated repeatedly with daughters' home-leaving styles. Specifically,
disaffiliations within the relationship linked to daughters' perceptions
of separation as being self-initiated. Self-descriptions by mothers
and daughters also correlated, in opposite directions, with home-leaving
styles.

Problems were identified in the social behavior measure
(Benjamin's SASB) and in the design of the mothers' projective stories.
There remains the need for more intensive studies of normative separation

in late adolescence.

ACKNOWLEDGMENTS

Many have supported me in my struggles to become a researcher.
Special thanks goes to my chairperson, John Hurley, for his belief in
me when I had stopped believing and his warm acceptance of my tears when
I was most frustrated. He generously gave encouragement, time, and
editorial assistance. To my other committee members, John McKinney and
Elaine Donelson, I also extend thanks for their suggestions, their support,
and the freedom to work in my own style. I am especially grateful to
John M. for his absolutely contagious fascination with separation in
adolescence.

Thanks too are due the many peers who listened, argued, and
helped. DeWayne Moore became my model-—a researcher who enjoyed his
work. The sophistication of this study's statistical analyses, far
exceeding my original expectations or skills, is a tribute to his knowl-
edge and teachings. Thanks also to Donna, Carrie, and Steve, the inter-
ested, creative, and "reliable" psychology majors who coded the stories.
And, of course, I am indebted to the mothers, daughters, and their
schools, without whom this project would not have been possible.

There are so many more I wish to acknowledge. To Dave, who
helped me explore my own feelings about separations. To Rick, who
tempted me to enjoy my outside life during most of this drawn-out project.

And to Wes, the next separation researcher, who applauded my thinking

ii

and repeatedly dared me to have fun with my research. Thank you all.

Finally, to my family. No treatise on separation can be complete
without them. I am especially proud of my father who at age 50 conquered
many obstacles and became a researcher. I also admire my mother's
characteristic intensity and sensitivity. They have made our struggles
to disengage harder but richer. I feel both now acknowledge my
separateness. And to myself, who asked for, risked, and accepted all
these peoples' help and who grew, intellectually and personally,

through this endeavor.

iii

TABLE OF CONTENTS

I. Introduction . . . . . . . . . . . . . . . . . . . . .
Separation in the Context of Adolescent Development
Separation in the Context of Family Development .
Pathological Styles of Separation . . . . . . . .
Normative Styles of Separation . . . . . . . . . .

Structural Analysis of Social Behavior as a

Measure of Perceptions of Interpersonal Behavior

Engagement Styles as a Measure of Perceptual

Approaches to Separation .
Statement of the Problem . .
II. Method . . . . . . . . . . . . .
Respondents . . . . . . . .
Instruments . . . . . . . .
Procedure . . . . . . . . .
Analysis . . . . . . . . . .

III. Results . . . . . . . . . . .

The Measure of Interactional

Styles of Perceiving Separation

Correlational Analysis . . .

IV. Discussion . . . . . . . . . . .

Relationships Hypothesized and Observed

In Review . . . . . . . . .
Looking Toward the Future .

List of References . . . . . . . . .

Behavior

18

24

28

35

35

37

42

44

45

45

54

6O

67

67

72

74

78

Appendices
A.
B.

C.

Student-Parent Letter . . . . . . . .
Structural Analysis of Social Behavior

Engagement Style Stories . . . . . .

LIST OF TABLES

Spearman Correlations Between the Affiliative and

Interdependence Axes of the SASB . .

Spearman Correlations Between Hemispheres Across

All Levels of Interaction Assessed by the SASB .

Spearman Correlations Between Selected Quadrants
Across All Levels of Interaction Assessed by the SASB

Spearman Correlations Between SASB Dyadic Ratings and

Daughters' Engagement Style Scores .

vi

47

49

52

63

LIST OF FIGURES

l. A Simplified Version of the Chart of Social Behavior
(Benjamin, l979b) . . . . . . . . 2O

2. Model of the Structural Analysis of Social Behavior

(Benjamin, l979d) . . . . . . . 21

vii

CHAPTER I

INTRODUCTION

Separation in late adolescence is a drama in which the entire
family participates. This study focuses on mothers' and their adolescent
daughters' perceptions of their experiences at the time when the young
woman is preparing to begin college. For some, it will be their first
major physical separation. For all, it marks the beginning of a new phase
in their lives.

As with any change, it is a period of loss and of challenge. The
adolescent is losing the person she was and the childhood bond to parents
that once gratified and sustained her. At the same time she is challen-
ging her future and what she will be at a more autonomous level. The
process entails feelings of pain and of pleasure. It builds upon both
struggles and dreams.

At adolescence, parent-child interactions are suffused with their
shared acknowledgement, conscious and unconscious, that the adolescent
is changing and must soon leave home. In each family, the preparations
for and the reactions to this task vary. However, the disengagement
is necessarily mutual: one in which both parents and child participate
and through which both_can develop.

Individuation need not destroy a positive parent-child bond. For
some, the relationship evolves beyond childhood dependency and parental

protectiveness into a new form of intergenerational pairing. Mothers and

2
daughers can relate as unique individuals who care for and love while
respecting each other's separateness and distance.

This research study examines mothers' and daughters' perceptions
of themselves, their relationship, and their approaches to home-leaving.
Broadly stated, the underlying interest is the interaction of patterns
of relating within the family with cognitive styles of separating. In
examining these two aspects of disengagement, the family that one is
leaving from and the manner of this going forth, both the mothers' and
daughters' perspectives are addressed. It is hoped that such data will
contribute to understanding the neglected area of separation in late

adolescence.

Separation in the Context of Adolescent Development

 

Disengagement is best understood from the perspective of the
life span. It starts with the beginning of life, with the neonate's
birth out from the mother's body. There are two periods, however, when
developing autonomy is of particular importance to personality develop-
ment: the first individuation process in the second and third years of
life (Mahler, 1963), and the "second individuation process" in adoles-
cence (B105, 1967). The first parallels the development of the child's
capacity to walk and to physically move away. The toddler, for the first
time, develops a sense of self that is separate from his or her image of
mother. Individuation is again the focus at adolescence. The adoles-
cent's libido must detach from its first loves, the parents, in order to
pursue new attachments and interests.

In adolescence, this separating is often played out within a

setting of drivenness and imbalance. The forces that fuel this inner

3
turmoil are genital maturity (the drives and desires are changing),
increasing cognitive capacities (the ability to envision and those
visions of futures are expanding), rapid physical growth (the body and
image of that body have and are changing), and social realignments (the
position within society, with family, and with peers is also changing).
Though often hard to live through, or to live with, these upheavals are
rehearsals of new ways of being and thus represent active work at growth.
In the adolescent's experimenting that acts as a bridge between childhood
play and adult adaptation, the focus is mainly internal. It is a time
for the soon—to-be-independent offspring to take inventory of his or her
talent, discover his or her uniqueness, and develop a personal dream
(Gould, 1978).

There are a range of perspectives that address this work of the
adolescent. The classic psychoanalytic view (Fenichel, 1945) stresses
the reestablishment of the psychic equilibrium, disturbed by the sexual
impulses arising at puberty, and the acceptance ofgenital sexuality.
The personality is basically formed, though the eruptions of adolescence
can possibly solve leftover infantile conflicts or mold them into their
final form. Ego analysts Anna Freud (1958) and Peter Blos (1967) also
focus on the disturbed equilibrium, but with greater emphasis upon the
flexibility of the personality. Through the loosening of infantile
object ties and the abandonment of earlier ego states,a new equilibrium
can be established. Rather than simply rework, the goal, in their eyes,
is to create a more mature and more flexible defensive structure that
comes to terms with the person's adult sexuality and independence. The
end product is a unique and autonomous sense of self.

Erik Erikson (1968, 1970) expands the definition of the work of

4
the late adolescent through his increased emphasis upon the youth's
place within the social structure. The task, as he perceives it, is
one of integration. The adolescent's ego actively recombines fragments
of the past with the social and historical realities of the present
into a new whole--a sense of self and of this self's place in the world.
The process requires consolidating the adolescent's potentialities and
dreams with societal and personal limitations. The goal is the develop-
ment of an appropriate identity in the world of peers and of work.

In yet another perspective, John Bowlby (1973; Ainsworth, 1969)
incorporates the ideas of psychoanalysts, ethologists, and developmental
biologists. According to him, the task of the growing person is the
transition from infantile dependence to a state of mature dependence.
Such growth depends upon internal expectations of an important other's
accessibility and responsiveness. In the end, the well-trained adoles-
cent balances self—reliance with trust of and dependence upon others.

Underlying these viewpoints is the consensus that the "new"
condition of adolescence is complicated by the fact that it is not really
new. The old impulses and fears of childhood, whether called psycho-
sexual (the analysts), the early stages of life (Erikson), or internal
working models of attachment figures (Bowlby), reappear. Baggage from
the past, these feelings lend intensity to the adolescent‘s current
apprehensions and fantasies about separation. At the same time, the
adolescent's tentative experiments with separating, whether in the form
of home-leaving or self-changing, most arouse these primitive feelings
(Gould, 1978; Schneider, 1980). To this conflict, though, the adoles-
cent brings new tools. Developing capacities for abstraction and formal

reasoning provide the potential for less rigid and less limiting

solutions to the infantile rages. Thus, growth takes a cyclic form:
a small step towards individuation, reawakened childhood conflicts, their
internal reworking, followed by yet another small step towards

individuation.

Separation in the Context of Family Development

 

To understand separation and its impact, the family unit must be
considered. Modify any member or their interactions, as in an adoles-
cent's preparations to leave home, and this change reverberates throughout
the entire unit. How the various members respond to the resulting dis-
equilibrium depends upon their past interactional styles, their cultural
expectations, and each individual's personal needs (Duvall, 1967;
Rubinstein, 1972).

An adolescent's separation, then, is most accurately conceptual-
ized as a family developmental task (Worby, 1971). Indeed, recent
empirical research using paper-and-pencil questionnaires (Sullivan and
Sullivan, 1980) demonstrated that home—leaving in late adolescence af-
fected the views and attachments of both generations separating. Simul-
taneous psychoanalyses of early adolescents and their mothers, under the
supervision of A. Freud (1960), also highlighted the reciprocal nature
of families' readjustments over time. When one member lagged behind,
perhaps a child unwilling to give up dependent patterns of relating or
a parent continuing to protect and cling in no longer helpful ways, both
parent and child had more difficulty separating and growing.

The family's task of promoting separation at adolescence is a
difficult one. The family must gauge not only the child's developmental

stage but also his or her unique interpersonal style. What for one child

6
would entail supportive caring while disengaging would feel to another
like suffocating oversolicitation and toytn;another like sudden rejection.

Added to this equation are the parents' own feelings. Their
responses to their offspring's changes may be colored by the re-emergence
of their own unresolved adolescent conflicts. Some parents unconsciously
enlist their child to act out or to atone for these old impulses
(Anthony, 1969; Johnson, 1959). A further complication is the very dif-
ferent nature of the changes the family members are undergoing. As
parents deal with their own aging, physical deterioration, and narrowing
opportunities, their adolescent faces expanding capabilities, blossoming
sexuality, and widening options. Some parents react in ways, such as
competing, seducing, or befriending as if a peer, that endeavor to deny
or undo these discrepancies between the generations (Elson, 1964).

However, a reactivation of adolescent emotions need not be nega-
tive. It can lead to a greater understanding by the parent, whether in
gaining insight within the self or in empathizing sensitively across the
generations. Some adults may even repeat, with their child, the steps
of their own past development from a new perspective. In this struggling,
the parents themselves can grow and further resolve the issues of adoles-
cence (Benedek, 1959).

To best facilitate their adolescent's separation, the parents'
own emotions cannot overwhelm the "reality" of their child. For whenever
parents' strong and poorly integrated feelings seriously warp their
perceptions of and relationships with their child, even in the bene-
volent attempt to undo deficiencies in their own growing up, they
respond to their child out of their own needs and not those of their

separate and unique offspring. In doing so, this burden from the

7
parents' past jeopardizes their child's present growth and individuation
(Anthony, 1969; Stierlin, 1974). Thus seen in its many layers, the task
of every parent and adolescent, to evolve a new level of relationship,

is indeed a challenging one.

Pathological Styles of Separation

 

The Mechanism: Distorted Perceptions

 

Family therapists have become attentive to the types of family
interactions that lead to serious psychopathology in the offspring.
Shapiro (1967, 1968) focused on 25 families with adolescents who became
severely disturbed following their first prolonged separation from home.
According to trained clinical observers, these parents held distorted
views of their offspring. Their expectations and attributions were
judged to be biased, exaggerated, inappropriate, or simply untrue.
Further, these distortions were not random. Rather, they gratified the
defensive needs of the perceiving parent.

Shapiro developed the concept of delineation to explain what was
happening within these families. Delineation referred to a person's
view of another as revealed overtly and covertly, in words and in be-
havior. His central hypothesis was that parental delineations of the
child influenced the adolescent's identity formation and individuation.
Specifically, in his research sample where a separation precipitated an
extreme emotional crisis, the parents had falsely projected their own
unacceptable feelings and reawakened adolescent conflicts onto their
child. Shapiro labeled these externalizations of the self defensive

delineations. Another research group (Wynne and Singer, 1965), studying

 

parents of schizophrenics, also reported gross distortions in parental

perceptions and communications.

When parental delineations are distorted, the question arises as
to why some adolescents do not see them as false and simply disregard
them. The emotionally disturbed offspring in Shapiro's and in Wynne and
Singer's samples could not. Rather, they seemed to have embraced their
parents' reality, at the expense of trust in their own experiencing.
This outcome seems to be multiply determined by: a) the nature of the
distorted material, b) the manner in which it was communicated, and
c) the adolescents' openness to others' view of themselves.

First, the material that is most often distorted in families
are feelings. Subjective, elusive, fluctuating, and often illogical,
emotions are particularly vulnerable to outside influence. Usually, they
must be asserted and shared before they can be confidently identified.
(This is the common experience of not knowing what you are feeling until
you spontaneously say it aloud.) In this attempting to gain validation
and self-understanding, an adolescent opens him or herself to others'
interpretations and labels. When those others' reactions are defensive,
serving the perceivers' own needs, they still often take. After all,
they probably contain "some kernal of truth" (Wynne, 1972, p. 183) and
leave the adolescent feeling dimly understood. Thus, at a time when new
cognitive skills could become tools of a developing ego, energies are
instead perverted to doubting gut feelings. An adolescent who embraces
the distorted feelings, whether in passive compliance, immobilizing con-
fusion, or rageful denial, remains bound to the parents' seemingly
stronger "reality."

A second factor often identified as operating within disturbed

families is an ambiguous and unclear communication style. Bateson's

9
(Bateson et a1., 1956) contribution, an outgrowth of theoretical and
empirical work with mothers and their disturbed children, is the con-
cept of the double bind. Contradictory messages are communicated
covertly along verbal and nonverbal channels of interaction. Any response
by the adolescent will clash with at least one of the simultaneous
messages sent. The crux of the bind is that the multiple messages
elicit a multitude of feelings, especially anger and distancing, in the
recipient. Moreover, these intense emotions somehow feel inappropriate in
light of the common facade of understanding and helpfulness maintained
by the parents. Again, the child is perplexed and distrustful of his
or her aroused feelings. A common way out of this confusion is to re-
press the spontaneous experiencing in favor of the sender's seemingly
stronger vision of the relationship. As before, the net effect is
binding.

Lastly, the vulnerability of an adolescent to parental attribu-
tions and expectations also appears to be a function of his or her depen-
dence upon those projecting:

Perceptions and expectations bearing on separation. . .carry
the greatest leverage because they determine[:the adolescent's]
susceptibility to all other parental perceptions and expecta-
tions. They entail a meta-message or meta-perception about
these other perceptions. They convey to the adolescent how his
parents perceive and judge his confidence and capacity either
to heed, or turn away from, these parents themselves! They con-
tain a message about his ability to liberate himself from his
parents and thereby to immunize himself against what they-~his
parents--think, say, and want. (Stierlin et a1., 1971, p. 413)
Thus, the more subjective the perceptions, the more ambiguous their
communication, and the more dependent the adolescent, the more powerful

and potentially harmful the parental distortions and attributions.

The impact of these parental perceptions must be integrated with

10
the recognition of the child's power within a family. The adolescents
in the cited studies were not passive victims. Bell (1968, 1974) has
documented how even the young infant plays an active role in initiating
and maintaining social and caregiving interactions with the parents.
As an adolescent, the child can more intentionally and skillfully play
upon the parents' vulnerabilities. It follows, then, that even within
a family where a child's autonomy is totally unacceptable, the adolescent
plays a part in the separation-delaying strategies. Forjhnnzas parents
bring their perceptions to bear on their child, so do they ordinarily
open themselves to this child's attributions and expectations. Although
the adolescent may feel "helpless, entrapped, and one-down, in reality,
[he or she] often engages in as much hostile power, double-binding, and
blaming of the parents as vice versa" (Benjamin, 1979b, p. 18).

When parental delineations are defensive and concentrate on
prohibitions against age-approprite individuation, both parents and child
embark upon a reinforcing and restricting interactional cycle. The
adolescent who lacks a core of autonomy remains extremely sensitive and
vulnerable to parental perceptions. To become free of their influence
the adolescent must separate. Yet, this is very difficult when the
parental messages upon which he or she has grown so dependent upon com-
municate that he or she Ean_not and mu§t_not leave. At the same time,
there are many advantages to staying. Such a child, underdeveloped and
unfulfilled, remains strategically placed within the family to blame and
shame the parents (Stierlin, 1977). Thus, both_parties in families where

separation has gone amiss control and are controlled, bind and are bound.

