Page 1
/79
/No, e6O7
PREDICTION OF VERBAL DOMINANCE BEHAVIORS
USING CONSTRUCTIVIST THEORY
THESIS
Presented to the Graduate Council of the
University of North Texas in Partial
Fulfillment of the Requirements
For the Degree of
MASTER OF SCIENCE
By
Caroline Curlin, B.A.
Denton, Texas
May, 1990
Page 2
Curlin, Caroline. Prediction of Verbal Dominance
Behaviors Using Constructivist Theory. Master of Science
(Clinical Psychology), May, 1990, 94 pp., 20 tables, 44
references.
This study assessed how Constructivist theory accounts
for verbal dominance. Conversations of rotating dyads were
tape recorded, then coded for measures of dominance.
Subjects completed a trait dominance scale and a
constructivist personality test. Interpersonal rankings of
dominance were found to be more consistent with observed
behavior than trait dominance scores. Extreme trait
dominance scores were associated with a constructivist
measure indicating maladjustment. Dyads identified as more
resistant to change were found to use fewer verbal control
strategies; male/male dyads were characterized by direct,
functional interactions. Dyads that were highly comfortable
with one another utilized fewer verbal control methods.
Lastly, interactions in which participants reported
unfamiliar self-experiencing utilized higher levels of
verbal control. Implications for group processing,
assessment of dominance and sex differences are discussed.
Page 3
TABLE OF CONTENTS
LIST OF TABLES . .
CHAPTER
I. INTRODUCTION
II. METHOD
SubjectsProcedureMeasures of resistanceDependent measures
III. RESULTS . . . . . . . . . .
IV. DISCUSSION. . . . . . . . .
APPENDIX. ........ .....
REFERENCES... .... ...... .
... 31
to change
38
51
. . 60
. .83
iii
.v
Page 4
LIST OF TABLES
Table Page
1. Interrater Reliabilities of ObservationalMeasures . . . . . .. . . . . . . . . . . . . . 36
2. Means and Standard Deviations of Independentand Dependent Variables: Monadic . . . . . . 38
3. Means and Standard Deviations for DependentVariables: Dyadic . . . . . . . . . . . . . . 39
4. Correlation Matrix of Verbal Measures andDominance Scale . . . . . . . . . . . . . . . 40
5. Correlations Between Peer Ranking of Dominanceand Verbal Measures. . . . . . . . . . . . . 42
6. Regression Results for Deviation of Dominanceand FIC . . . . . . . . . . . . . . . . . . . . 43
7. Regression Results for Total Talk Rate . . . . 78
8. Regression Results for Differential Talk Rate . 79
9. Regression Results of Total Interruption Rate . 79
10. Regression Results of Differential InterruptionRate . . . . . . . . . . . . . . . . . . . . . 79
11. Regression Results of Total Question Rate . . . 80
12. Regression Results of Differential QuestionRate . . . . . . . . . . . . . . . . . . . . . 80
13. Regression Results for Total Ah Rate . . . . . 81
14. Regression Results for Differential Ah Rate . . 82
15. Regression Results for Total Silence Rate . . . 82
16. Regression Results for Differential SilenceRate . . . . . . . . . . . . . . . . . . . . . 82
17. Correlations Between Resistance to Change,Sex, and Predictor Variables . . . . . . . . . 46
iv
Page 5
LIST OF TABLES--CONTINUED
Table Page
18. Correlations Between Resistance to Change,Dominance Scores and Proportion of IndependentSelves Brought to Interactions. ., . . . . . . 47
19. Regression Results for Total Comfort . . . . . 49
20. Correlations Between Comfort andPredictor Variables ........................50
V
Page 6
CHAPTER I
INTRODUCTION
When two people begin to develop a relationship,
apparently insignificant statements and behaviors have great
consequence for future structure and development of that
dyad. Each question, suggestion or gesture is a message
regarding the nature of the relationship that will exist.
Individuals who believe that their relationships evolve
without influences from one or both partners are unaware of
the process by which preferences and expectations are subtly
expressed in the initial stages. They may be aware -that the
relationship is platonic, not sexual, that practical jokes
do not occur or that one member usually decides where they
will eat dinner. However, it is not uncommon for
participants to be unable to describe the behavioral and
verbal communications by which such "rules" were
established.
Each time two people are engaged in an interpersonal
situation, communications are exchanged, both verbally and
nonverbally. Some theorists posit that it is impossible not
to communicate in the presence of others; even silence
expresses some meaning (Haley, 1972; Watzlawick, Beavin, &
Jackson, 1967). Danzinger (1976). argued that "a verbal
Page 7
2
message is never merely a neutral transmission of
information about the world outside, it is always also a
communication about the relationship between the speaker and
his audience" (p. 26). Researchers have regarded such
behavioral and verbal messages as requests (Haley, 1972),
attempts to impose behavioral change (Reusch & Bateson,
1951) or inquiries into the anticipated state of the
relationship (Kelly, 1955). Communications serve to
monitor, suggest, reply and delineate what future
interactions will entail. As each communication is reponded
to, either positively or negatively, the relationship is
further defined and structured. For example, if one member
talks animatedly about a recent baseball game, the other
person can either join in the conversation, gesturing
frequently, or can yawn and have nothing to say. In any
event, the message has been responded to, either directly or
indirectly, and the relationship further defined.
Several researchers have posited ways to conceptualize
this process. Foa and Foa (1980) developed a system to
classify what types of resources are exchanged, defined
along dimensions of concreteness and particularism (whether
any person can fulfill the need or if a particular person is
required). Reusch & Bateson (1951) emphasized the
interpretation or "codification" of messages and events that
take place in a relationship; such codification defines the
role and status of participants. Haley, 1972,
MUM
Page 8
3
conceptualized this process as answering two questions; 1)
what kinds of behaviors will take place and 2) who decides
what takes place. The first question refers to the content
of the behaviors existing in the relationship, the latter to
the structure, or the relative influenceability of each
member in determining what occurs between them. This does
not refer to control of the other person's behavior, but
rather the attempt to control the relationship's definition.
Interpersonal communication has also been divided into
similar poles. The content of messages has been refered to
as semantic (Danzinger, 1976) or denotative (Keisler &
Bernstein, 1974), that is, the actual verbal content used to
communicate information to the other participant. All that
is necessary to understand messages at this level is a
command of the language. Messages regarding the structure
of the dyad have been referred to as pragmatic
communications, "hidden agendas" regarding the way in which
the members interact with one another (Danzinger, 1976).
They are also known as connotative messages conveying
information regarding the attitudes or affect of one of the
participants (Keisler & Bernstein, 1974).
Although the denotative words spoken may indicate a
straightforward statement, connotative meanings may indicate
a quite different intent. Sarcasm is usually expressed by
tone of voice or gestures, and if interpreted by the
denotative content alone, would be greatly misunderstood.
Page 9
4
At this level, more than linguistic knowledge is necessary
for interpretation, for social norms and idiosyncratic
gestures and terms are utilized. Connotative messages
constitute much of the process of establishing the future
structure of relationships. By interpreting words merely at
the denotative level, as common in behavioral theory, one
may disregard important statements essential to resistance
and transference in therapy (Keisler & Bernstein, 1974), as
well as the everyday process of structuring relationships.
The focus of this study is upon this process of
connotatively communicating the differential influence of
participants, or the structure, of dyadic relationships.
Relationships can be structured in two ways according
to Haley (1972): complementary or symmetrical. A
complementary relationship is one in which both persons are
exchanging different types of behavior; this is typical in a
professor/student or physician/patient relationship. One
member has more influence in determining what shall occur
than the other. Symmetrical relationships are characterized
by an exchange of the same type of behavior, as is common in
friendships or peer relationships. Both members have
similar amounts of influence, and status differences between
the two are minimized. Watzlawick, Beavin & Jackson (1967)
posited that "all communicational interchanges are either
symmetrical or complementary, depending upon whether they
are based on equality or differences" (p. 70).
IMF-,AWA I , I I 11,111, m lim PrA-P .. .. -- -- -
Page 10
5
Entering a relationship with societally defined roles
lessens the need for establishing the structure of the
relationship. When one visits a physician, one expects that
the doctor will make the proper inquiries, label the
disorder and prescribe medicine, and, that the patient will
obediently conform to those suggestions. Generally, the
physician controls what occurs. Most individuals are aware
of societal expectations and so conform to physicians'
suggestions. It is unusual for a patient to diagnose the
illness and to suggest treatment to the physician, although
this may occur. In such a case of misunderstood roles, the
physician is likely to re-establish the fact that he or she
has attended medical school and is licensed to prescribe
medicine, not the patient. In this manner conflicting
expectations about appropriate behavior are resolved. When
individuals enter relationships where role expectations are
not prescribed, they must negotiate to define the structure
of influenceability. This process of arranging influence in
socially unstructured relationships is the focus of this
paper.
According to Haley (1972) the process of defining the
structure or level of influenceability of dyadic members is
mutual; "two people inevitably work out together what kind
of relationship they have by mutually indicating what kind
of behavior is to take place between them" (p. 10). The
process of equal participation in determining differential
Page 11
6
levels of influence may appear to be conceptually
incompatible. This process is one in which participants
mutually determine their relative influenceability. It
would seem logical that the more influential person
communicate their status to the other, whether acceptable or
not. Haley (1972) argued that it is impossible for one
person to define the relationship exclusively; for one
person to dominate, the other person must agree. It is just
as theoretically inconceivable for a passive person to have
no influence in determining their relative position. Haley
states (1972) "whenever a person tries to avoid controlling
the definition of a relationship he must at a more general
level be controlling what type of relationship this is to
be--one where he is not in control" (p. 10). Similarly,
Leary (1957) posited an interpersonal theory rejecting any
possibility of responding in a vacuum, for the behavior of
one person simultaneously affects the behavior of the other.
He classified behavior along two dimensions:
dominance/submission and love/hate. Behavior in one
quadrant "pulls" for or draws complementary behavior in the
opposite quadrant. In this manner, passive behavior both
verbally and nonverbally places expectations for dominant
behavior upon others.
The process by which both members mutually negotiate
their relative levels of dominance occurs via "maneuvers,"
communications that put the current status of the4
- pill-No, I i Mil
Page 12
7
relationship into question (Haley, 1972). Stable
relationships are those which have reached a state of mutual
agreement; thus, relatively few maneuvers are executed.
Unstable relationships occur when one or both persons wish
to change the existing system and behave in a manner so as
to suggest, question, or demand new roles. Each maneuver
must be responded to, either negatively, in which case
instability continues, or positively, leading to a stable
relationship once more. Typically beginning relationships
are characterized by instability, but relationships of any
length can exist in this unstable state.
This process of interpersonally maneuvering to define
the nature of interactions exists among participants in
larger social units as well as in dyads. Yalom (1970)
described several phases that groups must undergo to reach a
mature, profitably working state. The first stage is the
cocktail stage wherein group members as strangers attempt to
become familiar with one another, establish roles for
themselves and search for meaning in therapy. Groups at
this stage are typified by stereotyped, innocuous, social
chit-chat oriented to establish a sense of belonging and
understanding. The second stage consists of conflict and
competition for dominance. By expressing hostility toward
the leader and other members, each person strives to
establish a certain amount of influence in the group. Until
a mutually agreeable, stable dominance hierarchy is
Page 13
8
established, the group cannot progress to the third stage,
characterized by cohesiveness and intimacy, where most of
the valuable therapeutic work is accomplished. Just as
members must struggle, compete and negotiate the hierarchy
of influence in groups, they must also do so in dyads.
Similarly, Duck (1972) has characterized this struggle
for dominance in dyads as evolving from a superficial to a
more psychological basis. He reported that during the
formation of dyads, the salient factor in predicting
friendship development was similarity of concrete
constructs, those regarding physical attractiveness or other
superficial characteristics. As the relationship
progressed, the similarity of psychological constructs
influenced its evolution. As dyad members reach a cohesive
state, negotiation of more personally meaningful constructs
becomes central.
In addition to using the methodology of self-report
measures, researchers have attempted to determine behavioral
indicators of the process of maneuvering for influence.
Several fairly established behavioral measures of the level
of interpersonal dominance have been reported in current
literature. While many types of communication concomitant,
unorderly, or unstructured, conversation is considered by
communication theorists as a unique subset of sequential
communication defined as an "orderly exchange of speaking
turns" (Argyle, 1975, cited in Roger & Schumacher, 1983).
". l.,. -. , - -t wo"POINOWN"m
Page 14
9
The control of the flow and sequence of conversation has
implications for the relative influence and value of each
participant's input. Mishler and Waxler (1968)
differentiated two techniques of controlling conversation:
attention control and person control. Attention control is
a technique used to influence the course of interaction
while maintaining attentional focus on the speaker him- or
herself. One can influence the sequence and flow of
conversation while keeping the floor by directing statements
to one particular participant, participating frequently, or
by speaking in long, uninterrupted sequences. The second
technique, person control, an attempt to manipulate the
conversation by controlling the other person's behavior.
This is accomplished either by interrupting the person
holding the floor or by questioning. Mishler and Waxle
(1968) categorized person control strategies into direct and
indirect attempts at control: direct attempts control by
"explicit confrontation of another person in an attempt to
stop the speaker before he has finished his idea, to prevent
his being heard, or, more indirectly, to control the
direction and content of his interaction [by questioning]"
(p. 140); indirect attempts to attention control might
consist of a child's throwing a temper tantrum, or speaking
more loudly than others to draw attention. Mishler and
Waxler's hypothesized that more functional behavior is
Page 15
characterized by direct rather than indirect attempts at
influence, whether by person or attention strategies.
In dyads in which a dominance hierarchy has already
been established according to cultural or occupational
status, role differentiation of dominance behaviors is
readily seen. Wiens, Thompson, Matarazzo, Matarazzo, and
Saslow (1965) found that the interactional pattern described
above vary with status; those nurses in higher status
interrupted other nurses more, spoke in longer segments and
defended their positions more with physicians than lower
status nurses interacting with the same physicians.
Similarly, studies of interactional patterns of families
with schizophrenic children have assessed dominance by the
number of successful interruptions and the amount of
influential members' talking time. (Jacob, 1975; Riskin,
1972, cited in Ferguson, 1977). Mishler and Waxler (1968)
reported that in comparison to families with schizophrenic
children, "normal" families evidenced more role
differentiation in measures of verbal dominance between
parents and children.
Measures of verbal dominance behaviors have also been
used to determine the types of structure established in
relationships in which roles have not been culturally
defined. Nicola Ferguson (1977) transcribed seven-and-a-
half hours of spontaneous conversation between one
undergraduate and fifteen of her friends. Ferguson
Page 16
11
correlated several measures of verbal interruptions with two
measures of dominance within each of the fifteen dyads. The
two measures of dominance included a self-rating personality
questionnaire and the main subject's rating of her friends'
level of dominance. Ferguson concluded that only two
measures of interruptions (overlaps [unsuccessful
interruptions in which the first speaker finishes her
utterance], and silent interruptions [switches of speakers
during silence]) correlated significantly with either
measure of dominance.
Rogers and Jones (1975) studied 18 dyads composed of
one high dominant person and one low dominant person as
measured by Cattell's 16PF and Gough's self-descriptive
adjective checklist. Subjects were given a cooperative
problem-solving task to complete. Results indicated that
high dominant individuals talked more and attempted more
interruptions compared to their low dominant partners. In
addition, high dominant subjects appeared to have more
success at interrupting their partners, although this
comparison fell short of statistical significance. In a
similar study, Roger and Schumacher (1983) measured the
number of successful and unsuccessful interruptions by 18
male and 18 female dyad members when discussing issues on
which they were known to disagree. Subjects were grouped
into three dyads based on their scores on the Dominance
Scale of the Edwards Personal Preference Schedule; thus
Page 17
12
high-high, high-low and low-low dominant dyads were formed.
The researchers found that the rate of successful
interruptions increased over time for high-high dyads and
that significantly greater opinion change occurred in the
low-low dyads than the high-high dyads. Unsuccessful
interruption rate was reported to be an un reliable measure.
Aries, Gold and Weigel (1983) studied a trait scale of
dominance's ability to predict verbal and behavioral
measures of dominance in small groups of either the same or
mixed sex. Results indicated that in same sex groups, while
the personality measure accounted for an average of 10% of
the variance in single behavioral measures, 40% was
accounted for using all of the measures. However, in mixed
sex groups, personality-behavior correlations were
significantly reduced.
Fruge (1973) studied 32 males interacting in rotating
dyads to examine differences in self-report measures,
behavioral measures and peer rankings of dominance. His
attempt to predict overall dominance behaviors using
personality measures was unsuccessful; pretest measures of
influence were no better than chance in predicting future
behavior. However, dyadic analysis of data produced much
more discriminative data than monadic analysis. Also, the
subjects' use of indirect control strategies such as
questioning was accompanied by the subjects' mutually low
opinions of influenceability as well as low impressions of
Page 18
13
comfort and enjoyment. In particular, dyads in which one
individual exercised a proportionally larger amount of
indirect person control (questioning) reported the most
dissatisfaction. Fruge argued that questioning was an
attempt to influence the flow of conversation without
accepting responsibility. This corroborates Mishler and
Waxler's (1968) hypothesis that direct attempts to influence
conversation are more functional. As Fruge stated, dyads
who accept responsibility for directly influencing change
produced "well coordinated, spontaneous exchanges" (p. 37).