11

The Outcome: A Model of Problematic Separation Styles

 

Stierlin (1974, 1975; Stierlin et a1., 1971) in his research on
families with adolescents hospitalized for psychiatric problems upon
their first prolonged separation from home, developed three interactional
modes that describe ways separations go awry. This classification of
family forces that pull together or push asunder is characteristic of
all families at some time during the lifelong process of separation.

The modes only become pathological when they are too rigid or out of sync
with the participants' current needs; for in healthy separations, the
dynamic inward and outward forces balance more evenly. The three major
modes, that when used extensively by families inhibit growth through
separation, are described below.

The binding mode. When this mode dominates the forces within

 

a family are pulling always inward. The unspoken assumption shared by
all is that the outside world is hostile and dangerous. It is believed
that only within the family can one find satisfaction and security.

Boundness, then, sets the scene. All players in this drama react
to the crises and challenges of their personal development tasks with
fear and increased clinging. The parents, rather than help launch their
child, try to bind him or her even closer. The adolescent, at a time
when energies should be focused on forming separate and non-incestuous
attachments, responds by hanging on even tighter. Intense, valued ties
to parents contraSt with shallow, nonessential relationships with outside
others.

This insulation intensifies anxiety over separation. The adoles-
cent has no objects other than the parents to play out the re-activated

infantile rages or to practice adult behaviors. Without skills for

12
relating to anyone apart from the close family, the threat of a hostile
and unresponsive world outside becomes a self—fulfilling prophecy. The
tightly bound adolescent, lacking a sense of his or her efficacy and
having no one else to lean upon, often sees no alternative but to remain
home with parents, even at a high cost to him or herself.

The expelling mode. This interactional style also interferes

 

with age-appropriate individuation and development, but in a very dif-
ferent way. Expelling represents a push away rather than a pull inwards.
These parents feel that they best meet thei£_needs, resolve their growth
crises, by accelerating the impending separation. The parental message
is experienced by the child as, "You are no longer important to me. The
sooner you leave the better." Indeed, the child's needs at this point
of development are not considered. There is no embroiling conflict, no
shared struggles, only a steady drifting apart.

What may at first appear to stimulate development, separation
without a fight, in actuality impedes it. An adolescent who achieves
autonomy in this manner does so prematurely. The abrupt disengagement
denies him or her the opportunity to work through conflict and anger
within the safety of a caring and committed relationship with the
parents. Feeling shunned and rejected, the offspring instead looks to
peers and the outside world for objects to bind to. However, the
shallowness and exploitiveness that marked his or her family relation-
ships will most likely be repeated in new couplings.

The delegating mode. The delegating mode is a combination of

 

binding and expelling in the sense that autonomy is allowed and en-
couraged, up to a point and for certain purposes only. The image is

that of a long leash. A delegate is sent out on parental missions yet

13

expected to return home to share the experiences and the achievements.

The missions vary, as do the largely unconscious parental moti-
vations behind them. But as long as the delegate does not choose where
to invest his or her energies nor is allowed to complete his or her
age—appropriate individuation, growth is stunted. The effect, though, is
not as detrimental as the pull on the child bound nor the push upon the
child expelled. As the tasks assigned require travel and competencies
in the world outside the family, some distance from home is secured.
From this vantage point a child has the opportunity to more clearly see
the family as it is, rather than as it is represented by parents. Some
will take advantage of this opening.

The continuum of binding through expelling appears a useful
tool in conceptualizing growth-inhibiting responses to the developmental
pressures to individuate. This dimension is echoed in the work of social
psychologists, sociologists, small-group therapists, and anthropologists
looking at families and ad hoc groupings. Indeed, among family thera-
pists alone, one uncovers a wealth of concepts and distinct terminology
developed in the past two decades that relate primarily to the extreme
states of family cohesion. To name just a few, the family interactions
that Stierlin described as binding and expelling were foreshadowed in
Wynne's concepts of pseudo-mutuality and pseudo-hostility and reappear as
enmeshment and disengagement in Minuchin's theorizing, as diffuse boun-
daries dependence and rigid boundaries independence in Olson's conceptu-
alizing, and as undifferentiated emotional fusion and emotional divorce
in Bowen's framework. The interest from so many academic corners and
the parallel formulations among the different family clinicians lend

cross-validation to interdependence as an important variable in the

14
health and growth of families (Olson, 1979).

To summarize, separation demands the conflicted reworking and
shifting of past loyalties. In some extreme families of the types
described above, the fight will leave the adolescent weakened and de-
velopmentally arrested. Yet, within other families, this working through
will be a time of much change, growth, and reaching towards potential.

A new level of relationship between parents and child is laid
out slowly over time. There is no one best approach. The path to
autonomy is a personal solution molded from the strengths, fears, and
bonds of all involved. The family best facilitates the process by
balancing the opposing binding and expelling forces. In the end, whether
the new relationship between late adolescent and parent will be a mature
and interdependent one or the conflictful continuation of the regressive,
ambivalent, and dependent one depends to a great extent upon the mes-
sages sent out and the responses returned during the slow and often

painful struggle to disengage.

Normative Styles of Separation

 

Though there has been limited research on problematic separations
in adolescence, there has been even less on normative separations.
Attempts by McDill in the 19305 and by Sherman in the 19403 to relate
the scores of adolescent respondents' on an "emancipation" scale with
various subject characteristic (i.e., age, intelligence, physical
traits, social class, religion, and time at college) ended in frustration.
The correlations were consistently quite low (Hotch, 1979).

This research area was abandoned until the 19605 when Murphey,

Silber, Coelho, Hamburg, and Greenberg (1963) explored the reactions of

15
19 "competent," academically gifted, upper-middle class high school
seniors to the transition of leaving home to go to college. Competency,
their criteria of inclusion, was defined as academic effectiveness, close
peer friendships, and social group memberships. Their data were exten-
sive interviews with both the students and parents, before and after the
separation. Following a suggestion by Bowlby, they decided that the
variations within this group of adolescents was best captured by two
dimensions--one looking at manifestations of self-sufficiency and the
other at characteristics of the relationship to the parents. The re-
searchers then subjectively rated each adolescent as high or low in
autonomy and in relatedness.

Autonomy was defined as the adolescent's willingness "to make
separate responsible choices" (p. 645). Decisions about time, scholas-
tics, money, sexual mores, and social commitments were considered in
making this rating. The second dimension, relatedness, was evaluated
according to the student's stated satisfaction with his or her relation-
ship with parents. The criteria for this judgment revolved around posi-
tive aspects of the parent-child interactions: feelings of growing
equality, freely shared communication, emotional closeness, and pleasure
with and interest in the parents. In the end, the two components were
combined to determine the extent to which familial relatedness was
maintained in the company of growing autonomy.

Students fell into all four combinations of these two variables.
One adolescent was rated low in autonomy and high in relatedness. The
researchers did not address this style, as they later deemed its sole
representative not "competent" according to the study's original

standards.

16
Most (nine of the nineteen students) were rated high in both

autonomy and relatedness. Within these families, separation was exper-

 

ienced as a gradual loosening of the reins of control and accompanied by
a growing sense of equality. As these students began to function in more
autonomous ways, their parents responded with confidence, support, toler—
ance, and encouragement. These parents regarded separation as normal

and necessary for growth. Indeed, they served as models of autonomous
adults, with inner-directed values which they clearly and openly communi-
cated. Both parents and adolescent responded to the child's growing
autonomy with pleasure.

In contrast, the three families with students rated low in both

 

autonomy and relatedness exhibited poor boundaries and unclear communi—

 

cations. These adolescents sometimes acted, sometimes were treated, as
extensions of the parents. While the parents continued to accept much
responsibility for their youngsters' actions, the adolescents often
reciprocated by continuing to blame others for their problems. This
general lack of confidence in these students' abilities was compounded by
a lack of clarity. Feeling unsure of who they were and what they stood
for, these parents often relied upon disguised or manipulative com-
munication of their expectations. Their actions might not match their

verbally stated beliefs; their strong feelings might never be openly

 

1Interestingly, in a study of families with preschool children,
Baumrind (1971) cited many of the same traits in her descriptions of
those parenting styles most strongly linked to the development of tod-
dlers' independent and responsible behaviors. Called "Authoritative,"
these parents affirmed their children's present abilities while clearly
stating future standards and demands. They differed from the parents of
these high related, high autonomy teens in their emphasis on firm con-
trol and disciplined conformity, which may be a reflection of the de-
velopmental needs of the younger child.

l7
shared. As a result, neither students nor parents accurately recognized
the beliefs and needs of the other. Separation for these families
entailed more of a struggle.
Embracing another style of relating were six families with

students rated high in autonomy and low in relatedness. Here again,

 

separation was a fight. Although these parents did provide many of the
conditions for developing autonomy (such as recognizing some of their
offsprings' strengths and separateness), they were less willing to let
the adolescent play out the separation within the confines of a suppor-
tive and accepting family. They did not bother to temper their expec-
tations to better fit their adolescents' changing realities and lessening
dependence. The resulting conflicts, of rigid role expectations with new
values and skills, fueled a growing rift between parents and children.

Among both these last two sets of "low-relatedness" students
(whether high or low in autonomy), the parents had difficulty recognizing
or adapting to the changing needs of their separating children. Whether
this situation developed out of the parents' attributes and their re-
sponses to their child‘s growth or out of these adolescents' attributes
and styles of disengaging from the family circle was not explored by
Murphey and associates. However, this study did demonstrate a link be-
tween parents' perceptions of offspring and adolescents' responses to
separation. Among this sample of competent high school students, inac-
curate and inflexible parental perceptions of the child were related to
family conflict.

There has been some interest, of late, in continuing the work of
understanding normative separation at the end of adolescence. Hotch

(1979) picked up where Murphey et a1. left off 15 years earlier. Her

18

subjects were 94 high school seniors selected on the basis of classroom
membership. To operationalize the subjective ratings of relatedness and
autonomy, Hotch developed a questionnaire that assessed adolescent
perceptions of self-sufficiency (four items measuring a student's finan-
cial responsibility for personal expenses) and relatedness (nine items
tapping the frequency of positive familiy interactions). She hypothe-
sized a relationship between a student's score on each of these two
variables and his or her style of perceiving home-leaving (assessed via
a projective measure to be discussed later). Relatedness to parents
emerged as the most powerful predictor of perceptual approaches to
separation. For males only, self—sufficiency was also an important
predictor. However, the interaction of relatedness and self-sufficiency
did not correlate with styles of perceiving home-leaving for either sex.
As the preceding review clearly shows, much work remains to be done be-
fore even a simplified understanding of the relationship between familial

interactions and styles of separating can be developed.

Structural Analysis of Social Behavior as a

 

Measure of Perceptions of Interpersonal Behavior

 

Description of Model

 

Lorna Smith-Benjamin (1974, 1977, 1979a, l979b, l979d;
McLemore and Benjamin, 1979) has also explored normative relations be-
tween parent and child at various stages of the life cycle. Though she
worked independently from, and used very different research tools than
did Murphey et a1., in the final analysis both research groups reduced
the complexity of interpersonal behavior to two strikingly similar con-

cepts: one dimension relating to issues of autonomy and another dealing

19
with those of relatedness.
Benjamin's model has many advantages over other conceptualizations
of parent-child relationships. The bulk of studies and models reviewed

thus far reflected the researchers' perceptions, as gathered by trained

 

observers or developed through subjective summaries of interview materi-
al, of how family members perceived each other. In contrast, Benjamin's
Structural Analysis of Social Behavior (SASB) provides a means of tapping

the involved participants' personal perceptions of their own and other

 

family members' styles of interacting. In addition, the SASB's framework
for recording these perspectives is not aligned with any single theore-
tical bias.

The output of the SASB is a visual display that quantitatively
defines social interactions. Its structure is that of a topographic
map, unifying the many component behaviors of a dyadic pairing into a
larger, more comprehensive whole. In this form, the data are amenable
to objective evaluation and statistical manipulation. Thus, the SASB's
utility, and beauty, lies in its ability to achieve conceptual clarity
while addressing the richness of interpersonal relationships. What
follows is a description of this model.

Planes. The SASB is presented in its simplified version (Figure
1) and in its entirety (Figure 2). Its most basic division is that of
focus, which translates spatially into three distinct planes. The top
diamond, Other, depicts behaviors done to or for others. The middle
surface, Self, lists behaviors done to or for the self. These planes are
prototypes of familial behaviors. The Other plane describes behaviors
more characteristic of a parent while the Self plane those more charac-

teristic of a child. The bottom plane, Introject, portrays the

INTERPERSDXAL

 

Plane I - Other

(Parentlike

20

Endorse freedom

     
   
      
    

 

   

 

Tender sexuality

 

 

. . Invoke Encourage'
-¥- !
B‘”a‘l°rs’ Hostile Friendly
Autonomy Autonomy
Annih;lating attack

Hostile Friendly
Power Influence

.Managgiicontrol

Freely come and go

Plane 11 - Self

(Childlike

Behaviors)

Desperate protest

   

   
      
   
    
     
   

  

Take
HOStile
Autonomy

Enjoy
Friendly
Autonomy

 

Hostile
Comply

Friendly
Accept

   

 

Yield, submit, give in

Ecstatic response

 

INTRAPSYCHIC
Introject (I)
of
Plane I
to
Self

Torture, annihilate self

Figure 1.

Happy-go-lucky

II

   
 

 

   
     
 

Reject Accept,
Self Enjoy
Self
Oppress Manage,
Self Cultivate
Self

III IV

 

Control, manage self

(Benjamin, 1979b).

    
  
 

Love, cherish self

A simplified version of the chart of Social Behavior

INTERPERSONAL

Uncaringly let go 128

Forgt 127

Ignore. pretend not there 126
Neglect interests. needs 125
lilogrcal initiation 124

Abandon, leave in lurch 123
Starve. Cut out 122

21

120 Endorse freedom

 

118 Encourage separate identity

 

117 You can dart fine
116 Carefully, fairly consider

 

 

115 Friendly listen

 

114 Show empathic understanding
113 Confirm as OK as is

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

112 Stroke, soothe. calm
Angfv dismiss. men 121 111 Warmly welcome
r .
OTHER Annihilating attack 130 1 110 Tender sexuality
Approach menacingly 131 141 Friendly invite
Rip off, dfaln 132 142 Provide for. nurture
Punish, take revenge ‘33 143 Protect, back up
Delude. divert. mislead 134 ‘44 Sensible mall/5'5
Acetise, blame 135 145 Constructive stimulate
Put down, act superior 136 147146 Pamper, overindulge _
Intrude. block, restrict 137 Benevolentimonitor, remind
Enforce conformity 138 ‘48 Specify M1“ ‘ best
Manage. control 140
220 Freely come and go
Go own separate way 228 218 Own identity. standards
Defy. do opposite 227 217 Assert on own
Wall-off, nondisclose 226 216 “Put cards on the table"
Busy with own thing 225 215 Openly disclose, reveal
Noncontingenr reaction 224 214 Clearly express
Detach, weep alone 223 213 Enthusiastic showing
Refuse assistance. care 222 212 Relax, flow. enjoy
Flee. escape. withdraw 221 211 Joyful approach
SELF Desperate protest 230 210 Ecstatic response
Wifv. fearful 231 241 Follow, maintain contact
Sacrifice greatly 232 242 Accept caretaking
“9"". SCUM! 233 243 Ask, trust. count on
Uncomprehenqu agree 234 244 Accept reason
Whine. defend, justify 235 245 Take in. learn from
Sulk. act out upon 236 246 Cling, depend
Apathetic compliance 237 ‘ "‘ 247 Defer. overconform
Follow rules. proper 238 248 Submerge into role
Yield, submit, give in 240
INTRAPSYCHIC 320 Happvso-Iudcv
Drift with the moment 328 r—1 '1' 318 Let nature unfold _
Neglect options 327 —1 . 317 Let self do it. confident
Fantasy, dream 326 316 Balanced self acceptance
Neglect own potential 325 315 Explore, listen to inner self
Undefined. unknown self 324 314 Integrated. solid core
Reckless 323 313 Pleased with self
Ignore own basic needs 322 312 Stroke, soothe self
Reiect. dismiss self 321 311 Entertain. enjoy self
Torture, annihilate self 330 310 Love, cherish self
Menace to self 331 341 Seek best for self
Drain. overburden self 332 342 Nurture. restore self
Introj ect Vengeful self punish 333 343 Protect self
of OTHER Deceive.divert self 334 344 Examine. analyze g." ’
Guilt. blame, bad self 335 345 Practice. oecome accomplished
to SELF Doubt, put self down 336 346 Self pamper, indulge
Remain, hold back a," 337 347 Benevolent eye on self
Force proofing 333 348 Force ideal identity
Control, manage self 340
Figure 2. Model of the Structural Analysis of Social Behavior

(Benjamin , l979d) .

22
intrapsychic effect of behaviors directed toward the self, as when the
actions of others are turned inward.

Axes. Each of the three surfaces described above are built upon
two axes. The horizontal axis runs from affiliation to disaffiliation.
The vertical axis represents the continuum of maximal independence
through maximal interdependence. The actual behaviors described by
the end points of these axes depends upon the applicable interpersonal
focus. For example, the interdependence dimension spans from endorse

freedom to manage, control when attention is centeredcxithe other (top

 

diamond); but from freely come and go to yield, submit, give in when

 

 

concerned with the self's interpersonal behavior (middle diamond); and

from happy-go-lucky to manage, control self when the focus is inward

 

 

(bottom diamond, Benjamin, l979b).

The inclusion of the full range of interdependent behaviors makes
this a useful expansion of Leary's (1957) interpersonal circle. In
Leary's model, dominate was the opposite of submit. The implicit as-
sumption of this earlier system, that to not dominate meant to submit,
left no options for autonomous behaviors. The upper halves of Benjamin's
interpersonal surfaces allow for these additional choices.

Quadrants. The SASB collapses into sets of four complementary
quadrants, each representing one-quarter of an interpersonal surface
(see Figure 1). A quadrant is defined by its bounding affiliation and
interdependence poles.

Individual items. Within each quadrant are nine behavioral

 

descriptions. Adjacent ones are quite similar while, at the same time,
differing slightly in the direction of the pole being approached. Each

is a mathematically defined mixture of affiliation and interdependence.

23

For example, item 126, ignore, pretend not there is composed of -3 units
of affiliation and +6 units of autonomy.

Moreover, the nine behaviors making up a quadrant each represent
one of the nine basic tracks. They are (0) primitive basics; (l)
approach—avoidance; (2) need fulfillment, contact, nurturance; (3) at-
tachment; (4) logic and communication; (5) attention to self development;
(6) balance in relationship; (7) intimacy-distance; and (8) identity.
The number in parentheses corresponds to the last digit of the SASB item
number. Thus, all items ending with the numeral 2 concern different
forms of need fulfillment and nurturance behaviors. Together, the three
surfaces contain 12 forms of each track. As an example, the variety of
track 2 behaviors concerning needs, contact, and nurturance ranges from

112, stroke, soothe calm; 132, rip off, drain; 222, refuse assistance,

 

 

care; 242, accept caretaking; 332, drain, overburden self; to 342,

 

 

 

nurture, restore self.

 

Interpretational Strategies

 

The theoretical and statistical bases of the SASB lead to a
correspondence between model structure and interpersonal behaviors. For
example, opposite behaviors fall at 1800 angles within the same plane,

as in the opposites of 118, encourage separate identity and 138, enforce

 

 

conformity, and of 128, uncaringly let go and 148, specify what's best.

 

 

 

Complementary behaviors, the tendency to interact in reciprocal
ways, to induce in another behaviors complementary to those offered, are
expressed as equivalent topological positions across the Self and the

Other surfaces. Point 146, pamper, overindulge complements 246, cling,

 

depend, and 117, you can do it fine complements 217, assert on own.

 

 

24

Thus, the predictions that oversolicitude tends to elicit dependence and
that friendly encouragement of and belief in another is likely to elicit
that other's self-assertion are built into this model.

Another level of analysis is this model's ability to contrast two
persons' perspectives of the same interaction. In this study, the com-
parison of interest is that between parent and child. Some believe that
strain within a marital couple (Laing, Phillipson, and Lee, 1966) or
within a mother-daughter pairing (Allen, 1976) will be reflected in high
dissonance between the two participants' perspectives of their relation-
ship. Indeed, this was the case in a small exploratory study of SASB
responses of spouses (Benjamin, l979c). The perceptual agreement among
marital couples entering therapy was lower than that among couples not
seeking help (mean consensus of 0.60 compared to 0.86).

In conclusion, Benjamin's SASB records a person's own perceptions
of interactions with significant others. The manner in which the sum-
marizing picture is then constructed, a whole consisting of many inter-
relating component parts, makes this an attractive tool for capturing

familial relationships.

Engagement Styles as a Measure of Perceptual

 

Approaches to Separation

 

Separation has two parts: a leaving from and a going toward.
In adolescence, what is being left is the child's former place within
the family and what is being strived for is a more autonomous way of
life. The student's current View of self and family may be assessed by
the SASB. The study's second tool, a measure of engagement styles, aims

to capture the form of the venture outward. Its data are tlmaadolescent's

25
imaginings of leaving home and making a new place for the self separate
from the family of origin. I
Engagement style is a dimension developed by McKinney (1980)

that taps how individuals perceive themselves in interactions with others
and the environment. One end of its continuum is an agent perception,
which implies an engagement with the environment by "doing" or acting on
it. The opposite end, a patient perception, implies "being done to" by
the environment. A composite of these two extremes is called communion.

This perceptual style represents the interaction of "doing" and "being

done to." It is considered the optimal style of approach for it acknowl-
edges the reality that one is at the same time an actor and acted upon
by others.

At first glance, interpersonal engagement styles, which range
from "doing" to "being done to," seem quite similar to conceptualizations
of locus of control, which range from internal to external. Both con-
structs represent how individuals cognize their environment in that
changes in one's actual situation are mediated by internal conceptual-
izations and interpretations of these changes. Bothereialso processes,
dynamic dispositions, that interact with context, as opposed to stable
traits (Lefcourt, 1976; McKinney, 1980).

However, closer examination of these two constructs reveals that
their relationship is more complicated. Locus of control, derived from
social learning theory, applies to the source of behavioral control and
expectancy of reinforcement while engagement style, an outgrowth of
existential theory, applies to the source of behavior itself. An ado-
lescent may choosewto act or not to act, but perceiving the self as the

source of activity is not equivalent to perceiving those actions as

26
effective or powerful. Indeed, when measures of engagement style and
of locus of control were given to 51 college students they were found
to be unrelated (£_ = 0.01, McKinney, 1980).

Engagement styles also should not be confused with, nor made
dependent upon, value judgments. An agent sees him or herself as doing.
The construct does not take into account whether these actions are
kind, cruel, or ambivalent.

Hotch (1979) translated these general styles of self-engagement
into a situation-specific measure of perceptions of leaving home. She
felt that an adolescent holding a patient perspective might view home-
leaving as being forced upon him or her. When this "done to" sense
permeates the act of separation, a teen may feel abandoned. In contrast,
an adolescent with an agent orientation would view the home-leaving
process quite differently. The emphasis would be upon the self's
activity, upon "doing." For whatever reasons the home-leaving is initi—
ated, once undertaken, the focus would be a personal one. This extreme
centering on the self might result in the separating adolescent feeling
that he or she is the one leaving or, in the extreme, abandoning. The
last style is that of an adolescent holding a communal perspective. The
accompanying view of self might be one of initiative mixed with respon-
siveness. Here, separation becomes a developmental task that is neither
all one's own nor all another's. Rather, the emphasis is on balance.
The adolescent's movement away from home is matched by others' reactions
to that movement. Within a positive family system, such a child's out-
ward strivings would be met by others' support, prompts, space for
experimentation, and willingness to let go when appropriate. This

sensitive balancing of encouragement with comforting is called reciprocal

27
matching by clinicians. Indeed, McKinney and Hotch, focusing on the
developmental period of adolescence, hypothesized an optimal style of
parenting an adolescent that parallels, with age-appropriate behaviors
and expectations, the optimal style of parenting a toddler (Parens,
1979).

The data of another researcher, also studying late adolescents in
transition, can be re-examined in light of these perceptual dispositions.
Coelho, Silber, and Hamburg (1962) administered TAT-type projective
pictures depicting various problem college situations to ten randomly
chosen, ten extremely competent, and ten psychiatrically disturbed
college freshmen. Although Coelho and his associates presented examples
highlighting the variables that concerned them (problem solving skills,
active striving towards a goal, and efficacy of proposed solutions), their
stories can be reinterpreted in terms of self-perceived engagement
styles. For example, one competent student devised an agentic approach
to a problem situation:

. . .He's got 5 hours of homework for the afternoon and a test
tomorrow. He's not worried, but he can't help planning now how
he'll do his studying. He's a good student and prides himself
on his organization of time. (p. 360)
Note the lack of reference to any influence by another's behaviors or
attitudes. To another visual stimulus, this same student imagined a
communal coping response:
I had been asked to attend a meeting as a delegate of the student
council. . .My parents' consent was needed. . .I called them and
they hesitated with an answer. I had planned arguments though
--so they finally gave it. . . .(p. 360)
Here the student both acted in a self-initiated way and was acted upon

by his parents. To the same stimulus, a psychiatric patient created a

very different scenario, one that illustrated a patient style ofrelating:

28

. . .She wants to be a part of the modern dance group, which

means she will have to shell out more money, and she needs

her parents' "say so." Her parents are alert (?) and they

don't both agree with Sally. Especially father, who will

shell out the money. (p. 360)
This woman, in her passive request for parental approval and support,
displayed a reactive perception of herself in interactions with her
environment.

In sum, self-views of engagement style seem to tap an important
quality in the range of late adolescents' responses to a variety of
eliciting situations. Specifically, the willingness to act and to
accept one's potential for autonomous behavior represents an agent

component while the experience of being acted on and of being open to

what others offer suggests a patient component.

Statement of the Problem

 

The aim of this research is to examine perceptions of individuals
preparing to separate. The data are phenomenological. Young women in
transition (leaving high school and about to begin college) and their
mothers described their views of their relationships and created stories
about the upcoming separation. This study's focus is the relationship
between the participants' styles of interacting and their cognitive
approaches to home-leaving.

To date, most research on separation and attachment has concerned
the early years of life. With respect to adolescence, much has been
hypothesized but only a few investigators have attempted to empirically
respond to the issues raised. This study has evolved from those few
works that examined separation in late-adolescence. Just as an adol-

escent disengaging from the family uses experiences of the past to mold

29
a unique and "new" style of being, so has this study grown and "indi-
viduated" from those that inspired it. Murphey et al.'s (1963) broad
subjective judgments by outside observers are replaced by the respondents'
own perceptions, as quantified by Benjamin's SASB (l979b). An examina-
tion of "competent" teens and their parents becomes a more general look
at the possible range of relational styles among "normally" healthy
mother—daughter pairs. The measure of styles of perceiving home-leaving,
developed by Hotch (1979), is expanded to elicit longer stories by
mothers and adolescents about a wider range of separation situations.

In this exploratory work, the choice was made to sacrifice an
understanding of the entire family system for a deeper look at the
mother-daughter dyad. The decision to look at just females was a prac-
tical one. Still, some feel that the separation of daughter and mother
is unique and deserves separate consideration (Hammer, 1975). Probably
the inclusion of these particular family members most strongly reflected
this author's personal memories. My relationship to my mother was
central to the range of feelings, fantasies, and behaviors that were my
experience of separating from home.

Four hypotheses were formulated to be statistically analyzed
via correlations. These predictions relate the respondents' perceptions
of themselves and of their styles of interacting with each other, as
operationalized by the SASB, with their perceptions of their character-
istic approaches to home-leaving, as tapped by the agency-patient
stories.

The following predictions concern both the daughters' and
mothers' relational styles. In a family, views of patterns of inter-

acting need not be complementary. That a daughter is perceived as

30

submissive does not necessarily mean that the mother will be perceived
as controlling. However, by addressing the perspectives of both members
of the dyad the hypotheses become tools for teasing out the perceiver,
mother or daughter, and the perception, of mother or of daughter, that
will account for most of the variance in the range or the observed
perceptual approaches to home-leaving. The hypotheses are summarized
below, then discussed.

la. Adolescent daughters who are perceived as dependent, re-
gardless of the direction of their affiliation, will have a patient per-
ception of separation from the family.

lb. Mothers who are perceived as controlling, regardless of the
type of affiliation associated with this control, will have daughters
with a patient perception of separation from the family.

2a. Adolescent daughters who are perceived as autonomous and
disaffiliated will have an agent perception of separation fromtflmafamily.

2b. Mothers who are perceived as autonomous and hostile will
have daughters with an agent perception of separation from the family.

3a. Adolescent daughters who are perceived as autonomous and
positively affiliated will have a communal perception of separation from
the family.

3b. Mothers who are perceived as endorsing freedom and relating
warmly will have daughters with a communal perception of separation from
the family.

4. In families where mothers' and daughters' perceptions of
their relational behaviors are highly discrepant, daughters are expected
to endorse a patient perception of separation from the family.

Hypothesis 1 focuses upon families who display a high degree of

 

31
adolescent submission and/or maternal dominance. Such relational pat-
terns match Stierlin's (1974) binding mode. Here, a sense of security
and emotional support are dependent upon the adolescent's continued
acceptance of a family role that no longer fits the child's developing
skills nor acknowledges her changing needs and potential. At one extreme,
that of positive affiliation, the bond takes the form of loving symbiosis
(Benjamin, l979b) or, as Murphey et a1. (1963) would have called it,
high relatedness, low autonomy. Quadrant IV (Friendly Influence/Friendly
Accept) behaviors of the SASB illustrate this interpersonal situation, as
the following two questionnaire items demonstrate:

OTHER: 138 For her own good, I specify, tell her what
is best for her to do, he, think.

SELF: 248 I feel, become, what I think she wants.

Hostile control represents the other extreme of the binding mode.
This combination resembles the low autonomy, low related adolescents in
Murphey et al.'s framework. Such high degrees of control over, submission
to, and disaffiliation between are embodied in Quadrant III (Hostile
Power/Hostile Comply) of the SASB. The items below highlight first the
entrapping behaviors and then the experience of entrapment:

OTHER: 138 I enforce conformity to the norms I prefer,
insist she be "proper."

SELF: 238 I follow her preferred rules, standards,
routines.

Whether the bond be loving or hostile, if it tightly binds, as in
the above examples, the resulting lack of autonomy is expected to be
associated with a highly reactive style of perceiving separation. Such
an adolescent's movements away from home would be highly dependent upon

the parent's desires, attitudes, or needs. This extreme reactiveness is

32
the definition of a patient perspective of home-leaving.

Hypothesis 2 explores the effects of autonomy mixed with hos-

 

tility upon the disengagement process. This combination is descriptive
of Stierlin's expelling mode. Here, the experience, depending upon the
perceiver's perspective, would be one of rejecting and pushing away or
of being rejected and ejected. These Quadrant II (Invoke Hostile
Autonomy/Take Hostile Autonomy) SASB behaviors parallel the low related-
ness, high autonomy pattern in Murphey et al.'s schema. The presumed
reaction to such a situation, "I am on my own so I will just have to
make it," seems conducive to an agent perspective of separation. Indeed,
to perceive all behavior as originating from the self would serve to
attenuate any pain arising from the acknowledgement of the lack of
loving support from the parentschnfiJm;home-leaving. By necessity, and
perhaps also by choice, such an adolescent will probably act on her own.
SASB items maximally representative of this dyadic interactional pattern
are:

OTHER: 128 I uncaring let her go, do what she wants.

SELF: 228 I go my own separate way.

The focus of Hypothesis 3 is autonomy, freely given and/or taken,

 

combined with positive affiliation. Stierlin, in his work with extremely
maladjusted adolescents, had no name for this interactional style. Those
who studied normal population did. Benjamin called these Quadrant I
(Encourage Friendly Autonomy/Enjoy Friendly Autonomy) behaviors success-
ful, friendly differentiation. Among Murphey et al.'s sample of compe—
tent high school graduates, this high autonomy, high relatedness pattern

was most frequent. Representative items of this interactional style are:

33

OTHER: 118 I "give her my blessing" and leave her to
develop her own identity separate from me.

SELF: 218 I act according to my own identity,
internal standards.

This coexistence of seeming opposites, dependence upon self with
closeness to others, is the bedrock of Bowlby's (1973) conceptualization
of true self-reliance. Unlike present cultural stereotypes, the healthily
self—reliant person maintains the ability to exchange roles. She balances
self—initiative with the capacity to seek out and accept others' caring
and helping. Not only is such self-reliance compatible with the capacity
to rely upon others, it, in fact, grows out of and is complementary to
such interdependence. The family that supports its child while respecting
her separateness and individual style of coping fosters this inner
security. The adolescent who separates from such a family base does so
feeling the impact of others upon herself while experiencing room and
encouragement to act on her own. It is predicted that this combination
of having felt done to and doing when in the family will be associated
with a communal perspective of home-leaving.

The three hypotheses above are non-reciprocal. At issue in each
are the respondent's viewpoint, whether mother's, daughter's, or both.

In contrast, Hypothesis 4 revolves around reciprocally-matched compari-

 

sons (Laing et a1., 1966). Rather than focusing on whether a daughter
saw her mother as loving or hostile, it compares these two members'
perspectives of the same behavior. The prediction: when mother's and
daughter's perceptions differ greatly, the daughter is expected to hold
a more reactive perspective of home-leaving.

The rationale for this hypothesis is that discrepancies in

perspectives may reflect distorted perceptual and/or misleading

34
communication styles within the family. Shapiro (1967, 1968) and Wynne
(1970) observed that a parent's tendency to defensively delineate (the
need-motivated attribution of a person's own feelings and attitudes onto
another) binds the adolescent homeward. To the extent that these per—
ceptions are communicated covertly, via injection of meaning, conceal-
ment of meaning, and double binding, the tie to the family will be in-
tensified. It is suggested that an adolescent maturing in such a family
will be left feeling mistrustful of her own perceptions and without
confidence in her own developing abilities to live apart from her
family. This depreciation of self suggests deference to another's
seemingly stronger world-view and is characteristic of a patient percep-

tion of home-leaving.