In mixed sex dyads, men have been observed to be much
more likely to interrupt than women. Zimmerman and West
(1975) reported that in mixed sex dyads, men contributed
100% of the overlaps (attempted interruptions) and 96% of
the interruptions. It appears that males are much more
likely to exert direct person control when interacting with
females than with males. Both Rogers and Jones (1975) and
Roger and Schumacher (1983) studied same sex dyads. Neither
found significant sex differences in interruption attempts
or successes between male or female dyads; Rogers and Jones
found no difference in talking time. However, Roger &
Schumacher found that females gave significantly more back
channel utterances, brief interjections such as "uh-huh"
that convey agreement and encouragement to continue
speaking. It appears that males are much more likely to
iM1+'^ ' ' .,.. ., .. ., .. , . zr ,i+t + k fi1sF At k :r., r:...:. ... :::.:
Page 19
14
exert direct person control when interacting with females
than with males.
Current literature has predominantly used trait theory
to explain the occurrence of dominance behavior.
Psychologists and lay people alike have relied on trait
theory to explain and predict behavior for years (Allport,
1937). Psychologists' study of personality has been based
upon the existence of enduring traits in human behavior
(Wiggins, 1979, cited in Aries, 1983), dispositions to act
in a certain manner that are consistent, stable and evident
across time (Mangusson & Endler, 1977). Unfortunately, the
reliance on such consistencies has not been demonstrated to
be of empirical value. The highest consistent correlations
between personality measures and subsequent behavior has
been .30 (Mischel, 1968). Mischel (1968) has been an
outspoken proponent of the use of situational variables,
rather than personological variables to predict and explain
behavior. In Personality and Assessment, 1968, Mischel
argued that the concept of traits has been proved
"untenable," that behavior is highly situationally specific,
and that social learning theory could be a better approach
to personality. Epstein and O'Brien (1985) retorted that
while individual item measures of behavior have little
reliability regarding trait measures, aggregating
observations over many occasions will result in a more
accurate measure. Epstein and O'Brien (1985) state that "the
Page 20
15
major value of traits lie, however, not in their usefulness
in predicting specific behavior, but in their value as
predictors of aggregated behaviors, that is, of behavior in
the long haul averaged over many situations, occasions and
responses" (p. 532). Thus researchers have attempted to
measure a variety of behavioral traits across varying
temporal and spatial settings.
Mischel (1973) posited that instead of viewing
behavioral variablity as evidence of human inconsistency and
unreliability, it could be considered a discriminative
ability necessary for coping. Variable behavior results
from people's sensitivity to situational cues and their
ability to adapt functionally to fluctuating demands and
contingencies. In fact, behavior that is extremely "trait,"
that is, not varying in response to situational variables,
may be maladaptive. Mischel (1973) further states that
"indiscriminate responding (i.e., 'consistent' behavior
across situations) tends to be displayed more by
maladaptive, severely disturbed, or less mature persons than
by well functioning ones" (p. 258). Pathology can also
exist in the other extreme; persons who have few consistent
ways of behaving or who cannot meaningfully relate
situations to ways of behaving may act chaotically. This
will be described in further detail later.
Constructivist theory was used in this study to predict
dominance structure in dyads. Kelly (1955) first proposed
Page 21
16
his view of humans as scientists, hypothesizing meanings in
the world and then validating, adapting and
reconceptualizing these constructs. One metaphor employed
within constructivist theory is the view of humanity
existing within a "community of selves." Contrary to
Western culture which values individuality and individual
consciousness, humans may be understood by considering the
self as "an organised grouping of sub-selves or tendencies"
(Mair, 1977a, p. 137). Each self has differing goals,
values, potentials and motivations, thus each regulates
behavior differently (Cantor, Markus, Niedenthal, & Nurius,
1986). The self that is most influential in each particular
situation determines the attitudes and behavior of the
individual.
Some theorists have attempted to describe the same
constellation of selves for each person. Breckler and
Greenwald (1986) listed four motivational facets of the
self:, the Diffuse self interested in hedonistic
gratification, the Public self, the Private self, and the
Collective self. Situational conditions determine which
self engages; the Diffuse self may become active under the
influence of drugs, the Public self when powerful others are
present, the Private self when alone, and the Collective
self when reference group issues are important. However, it
is a tenet of this research that the community of selves
varies substantially from person to person, according to
Page 22
17
each one's idiographic constellation (Gergen, 1971; Cantor
et al., 1986). Mair (1977b) gave the example of Peter who
determined that among his characters of selves existed the
"Conversationalist" who loved interacting and relaxing with
others with no other goal than enjoyment. Peter's
"Businessman" self was very organized, constructive, and did
not like to waste time. Usually these two selves could
cooperatively alternate in dominance, but occasionally one
or the other would become greedy and "hog the limelight" (p.
135), causing Peter's behavior to become one-sided. Selves
also vary in the extent to which they are salient; some
selves may be applicable in only a few situations while
others are used in the majority of one's interactions.
Selves may represent a single period in one's past or future
(Cantor et. al., 1986). Individuals may also vary in how
they differentiate between selves. Just as one person can
differentiate between two sizes of objects, large or small,
another may distinguish fine gradations between several
sizes.
Scott (1988) proposed that individual variations in
constructing and organizing one's community of selves
influence propensity for constructs which admit new elements
into the system. Scott based some of his thoughts upon
Hinkle (1965), who determined that the range of implications
in changing a construct (the amount of connectedness) and
the superordinance, or personal importance of the construct
Page 23
18
to be changed, influenced how likely that construct was to
permit change. Kelly's (1955) original thesis stated that
the purpose of constructs was to predict and understand the
world around us. Awareness of imminent reduction in the
predictive utility of one's constructs (threat) or awareness
of an existing state of lack of predictive implications
within the system (anxiety) are noxious states, thus are
generally avoided. Logically, if construct change leads to
either threat or anxiety, such change is unlikely to occur.
Scott (1988) expanded and modulated Hinkle's hypothesis to
determine four circumstances that predict the likelihood for
selves to accommodate new behavior based upon this
principle.
Kelly's Modulation Corollary (1955) defines
permeability as a construct's capacity to assimilate new'
elements. Taking in new information may or may not alter a
construct (Kelly, 1955). Moreover, if two constructs
predict incompatible events, the constructs are considered
inferentially incompatible (Fragmentation Corollary). When
this occurs, the degree of incompatiblity of the event with
the pre-existing construct determines how much the construct
will have to change maintain the new element. If a
construct is applied in a rigid, inflexible manner, it will
more likely be threatened by incompatible external events.
Changes that reduce the predictive utility of a construct,
particularly those that permit little flexibility, produce
Page 24
19
threat or anxiety, so that further changes will be less
likely to occur.
In addition to the measure described above, Scott
(1988) also proposed that "the more superordinate a self is
and the greater the number of selves with which it is
functionally dependent, the greater implications of any
change" (p. 11). That is, for a superordinate construct to
change, it would have to influence many of its dependent
constructs. Similarly, the more connected (functionally
dependent) the construct, the greater the number of
dependent constructs must also change. Just as
inferentially incompatible constructs increase threat and
anxiety because they cannot serve to predict consequences,
superordinate and connected constructs resist change- as
well.
The last characteristic proposed is in relation to
propensity for change is predominance. It is hypothesized
that a self's range of convenience will more likely be
explored and delineated if that self is frequently utilized.
When an opportunity for change arises, predominant selves
are more likely to undertake the new role because their more
familiar ranges of convenience have already been extended to
include role changes. In summary, the four construct
measures hypothesized to predict propensity for change as
modified from Scott, 1988, are constriction,
Page 25
20
superordinateness, number of functionally dependent selves,
and predominance.
In a similar vein, Kelly (1955) proposed that to an
extent, use of propositional constructs, those that are
relatively loose and unrelated to other dimensions, is more
adaptive than using rigidly related constellatory
constructs. Bieri (1955, cited in Landfield & Barr, 1976)
first coined the term "cognitive complexity" to refer to the
use of many unrelated, independent constructs. A person
displaying cognitive complexity would have a choice of ways
to conceptualize information, thus making sense of a greater
variety of conflicting, confusing events. A person with
only few available uncomplicated constructs could merely
deal effectively with situations congruent with his or her
existing system. Anything outside that person's limited
range of convenience would be seen as complicated,
mysterious and too different to be understood. On the other
hand, too much complexity without a method for organizing
constructs' interaction would result in a very original and
creative way of seeing things, but in an inability to
effectively process the knowledge. Schizophrenics
demonstrate extreme dysfunction in this manner: they may
have sufficiently complex, varied ways of construing their
experience, but cannot adequately order, integrate or
connect their thoughts within the myriad of their
unorganized, separate constructs. These individuals lack
Page 26
21
what Kelly (1955) described in his Organization and
Fragmentation Corollaries as ordinal relationships and
hierarchical organization. Assigning superordinate and
subordinate status to constructs produces essential,
functional orderliness and organization within the system.