CHAPTER II

METHOD

Respondents

 

This study focused upon mothers' and daughters' perceptions of
their relationship and of home—leaving. It was set in the months between
high school and college. To recruit participants, letters were distri-
buted to all about-to-graduate female seniors at two local high schools
in late May, 1979. Both schools were in communities adjacent to a
large mid-western university (MSU), and they served middle to upper-
middle-class neighborhoods. The letters (reproduced in Appendix A)
briefly described the study and its potential relevance to the adoles-
cents' and mothers' lives. A prestamped return postcard was enclosed to
facilitate expressions of interest. This procedure, letter with post-
card, is the standard technique for university research in local school
systems.

The response rate was dismal. Only four mother-daughter pairs
responded to the 200 letters distributed. Yet, what then seemed limited
to a pilot study soon evolved into the present report. The initial four
families, finding the tOpic and tasks interesting, spontaneously recom-
mended friends and neighbors who might also be willing to participate.

A referral network was thus established. Each agreeable subject added
a few more names to the list. In the end, telephone calls by this re-
searcher to approximately 120 local high school seniors yielded the total

sample of 26 mother-daughter pairs.
35

36

As illustrated by this beginning, the data collection entailed
many compromises. The initial criterion for inclusion, plans to live at
a college at least 30 miles from home, were relaxed to cope with the
recruitment realities. Any willing mother-daughter pair where the
daughter had just graduated from high school and had lived atluxmaduring
her senior year was included. This loosening of standards had no ob-
vious effects upon the study's main focus. Even among those few families
where the daughter anticipated continuing to live at home, informal poSt-
data collection discussions confirmed that mother and daughter felt that
the upcoming school year marked a significant change in their relationship
and their lives. Like the families in which the daughters readied to
leave home, they prepared for September with discussions of expectations,
thoughts of increased independence, and conflicts over projected house
rules.

The process of subject self-selection did produce an obvious bias
in the sample. For example, the high school senior most helpful in dis-
tributing the initial solicitation letters did not take part in the
study. Her mother, citing their strained relationship, refused.
Avoidance or withdrawal by families manifesting pronounced discord was
often repeated. When there was open intergenerational conflict,
daughters seemed unwilling to ask their own mother to participate or
even to suggest the names of otherdissonant mother-daughter pairs.

The end result was a sample skewed toward compatibility. In addition,
the timing of the data collection tended to exclude families with ex-
tensive pre-college separations. If either mother or daughter lived
out-of—town much of the summer, that dyad cited inconvenience and

usually declined.

37
Instruments

Structural Analysis of Social Behavior

 

This study employed a three-part questionnaire designed by
Benjamin (personal communication). It consisted of five-repetitions of
36 randomly-ordered statements corresponding to the points on the SASB's
behavioral maps. Series A instructed the subject to describe how a
significant other (mother or daughter) usually behaved toward her.

Series B requested that the subject describe her own behavior toward

that other. Series C asked the subject to describe how she behaved toward
herself. All questionnaires are fully reproduced in Appendix B. The
accompanying response alternatives ranged from 0% (never, not at all)

to 100% (always, perfectly) of the time.

Both inter- and intra-personal perceptions of the mother-daughter
relationship were assessed. When the focus was upon interpersonal be-
haviors, as in Series A and B, the transaction was subdivided into the
actions and the reactions of the person being rated. For example, in
completing Series A a daughter detailed both how her mother treated her
(expressed on the Other surface) and her view of how her mother re-
sponded to her (expressed on the Self surface). In all, the three
series yielded five behavioral diamonds: other as treating and re-
sponding to self, self as treating and responding to other, and self
as treating self.

One common criticism of direct self-report measures, of which the
SASB is one, is the potentially distorting influence of socially desir-
able response sets. This study strove to minimize such bias by emphasizv
ing confidentiality and encouraging honesty. Even so, social desir-

ability often is not a conscious behavior. Such response patterns most

38
likely operated, especially in the form of extremely positive views of
the mother-daughter interaction.

The distorting effect of such response sets may commonly be
overvalued, though. Benjamin (1977) suggests that such tendencies tap
meaningful personality dispositions. Zlcase in point was her Clinical
experience with a father whose supposed positive "distortions" on the
SASB reflected his actual interactional style. When his behavior changed,
so did his questionnaire ratings. It seems premature, then, to auto-
matically dismiss "fake—good" protocols as invalid. This supposed
source of error variance may accurately reflect a respondent's actual
defensive style of relating or at least his/her perceptions of this
interpersonal style.

To analyze these data, participants' responses were structured
into behavioral maps. Benjamin has studied the validity of these maps
via factor analysis, dimensional ratings, and circumplex analyses. Her
tests spanned diverse populations: (1) self and interactional ratings
of mothers with normal children being seen at a pediatric outpatient
clinic (§_= 171), (2) self and interactional ratings of mothers with
disturbed children being seen at a child psychiatric clinic (§_= 50),

(3) self-ratings by volunteer undergraduate students (§_= 200), and

(4) complete series of ratings covering perceptions of self, significant
other, and memories of relationship with mother and father in early

and middle childhood by adult normals (§_= 60) and psychiatric volun-
teers (§_= 50).

Factor analysis (Benjamin, 1974, l979c) established some, but
not total, support for this model. Four primary factors emerged that

together accounted for about 64% of the variance. Two factors, named

39
"disaffiliation" and "affiliation," corresponded to the extremes of the
models' horizontal dimension. These factors appeared consistently. In
contrast, the two factors labeled "emancipation" and "power," definers
of the vertical dimension, were less clear. Most of the discrepancies
between the SASB's theoretical structure and empirical factor loadings
concerned misplacements along the vertical dimension.

This finding is consistent with the SASB's developmental history.
Affiliation, the more valid dimension, has long been used in inter-
personal measures. But it was not until 1959 that Schaefer (1965)
introduced the Autonomy versus Control dimension to established behav-
ioral circumplexes. Even in his model, this new addition had lower in-
ternal consistency reliability than did the better-established Love
versus Hostility (Affiliation) axis.

Model validity was also tested via judges' ratings of the plane
and quadrant placement of the questionnaire items. Their decisions more
closely matched the model structure than did the factor analytic re-
sults. Items were placed in the correct quadrant most of the time,
especially on the affiliative side of the map. Within the quadrants,
however, the raters' intuitive item-orderings often did not follow the
model's theoretical format (Benjamin, 1979c).

Inter-item correlations were also examined. Working from
Guttman's (1966) conception of a circumplex, Benjamin (1974) constructed
an intercorrelation matrix using the 221 ratings of maternal behavior
by mothers of normal and of disturbed children. The resulting table
of correlations was quite suggestive of the circumpial ordering of the
points, where adjacent items highly positively correlated, orthogonal

items did not correlate, and opposite items highly negatively correlated.

 

40
Yet here again, the match was far from perfect. Isolated items seemed
out of place. These correlated highly with items in other quadrants
while not correlating, or even negatively correlating, with items close by.
In empirical tests of her model, Benjamin also addressed issues
of reliability. Test-retest reliability among blind judges, for plane
and quadrant orientation, fell between 0.85 and 0.93.

Reliability, a measure of behavioral consistency, applies to

people as well as to scales. Generally, people behave in ways that
contribute to personality homeostatis. Less common are sudden dis-
continuities or great variations, which suggest internal stress or
growth (Caplan, 1964). This clinical observation was congruent with
SASB analyses. The coefficient of internal consistency was higher
among normal samples (with an average of 0.92 for normal children and
0.97 for normal adults) than comparable psychiatric samples (0.68 for
psychiatric children and 0.81 for psychiatric adults). These data also
matched the phenomenological experiences of respondents and their
therapists. Less consistent maps corresponded to descriptions of
incongruity and transition (Benjamin, 1974).

In all, the degree of match between Benjamin's theoretical model
and the empirical tests was at times impressive, at other times limited.
In 1974 Benjamin acknowledged that the diamond surfaces may profit from
post hoc rearrangements and questionnaire revision. Minor revisions

have already been made. Further model changes are needed.

Engagement Style

 

This study's measure of imagined responses to separation was a

direct outgrowth of work by Hotch (1979). She asked a class of

41
about-to—graduate high school seniors to write short separation scenes.
Undergraduate research assistants then scored the stories along the
agency-patient continuum. In evaluating each projection, five points
were assigned to fictional heroines who acted in a completely self-
initiated (agent) way; three points were assigned to descriptions of
adolescents separating in a both "doing" and "done to" (communal) sense;
one point was give to figures with a totally reactive (patient) stance;
and four and two points, respectively, were awarded to those heroines that
fell midway between the communal balance and the agent and patient
endpoints.

Of the 64 young men and women who completed both stories, Hotch
found that their responses distributed across the entire range of the
scale, with some skew toward the agency end. Given the stories' brief,
open-ended format, inter—coder reliability was sufficiently high (Story
1, £_= 0.80; Story 2, £_= 0.75), as was the two-item index's reliability
coefficient (0.47). Additionally, these scores were unrelated to story
length or affective tone. Hotch viewed this engagement style measure
to be a useful and reliable tool for distinguishing among cognitive
approaches to home-leaving in adolescence.

Hotch's situations were repeated as the first two stories in
this study. Three more stories and expanded instructions, stressing the
inclusion of the characters' feelings, were added. All can be found in
Appendix C.

For this project daughters were simply asked to create stories.
The mothers' task was more difficult. They were instructed to write from
an adolescent's perspective. Instructions "to pretend that you are now

a teenager" were repeated throughout the task as mothers quickly

42
reverted to identification with the parental figure in their projections
unless strongly cued to do otherwise. The decision to tap the mothers'
adolescent memories was motivated by the desire to assess the effects of
modeling across the generations. Would how the mother left her family
of origin influence how her daughter anticipated leaving home?

A projective approach was chosen for it maximally evoked the
richness of the participants' perceptions, emotions, and conflicts. As
such, these stories revealed their conscious and unconscious feelings .
about separating. Stimuli geared specifically to the experience of
home—leaving in adolescence were selected over more popular projective
measures as the state of the art seemed such that one would have been
unable to generalize confidently from broad personality traits to

specific styles of approaching separation.

Procedure

Following intial contact via postcard or telephone call, this
investigator visited potential participants in the summer months after
high school graduation. Their homes were the experiment's setting--a
technique to increase their willingness to participate and their comfort
while doing so. This also insured confidentiality, a concern raised
repeatedly, especially by the adolescents. The researcher's physical
presence guaranteed that mother and daughter sit separately (usually in
different rooms) and minimized pressures to share reponses. Hopefully,
this concrete assurance of privacy also increased these respondents'
openness and honesty. In almost all cases, mother and daughter were
tested simultaneously.

As this research focused upon a normative life change, the

 

43
study's introduction was straightforward. The investigator referred to
the letters distributed earlier and reiterated the study's purpose: to
learn more about how mothers and daughters approached this important
transition. Confidentiality, from outsiders, school personnel, and other
family members, was stressed. Participants were also promised a short
summary of the results. Before beginning, both mothers and daughters
gave their informed, written consent to participate.

This study had three main parts. The SASB questionnaire and the
five-story engagement-style measure were presented in a counterbalanced
order. Afterwards, respondents could elect to complete the Separation
Anxiety Test (Hansburg, 1972). This pilot for further studies had a
semi—projective format, 12 separation scenes followed by a list of
feeling phrases. Of the 52 participants, 33 chose to complete this
test. Closing the study was a short questionnaire asking for broad
demographic and family information. Time needed to complete all tasks
ranged from one hour to two hours, ten minutes. The mean times were
1.4 hours for mothers and 1.6 hours for daughters.

After data collection mother and daughter were encouraged to ask
questions and vent feelings. Usually, mothers were more open in ex-
pressing themselves. They shared philosophies of child-rearing,
memories of home-leaving, and strategies for helping their daughters
through this transition. Some directly addressed the impact of having
taken part in this study itself. Common among these responses were
feelings of having learned through their participation and of gratitude
for the opportunity to focus on the upcoming changes. One mother tele-
phoned a week after our meeting to thank me. The questionnaires had

inspired a long talk between herself and her daughter that left both

 

 

44
feeling close and more aware of how this transition was experienced

by each of them.

Analysis

To answer the questions raised by the first three hypotheses,
data analyses centered upon the correlations between perceptions of the
mother-daughter interaction and cognitive styles of home-leaving. The
first aspect was tapped by the SASB. This complex instrument yielded
many, often redundant, variables. Those that were most meaningful, in
terms of internal consistency and model structure, required identifi-
cation. The second aspect was measured by the stories scored along the
agency-patient dimension. As this index expanded upon previous work,
an investigation of its reliability necessarily preceded the correla-
tional studies.

The fourth hypothesis focused upon the similarity of mother's
and daughter's perceptions, as assessed by the SASB, of the same dyadic
interactions. Again, correlations were the statistic of choice in
determining the degree of consensus between these two members' perspec-
tives. Extent of perceptual agreement was then compared with endorsed

engagement style of home-leaving.

CHAPTER III

RESULTS

The Measure of Interactional Behavior

 

A nonparametic correlation, Spearman's rho, was used to assess
the SASB's structure. Rho is the ordinary Pearson product-moment cor-
relation for ranked data and ranges from -1 to +1. Rho was selected
because of frequent outlyers (extreme scores by a single subject) and
occasional marked skews (in the direction of disaffiliation on the
horizontal axis and submit on the vertical one) in the SASB ratings.

With complete data from only 25 mothers and 26 daughters, out-
lyers could substantially influence the Pearson correlations. Sometimes
a strong linear relationship, visible on scatter plots, was diluted by
a single case's value. In other instances, one extreme rating markedly
enhanced an otherwise weak correlation. Though these outlyers may have
represented a valid sampling of the population extremes and, thus, exer-
ted a meaningful influence, the small sample size precluded a reasonable
assessment of their representativeness. The solution, a nonparametic
approach, attenuated the effects of both outlyers and skews (Glass and
Stanley, 1970; McCall, 1970; Marascuilo and McSweeny, 1971; Siegel,
1956). On the whole, the rho correlations reported here were quite Close
to their Pearson counterparts.

The SASB sampled a multiplicity of perspectives. In all, this

study tapped ten viewpoints: both the mothers' and their daughters'

45

 

 

 

46
perceptions of how mother acted toward daughter, how daughter acted
toward mother, how mother reacted to daughter, how daughter reacted to
mother, and how each acted inwardly. Thus, each respondent gave two

active, two reactive, and one introspective impression. Respectively,

 

these were represented by the Other, Self, and Introject surfaces. To
simplify the following analyses, only the range and average correlation
across all surfaces will be reported.

The SASB is structured around two independent dimensions. But,

 

within this sample of mothers and daughters, these basic assumptions of
independence and of dimensionality simply did not hold. The first issue,
that of independence of the supposedly perpendicular axes, was assessed
by correlating the two axial scores within a single map. As discussed
earlier, each of the SASB's behavioral descriptions consisted of mathe-
matically defined proportions of affiliation and interdependence. By
multiplying an item's affiliative weight (from -9 to +9 according to
how close it is to the hate or love poles) by the respondent's rating
(from 0 to 100% of the time) and then summing the products across all
36 items on a diamond, a weighted affiliation score is obtained. The
weighted interdependence score is similarly derived, but in this case
the highest weights are assigned to the autonomous behaviors and the
lowest to the controlling behaviors. Together, these two weighted
scores summarize a subject's responses. The results of the correlations
between these global weighted axial scores are displayed in Table l.
Across the daughters' five perspectives, the correlations be-
tween their affiliation and interdependence axes ranged from 0.06 to
0.73, and averaged 0.37 (p_ < .10). Among the mothers' ratings, the

interaxial correlations were lower. They ranged from 0.12 to 0.68, and

47

Table l.--Spearman Correlations Between the Affiliative and Inter-
dependence Axes of the SASB.

 

 

 

 

Perceiver

t' L l
Interac ion eve Mothers (§_= 25) Daughters (3.: 26)
Mother treats .22 .73a
Mother responds .68C .23
Daughter treats .12 .52a
Daughter responds .15 .06

. b
Self View .21 .33

b
Mean r .28 .37
—S

soc .23 .26

 

ap_< .001, two-tailed test.
E.< .10, two-tailed test.

SD = standard deviation.

48
averaged 0.28. The underlying structural concept of two independent
axes was then, among the more extreme surfaces, incongruent with these
data. The negative evidence was strongest among daughters' ratings.

In addition, the assumption of dimensionality must also be
questioned. Aspects of the model's development simply were not dimen-
sional. Benjamin's axes joined two factors end to end, the positive
extreme of one paired with the negative extreme of the other. But a
factor is a bipolar variable. Thus, the union of two factor ends
created less a continuum than the juxtaposition of two variables'
extremes, which may or may not have related in some predictable fashion.
Indeed, a model of different types of behavior at each axis end often
fit the data better than the hypothesized dimensionality.

Benjamin's (1974) working definition of dual dimensionality was
that adjacent items would highly positively correlate, items 90 degrees
apart would not correlate, and opposite items would highly negatively
correlate. Translating these item trends into larger model relation-
ships generated the expectation that the hemispheres would negatively
correlate.

Hemispheres represented opposite ends of a single axis while
neutralizing the influence of the other axis. The affiliative hemi-
spheres were the right (affiliative) and left (disaffiliative) sides
of the behavioral maps. The interdependent hemispheres were the top
(independent/endorsing freedom) and bottom (dependent/controlled) parts
of the map. To determine the hemispheric score, the respondent's
endorsements (0 to 100) simply were added across the 18 behavioral de-
scriptions making up the hemisphere in question. No weightings were

included. As shown in Table 2, the expectation of negative correlations

49

Table 2.--Spearman Correlations Between Hemispheres Across All Levels
of Interaction Assessed by the SASB.