Thus an individual with adequately complicated and diverse
constructs may be unable to use the system practically
because his or her constructs are not interrelated. Another
individual may not have such a range of complexity, but
because of adequate hierarchical organization, is able to
effectively use his or her available constructs.
Landfield and Barr (1976) define hierarchical
organization; or ordination, as the ability to differentiate
and meaningfully use gradients within a construct; that is,
when evaluating acquaintances, a person has the ability to
finely distinguish high and low measures of the descriptor.
The ability to discriminate levels of meaningfulness within
a construct is one task necessary to structure, compare and
meaningfully order, that is, to hierarchically organize
constructs. An individual may have many complex constructs,
but to be unable to make discriminations within the system.
Landfield and Barr (1976) categorized subjects into
four quadrants based upon their cognitive complexity, or
number of Functionally Independent.Constructs (FIC), and
their ordination (0). Subjects classified into Quadrant 1
had low integration and low differentiation, thus would have
Page 27
22
few ways of understanding their worlds and would have
restricted levels and ranges of meaningfulness. These
subjects are thought to be easily understood by others due
to their simplicity, but they can only meaningfully construe
a limited range of information. Subjects classified into
Quadrant 2 have high integration and low differentiation.
They are thought to have limited ways of understanding
others but have more levels and more deeply explored ranges
of meaningfulness. These persons would be easily understood
by others due to their simple constructs, but would be
better able to understand others and could expand their
existing system because of their greater sensitivity to
using their existing constructs. Subjects in Quadrant 3
have low integration and high differentiation, and are
thought to have a variety of types of meaning but limited
levels and ranges of meaningfulness. Because of their
inability to integrate, they would be unable to understand
others or be understood themselves. Schizophrenics and
persons with thought disorders or personal confusion would
be placed in this quadrant. Quadrant 4 consists of subjects
with both high integration and high differentiation, and
would have the greatest construct complexity and the highest
level and range of meaningfulness so they could comprehend
others easily. However, due to their complexity, others may
find them less easy to understand.
Page 28
23
Landfield and Barr (1976) supported these hypotheses in
a study of four ten-person groups who interacted for several
sessions. The authors concluded that simplicity and
restriction of construction leads to impairment in
meaningfully understanding those more complicated or diverse
than oneself but concurrently produces little difficulty in
being understood by others. Similarly, those who were
complicated enough to understand others were difficult to
understand themselves. Of particular interest was the
finding that those in Quadrant 3 exhibited a "fragmenting
orientation to people" (p. 25) and even had difficultly
understanding themselves.
Researchers have used constructivist measures of
ordination and complexity of construction to explore and
explain group therapy processing. Landfield and Rivers
(1975) found a decrease in FIC and 0 scores at the end of
twenty-week group sessions. Barr (1977, cited in Landfield,
1978) also reported a gradual lessening construction
complexity as the sixteen-session group progressed. The
level of meaningfulness also decreased at the half-way mark
but increased somewhat by the end of group. Landfield
(1978) concluded that in the initial stages of group
therapy, members tend to assume they have a greater
understanding of other members than is accurate. As
continued interaction invalidates those assumptions,
constructs begin to have less meaning; a reconsolidation
Page 29
24
process may eventually occur leading to increased
meaningfulness of interactions with others.
Viewed in a context of the first of several group
meetings, this study may provide insight into the process of
gaining an initial useful understanding of others and
negotiating for positions of dominance that occur in group
therapy. The use of rotating dyads, such as employed here,
is commonly used in the initial stages of long-term group
process or as an introduction to topics later discussed more
completely (Landfield, 1978). Rotating dyads has been found
to help members become more familiar and to increase
interaction levels in later large meetings (Neimeyer, 1978,
cited in Landfield, 1978; Fairbairn, 1978, cited in
Landfield, 1978). Additionally, information about how
easily group members cope with change suggested by other
members, how they assimilate new constructs encountered in
dissimilar individuals, how they adjust to a previously
existing structure of dominance and make sense out of new,
challenging ways of viewing the world can aid in both
explaining and predicting group process. One may be able to
better predict which members can acclimate to an existing
atmosphere, which members might be able to terminate
prematurely, or which ones may be better suited for a
particular group. In their study of encounter groups,
Lieberman, Yalom and Miles (1973) reported that the group
members' individual needs and conflicts could be used to
Page 30
25
predict assimilation into pre-existing groups. For example,
those who have an unmet need for intimacy may not adjust to
a group led by an aloof, impersonal leader. However, while
similarity is necessary for meaningful understanding of
others, a certain degree of uniqueness provides the new
perspective needed for change to occur. Landfield (1975)
suggested that although congruence of content seems to aid
communication and that lack of construct congruency between
therapist and client may lead to premature termination, some
amount of incompatibility in organization facilitates the
emergence of new ideas. It seems that a balancing act with
the degree of similarity is required for proper match of
patient to therapist and/or group.
The purpose of this study was to examine how
constructivist theory accounts for behavioral measures of
the dominance structure established in dyads in comparison
to a replication of recent literature utilizing trait
theory. Six general hypotheses from research on dominance
behavior in dyads, trait and constructivist theory shall be
described with a brief synopsis following.
Hypothesis One: Trait theory assumes that an individual
will behave in a consistent manner throughout most of his or
her interactions, particularly if many behavioral measures
are used instead of one. Accordingly, subjects who score
high or low dominant on a pre-test measure should behave
consistently in most interactions. Although some studies
'dYrb 2& ... ,n l't GS3wcA'. i: ;...,.-,,; «:N,.:n. -. ,. .. _ .5. irn.M..,. .,.
Page 31
26
have not used level of questioning as a measure of
dominance, Mishler & Waxier (1968) and Fruge (1973)
classified questioning as a way to influence the flow of
conversation without taking responsibility. If a person is
dissatisfied with the existing dyad structure but is
hesitant to appear influential to others, he or she is
likely use a numerous questions in an attempt to control the
dyad without receiving such an attribution.
Hypothesis One: High dominant individuals were predicted to
display more interruptions, longer time talked and fewer
questions; low dominant individuals were predicted to
display the opposite: fewer interruptions, a shorter time
talked, and more questions.
Hypothesis Two: According to Mischel, individuals who are
extremely "trait" in their behavior, rigidly responding in a
consistent manner without regard to situational cues, are
more likely to be maladaptive. Both extremely high and low
dominant individuals demonstrate inflexible and rigid
behavior, despite circumstantial demands. If these
individuals are indeed maladaptive, they would be classified
in Landfield's system as Quadrant 3. However, the dynamics
underlying this coping inability are not clearly delineated.
Although their dysfunction may be due to an inability to
effectively respond to cues they pick up socially (as would
be the case of Quadrant 3), it may also be due to a basic
inability to screen those cues. The latter case is better
Page 32
27
described by individuals classified into Quadrant 1--they
have few ways of understanding the world around them thus
have difficulty understanding anyone or anything outside
their limited range of convenience. Perhaps high dominant
persons are Quadrant 1, unable to understand cues
incompatible with their limited systems and low dominant
persons are Quadrant 3, picking up cues but unable to modify
their behavior because of a lack of understanding.
Hypothesis Two: Individuals with extremely deviant
dominance scores were predicted to evidence more extreme FIC
and Ordination scores than those with average dominance
scores.
Hypothesis Three: In constructivist theory, using questions
rather than interrupting or talking for long stretches was
hypothesized to be more characteristic of those who were
highly resistant to change. Mishler and Waxler (1968) and
Fruge (1973) considered direct attempts to control, whether
person or attention control, to be more interpersonally
adaptive. Rigid individuals are unable to assimilate those
behaviors, even if they are more functional. Individuals
who are resistant to change within their internal structure
may also be resistant to changing external structure, and so
may prefer to maintain the status quo in relationships.
These individuals were hypothesized to use indirect rather
than direct control strategies to communicate their needs.
Similarly, highly flexible individuals would not have an
Page 33
28
aversion to accepting responsibility for the status of
structure and would utilize more direct, functional
strategies of control.
Hypothesis Three: Dyads identified as highly resistant to
change were predicted to utilize fewer interruptions, more
questions and shorter speech durations than dyads identified
as permeable to change.
Hypothesis Four: Permeability to change was hypothesized to
be related to the number of functionally independent selves
subjects reported after each dyadic interaction.
Individuals permeable to change would more easily adjust
those selves brought to different dyads. As flexible dyadic
members present different constructs and selves in their
interactions, these members would have less difficulty
bringing selves that were appropriate for each separate
interaction. Similarly, individuals resistant to change
would have been more likely to present the same selves
regardless of their partner.
Hypothesis Four: Individuals classified as resistant to
change were expected to bring fewer independent selves to
interactions in comparison to those more permeable to
change.