 

 

HEMISPHERE PAIRS

 

 

Interaction .
Level Top- R1 ght- Top- Bottom- Top- Bottom-
Bottom Left Right Left Left Right
Mothers (N = 25)
b c b
Mother Treats 24 -24 57d 45d 04 59
Mother Responds 41C -63: 39 39d 14 18
Daughter Treats 05 —66 40C 37 -08 24
Daughter Responds —18 -712 30 21C 13 24d
Self View 24 -35 71a 44 16 35
Mean £15 15 -52b 47C 37d 08 32
so 22 21 16 10 10 16
Daughters (N = 26)
Mother Treats —13 —67: 61a 86a -18 -363
Mother Responds 16 -36d 45: 44: 27 37
Daughter Treats 26 -34 47d 80c 32b 05
Daughter Responds 09 -20C 36b 42b 58 30
Self View 19 -42 60 51 17 24
Mean £8 11 -40C 50C 61a 23 12
so 15 17 11 21 28 29

Average for Mothers and Daughters (N = 25)

Combined Mean £15 13 -46C 49C 49 16 22
so 18 19 13 20 21 25

 

*All decimals omitted.

ap < .001, two-tailed test.

bp'< .01, two-tailed test.

Cp_< .05, two-tailed test.

dp.< .10, two-tailed test.

50
between opposite hemispheres often was not borne out by the data.

For mothers and daughters combined, the correlations of the
independent (top) and dependent (bottom) hemispheres, the two sides of
the interdependence dimension, ranged from -0.18 to 0.41, and averaged
0.13. Indeed, for eight of the ten perspectives the observed correlations
were slightly to moderately positive. These data suggest that the inter-
dependence dimensions represented the union of two relatively unrelated
types of behavior rather than the postulated continuum between opposite
extremes.

The relationship between the sides of the affiliative dimension
better fit the SASB's theoretical structure. The correlations for
mothers and daughters between the affiliative (right) hemisphere and
disaffiliative (left) hemisphere spanned from -0.71 to -0.20, and
averaged -0.46 (p_< .05). Among five of the ten perspectives, this in-
verse relationship was statistically significant. Though a true dimension
should have even stronger negative correlations, the affiliation axis
approximated a dimension much better than did the interdependence axis.

An alternative method of assessing dimensionality examined the re-
lationships among the quadrants. Similar to hemisphere scores, quadrant
scores are simple sums of the respondents' endorsements of the nine
subsumed behaviors. (Figure 2, shown earlier, provides a diagram of
quadrant placement.) Of interest were adjacent quadrants, sharing the
Same value on one dimension while having opposite values along the other
dimension. Quadrants with the same affiliative values but falling on
different ends of the interdependence axis were observed to measure

Quite similar behaviors. An example was the relationship between

51
Quadrant I (Encourage/Enjoy Friendly Autonomy) and Quadrant IV (Friendly
Influence/Accept). Across these mothers' and daughters' many perspec-
tives, the correlations between these two interdependence opposites on
the affiliative side of the model extended from 0.21 to 0.66, with a mean
of 0.46 (p_< .05). (Refer to Table 3.) This interdependent "dimen—
sion's" discriminative ability was even weaker on the disaffiliative
side. There, the correlations between Quadrant II (Invoke/Take Hostile
Autonomy) and Quadrant III (Hostile Power/Comply) reached from 0.37 to
0.89, with an average value of 0.61 (p_< .01). These quadrants were much
more highly correlated than the model's theoretical underpinnings, which
held that items 90 degrees apart should be uncorrelated, suggested.

As with hemispheres, only the affiliative axis seemed able to
consistently discriminate between quadrants. At the upper half of the
Inap (minimal interdependenceh Quadrant II (Invoke/Take Hostile Autonomy)
<and II (Encourage/Enjoy Friendly Autonomy) were inversely related. The
spread of correlations was from -0.07 to -0.83, with a mean of -0.48
(p_< .02). At the dominate/submit bottom, Quadrants IV (Friendly
Influence/Accept)and ILI(Hostile Power/Comply) were unrelated, with a
range of -.30 to 0.18 and a mean of -0.07. Not only did degree of
affiliation truly vary across the horizontal axis but, in the upper
(maximal independence) part of the map, it seemed the only factor
operating.

Other interhemisphere relationships further clarified the actual
'meanings of the two axes. As indicated by the data in Table 2, socially
desirable behaviors (independence and affiliation) and socially undesir-
able behaviors (dependence and disaffiliation) tended to cluster to-

gether. Across all ten perspectives, the independent (top) and

52

Table 3.—-Spearman Correlations Between Selected Quadrants Across
All Levels of Interaction Assessed by the SASB.

 

 

QUADRANT PAIRS

 

Interaction Across Across
Level Affiliative Axis Interdependence Axis

 

 

I-II III-IV I-IV II-III

 

Mothers (N = 25)

Mother Treats -59: 12 43: 60b
Mother Responds -61 -27 53b 74a
Daughter Treats —68: -15 53b 72a
Daughter Responds —60 -30 61a 46C
Self View -17 -10 65 46C
Mean Els -53b -14 55b 60b
SD 20 17 08 13
Daughters (N = 26)
Mother Treats —83: -06 21a 89:
Mother Responds --49C 01 66 66a
Daughter Treats -41 08 22 81d
Daughter Responds -07d 18 25b 37c
Self View -35 -22 51 40
Mean 5. -43 00 37d 63b
so 5 27 15 20 23
Average for Mothers and Daughters (N = 25)
Combined mean £_ -48C -07 46C 61b
so 5 23 17 17 18

 

*All decimals omitted.
ap_< .001, two-tailed test.

bp < .01, two-tailed test.

Cp_< .05, two-tailed test.

p < .10, two-tailed test.

53
affiliative (right) hemispheres correlated at an average level of 0.49
(range = 0.30 to 0.71, average p_<: .02). Among more negative behaviors,
the dependent (bottom) hemisphere related to the disaffiliative (left)
hemisphere with the same overall strength (range = 0.21 to 0.86, average
£4; = 0.49, average p_< .02). Within just the daughters' ratings, this
observed dependence-d1saffiliation pairing was strongest (average £15=
0.61, p_<: .001). Whether these pairing represent a bias in the SASB
or the actual behavioral correspondence among these respondents is
unknown.

In all, the two principal axes were neither independent nor
truly dimensional. Particularly, the SASB did not assess interdepen-
dence adequately. This limitation was especially clear among daughters'
ratings. To date, this problem has never been directly addressed in
the literature. At the same time, a careful look at the data within the
tables of published research utilizing the SASB, particularly Benjamin's
1974 analysis of her model, revealed problems with the vertical dimen-
sion like those outlined here. Also, in Freedman's (1978) unpublished
dissertation research, similar weaknesses in the vertical axis were
encountered. That researcher, working in a totally different area of
study, found "support on the affiliation dimension but not. . .on the
autonomy dimensions" (p. 63). Possibly, an analysis of the internal
reliability of that study's male and female college students' ratings
'would have yielded an analogous lack of interdependence dimensionality.

These internal SASB problems made it more difficult to evaluate
'the original hypotheses. The meaning of the respondents' global
‘Meighted axial scores were murky. Their weightings, as demonstrated

above, were of questionable value for the two ends of an axis were not

[H

(I)

,._.J

'(3

[CT

m:

o

a:

54
necessarily opposite. In contrast, hemisphere scores measured only one
end of the theoretical behavioral continuum. More limited in scope,
they proved to be more consistently interpretable. Each differed from
its mate in a meaningful way. Though not true opposites, the right and
left sides of the affiliative axis were inversely related (average
£15: -0.46). The top and bottom interdependent hemispheres differed
as two unrelated types (average E15: 0.13). Hemispheres also seemed
more powerful and interpretable than did quadrant scores, for these

smaller circumplex units often related to each other in unexpected and

irregular ways.

Styles of Perceiving Separation

This study's dependent measure was styles of perceiving home-
leaving, as tapped by five projective stories rated along the agency-
patient continuum. The coding schema was developed out of Hotch's
(1979) experience with the measure and modified slightly by this
researcher.

The focus of the rater's judgments revolved around locus of
behavior. In stories scored five, the scale's agent extreme, the heroine
made no reference to others' needs, attitudes, or behaviors impacting
on or influencing her. This overall sense of self-initiated action was
not necessarily equated with detachment from others. A woman who ex-
.pressed happiness over her return home could still be rated as an agent
as long as no explicit reference to others' influence accompanied the
raffect. Below is an example of an agent response to the first situation:
isomeone who has just recently finished high school is living away from

home, away from her family.

‘3

55

She is attending a university. Her days consist of classes
and studying. She enjoys school, living in a dorm, and the
freedom to make responsible decisions. She likes the new
capability to decide where, when, and how she wants to do
things and get them accomplished. After making a few mis-
takes (slept in and missed an important class, forgot to do
the laundry and wears socks twice) she realizes how important
her decisions are and the responsibility they entail.

A score of four was assigned to those stories in which the
heroine was to some limited degree influenced by others yet continued to
act in a basically self—determined manner. Such heroines typically

experienced more unsureness, self-doubt, and dependency than the extreme

 

agent.

Continuing this schema, ratings of three, a communal response,
corresponded to an equal emphasis on the self as doing and being done to,
whether by others or by circumstances. In these stories the pronoun "I"
commonly gave way to "we." This style was characterized by heroines who
both initiated and reacted. The flavor of this communal approach to
home-leaving is evident in the story below, again to the first situation:

Amy, with diploma in hand, is ready for the "real world." She
no longer must go by the standards set for her or obtain the
goals that are proper. Instead she now has a choice of which
path to follow. Her parents no longer have the say of what
she can do and when she can do it. Being "on her own" sounded
fantastic during the last few weeks of her senior year, but

as things get closer she realizes it's not as easy as it
sounds. There will be the hassle of getting ready to move and
the expenses that go along with it. Friends will be left be-
hind and Mom and Dad won't be around for the little things,
although all this is just an initial impact. Soon the ex-
penses will taper off, new friends will be made, and there's
always the phone for minor problems.

Similarly, scores of two represented a shift in the balance.
Iiere, more emphasis was upon being done to as opposed to doing. Although
1:hese heroines had some thoughts of their own, they tended to defer to

tflue ideas and desires of others. Where the more agentic styles had been

56
characterized by acting, trying, and considering realistic alternatives,
these stories were more often filled with hOping, wishing, and
fantasizing.
At the scale's lower end, the score of one represented the
patient extreme. Heroines of these stories played only a minimal role
in the separation dramatizations. Their actions were dictated by dom-
inating others. The adolescents simply reacted, whether compliantly or
rebelliously. Such responses were often extremely emotional. The fol-
lowing reply to the first Situation imparts the reactive flavor of
these patient stories:
The excitement and tension of graduation night has subsided.
Mary thinks back to her mother's anger in the weeks pre-
ceding graduation, bemused at the traditional apron-strings'
pulling. She sets the thoughts aside. Forrumnshe has to
worry about keeping up withrmu:mother's expectations. If
she fails it will be a devastating blow to herself and her
mother. Tomorrow she must go for some more job interviews,
those awful things that mother says builds character. Ha,
Mary thinks, all they do is give you a restless night.

In sum, the cognitive thread that underlaid this coding progression was

the relative emphasis of acting and of being acted upon.

In this study, two undergraduate research assistants, unfamiliar
with the hypotheses, coded all stories. To assure independent scoring,
stories were read, rated, and recorded in a completely random order.
Each assistant worked separately.

Their inter-coder reliabilities were encouraging. Among
daughters' stories, raters' judgments correlated from 0.69 to 0.91 with
a mean of 0.82 across all five. Inter-coder agreement increased to
0.85 when Story 3, the lowest, was set aside for reasons that will be

¢axplored shortly. The correlations for ratings assigned to mothers'

Estories were even higher. Ranging from 0.88 to 0.97, the average

57
correlations between raters' decisions was 0.92. These statistics
indicate that the coding schema adequately communicated the scale's
delineations.

An examination of the internal reliabilities and correlations of
this engagement style index further identified the components of this
measure. Among the 21 daughters who completed all five stories, the
most reliable rater's alpha measure of internal consistency was only
0.39. After Story 3's removal, alpha rose to 0.63. Reliability levels
remained stable when the data from the five young women who wrote only
four stories were included. In the end, then, all 26 adolescents could
be used in the analyses.

Clearly, Story 3 with its low inter-coder agreement and internal
consistency, contributed much error to the daughters' measure of engage-
ment styles. Not surprisingly, it correlated negatively with the other
stories. Accordingly, this third projective situation, of having to
liveanzhome although desiring to leave, was removed from the overall
index. A general index of engagement style, then, was the mean score
across Stories 1, 2, 4, and 5.

Next, an intercorrelation matrix was constructed using the best
coder's engagement style ratings across all adolescents for each of these
four stories. From visual inspection, it appeared that Stories 1 and 4
went together (£_ = 0.46, p_< .05) as did Stories 2 and 5 (£_= 0.42,

2 .
p_= .05). Correlations between clusters were usually much lower.

 

2In following the example of McKinney and Hotch, the scale's
developers, this measure's internal reliability was examined via
Pearson correlations

 

58
Moreover, these correlational pairings made intuitive sense.

The first pair focused upon leaving home. The task of Story 1
was to imagine "Someone who has just recently finished high school and
is living away from home," and that of Story 4 wastx>write about a
"Daughter and mother who have just arrived on a college campus together.
The first term is about to start." Internal consistency, corrected by
the Spearman-Brown Prophecy formula, of this Going subscale (Stories 1
and 4) was 0.63. The second pair had as its core the experience of

renewing contact with the family after having left home. It consisted

 

of Story 2, "Someone who has been living away from home for a while is
now back for her first visit home," and Story 5, "After three months of
being away at college, a daughter receives an unexpected telephone call
from her parent." This Returning subscale (Stories 2 and 5) had an
alpha value, again corrected by the Spearman-Brown formula, of 0.59.

Both subscales overlapped to some degree. The Going and Return-
ing subscales had approximately 10% of variance in common (£_= 0.32,
as). Each shared at least 50% of its variance (£_= 0.83, 0.77, respec-
tively) with the General Index (comprised of Stories 1, 2, 4, and 5).
In effect, both subscales had logical and empirical bases. They seemed
to measure perceptions of leaving home, though slightly different as-
pects of this.

Next, the mother's stories were examined. They were asked to
participate in this aspect of the study because separation is a process
in which the entire family participates. It was thought that the
daughters' preparations would elicit the adults' own memories and un-
conscious conflicts. Indeed, this seemed the case. However, the pro—

jective instrument, as used here, failed to capture these feelings.

 

59
The internal reliabilities of the mothers' stories were poor.
The alpha value remained at 0.30 for the last 20 mothers who completed
all five stories even when the worst story was removed. But these low
internal consistencies illuminate a flaw in this study's design.
The mothers' directions matched their daughters in all respects
but one. While daughters were allowed to imagine freely, mothers were

repeatedly reminded to write from an adolescent perspective. Most

 

often, they did so. But a quirk in this test's development highlighted
the misjudgment of forcing a viewpoint other than the storyteller's own.

At the start of data collection, only the general introduction
to the storywriting task included instructions to the mothers to pretend
that they were again adolescents. The first four adult respondents
quickly reverted to identifying with the parental characters in their
creations. To "correct" this, prompts to write from a teenage perspec-
tive were inserted after the description of each scene.

When the data of these first four mothers (each completed only
four stories) were included in the mothers' figures, the alpha measure
of internal consistency jumped to 0.66. To allow mothers to write from
their own perspective obviously made for a more reliable test. These
numbers agreed with the mothers' spontaneous comments. They felt that
time storywriting task would have been easier and more meaningful if they
had been allowed to write from an adult perspective. It appears that
such an approach would also have been more valid. In light of the actual
instructions and the resulting lack of internal consistency, I judged it
futile to include the mothers' scale of engagement styles in further
analyses. I did learn, though, that a projective properly geared to

parents' own feelings would yield valuable information about their

60
approaches to separation.

In the end, only the daughters' stories allowed an intelligible
assessment of self-engagement styles. Their perceptions of home—leaving
could be subdivided into three related yet distinctly different measures: a
General Index of engagement style (their mean scores across Stories 1,

2, 4, and 5); a Going score (their means across Stories 1 and 4); and
a Returning score (their means across Stories 2 and 5). The 26 adoles-
cents' General Index scores, ranging from 1.7 to 5.0, covered most of the

five—point agency-patient continuum. With a mean of 3.2 and a standard

F‘mrfirt :1

deviation of 0.72, its distribution was skewed slightly toward the upper
(agent style) end of the scale. On the two subscales, the daughters'
scores ranged from 1.5 to 5.0, with means near the communal midpoint and
little skew. In addition, engagement style score and story length were
unrelated. (With Story 4 as the randomly chosen test case, r = 0.14).

In sum, it has been shown that this measure of engagement style
was capable of describing, in both a meaningful and discriminating
fashion, these late adolescents' perceptions of home-leaving. As used
in this study, the measure did not succeed in this task among the

mothers.