Hypothesis Five: According to Leary's (1957) theory,
interpersonal demands or expectations placed on dyadic
partners influence behavior. Behavior in one category such
as dominance naturally "pulls" for complementary behavior
Page 34
29
such as submission. When the behavior of others does not
conform to what is expected and requested, discomfort,
ineffectiveness and tension result. As Fruge (1973)
reported, interpersonal dissatisfaction, disrespect for
another and discomfort correspond to the use of indirect
control methods. When one is unable for whatever reason to
meaningfully construe an interaction that interaction is
experienced as unpleasant, thus more dysfunctional, indirect
methods of control predominate. In addition, verbal
expressions indicative of tension would also increase.
Occurrences of "Ah's," non-meaningful utterances, and
silence in conversations have been found to reflect tension
and discomfort (Sigmund & Pope, 1968; Doster, 1975).
Because silences were hypothesized to be due to the next
speaker's unability or unwillingness to continue the current
line of thought or to elaborate another line, silences were
attributed to the next speaker following the pause.
Hypothesis Five: Dyads describing themselves as
uncomfortable were predicted to exhibit greater rates of
"ah's," silences, and questions, fewer interruption rates
and shorter speech durations than dyads with comfortable
self-descriptions.
Hypothesis Six: The same experience of anxiety, discomfort,
and inability to directly control were expected to occur
when persons were not able to behave in well-established,
familiar ways. When confronted with another whose thinking
Page 35
30
is very alien to our own, according to the Sociality
Corollary (Kelly, 1955) we are unable to meaningfully
construe the behavior or constructs of another person and
find it difficult to bring an appropriate self to that
interaction. Instead of meaningfully understanding the
interpersonal process and therefore behaving in an
appropriate way, we merely respond without understanding.
As discussed regarding the similarity of therapist and
client affecting treatment, some uniqueness may expand one's
current system, but a large amount of dissimilarity hinders
communication. When for some reason one is unable to behave
in familiar, understandable ways, then anxiety, tension and
inability to utilize direct control strategies were
hypothesized to occur.
Hypothesis Six: Self-descriptions considered non-prototypic
were predicted to accompany greater rates of silences,
"ah's," questions, lesser interruption rates, and shorter
speech durations than those evidenced by prototypic selves.
Page 36
CHAPTER II
METHOD
Subjects
Forty-one subjects were recruited from an introductory
psychology course in return for extra credit. Informed
consent was obtained from all subjects (see Appendix A)
before testing began.
Procedure
In the first testing session, subjects were pre-tested
as a group using the Dominance Scale of the California
Psychological Inventory and the Community of Selves Repgrid.
The Dominance Scale of the California Psychological
Inventory consists of 44 empirically derived true/false
items. Scores were converted to standard scores using
separated norms for sex. A high scorer is described as
confident, self-reliant, independent and displaying
leadership potential. Low scorers are reported to be
unassuming, retiring and indifferent. Gough (1957) reports
test-retest correlations for this scale ranging from .72 to
.80. Several cross-validations studies indicate acceptable
amount of validity. Standard scores ranged from 19 to 68.
The Community of Selves Repgrid (Doster & Watson, 1987)
consists of two sections. First, constructs are elicited by
31
. ...
Page 37
32
Landfield's (1971) modification of Kelly's (1955) Role
Construct Repertory Grid (see Appendix B). Secondly, these
constructs are applied to a series of potentially relevant
self-experiences, including the "Usual me" in order to gain
information regarding the self most predominant within the
construct system (see Appendix C).
At the second testing session, subjects participated in
pre-arranged groups of eight to nine persons, arbitrarily
assigned. The last four digits of their social security
numbers identified them. Subjects successively interacted
one-on-one with. other group members for five minutes until
each possible pair had interacted, the pairing having been
determined randomly. All interactions were tape recorded.
After each five minute conversation, subjects rated their
self experiencing for that interaction on their constructs
elicited from the Rep-test. After all pairs had interacted,
group members ranked one another according to perceived
influenceability and the degree to which they felt
comfortable with each partner (see Appendix E).
Peer rankings of dominance, consisting of seven or
eight scores (depending upon the group size) were averaged
for each person. Higher numerical rankings indicate lower
perceived dominance by their peers. Scores ranged from 2.00
to 7.88. Total and differential comfortability were
regarded in a dyadic manner. These measures consisted
respectively of the total or absolute difference of each
Page 38
33
participant's ranking of self-comfortability within the dyad
divided by the total group membership, to control for
variance due to unequal sizes. Higher scores indicated that
greater unease was experienced. Total comfortability scores
ranged from .29 to 2.0; differential comfortability ranged
from .00 to .88.
Measures of Resistance to Change
Four factors hypothesized to be related to propensity
to change were derived from the Rep-test. Monadically, each
of the four measures were computed before adding dyadic
information, obtaining an overall measure of the subjects'
propensity for resistance to change. This monadic
resistance to change measure was computed by separately
ranking the scores of all thirty-seven subjects for each of
the four measures. The rankings were then computed to
achieve a total resistance to change score. Scores varied
from 33.60 to 112.50, with lower scores indicating less
resistance to change.
Each of the four indicators had dyadic permutations as
well, consisting of conceptually similar indicators specific
to each interaction, seven or eight per person. These four
factors were ranked separately for all 284 available
incidences; thus each dyad had two separate rankings for
each of its four factors: both total and differential
resistance to change. Total resistance to change was
derived by totalling all eight scores ranging from 264.50 to
Page 39
34
1,548.00. Differential resistance to change consisted of
the total of the absolute difference between each of the
four sets; scores ranged from 22.50 to 628.50.
Constriction. A construct or self is thought to be
constricted or rigid when it is rated in extremes, either
positively or negatively. This implies that there is no
maleability or flexibility in that self, but either-or
thinking. Such selves or constructs would be impermeable to
external influence. Specifically, a construct is determined
as constricted monadically either when one rating level was
used eight or more times or when two levels together were
used eleven or more times. Dyadically, each self rating
after interactions was defined as constricted either if a
zero or six rating was used eight or more times or if the
total was eleven or more.
Superordinateness. Landfield and Barr (1976) proposed
that superordinateness was the ability to differentiate
between rating levels used in describing a self. This
superordinateness score was computed by multiplying the
number of different rating levels used to describe each self
by the difference between the lowest and highest rating.
This measure was computed monadically with original rep-test
data and dyadically for each of the seven or eight elicited
selves.
Connectedness. A self is thought to be connected
within a system when using that self affects others. It is
Page 40
35
operationalized monadically as the number of functionally
independent selves for the self-rep, and dyadically as the
number of selves functionally dependent upon each of the
seven or eight elicited selves.
Predominance. To tap the frequently used self, a usual
self was provided on the Rep-test. Predominance was defined
monadically as the degree of functional dependence of each
of the seven elicited selves with the usual self, totalled
across all selves. Dyadically the information was not
totalled but maintained separately for each of the elicited
selves.
Dependent Measures
Two independent raters examined a sample of tape
recordings for total time talked, number of interruptions,
and number of questions for each subject. Raters were given
definitions of each measure based upon Aries, Gold & Weigel
(1983) and Fruge (1973) (see Appendix D). All measures met
acceptable standards of interrater reliability except for
interruptions, which was initially .68. Coding
discrepancies were reevaluated by the raters and the rating
criteria solidified. Subsequent interrater reliability of
interruptions met acceptable levels ( r =.92). After coding
standards were thus established, the principle examiner
completed the rest of the coding. Interrater correlations
of observational measures are available in Table 1 below.
Page 41
36
Table 1
Interrater reliabilities of Observational Measures
Variables Correlation
Speech .92
Interruptions .92
Questions .92
Ahs .81
Silences .83
Because dyads varied in conversational time allowed,
observational measures were converted to obtain more
proportionally representative scores (Fruge, 1973). In
addition, to utilize both numerator and denominator in the
same scale, all time-related measures such as speech
duration, total and differential talk time, plus both
monadic and dyadic silence rates were all converted to
minutes. For monadic analysis, speech duration was computed
by dividing time talked by the total interactional time of
the dyad, including pauses; thus a score of .50 would
reflect that the subject talked for 50% of available
interaction time. A Pearson-moment correlation conducted
between the actual speech time of a dyad and the total
interactional time was .99, indicating that the proportional
measure of speech duration was not inflated due to excessive
Page 42
37
silences. Silence rate was computed by dividing the
combined length of silences by the total interactional time
of the dyad. Interruption rate, question rate, and "ah"
rate consisted of the incidence of each variable divided by
the total available time of that dyad.
Measures for dyadic analyses were computed separately.
Total and differential talk time were computed respectively
by totalling and subtracting the speech duration of both
participants, excluding silences, and dividing the results
by the total interaction time. Total and differential
interruption, question, "ah," and silence rates were
similarly obtained by dividing the total or the absolute
difference of each occurrence by the total interaction time.
Combined length of silences was used for dyadic silence
rates.
Page 43
CHAPTER III
RESULTS
The means and standard deviations of the monadic self
report and observational measures are listed in Table 2;
dyadic means and standard deviations are listed in Table 3.
Table 2
Means and Standard Deviations of Independent and
Dependent Variables: Monadic
Males Females
(30) (11)
M SD M SD
Dominance 44.81 11.17 47.91 10.01
Speech Duration .50 .08 .47 .09
Interruption Rate .65 .53 .94 .60
Question Rate 1.39 1.44 1.39 1.10
Ah Rate 1.01 .48 .89 .50
Silence Rate .02 .01 .03 .03
Resis. To Change 64.31 14.24 65.99 19.76
Peer Rankings 3.91 1.18 4.71 1.29
of Dominance
38
. w. _:. :.. , --
Page 44
39
Table 3
Means and Standard Deviations for Dependent Variables: Dyadic
M/M M/F F/F
(77) (55) (10)
Variables M SD M SD M SD
Comfort
Total
Differential
Speech Duration
Total
Differential
Question Rate
Total
Differential
Interruption Rate
Total
Differential
Ah Rate
Total
Differential
Silence Rate
Total
Differential
.17
.:29
.41
.21
.98 .03
.23 .15
.:L4
.51
.99
.59
1.13
.29
.99
.22
1.43
.70
.51 .49
.24 .30
.04
.45
.02
.64
.42
.28
.01
.16
1.05
.79
.76 .56
.31 .26
.90 .65
.36 .34
.01 .01
.47 .55
.65
.36
.03
.66
.85 .40
.28 .25
.99 .02
.24 .16
1.26 .94
.79 .88
.89 .64
.31 .37
.56 .41
.29 .25
.01 .02
.35 .61
Page 45
40
Table 4
Correlation Matrix of Verbal Measures and Dominance Scale
Variables'z B C D E F G
A .28* -.33* -.11 .10 .00 -.15
B -. 50*** -. 39** .07 -. 39** .16
C .81***-.29* .08 -.24+
D -. 32* -.01 .02
E -. 09 .10
F -.07
'A = Dominance Score; B = Speech Duration; C = Interruption
Rate; D = Question Rate; E = Ah Rate; F = Silence Rate;
G = Sex.
+ 2 < .10; * 2 < .05; ** 2 < .01; *** P < .001.
The first hypothesis proposes that persons higher on
the dominance scale would evidence higher levels of
interruptions, longer speech duration, and fewer questions.
Pearson product moment correlations were performed upon
dominance scores, sex, and verbal behaviors based upon
subjects' interactions. The dominance scale correlated
significantly with speech duration (r (36) = .28, 2 <.05)
and interruption rate (r (36) = -.33, 2 <.05). More
dominant individuals spoke longer and were interrupted less.
Sex was not significantly correlated with dominance or any
-Now.
Page 46
41
of the dependent measures, although there was a tendency for
females to have a higher rate of interruptions (r (41) =
-.25, <.10). Pearson correlation coefficients for these
measures are available in Table 4. These results gave
conflicting support for the hypothesis.
Additionally, peer rankings of influenceability were
used to predict the same characteristic verbalizations.
Other Pearson product-moment correlations were computed
between rankings of influentiability and the dependent
variable rates. These correlations are available in Table
5. Speech duration was negatively correlated (r (41) =
-.47, 2<.001), silence rate positively correlated (r (41) =
.50, <Q.001), and "ah" rate negatively correlated (r (41)
=-.36,2 < .01) with peer rankings of dominance.
Individuals characterized by their peers as being more
influential talked for relatively longer periods of time,
had fewer silences, but more "ah's." Sex was not
significantly related to peer rankings of dominance although
there was a tendency for females to be ranked as more highly
dominant. This provides greater support for the hypothesis,
although again, the sex correlate was not predicted.
Lastly, to compare peer opinion with self-rating, a
Spearman's rank order correlation was computed between
ranking according to the dominance scale and ranking of
influentiability by subjects. These two measures were
highly positively correlated (r (36) = .45, <.01); those
n 't k uS JSdi x ki*m ijii;+.; .. :.i n.:n., ;,r:.. .. : - .. ,::.. : z .:; l, c.: ;u
Page 47
42
considered dominant by their peers usually experienced
themselves as dominant on the trait scale.
Table 5
Correlations Between Peer Ranking of Dominance and Verbal
Measures
Variables' B C D E F G
A -.47*** .19 -.13 -.36** .50*** -.16
B -.50*** -. 39** .06 -.39** .14
C .81*** -.29* .08 -.21+
D -.33* -.01 .01
E -. 09 .11
F -. 08
'A = Average Peer Rankings of Dominance; B = Speech
Duration; C = Interruption Rate; D = Question Rate; E = Ah
Rate; F = Silence Rate; G = Sex.
* 2 < .05; ** < .01; *** 2 < .001.
It was next hypothesized that extreme trait dominance
scores, whether very high or low, would indicate pathology
as evidenced by extreme FIC and/or Ordination scores
(Landfield & Barr, 1976). A forward multiple regression
analysis was conducted to determine the ability of FIC,
Ordination scores, and sex to predict deviation of
dominance. The absolute value of the difference between
. ,_ :.::.,.. ..,.ter,,,.,. , , ;
Page 48
43
dominance scores and the overall mean of the sample was used
as the dominance deviation score. FIC significantly
predicted dominance deviation ( F (1, 58)=4.07, 2 < .05, R
2=.06). Individuals who reported dominance levels deviating
significantly from the mean, whether high or low, evidenced
lower FIC scores. Ordination did not reach minimal
tolerance levels and thus was not entered. This partially
supported the hypothesis.
Table 6
Regression Results for Deviation of Dominance and FIC
Predictor Value Beta Value R2
FIC -.25* .06
Ordination -- --
* 2 < .05.
Individuals highly resistant to change were expected to
favor questioning over interrupting or holding the floor to
direct conversation flow. Separate multiple regressions
analyses were run between resistance to change and sex for
each of the dependent variable rates. This analysis was
done at the dyadic level, so that the total and differential
resistance to change scores between each couple were treated
as a case, rather than as individuals. The cross product of
the total and difference scores was used as a predictor to
Page 49
44
determine if there was a significant interaction between the
two separate factors. The criterion variable rates were
also computed as total and difference rates within each
interaction. Gender was partialled into male/male,
male/female and female/female dyads. There was a
significant difference in ratio: 77 male/male dyads; 55
female/male dyads; and 10 female/female dyads. Dummy coding
of sex was used to separate dyads so that when one sex was
considered as a predictor variable, the other two sex groups
were combined for comparison. For example, the criterion
variables were predicted by male/male dyads in comparison to
combined male/female and female/female groups. The
regression analyses were done in a forced entry manner so
that Total Resistance to change, the Differential Resistance
to change and the cross-product were entered sequentially.
Results are contained in Tables 7 through 16 and are
available in Appendix G. All tables report R 2 as the
increment of explained variance for each unique variable,
and the significance level of Beta values refer to the
significance of the added predictability of that variable,
not of the overall model. Correlations between predictor
variables and criterion variables are available in Table 17.
Total resistance to change was a significant predictor
of total interruption rate (F (1,113) = 8.39, 2 <.005, R 2 =
.07), total question rate (F (1, 113) = 20.45, <.001, R 2
= .15), and the difference in question rate (F (1,113) =
Page 50
45
6.08, p <.05, R 2 = .05). Dyads evidencing high total
resistance to change were associated with lower overall
interruption rates, and with both members asking few
questions. The differential resistance to change and the
product between total and differential resistance to change
did not contribute substantial increments in predictability.
Overall it seems that interruption and question rate were
valid predictors, accounting for a combined 27% of the total
of the total variance. These results partially support the
hypothesis.
Sex was also a significant predictor of control
methods. Differential resistance to change and the cross-
product resistance to change caused suppression of the
explanatory model for each criterion variable, so the
analyses were re-examined without these two variables. Beta
values and R2 reported in Appendix G refer to analyses
performed without these two factors entered into the model.
Male/male dyads were characterized by lower total rates of
silences (F (2,112) = 6.47, p <.001, R 2 = .10), higher
total rates of interrupting (F (2,112) = 7.58, p <.05, R 2 =
.05) and longer overall speech duration (F (2,112) = 6.67, p
<.001, R 2 = .10). Overall, 25% of the variance was
predicted using these variables. Female/female and mixed
sex dyads were not predicted by any of the variables.
t auf a : r. t¬; - " . , - .. _.... -si iW+Meiti^,' . - i :.::.,-..YZ .:,cc,..,: cAGR g.
,..;,v: _
Page 51
46
Table 17
Correlations Between Resistance to Change, Sex and
Criterion Variables
Variables 1 A B C D E
Total Talk Rate .01 .04 .01 .32 -.27
Differential Talk Rate .10 .03 .05 .03 -.02
Total Interruption Rate -.26 -.07 -.17 .26 -.19
Differential Interruption -.16 .11 .04 .13 -.12
Rate
Total Question Rate -.39 -.06 -.20 .15 -.15
Differential Question Rate -.23 -.08 -.15 .10 -.06
Total Ah Rate .05 .10 .14 -.16 .05
Differential Ah Rate .03 .09 .11 -.14 .09
Total Silence Rate -.01 -.04 -.01 -.32 .27
Differential Silence Rate -.09 -.06 -.05 -.16 .11
'A = Total Resistance to Change; B = Differential Resistance
to Change; C = Cross-product Resistance to Change;
D = Male/male dyads; E = Male/female dyads.