Correlational Analyses

 

Hypotheses l, 2, 3

 

The first three hypotheses, relating patterns of mother-
daughter interaction with styles of perceiving home-leaving, were ex-
amined together. Interpersonal behaviors were assessed by the SASB.
The engagement style of separation was tapped by daughters' projective

stories. Briefly, Hypothesis I predicted that maximal dependence,

61

regardless of the associated type of affiliation, would be related to
a patient perception of separation. Hypothesis 2 stated that the auto-
nomous, disaffiliative combination would be related to an agent percep-
tion. Finally, Hypothesis 3 paired autonomous, affiliative behavior with
a communal perception of separation.

As demonstrated earlier, the axes of Benjamin's behavioral model
were neither independent nor truly dimensional. Accordingly, the meanings
of the global dimensional scores and other scores were not just as

claimed, nor were they always easily interpretable. Therefore, the

correlations performed to test the hypotheses involved many levels of

the SASB model: 1) global scores (a map's median and mean), 2) the
axes' dimensional scores, 3) the hemisphere scores--a total of twelve
measures.3 In addition, there were three engagement style measures. Due
to this large number of overlapping variables and the resulting increased
risk of Type I errors, the decision was made to exclude isolated statis-
tically significant correlations from this report. Also, given this
study's exploratory nature, the examination of both significant findings
and related trends seemed appropriate. And again, to protect against the
inordinate influence of outlyers in the SASB ratings, Spearman's rho was
the statistic of choice.

Mothers' perceptions of the dyad. AS depicted on the left side

 

of Table 3, there were numerous statistically significant relationships

between mothers' perceptions of family interactions and daughters' styles

 

3There were also quadrant and smaller cluster scores. To avoid
confusion and random post hoc hunts for findings, these were reported
lonly when they shed light upon possible meanings of the dominant
findings.

62
of perceiving home-leaving. As the components of this relationship
were derived from different perspectives (mothers' and daughters')
and instruments (direct and projective), this prevalence of findings
was especially exciting.

With the SASB scores, it did not matter whether mother was
rating how she treated daughter, how she responded to daughter, how
daughter treated her, how daughter responded to her, or how mother
treated herself. Across all perspectives, the general thrust of the
findings remained constant. The manner of measuring engagement style,
though, did matter. The Going subscale, made up of stories 1 and 4, was
most sensitive to the observed linkages. It contributed 11 of the 16
0.05 significance level correlations. The General Index accounted for
the remainder. The findings are outlined below.

1. Disaffiliation correlated with agency: Mothers' perceptions
of disaffiliative behavior related to the daughters' agent projections to
the home-leaving situations, as measured by the Going subscale. This
finding was consistent across all four levels of interpersonal behavior.
The relationship took the form of high positive correlations between the
mothers' disaffiliative (left) hemisphere ratings and the daughters'
stories. These correlations ranged from 0.46 to 0.55, with a mean of
0.51 (average p_< .02). Strong relationships between both quadrants
making up this hemisphere and the daughters' agency scores reinforced
this finding.

When mothers focused on their own interpersonal behaviors (how
they treated and responded to their daughters), this relationship between
disaffiliation and the daughters' styles of perceiving separation ex-

tended to the global weighted affiliation scores (for mothers‘ active

63

S. v cu

 

 

 

 

 

 

. r z .mutbcozcc "mm m z .MLS yo:
NN r s we. v as
mm H z .upottuswo new h z .muSZCZ .
6 uc V as
.u o - c: » fl -
sN z aster no «N z ustroozp .poooaec «.osauop S_<.
.N dN- 4N- S. oN NN- do- Ho- oN So LN o. co Sc- rs ca oracssotr
c.- Na cN- rc NS N. ca- No- use So r_o a. N_- do- oN ooS oc_oc
S_- _o- are- Na nSv cc SN- c_- a. SN son oN cc N_ urn nSe austere
zo_> Naom
,_- Nc- cc- ae- oc S_ ~_- N. as- no- Sa- mo son as. mo- So assessor:
No q. So- so- So do m. as new :N- SN RN oc 66S- NS new oraou
N- Sa SN- NH- o“ NN so- we rN No. so 6mm 6mm N- cm oN Nanette
anaconda umuzosoo
so- rN :: as no S_ aN Na S: S_ oa- see as so no So oracssoom
.o- .o- N_- ro- a~ so me. so see c_- SN as SN- cN- oN new assoc
ac- cN vc- om RN SH Nd me as as No see co co SN oN autocto
museub uuusvsuo
so- cm on oi- SN- mN a. N. ea No- «a so no so- NN Sa usacusuom
r: ca oi as. m_- as so Na see ee- So an Ss- rim- SN nos assoc
__- rN n. as. do- SN so #0 SN we- as oSn e_ Na- an mm Houocoo
mvcoamom nocuo:
as- SN S_- S_ rN mN “a- N. «N as- so- a. mN ca- ao- Sa «ocacusuom
SS Sc- N: ca Na Sc- N" N. ch eN- ma so- mn- an- no see oocaoc
m_- NH cN- mo pen NN n.. as an o_- no Ma mo vN- no em assures
mWflOhfi Hus—.30:
smog atria .:z ;C& accucuz~ ~wuu< new: cc: uufi_ yzcaa uoz do? docueucm kuu< and! co:
JII -Hn- .-|Ii- a. - 11l|119d.-fiwt-1iilllv o—xum ucoEouwwcm
Suez. "L56: , e. .11- rt-N<- cu in v SEQ: muse 0cm ~0>S :0«uom.~ouc~
m:0..;cozta .mttrzcza: I mcomuacUuod .muuzuoz

 

 

 

 

.mcuoom 6~>bm accfitcctcm .muouzczmo vac mocfiuom ompc>c mmdm coozucm mcowuwuouuoo :thcon-I.v wanna

64
behaviors, E15: —0.42, p_< .05; for mothers' reactive behaviors,
51;: -0.51, p_< .02). This suggests that mothers' views of warm,
loving interactions possibly relate to less agentic (i.e., communal or
patient) perceptual styles. However, there was no support for this ex-
planation among the affiliative (right) hemispheric correlations. The
dominant finding, then, remains one of maternal perceptions of dyadic

disaffiliation corresponding to daughters' agency in home-leaving pro-

jections.

 

2. Mothers' median rating correlated with agency: An inter-
esting and unexpected finding consistent across all levels of focus was
that the mothers' median scores (numerically ranked midpoints) corre-
lated strongly and positively with their daughters' engagement style.

Here again, the Going subscale was most sensitive to this relationship.
When the mothers focused on their daughters' behaviors, the Spearman
correlation was 0.54 (p_< .01). When the mothers focused on their own
interactional behaviors, rho was 0.46 (p’< .05). When mothers looked
inward, the magnitude of the relationship with their daughters' Going
subscale approached significance (_r_s= 0.39, p < .10) , while that
with the General Index of engagement styles was significant (£15= 0.47,
p.< .05). Overall, the higher a mother's median rating on the SASB the
more likely her daughter created agent responses to the separation stim-
uli. In contrast, the mean scores (the statistical averages) of mothers'
interactional perceptions were unrelated to their daughters' engagement
styles.

3. Mothers' self-control correlated with agency: One other

IDattern of interest was the positive relationship between mothers' per-

<2eptions of high self-control (bottom hemisphere of the intrapsychic

65

surface) and daughters' agent perceptual styles, as measured by both
the General Index and Going subscale (average 515: 0.42, p_< .05). It
appears, then, that the mothers who "protected self" were likely to have
daughters who imagined leaving home in a self-sufficient manner.

This linkage was not, however, repeated with the interpersonal
bottom hemisphere equivalents (dominate, submit) of the "control, manage
self" intrapsychic mode. Indeed, at the interactive level, the relation- ..
ship suggested by the data was of a very different nature. In describing 4‘
how daughter treated, how daughter responded, and how they themselves
responded, mothers' top hemisphere (endorse freedom, take autonomy)
rating correlated to daughters' agentic styles (in order, £15= 0.44,
p_< .05; £15: 0.35, p_< .10; E15: 0.37, p.< .10). Because these
findings were statistically weaker and linked with conflicting trends
among the smaller model parts of the SASB, it is not feasible to inter-

pret them with any confidence.

Daughters' perceptions of the dyad. In contrast to mothers,

 

daughters' familial perceptions were almost totally unrelated to their
styles of perceiving separation. The right half of Table 3 provides an
overview of these correlations. There was, however, one exception:
daughters' self—control correlated to a patient style of home-leaving.
In general, daughters who kept a close eye on themselves,
controlled themselves, and restrained themselves tended to write patient
stories. This relationship was strongest among daughters whose self-
control appeared intrapunitive. Evidence included a strong negative
correlation between the "control self" (bottom) hemisphere and the
<3eneral Index of engagement style (££5= -0.48, p_< .02). Within this

Ilemisphere, Quadrant III (Oppress Self) inversely related to engagement

66
style (£15: ~0.40, p_< .05). Remaining at the intrapsychic level, the
global weighted interdependence score positively correlated (£13= 0.47,
p_< .05) with the stories' ratings. Behaviorally, this again translates
to a correspondence between greater self-control and a patient perception

of home-leaving.

Hypothesis 4

 

This last prediction focused upon the consensus of mothers' and
daughters' perceptions of their relationship. This was measured by
correlating mothers' and daughters' views of the same behaviors, e.g.,
the degree to which both family members saw the mothers' treatment of the
daughter in the same way. Specifically, it held that when their per-
spectives differed greatly the daughter would have a patient perception
of home-leaving. Within this group of respondents, the degree of per-
ceptual match varied from -0.54 (quite discrepant) to 0.94 (nearly
identical). Overall, there was a fair amount of mother-daughter con-
gruence (average £_= 0.52). However, variations in perceptual con-
sensus did not correlate with daughters' engagement styles of home—

leaving.

Summary

In this limited sample, mothers' perceptions of dyadic inter-
actions consistently correlated to daughters' styles of home-leaving.
Disaffiliation between mother and daughter, median SASB response choices,
and mothers' self-control were the primary foci of this relationship.
All related with the agent end of the engagement style continuum.

.Among daughters, self-restraint linked with a patient stance.

CHAPTER IV

DISCUSSION
While these data did not conform to the model suggested by the
initial hypotheses, a number of results provide insights into, and raise
many questions about, the experience of growing and separating in late
adolescence. In this chapter, the nature and possible implications of

these findings are explored.

Relationships Hypothesized and Observed

The first hypothesis held that maximal interdependence, whether

 

in the form of a mother controlling or a daughter yielding, would cor-
respond to a patient orientation toward separation. Although unsupported
at the interpersonal level,there were intrapsychic linkages. Behaviors
offered the self, as measured on the Introject surface, related to
daughters' perceptual styles of separation. Daughters who controlled

and managed themselves (bottom hemisphere behaviors) tended to embrace

a patient perceptual style. In contrast, an agent style was common to
daughters with mothers who controlled and managed themselves.

On the surface, these findings of strong self-control relating
to opposite engagement styles may seem contradictory. It is important
'to note, though, that these self-managing daughters and mothers often
inere not related. A highly controlled mother did not necessarily raise
63 highly controlled daughter. Indeed, the average correlation between

Inothers' and daughters' perceptions of the behaviors they offered

67

68
themselves was only 0.44 (range = -0.15 to 0.81, SD = 0.27), under 20%
shared variance.

The reason for the opposite linkages may lay in the dynamic
impact of these different family members' self-control upon the adoles-
cent story-writer. Daughters who keep a "benevolent eye on self,"

"force ideal identity," "force propriety," and "hold self back" would
appear continually vigilant of and sensitive to others' impressions.

This is the definition of a reactive stance. On the other hand, mothers
who put such energy into managing themselves may well have little energy
left for their offspring. Left more to their own devices, these daughters
could more likely see themselves as initiators. Or it may be that
daughters of highly controlled mothers chose to reject that constricted
approach, perhaps as out-of—date or unnecessarily rigid. In turning away
from their mothers' life-styles, these daughters may also have cut them-
selves off from influence by others within the family. This latter

case would yield a more agent style.

The second hypothesis stated that adolescents who acted in a

 

disaffiliated and autonomous manner, or were treated by their mothers in
a hostile yet endorsing-freedom style, would have an agent perception of
separation from the family. Half of this autonomy-hostility combination
clearly linked to specific engagement styles. Among mothers' perceptions,
there was a strong and consistent relationship between maternal percep-
tions of dyadic disaffiliation and daughters' agentic Projections. This
relationship held whether mothers described themselves or their daughters,
their actions or their reactions.

As correlations are not evidence of causality, one is left won-

dering whether these daughters' agent perceptions of home-leaving were

69
the precipitant or the product of hostility in the parent-child inter-
actions. It may have been that within these relatively hostile families
the adolescents rushed to separateknrclosing themselves off from family
influence. Equally likely, however, is a scenario where the daughters'
self-initiated (agentic) approach to home-leaving elicited feelings of
rejection and anger from the mother. In either case, this disaffiliation-
agency connection suggests that family disaffiliation pairs with premature
self—sufficiency and interpersonal isolation in these adolescents' home-
leaving styles.

The third prediction associated affiliation and autonomy with

 

communal perceptions of home-leaving. There was no support for this
hypothesis. However, there were also faults in this research design
that would have obscured such a relationship even if it existed. The
communal midpoint, wherermxnzof these respondents' scores clustered,
could not be isolated by this study's correlational analyses. Alternate
statistical tests that would have highlighted relationships with the
scale's middle values were prohibited by the small sample size.

The fourth hypothesis paired perceptual discrepancies with a

 

patient stance. Again, these data did not support this relationship.
One strong and totally unexpected finding was that mothers'

median interactional ratings correlated with daughters' agentic engage-

 

ment style, as tapped by the Going subscale. This occurred whether
mothers described their own or daughters' behaviors. It is unlikely that
this simply reflected shared response bias as the linkage spanned two
family members' perspectives and two very different tasks. Also, the
relationship was not duplicated by the mean, another global measure.

lIt seems, then, that the median behavioral rating has tapped some

70
important aspect of these mothers' perceptual sets.

The median is simply the middle score when ratings (frequency
of a given behavior in this case) were arranged from lowest to highest.
In correlations with other model parts, the median highly related to
disaffiliation (left hemisphere) and submission/dominance (bottom hemis-
phere) and was unrelated to affiliation (right hemisphere) and autonomy
(top hemisphere). In a sense median scores seem to have tapped mothers'
willingness to share the incidence of socially undesirable dyadic inter-
actions. Indeed, when disaffiliation was partialled out of the median
rating-agency style linkage, the correlations dropped sharply (range =
0.18 to 0.44, average £fi= 0.29). Though no longer statistically sig-
nificant, there remained a consistent if weak contribution. This suggests
that mothers who tended to see their relationship as being made up of
many different behaviors, both negative and positive, generally had
higher medians and raised daughters with more agent than patient per-
ceptions of home—leaving. The median may symbolize more about the
mothers' interactional modes, but from these data it is impossible to
tease out the meanings of the residual. It certainly warrants further
study.

Another interesting and surprising finding was that mothers'
behavioral ratings linked more closely to daughters' perceptual ap-
proaches to home—leaving than did daughters' own interactional ratings.
This disparity is perplexing. It may reflect difficulties in the be-
havioral measure or aspects of these different family members' personal
responses to the upcoming separation.

From an empirical perspective, problems with the SASB's basic

sstructure were magnified in the daughters' ratings. Among these younger

71
respondents, the two axes exhibited less independence (0.37 mean cor-
relation between the affiliation and interdependence axes versus 0.28
for the mothers). More specifically, daughters confounded maximal
interdependence (whether in the form of dominance or dependence) with
disaffiliation, for in all five perspectives their bottom and left
hemispheres correlated significantly (average r15: 0.61, average p_ <
.001). These negatively valenced behaviors were more loosely linked
(average £S= 0.37, average p < .10) for mothers. As a result, the
most powerful predictors among the mothers' ratings, disaffiliation and
dominance/dependence, were the very variables that were most difficult
to measure in the daughters' data.

Psychological explanations for the dearth of correlations between
daughters' interactional ratings and their engagement styles also exist.
Data collection took place just before the separation. Daughters, when
assessed here via self-reports, most often described their mothers as
affiliative and encouraging of autonomy. In clinical work with college
students, this researcher has learned that such initial pictures of home
often cloak very different feelings. A tentative hypothesis that at
times of separation young adults often conceal, distort, or repress
negative feelings and memories is supported by an unpublished study that
compared the content of intensive, individual interviews with Thematic
Apperception Test stories. Using a parallel sample of recently sepa-
rated freshmen and women, Hurley (personal communication) noted that in
the interviews adolescents usually portrayed close, loving families.
However, when given the projective distance of the story-telling task,
anger, distress, conflict, and sadness were typically added to their

superficially more positive portraits.

 

72

Even if this understanding were valid, would not these same
processes operate among mothers? Perhaps not. Raising a child, no less
an adolescent, is at times a frustrating, conflicting experience.
Society, to some extent, sanctions the expression of such feelings. Ad-
ditionally, some mothers, invested in other children, satisfying careers,
or healthy attitudes toward this transition, may have had less psychic
energy tied up in the separation than their daughters whose entire life-
styles would soon drastically change. Feeling less personally threatened
by the upcoming transition, these mothers may also have felt less de—
fended and, thus, better able to maintain a more realistic view of both
the good and bad in their relationship. Another possibility is that
mothers have greater perspective, by nature of their age, experience,
and stability, of family interactions. I would not be surprised if

these mothers knew their daughters better than their daughters knew them.