Individuals identified as more resistant to change were
predicted to display fewer independent selves across
interactions. A Pearson-moment correlation was conducted
between the monadic resistance to change score, the ratio of
functionally independent selves brought to the interactions,
_ it82k illwinFa"+i 'F t -... - .-. +,zx:ar +. :-,r:- , . ._ ... ... ,. :. tia.+ s::,i ., .. rn ::.,w;.
Page 52
47
and the dominance scores. The number of functionally
independent selves was divided by the number of group
members to control for variance in the number of different
selves assessed. Resistance to change was not significantly
correlated to the occurrence of independent selves but was
positively correlated to dominance (r (34) = .35, p <.05).
These results are available in Table 18. Individuals
identified as highly resistant to change appeared more
dominant on the trait measure but did not vary significantly
in the proportion of independent selves evidenced. This
does not support the hypothesis except by inference; highly
"trait" individuals were considered rigid according to the
constructivist measure.
Table 18
Correlations Between Resistance to Change, Dominance Scores
and Proportion of Independent Selves Brought to Interactions
Variables' B C
A -.19 .35*
B -. 18
'A = Resistance to Change; B = Ratio of Functionally
Independent Selves; C = Dominance.
* 2 <.05.
Comfortability of a dyad was hypothesized to be
evidenced verbally by "ah's" and silences. A multiple
Page 53
48
regression was computed for how ably each of the verbal
measures predicted the reported comfort of the dyad. These
analyses were computed in a dyadic manner; the total and
difference levels of comfort within each interaction were
compared to the total and differential rates of speech
duration, questioning, interruptions, silences and "ah's."
This model was significant in predicting total
comfortability using the total interruption rate (R 2 = .07,
F (1,139) = 10.66, 2 <.01) and the difference in question
rate (R 2 = .10, F (2,138) = 7.69, 2 <.001). Only these two
factors correlated highly enough to be entered into the
equation. Table 19 contains a listing of these results;
Table 20 contains correlations between total comfort,
differential comfort, the cross-product comfort and the
predictor variables. The factors correlated such that
higher total comfort was associated with fewer total
interruptions and mutual levels of questions. Differential
and cross-product comfort were not significantly predicted
by any of the variables. Although silences and " ah' s " did
not vary consistently, there appears to be some influence
upon direct control strategies by combined comfort.
. .'R;:YYf e V179V vA - iUir7r ".:: «4}iK? ifiblNt. i: ._: ._: .:.:.. .. t .. : :r::;r ;.w:r. -. ?b7Yx' __ -
Page 54
49
Table 19
Regression Results for Total Comfort
Predictor Variables Beta Values R2
Total Interruption Rate -.27** .07
Differential Question Rate .18* .03
* 2 < .05; ** 2 < .01.
Interpersonal incidences of anxiety, discomfort and
inability to exercise direct control strategies were
hypothesized to accompany nonprototypic, or nonfamiliar,
self-experiencing. Ratings of self characterizations after
each conversation were separated into those that were
functionally dependent to selves elicited previously and
those that were unique. A one-way MANOVA was computed
between these two levels of selves with the five verbal
measures and self-reported comfort after each interaction.
The overall main effect was significant (F (6,236) = 3.34, ?
<.01) and subsequent univariate analysis indicated that
question rate (F (1,241) = 11.23, 2 <.001) and interruption
rate, (F (1,241) = 4.68, 2 <.05) were significantly higher
in nonprototypic selves. The hypothesis was questionably
supported; while higher question rate was supportive, the
higher interruption rate was incongruent, and additional
Page 55
50
indices of higher total rate of silences and " ah' s " were
missing.
Table 20
Correlations Between Comfort and Predictor
Variables
Variables1 A B C
Total Talk Rate .09 .06 .06
Differential Talk Rate .03 -.04 -.05
Total Interruption Rate -.27 -.04 -.10
Differential Interruption - .22 -.02 - .09
Rate
Total Question Rate -.02 .06 .04
Differential Question Rate .07 .08 .07
Total Ah Rate -.01 .05 .04
Differential Ah Rate -.00 .07 .05
Total Silence Rate -. 10 -.06 -.06
Differential Silence Rate .05 .01 .04
1 A = Total Comfort; B = Differential
product Comfort.
Comfort; C = Cross-
Page 56
CHAPTER IV
DISCUSSION
The purpose of this study was to assess how useful
constructivist techniques proved in predicting verbal
indicators of dominance in briefly interacting dyads. Six
hypotheses were proposed and tested. The first hypothesis,
of correlating scores on a trait measure of dominance with
verbal indicators based on subjects' interactions, -elicited
conflicting support. Higher trait-dominant subjects
evidenced longer speech durations and lower interruption
rates while interpersonally ranked dominant subjects spoke
for longer periods of time and evidenced fewer silences and
more "ah's." Interpersonal measures of dominance
demonstrated a more consistent behavioral characterization
than did the trait predictor.
Secondly, subjects who appeared highly consistent or
rigid in a trait manner were hypothesized to be less well
adjusted in constructivist terms. A multiple regression
analysis was computed using Ordination, FIC, and sex of
subjects to predict more extreme trait indicators of
dominance. This analysis was partially supported. Extreme
trait dominance was significantly associated with lower FIC
scores, corroborating Mischel, 1973. These individuals
51
- OW-klk*W,
Page 57
52
would be classified in Quadrants 1 or 2 of Landfield and
Barr's (1976) quadrant system, depending upon their
ordination scores. Such individuals are described as being
limited in conceptualizing the world and uncomfortable with
situations outside of their viewpoint. Such persons could
easily be considered rigid interpersonally, constantly
behaving either deferentially or autocratically despite
mediating situational variables. Conceptually similar is
Phillips' (1981) study which reported that people
hospitalized with schizophrenia who showed extreme FIC
scores, either high or low, evidenced more extreme thought
disorders. This current study found maladjustment at one
end of this continuum. Perhaps difficulty of the opposite
kind, instability and erratic behavior, would be
characteristic of normals at the opposite extreme. Phillips
(1981) did not utilize Ordination as a predictive variable.
Individuals who were highly resistant to change were
thought more likely to use indirect methods of influencing
their verbal interactions. Several multiple regression
analyses were conducted using resistance to change to
predict verbal measures of subjects' interactions which
supported the hypothesis. Dyads similarly higher in
combined resistance to change were found to have lower rates
of interruptions and mutually low levels of questioning.
Resistance to change was successful in predicting reluctance
to utilize not only direct interventions, but indirect
Page 58
53
methods as well. It would be useful to discover how dyads
resistant to change expressed their conversational
preferences, if at all. Sex was even a more significant
predictor of control strategies. Male/male dyads had fewer
silences, more interruptions and longer speech duration for
both participants, the prototype of assertive, direct verbal
interaction. Sex of dyads was more predictive of verbally
dominant behavior than either trait or constructivist
theories.
Individuals more resistant to change were expected to
display fewer independent selves across their series of
interactions. A Pearson Product-moment correlation was
computed between these two factors, without resulting
significance. It may prove less difficult to elicit
autonomous selves in future tests if conversations were
allocated more time. However, those subjects more resistant
to change were associated with higher dominance levels,
giving added support to the second hypothesis: highly
"trait" individuals may be interpersonally inflexible.
Dyads reportedly less comfortable with one another were
suggested to evidence more indirect methods of control as
well as verbal indices of discomfort. A multiple regression
analysis was conducted using the verbal measures to predict
comfort. Dyads with higher combined comfort had fewer
interruptions and mutual levels of questions, partially
corroborating this hypothesis. This replicates Fruge's
Page 59
54
(1973) finding that dyadic dissatisfaction rises when one
person asks a disproportionate number of questions. These
dyads seem to be avoiding direct interventions but utilizing
indirect methods in proportion to their partners, passively
following their lead. It seems unusual that "ah's" did not
rise with comfort. Perhaps this is due to the experience of
discomfort as one of tolerable incongruence or unease with
the other person's overall perceptions, rather than an
agitated, "state" experience of emotional unease.
Lastly, interactions in which participants were unable
to utilize a familiar role were expected to evidence
indirect control strategies. Results of the MANOVA were
that nonprototypic selves were associated with higher
questioning and interrupting, providing some degree of
endorsement. It seems apparent that nonprototypic
experiences were not merely confusing or anxiety provoking,
but that they motivated participants to become more involved
in their processing. This seems somewhat similar to what
uncomfortable dyads were evidencing. While these
experiences do not produce typical anxiety, they seem to be
discomforting and activating at a behavioral level.