In Review

Aspects of subject selection limit generalizability of these
findings. Particularly, the sample was small, basically middle-class,
and highly verbal. Few families with open conflict or prolonged pre-
college separations volunteered. Separation in late adolescence may
be quite a different experience among various other populations. To
their credit, though, the participants displayed much honesty, openness,
and thoughtfulness. Their stories, especially, were quite revealing.

As discussed previously, the study's measures had both limits
and strengths. With the story writing task, the engagement scale's
<:ommuna1 midpoint, an interactional style qualitatively different from

‘the agent and patient extremes, eluded this study's correlational

 

73
analyses. Also, the mothers' version, probably due to restrictions on
the story-teller's perspective, was unreliable. In the index's defense,
it successfully differentiated and reliably assessed these adolescents'
styles of perceiving home-leaving. This scale was also sensitive to
the different aspects of separation, the leaving and the returning home
again. Another important by-product of the story-telling technique were
the fantasies, fears, conflicts, and adaptational styles exposed. Reading
the stories was fascinating. This researcher plans to later return to
the projections, to pull from them further aspects of the disengagement
processes.

The other measure, the SASB, promised more than it delivered. But
it appears to have been unrealistic, given the present state of psy-
chological assessment, to expect a single paper-and-pencil questionnaire
to reliably assess the complexity of behaviors contributing to dyadic
affiliation, interdependence, and their interaction. Indeed, the SASB's
greatest limitation is a problem true of all direct, self-report instru-
ments. They, despite researchers' hopes, seem to elicit only the more
conscious layers of behavior. The SASB's structured response choices
permit easy defenses, especially of the more negative aspects of family
interactions. The unconscious, an important contributor to feelings and
behaviors, appears to be more clearly tapped by the projective task.
There, the conscious and unconscious naturally combined into scenarios
of binding and expelling, of conflict and support.

In addition, the instrumentation of the SASB's model was flawed.
In this sample, the interdependence dimension was not clearly and con—

sistently measured. As a result, the global weighted scores were not

74
an accurate standard for comparison nor did the quadrants differ from
one another in consistent and meaningful ways. Still, in research with
adolescents to date, there has not yet been a quantifiable, reliable,
and encompassing measure of dyadic interdependence. Stierlin (1974),
Shapiro (1968), and Murphey et a1. (1963) relied on subjective clinical
judgments. Hotch (1979) and Sullivan and Sullivan (1980) developed
reliable and quantifiable instruments but in doing so were forced to
reduce the concept to ten or less multiple-choice items. More work is
needed before a meaningful and objective measure of dyadic interdependence
can be developed.

These criticisms do not negate the SASB's many merits. The
affiliation axis clearly discriminated degrees of closeness and conflict.
Even more striking was the SASB's attention to the multiplicity of per-
spectives within a single dyadic interaction. Paralleling the study's
theoretical emphasis upon participant's interlocking contributions to
the separation, this instrument integrated the active, reactive, and
introspective components of each individual's personal reality. Indeed,
one important outcome of this project was the finding that a mother's
behavior toward self or her reactions to her offspring, not just the
commonly studied treatment of her daughter, related to the daughter's
perceptions of home-leaving. The dataq'then,echoe the emphasis of

contemporary theory upon the interactive push-pull within the family.

 

Looking Toward the Future

 

At this study's end, I am left wondering about current concep-
tualizations of separation in late adolescence. The most creative and

seemingly valid theoretical models have come out of the work of family

75
therapists with extremely disturbed youths and their parents. This
study's hypotheses postulated that the same psychological principles
would apply, though in lesser degrees, among healthier families from
which adolescents were about to successfully separate. This may not be
the case. Both the process and its determinants may differ, in degree
and in nature, among these two classes of families. Accordingly, the
patterns of the results may have diverged dramatically had these same
tasks been given to psychotic daughters and their mothers.

Clinical psychology has a long and distinguished history of
starting with pathological behavior and building from there. One example
is Freud's invaluable conceptualizations of psychosexual development in
early life. Yet later work by Mahler (Lax, Bach, and Burland, 1980)
observing the daily development of healthy toddlers, has expanded, in
important ways, analytic understanding of the first separation-
individuation period.

Let this study's close,then4 be a call for more observational
and theoretical work on normative separation responses in late
adolescence. Presently, psychologists have developed a list of factors
that seem to interfere with disengagement. But we are not yet at the
next step: understanding the factors that promote healthy separation
and individuation. Murphey and associates (1963) started this work
through their conversations with competent adolescents and their families.

But their end-product, fascinating descriptions of reciprocal patterns,

 

falls short of a unifying conceptualization of relational styles that
lead to growth at this point in the life cycle. There is a need now
to replicate,with normal families and adolescents separating, the

.in-depth observational and broad theoretical work of Stierlin (1974).

76

Equally important is the need to place this separation process
in the context of the life span. From this and other studies, frozen
at one point in adolescence, we know neither how certain disengagement
strategies have evolved nor the various adult life-styles to which they
contribute. It just could be, from the perspective of individual growth,
that the seemingly ideal affiliative autonomous response is more re-
strictive in terms of adult development than the disaffiliative auto-
nomous one, with its (possibly temporary) lack of investment in the
parental bond. Adolescent separation is a major rite of passage on the
way to mature interdependence. To truly understand its meanings and
ramifications, one must trace the path chosen over at least the next
decade of the adolescent's life.

This call for a broad perspective certainly in no way negates
the need for careful empirical work on the smaller parts of the process.
Still, putting all our research energies into studying broad group
differences seems premature when we have not yet discovered the nuances,
the ranges, the intricacies of individual styles of home-leaving in
adolescence. The scope of future work should also be extended by
expanding the dyad to at least a triad. The choice to study just
mothers and daughters was often challenged by the participants. They
clearly felt that a father or even an older Sibling play a prominent role
in their responses to this life change.

Finally. there is a need to make better contact with the family's
unconscious processes. At times of transition, experienced by most as
times of stress, much mental activity is conflicted, unclear, covert.

The engagement style stories were a first step, and the recommended

changes to make them consistently sensitive to parental feelings is a

77
good follow-up. Additionally, it would be both fascinating and useful
to supplement direct-response questionnaires of family interactions with
projective tests. Possible tools for this task include the Patho-
genesis Index (Robbins and Karon, 1980) and the Family Rorschach

(Loveland, Wynn, and Singer, 1963).

LIST OF REFERENCES

LIST OF REFERENCES

Ainsworth, M. D. S. Object relations, dependency, and attachment: A
theoretical review of the infant-mother relationship. Child
Development, 1969, 49, 969-1025.

 

Allen, J. Identity formation in late adolescent women. Unpublished
doctoral dissertation, City University of New York, 1976.

Anthony J. The reaction of adults to adolescents and their behavior.
In G. Caplan and S. Lebovici (Eds.), Adolescence: Psychosocial
perspectives. New York: Basic Books, 1969.

 

Bateson, G., Jackson, D., Haley, J., & Weakland, J. Toward a theory
of schizophrenia. Behavioral Science, 1956, 1, 251-264.

 

Baumrind, D. Current patterns of parental authority. Developmental
Psychology Monograph, 1971, 4(1, Pt. 2).

 

Bell, R. Q. A reinterpretation of the direction of effects in studies
of socialization. Psychological Review, 1968, 25, 81-95.

Bell, R. Q. Contributions of human infants to caregiving and social
interaction. In M. Lewis, & L. A. Rosenblum (Eds.), The effect
of the infant on its caregiver. New York: Wiley & Sons, 1974.

Benedek, T. Parenthood as a developmental phase. Journal of American
Psychoanalytic Association, 1959, 2, 389-417.

Benjamin, L. 8. Structural analysis of social behavior. Psychological
Review, 1974, 81, 392-425.

Benjamin, L. S. Structural analysis of a family in therapy. Journal of
Consulting and Clinical Psychology, 1977, 45, 391-406.

lBenjamin, L. S. A manual for using SASB questionnaires to measure cor-
respondence among family history, self-concept and current
relations with significant others. Unpublished manuscript,
l979a.

Ekanjamin, L. S. Structural analysis of differentiation failure.
Psychiatry, l979b, 43, 1-23.

 

Benjamin, L. 8. Update: Validity of the SASB model. Unpublished
manuscript, l979c.

78

79

Benjamin, L. S. Use of structural analysis of social behavior (SASB)
and Markov chains to study dyadic interactions. Journal of
Abnormal Psychology, l979d, 88, 303-319.

 

 

Benjamin, L. 8. Personal communication, May 3, 1979.

Blos, P. The second individuation process of the adolescence. Psycho-
analytic Study of the Child, 1967, 22, 162-186.

Bowlby, J. Attachment and loss, Volume II: Separation anxiety and
anger. New York: Basic Books, 1973.

 

Caplan, G. Principles of preventive psychiatry. New York: Basic Books,
1964.

 

Coelho, G. V., Silber, E., & Hamburg, D. A. Use of the student-TAT
to assess coping behavior in hospitalized, normal, and exception-
ally competent college freshmen. Perceptual and Motor Skills,
1962, 14, 355-365.

 

Duvall, E. M. Family development (3rd ed.). New York: Lippincott, 1967.

 

Elson, M. The reactive impact of adolescent and family upon each other
in separation. Journal of American Academy of Child Psychiatry,
1964, 3, 697-708.

 

Erikson, E. H. Identity: Youth and crisis. New York: Norton, 1968.

 

Erikson, E. H. Reflections on the dissent of contemporary youth.
Daedalus, 1970, 92, 154-176.

Fenichel, O. The psychoanalytic theory of neurosis. New York: Norton
and Co., 1945.

Freedman, S. M. Perceived interpersonal styles and preference for inter-
action: A pheomenological study of the complementary hypothesis.
Unpublished doctoral dissertation, Michigan State University,
1978.

Freud, A. Adolescence. Psychoanalytic Study of the Child, 1958, 1},
255-278.

 

Freud, A. Introduction to: Levy, K. Simultaneous analysis of a mother
and her adolescent daughter. Psychoanalytic Study of the
Child, 1960, 16, 378-380.

 

Glass, G. V., & Stanley, J. C. Statistical methods in education and
psychology. Englewood, N. J.: Prentice-Hall, 1970.

 

Gould, R. L. Transformations: Growth and change in adult life. New
York: Simon & Schuster, 1978.

 

80

Guttman, L. Order analysis of correlation matrices. In R. B. Cattell
(Ed.), Handbook of multivariate experimental psychology.
Chicago: Rand-McNally, 1966.

 

Hammer, S. Daughters and mothers: Mothers and daughters. New York:
New American Library, 1975.

 

Hansburg, H. G. Adolescent separation anxiety: A method for the study
of adolescent separation problems. Springfield, IL: Charles
C. Thomas, 1972.

 

 

Hotch, D. F. Separation from the family: A study of perceptions in late
adolescence. Unpublished doctoral dissertation, Michigan State
University, 1979.

Hurley, J. R. Personal communication, July 2, 1980.

Johnson, A. M. Juvenile delinquency. In S. Arieti (Ed.), American
handbook of psychiatry (Vol. 1). New York: Basic Books, 1959.

 

Laing, R. D., Phillipson, H. & Lee, A. R. Interpersonal perception:
A theory and a method of research. New York: Springer Pub-
lishing Co., 1966.

 

 

Lax, R. F., Bach, S. & Burland, J. A. (Eds.). Rapprochement: The
critical subphase of separation-individuation. New York:
Jason Aronson, 1980.

 

 

Leary, T. F. Interpersonal diagnosis of personality. New York: Ronald
Press, 1957.

 

Lefcourt, H. H. Locus of control: Current trends in theory and re-
search. New York: John Wiley, 1976.

Loveland, N., Wynne, L. C., & Singer, M. T. The family rorschach: A new
method of studying family interaction. Family Process, 1963,
_§- 187-215.

 

Mahler, M. S. Thoughts about development and individuation. Psycho-
analytic Study of the Child, 1963, 18, 307-327.

 

Marascuilo, L. A., & McSweeney, M. Nonparametric and distribution-
free methods for the social sciences. Monterey, Ca.: Brooks/
Cole, 1977.

 

 

McCall, R. B. Fundamental statistics for psychology. New York:
Harcourt, 1970.

 

McKinney, J. P. Engagement style (agent vs. patient) in childhood
and adolescence. Human Development, 1980, 33, 192-209.

 

McLemore, C. W., & Benjamin, L. S. Whatever happened to interpersonal
diagnosis? A psychosocial alternative to DSM-III. American
Psychologist, 1979, 34, 17-34.

 

81

Murphey, E., Silker, E., Coelho, G., Hamburg, D., & Greenberg, 1.
Development of autonomy and parent-child interaction in late
adolescence. American Journal of Orthopsychiatry, 1963, 33,
643-652.

 

Olson, D. H., Sprenkle, D. H., & Russell, C. S. Circumplex model of
marital and family systems: I. Cohesion and adaptability
dimensions, family types, and clinical applications. Family
Process, 1979, 18, 3-28.

Parens, Henri. Developmental consideration of ambivalence: An explo-
rationcfifthe relations of instinctual drives and the symbiosis--
separation-individuation process. Psychoanalytic Study of the
Child, 1979, g3, 385-420.

 

Robbins, D. S., & Karon, B. P. The pathogenesis index: Summary of
research. Unpublished manuscript, Michigan State University,
1980.

Rubinstein, D. Clinical issues in family therapy of schizophrenia. In
D. Rubinstein and Y. O. Alenen (Eds.), Psychotherapy of schizo-
phrenia. Amsterdam: Excerpta Medica, 1972.

 

Schaefer, E. S. A configurational analysis of children's reports of
parent behavior. Journal of Consulting Psychology, 1965, 22,
552-557.

 

Schneider, J. M. Loss and change. Unpublished manuscript, Michigan
State University, 1980.

 

Shapiro, R. L. The origin of adolescent disturbances in the family:
Some consideration in theory and implications for therapy. In
G. H. Zuk and Boszormenyi-Nagy (Eds.), Family therapy and dis-
turbed families. Palo Alto: Science and Behavior Books, 1967.

 

Shapiro, R. L. Action and family interaction in adolescence. In J.
Marmor (Ed.), Modern psychoanalysis. New York: Basic Books,
1968.

 

Siegel, S. Nonparametric statistics for the behavioral sciences. New
York: McGraw-Hill, 1956.

Stierlin, H. Separating parents and adolescents. New York: Times Book,
1974.

 

Stierlin, H. Psychoanalysis and family therapy. New York: Jason
Aronson, Inc., 1977.

Stierlin, H., Levi, L. D. & Savard, R. J. Parental perceptions of
separating children. Family Process, 1971, 19, 411-427.

 

82

Sullivan, K., & Sullivan, A. Adolescent-parent separation. Develop-
mental Psychology, 1980, 16, 93-99.

 

Worby, C. M. The family life cycle. Journal of Medical Education.
1971, 46, 198-203.

 

Wynne, L. C. The injection and the concealment of meaning in the family
relationships and psychotherapy of schizophrenia. In D.
Rubinstein and Y. Alenen (Eds.), Psychotherapy of schizophrenia.
Amsterdam: Exerpta Medica, 1972.

Wynne, L. C. & Singer, M. T. Thought disorder and family relations
of schizophrenics: III and IV. American Medical Association
Archives of General Psychiatry, 1965, 13, 187-212.

 

APPENDICES

APPENDIX A

STUDENT-PARENT LETTER

MI CH IGAN STATE UNIVERS ITY

 

Department of Psychology East Lansing ° Michigan -48824
Dear College-Bound Senior and Mother:

We would like your help! We are psychologists at Michigan State
University who want to learn about what it feels like to be leaving home
--both for the young woman who is going away to college and for the
mother whose daughter is leaving for school. We recognize this transition
is an important event in each of your lives. Because of our desire to
learn more about this frequently occurring but little understood transi-
tion v“: are writing and asking for your assistance.

Families of female high school seniors throughout this area are
being contacted via a letter distributed to all female students. We are
looking for interested daughter-mother pairs in which the daughter plans
to live at a college in the fall of 1979. Though the High
School Administration does not formally endorse this research project,
they have been most helpful to us and we would like to take this oppor-
tunity to acknowledge publicly their aid.

 

What would your participation involve? One of us would meet
separately with each of you during the summer for about 1 l/2 hours. We
would likeymnzto complete three simple questionnaires. Afterwards, we
would like to talk to you briefly to learn more about what you think and
feel about this experience. We realize that your are busy so individual
appointments willlx:made that best fit your schedule.

Your participation is totally voluntary. Only be returning the
enclosed postcard will we know your name and address. We hope, of course,
that you will choose to join us in exploring this transition. We also
hope that you will find your experience both pleasant and enlightening.

If you desire, we will share the outcome of this project with you.

All of your individual responses to the questionnaires and
interview will be kept strictly confidential. Whatever information you
provide is for research purposes only and will not be shared with your
school or other family members. Though mothers and daughters may want to
discuss their thoughts with each other after participating, we will not
be able to let one see the other's responses.

If you are interested, or would just like more information, we
hope you will return the enclosed postcard. That will not mean a
commitment. It simply provides us with the opportunity to talk with you
by telephone to give you further information and to answer all your
questions. Of,iJEyou prefer to talk before you return the postcard, feel
free to call Devorah Smith at 337-9330 (evenings) for more information.

83

84

While we may not be able to use all who volunteer, as many
interested families as possible will be contacted and offered the op-
portunity to participate. We appreciate your taking the time to read this
letter.

Sincerely,

John R. Hurley, Ph. D. Devorah Smith
Professor of Psychology Psychology Graduate Student

APPENDIX B

STRUCTURAL ANALYSIS OF SOCIAL BEHAVIOR

Instructions

The idea of this questionnaire is to look at your perceptions of
yourself and of your interactions with another who is important to you.
A study of the responses of a large group of individuals to this ques-
tionnaire could contribute further to the understanding of this transi-
tion in one's life. All of your responses will, of course, be held in the
strictest confidence. Please respond as candidly as you can because the
questionnaire is not at all trickly. The impression you create will
correspond closely to the responses you choose. So if you try to create
a "rosy picture," that's what will happen. On the other hand, responding
honestly will permit you to candidly explore yourself and this relation-
ship. DON'T WORRY ABOUT WHETHER YOUR RESPONSES SEEM CONTRADICTORY NOR
WHETHER THEY ARE "ACCURATE, REALISTIC AND FAIR." The idea is simply to
measure your perceptions and ng_gge_else (besides this researcher) is
going to see it and no one will argue its truthfulness.

Try not to fuss a long time over your answers. Simply write down
what impulsively first comes into your head. I appreciate your help and
hope you find this experience, as some others have, interesting and
perhaps even thought provoking.

Also, to help me read all your responses correctly, a chart of
your typical way of writing each number would be useful. So just write
each digit as you usually do in the space provided.

 

 

 

 

l 2 3 4 5 6 7 8 9 0
Below are the definitions of a few words that others have found
difficult to understand. Feel free to refer to these if you are also
unclear as to their meanings.
Annihilate: to destroy completely
Ecstatic: intensely delightful, enraptured
Stimulate: to rouse to activity or heightened action

Vehemently: forcefully, vigorously, intensely

Writhe: to twist or contort, as in pain

85

 

Series C*

86

c 1978, Lorna Smith Benjamin

Please write a number in the blank indicating how well each of the fol-
lowing phrases describes your feelings about yourself.

NEVER ALWAYS
NOT AT ALL PERFECTLY
O 10 20 3O 4O 50 6O 7O 80 90 100

 

A rating of 50 or above indicates "true"; a rating of less than 50 indi-

cates

10.
11.
12.

13.
14.
15.

16.
17.

18.
19.

20.
21.
22.

23.

24.
25.

26.
27.
28.

 

"false."

I neglect myself, don't try to develop my own potential skills,
ways of being.

I examine, analyze myself sensibly, carefully, realistically.

I let myself daydream and fantasize instead of actually doing
what would be good for me.

I let important choices, thoughts, issues, options slip by me
unattended.

Knowing both my faults and my strong points, I comfortably ac-
cept myself as I am.

I am pleased with, glad about myself.

I tell myself things to make me feel bad, guilty, ashamed,
unworthy.

I practice, work on developing worthwhile skills, ways of being.
I love, cherish, adore myself. ~

I nurture, care for, restore, heal myself as needed.

I vehemently reject, dismiss myself as worthless.

I let unwarranted, illogical ideas I have about myself go
unexamined and unchallenged.

I entertain myself, enjoy being with myself.

I am very careful to restrain myself, to hold back.

I control, manage myself according to my carefully thought out
goals for myself.

I torture, kill, annihilate myself just because "I'm me."

I deprive, deplete myself, make myself sacrifice for others even
if it means harming myself greatly.

I stroke myself, pat myself on the back for "just being me."

I keep an eye on myself to be sure I'm doing what I think I
should be doing.

I try very hard to make myself be as ideal as I can.

I listen to and follow what I find deep within myself.

I don't care if I harm myself by ignoring my own sickness or
injury.

I put a lot of energy into making sure I conform to standards,
am proper.

I vengefully, viciously punish myself, "take it out on myself."
I "sell out," make myself do and be things which I know are not
right for me.

I am happy-go-lucky, content with "here today, gone tomorrow."
I protect myself, take constructive steps on my own behalf.

I drift with the moment, have no particular internal directions,
standards.

87

29. I put a lot of energy into getting myself absolutely everything
I need or want.
30. By just letting myself do what flows naturally and easily I do

everything well enough to suit myself.

31. I feel solid, integrated, "together," acceptant of my inner core.

32. I am comfortable letting my basic nature unfold as it will.

33. I am reckless, carelessly end up in self-destructive situations.

34. I seek, try to find situations which will be very pleasant and
good for me.

35. I tell myself to be unsure, that I am inadequate because others
are better than me.

36. I approach myself with a negative, destructive attitude; I am

my own worst enemy.

88
Series A* c 1978, Lorna Smith Benjamin
Please place a number the blank indicating how well thephrasesdescribe

the behavior of your (mother/daughter) in relation to you. Use the
following scale:

 

NEVER ALWAYS
NOT AT ALL PERFECTLY

O 10 20 3O 40 50 6O 70 80 90 100

 

A rating of 50 or above indicates "true"; a rating of less than 50
indicates "false."

1. Constructively, sensibly, persuasively analyzes situations
involving me.
2 Has her own identity, internal standards.
3 Enforces conformity to norms she prefers, insists I be "proper."
4 Puts me down, tells me that I do things all wrong, acts superior.
5 Looks to me as an advisor because she feels she can learn from
what I suggest.
6. Complies with my wishes without much feeling of her own, is
apathetic.
7. Angrily rejects, dismisses, tells me to get the "H" out.
8. Comfortably accepts help, caretaking when I offer it.
9. Defies, does the opposite of what she thinks I want her to do.
10. Lets me know her views so I can give them due consideration.
ll. Enthusiastically shows, shares herself or "thing" with me.
12. Murderously attacks, annihilates me.
13. Picks up on what I say/do in an irrelevant or only distantly
related way; goes on her "own trip" with it.
14. Reacts to my sexual touch with ecstatic joyful love.
15. Invites me to be with her, to be in touch as often as I can.
16. Lets me know where she is so I can maintain friendly contact with
her if I want to.
17. Freely comes and goes without special regard for what I might
have to say about it.
18. Tenderly, lovingly touches me sexually if I seem receptive.
l9. Constructively stimulates me, shows me how to understand, do.
20. Accuses, blames me, tries to get me to admit I am wrong.
21. Gladly, enthusiastically, warmly welcomes me.
22. Depends upon me to take care of everything for her.
23. Punishes me, takes revenge.
24. Shows understanding of my view, has empathy for me.
25. Asks trustingly, vulnerably; counts on me to respond to her with
kindness and consideration.
26. Willingly accepts, yields to my reasonable suggestions, ideas.
27. Desperately writhes in agony as she protests that I am destroying,
killing her.
28. Gives "strokes," soothes, calms me.
29. Intrudes on me, blocks, restricts me.
30. Even if she feels suspicious and uneasy, she gives in to my ar-

guments, ideas.

89

Series A continued

NEVER ALWAYS
NOT AT ALL PERFECTLY

O 10 20 30 4O 50 60 7O 80 90 100

 

A rating of 50 above indicates "true"; a rating of less than 50
indicates "false."

31. Follows my preferred rules, standards, routines.
32. Rips me off, drains me, takes my vital supplies.
33. Is so eager that I be pleased with her that she defers, checks

with me on every little thing.

34. Is terrified, extremely wary, very fearful of me.

35. Deludes, deceives, diverts, misleads me.

36. Resentfully allows my needs and wants to prevail over hers at
her own enormous expense.

37. Provides for, nurtures, takes care of me.

38. Carefully considers my side of things, treats me fairly.

39. Ignores me, acts on her own as if I were not there.

40. Uncaringly lets me go, do what I want.

41. Vehemently refuses my caretaking, my offers to assist.

42. Tries as hard as she can to escape, to flee from me.

43. Benevolently checks on me and reminds me of what I should do.

44. "Gives me her blessing" and leaves me to develop my own identity
separate from her.

45. Forgets me, fails to remember and keep agreements or plans made
with me.

46. Does things the way I want but sulks quietly with resentment and
anger.

47. Yields, submits, gives in to me.

48. Approaches me very menacingly, gathers materials she can use to
hurt me.

49. Manages, controls, oversees every aspect of my existence.

50. Tells me that she thinks I am competent to do things on my own.

51. Expresses her thoughts and feelings in a clear and friendly
manner so I have every opportunity to understand her well.

52. Feels, becomes what she thinks I want.

53. Starves me, fails to give me my "due," cuts me out.

54. Tries to truly understand me; actively listens in a non-
judgmental and friendly way.

55. Detaches from me, doesn't ask for anything, weeps alone about me.

56. Tries to anticipate my every need so I don't need to do anything
for myself.

57. Hides her resentment and anger and scurries to avoid my
disapproval.

58. Asserts, holds her own without needing external support.

59. Walls herself off from me, doesn't hear, doesn't react.

60. Confirms, tells me she likes and appreciates me just as I am.

61. Avoids me by being busy and alone with her "own thing."

62. Relaxes, enjoys, flexibly flows, feels good about being with me.

63.

64.

65.

66.

67.

68.

69.

70.

71.
72.

90

For my own good, she specifies, tells me what is best for me to
do, be, think.

Whines, squirms, painfully tries to account for, defend and
justify herself.

Regardless of what I say or do, she treats me according to her
own unwarranted and illogical assumptions about me.

Goes her own separate way.

Looks after my interests, takes steps to protect me, actively
backs me up.

Freely and openly discloses her innermost self when I am listening.
Expects to have wonderful fun with me and so approaches me
joyfully.

Just when she is needed most, she abandons me, leaves me "in the
lurch."

Neglects me, doesn't attend to my interests, needs.

Tells me I am on my own; I can do and be whatever I want.

Series B

91

c 1978, Lorna Smith Benjamin

Please place a number in the blank indicating how well the phrases de-
scribe you in relation to your (mother/daughter). Use the following
scale:

 

NEVER ALWAYS
NOT AT ALL PERFECTLY

O 10 20 30 40 50 60 70 80 80 100

 

A rating of 50 or above indicates "true"; a rating of less than 50
indicates "false."

1.

DOOM

(Dd m Ul

g...-
OKD

11.

H
N
O

13.

F4
Jr.

15.

|-‘
O"

17.

18.
19.
20.
21.
22.
23.
24.

N
U"!
0

26.

N
\l
O

28.
29.

30.

I constructively, sensibly, persuasively analyze situations
involving her.
I have my own identity, internal standards.

I enforce conformity to the norms I prefer, insist she be "proper."

I put her down, tell her that she does things all wrong, I act
superior.

I look to her as an advisor because I feel I can learn from what
she suggests.

I comply with her wishes without much feeling of my own, am
apathetic.

angrily reject, dismiss, tell her to get the "H" out.
comfortably accept help, caretaking when she offers it.

defy, do the opposite of what I think she wants me to do.

let her know my views so she can give them due consideration.
enthusiastically Show, share myself or "thing" with her.
murderously attack, annihilate her.

pick up on what she says or does in an irrelevant or only
distantly related way; I go on my "own trip" with it.

I react to her sexual touch with ecstatic joyful love.

I invite her to be with me, to be in touch as often as she can.
I let her know where I am so she can maintain friendly contact
with me if she wants to.

I freely come and go without special regard for what she might
have to say about it.

tenderly, lovingly touch her sexually if she seems receptive.
constructively stimulate her, show her how to understand, do.
accuse, blame her, try to get her to admit she is wrong.
gladly, enthusiastically, warmly welcome her.

depend upon her to take care of everything for me.

punish her, take revenge.

show understanding of her view, have empathy for her.

ask trustingly, vulnerably; I count on her to respond to me
with kindness and consideration.

I willingly accept, yield to her reasonable suggestions, ideas.

HHHHHHH

HHHHHHHH

I desperately writhe in agony as I protest that she is destroying,

killing me.
I give "strokes," soothe, calm her.
I intrude on her, block, restrict her.

Even if I feel suspicious and uneasy, I give in to her arguments,

ideas.

92

Series B - Continued

NEVER ALWAYS
NOT AT ALL PERFECTLY

O 10 20 30 40 50 60 70 80 90 100

 

A rating of 50 or above indicates "true"; a rating of less than 50
indicates "false."

31. I follow her preferred rules, standards, routines.
32. I rip her off, drain her, take her vital supplies.
33. I am so eager that she be pleased with me that I check with her

on every little thing.

34. I am terrified, extremely wary, very fearful of her.

35. I delude, deceive, divert, mislead her.

36. I resentfully allow her needs and wants to prevail over mine at
my own enormous expense.

37. I provide for, nurture, take care of her.

38. I carefully consider her side of things, treat her fairly.

39. I ignore her, act on my own as if she were nOt there.

40. I uncaringly let her go, do what she wants.

41. I vehemently refuse her caretaking, her offers to assist.

42. I try as hard as I can to escape, to flee from her.

43. I benevolently check on her and remind her of what she should do.

44. I "give her my blessing" and leave her to develop her own
identity separate from me.

45. I forget her, fail to remember and keep agreements or plans made
with her.

46. I do things the way she wants but sulk quietly with resentment
and anger.

47. I yield, submit, give in to her.

48. I approach her very menacingly, gather materials I can use to
hurt her.

49. I manage, control, oversee every aspect of her existence.

50. I tell her that I think she is competent to do things on her own.
51. I express my thoughts and feelings in a clear and friendly man-
ner so she has every opportunity to understand me well.

52. I feel, become what I think she wants.

53. I starve her, fail to give her her due, cut her out.

54. I try to truly understand her; I actively listen in a non-
judgmental and friendly way.

55. I detach from her, don't ask for anything, weep alone about her.

56. I try to anticipate her every need so she doesn't need to do
anything for herself.

57. I hide my resentment and anger and scurry to avoid her
disapproval.

58. I assert, hold my own without needing external support.

59. I wall myself off from her, don't hear, don't react.

60. I confirm, tell her I like and appreciate her just as she is.

61. I avoid her by being busy and alone with my "own thing."

62. I relax, enjoy, flexibly flow, feel good about being with her.

63.

64.

65.

66.
67.

68.

69.

70.

71.
72.

lllllillll

93

For her own good, I specify, tell her what is best for her to
do, be, think.

I whine, squirm, painfully try to account for, defend and
justify myself.

Regardless of what she says or does, I treat her according to my
own unwarranted and illogical assumptions about her.

I go my own separate way.

I look after her interests, take steps to protect her, actively
back her.

I freely and openly disclose my innermost self when she is
listening.

I expect to have wonderful fun with her and so I approach
joyfully.

Just when I'm needed most I abandon her, leave her "in the
lurch."

I neglect her, don't attend to her interests, needs.

I tell her she is on her own; she can do and be whatever she
wants.

APPENDIX C

ENGAGEMENT STYLE STORIES

(NOTE: THOSE INSTRUCTIONS THAT WERE INCLUDED ONLY IN THE MOTHERS'
TASK ARE CONTAINED IN BRACKETS.)

This task taps your imagination. Each of the following short
descriptions sets a scene. Try to make up a short story about what you
think might happen in each of these situations. Concentrate on develop-
ing your story and the thoughts and feelings of your character(s) and
don't worry about your grammar and spelling. If you run out of space
feel free to continue your story on the back. This is an opportunity to
be creative and carefree, so relax. Whatever stories come to your mind
will be fine.

[In creating these stories try to pretend that ygg are now a
teenager! Call on your 933 feelings, perceptions, experiences and £95_
those of your daughter. Let your personal imagination guide you. You
need not be limited by the actual experiences of your children.]

 

 

Story #1

Think of this situation. Someone who has just recently finished
high school is living away from home, away from her family.

Write a story about this situation in the space below. Tell
what has happened before, what is happening 293, and what will happen in
the future. Say what the characters are thinking and feeling. Make your
story at least six_sentences long.

 

 

 

 

 

 

 

 

 

Story #2

Next, think of this situation. Someone who has been living away
from home for a while is now back for her first visit home. [Remember,
pretend that ygu are now a teen.]

Write a story about this situation in the space below. Tell what
has happened before, what is happening 22!! and what will happen in the
future. Say what the characters are thinking and feeling. Make your
story at least si§_sentences long.

 

 

 

 

94

95
Story #3

Now, consider this situation. This young woman has recently
finished high school and will be living at home this coming year despite
her desire to live on a college campus. [Try to call on your own feelings
and not those of your children in creating this story.]

Write about this situation in the space below. Tell what has
happened before, what is happening 22!, and what will happen in the
future. Say what the characters are thinking and feeling. Make your
story at least six_sentences long.

 

 

 

 

 

 

 

 

 

Story #4

Now think of this situation. Daughter and mother have just ar-
rived on a college campus together. The first term is about to start and
they are looking for a certain building. [Try to pretend that you are
now a teenager.)

Write a story about this situation in the space below. Tell what
has happened before, what is happening £93, and what will happen in the
future. Say what the characters are thinking and feeling. Make your
story at least six_sentences long.

 

 

 

 

 

 

 

 

 

 

Story #5

Next, imagine this situation. After three months of being away
at college, a daughter receives an unexpected telephone call from her
parent. [Think back to when you were a teen in making up this story.]

Write a story about this situation in the space below. Tell
what has happened before, what is happening new, and what will happen in
the future. Say what the characters are thinking and feeling. Make
your story at least six_sentences long.

 

 

 

 

 

 

 

 

 

 

 

"I111111171111'1111