One of the assumptions in this study was that direct
and indirect control strategies were identifiable, had
different functions in conversation and would differentiate
the types of individuals who prefered one type of strategy
over the other, based upon Mishler and Waxler, 1968.
Page 60
55
However, this study did not support this distinction;
interruptions and questions were primarily used in a similar
manner. Interruptions and questions correlated at a very
high level, r (41) = .81, < .001 (see Tables 4 or 5).
Rarely did interruption rate rise when question rate did not
follow, particularly in dyadic analyses. When dyads were
less resistant to change, uncomfortable or were unable to
utilize familiar ways of behaving, both questioning and
interrupting rose. It may be more useful to collapse direct
and nondirect categories of control strategies into the
existence of any verbal methods at all. When verbal
manipulations are not used, nonverbal messages may become
more influential. Eye contact, tone of voice, or posture
could convey interest level, dissatisfaction or aloofness.
Similarly, analyzing the content of questions and
interruptions could further delineate the intent of such
strategies. Questioning to introduce a new topic of
conversation seems to be qualitatively different from
further delving into a topic one's partner introduced.
Similarly, skillful questioning can perhaps convey
disinterest as deftly as interrupting can. Due to the
insignificance and unreliability of differentiating
successful from unsuccessful interruptions in the literature
(Roger & Schumacher, 1983), this distinction was not
employed. If a more reliable method of separating the two
were available, such a measure might be meaningful in
Page 61
56
delineating truly dominant individuals from those who are
merely uncomfortable with their relative influenceability.
The finding that interruptions were negatively
correlated with trait dominance is very difficult to
explain. It is in direct contradiction to past literature.
Trait dominance seems to be measuring a construct
qualitatively different from interpersonal dominance,
although the two were very highly correlated. The
behavioral correlates of interpersonal dominance were more
clearly related to "text book" definitions of dominant
behavior, long speech duration and fewer silences (see Table
5). Trait dominance somewhat resembled resistance to change
(lower interruption rates) which was also linked to
maladaptive rigidity.- As is common in self-report measures,
there is often a discrepancy between self-presentation and
actual behavior. Perhaps trait dominance was tapping into
more inflexible persons' wishes to be characterized as
influential. In addition, there were other factors found to
covary with verbal behaviors, dyadic comfortability and
prototypic experiencing. These variables could have
moderated the appearance of dominance as a trait. It has
also been suggested that self monitoring may mediate the
expression of control strategies (Soldow & Thomas, 1984), so
that the particular social circumstances underlying each
situation would affect whether direct or indirect methods
are employed.
Page 62
57
The high correlation ( r (36) = .45, p < .01) between
interpersonally ranked dominance and trait dominance has
implications for measurement of personality variables. It
seems evident that whatever process individuals use to
screen and interpret behavior is a fairly accurate one.
Since there was an indirect correlation between those
rankings and pathological rigidity, future research might
elaborate more directly the ability of laypeople to detect
dysfunctional behavior in others. Distinguishing the exact
mechanisms that persons use in impression formation might be
used to improve existing trait measures.
Resistance to change seems to have made a fairly good
showing in this study. Except for its inability to predict
the number of independent selves actually observed, it
behaved as expected with regard to verbal control methods.
This variable may prove useful in psychotherapy.
Individuals who exhibit few questions and/or interruptions,
who avoid initiating change in the status quo, electing to
mirror the amount of control mechanisms their peers use, may
be considered resistant to change. Such clients may be
anticipated to be resistant to therapeutic intervention, so
that less offensive, incongruent strategies may be chosen to
gain initial access to inflexible schemas. In a group
psychotherapy setting, this construct may be used to
identify potential peer leaders that would be active during
Stage II rebellion (Yalom, 1970), or those who would be
Page 63
58
unable to assimilate into an existing group structure,
especially those considerably foreign to the group's
perceived socio-economic status, marital status or gender,
as well as more intrapsychic differences. There are several
cautions, however. Sex findings indicate that such a
construct may have more meaning for male/male dyads. Women
appear to be disadvantaged in this respect, appearing more
rigid and inflexible perhaps due to sociological rather than
personological causes. Similarly, these verbal indicators
also may rise temporarily due to uncomfortable or unfamiliar
settings. Such situations might temporarily increase
control mechanisms due to internal emotional states rather
than permeability to change.
Discouraging to the utility of constructivist or trait
theory was the superior utility of dyadic sex in predicting
verbal behaviors. Male/male dyads more fully characterized
dominant interactions than any monadic analyses or dyadic
contributions of resistance to change. Female/female dyads
or mixed sex dyads were not as easily predictable, as Aries
et al., 1983, reported. The factors contributing to the
emergence of functional, direct conversation in males has
usually been attributed to socialization (Zimmerman & West,
1975). It would be interesting to consider if this type of
verbal asset could explain instances of "networking" that
help men climb the business ladder. Not only may male
subordinates have common backgrounds with their superiors,
Wi t,.-ti6'e h:5w. aSka.-, ,,i.n.,,. ....-. ,. i, ix _.,;. -v: i!m"n
Page 64
59
but they may also be able to communicate in a more direct,
functional manner than male bosses do with their female
subordinates.
It is recommended that further research consider
nonverbal indicators of dominance as well as verbal ones.
Although it may be more tedious and logistically difficult
to obtain and code behavioral measures, there is a wealth of
information overlooked at that level. In particular, which
person is addressed more frequently could be a meaningful
indicator of perceived interpersonal dominance. In
addition, further personality measures such as self-
monitoring may further elucidate sex differences and other
inconsistencies.
Page 65
APPENDIX A
INFORMED CONSENT
60
Page 66
61
Name of Subject:
Principle Investigator: Caroline Curlin
The purpose of this study is to obtain information
regarding problem-solving and individual differences.
Possible benefits include positive interpersonal experiences
with group members and a further knowledge of the experience
of topics of psychology.
I understand that I will be asked to fill out a number of
personal questionnaires related to topics of psychology
discussed in class. Group discussions will be led by
graduate students in the Psychology department, and each of
the sessions will last approximately fifty minutes. The
first meeting will differ from the other four sessions in-
that in order to facilitate getting acquainted, each member
will talk one-on-one with every other group member for ten
minutes. Group meetings shall be tape recorded or video-
taped and the tapes will be erased after the necessary data
has been collected.
The results of this study will hopefully be published in
a psychological journal, but the identity and information
about individual participants will not be revealed.
I have heard a clear explanation and understand the
nature of this procedure. I understand that there are few
if any potential risks involved. I have heard a clear
Page 67
62
explanation and understand the benefits to be expected. If
I have any questions about this study, the investigator or
the instructor will be glad to answer them for me. I
understand that I am free to withdraw my consent for the
utilization of my data at any time without any negative
consequences.
I have read the above and understand it and hereby
consent to the procedures described above.
Date
Subject
Page 68
APPENDIX B
INSTRUCTIONS FOR THE ROLE CONSTRUCT TEST
63
:x'k ?" A+X 75'.:1uctGrA+.F '.; ti F<nd i . ... .. :: ,.:.. .,w"w .. ,. "y",,t, ..+w'.,..' a
Page 69
64
First Step:
Find the slanted lines in the upper left-and corner of the
RESPONSE SHEET.
1. Write the first name of your mother or the person who
has played the part of .your mother where it says mother.
2. Write the first name of your father or the person who
has played the part of your father where it says father.
3. Write the name of your brother nearest your own age,
or the person who has played the part of such a
a brother.
4. Write the name of your sister nearest your own age,
or the person who has played the part of such a
sister.
5. The happiest person you know personally.
6. The teacher whose point of view you have found most
acceptable.
7. Your sister nearest your own age, or the person who
has played the part of such a sister.
8. Your closest present friend of the same sex.
9. The unhappiest person you know personally.
10. The person you have met whom you would most like to
know better.
11. The most unsuccessful person you know personally.
12. The person with whom you usually feel most uncomfortable.
13. Your father or the person who has played the part
of your father.
Page 70
65
14. The most successful person you know personally.
15. Your husband (wife) or closest present boyfriend
(girlfriend).
16. Re-enter the name of the person list in #1.
Below your list of names, find row A. Notice that row
A has two highlighted squares. Think carefully about
yourself and about these two people. As you think about the
three of you, try to find one way in which two of you are
alike and different from the other.
Write the way in which two of these three people are
alike in the space provided at the right under Similarity.
Write the way in which one of these three people is
different from the two who are alike in the space provided
at the right under Different. If you cannot - see a
similarity or a difference among the three people leave
blanks.
After you finish row A, complete row B, row C, etc.
Follow the same instructions given above.
Page 71
APPENDIX C
INSTRUCTIONS FOR THE COMMUNITY OF SELVES REP-TEST
66
Page 72
67
Turn the answer sheet over to the reverse side. Notice
the dotted line that extends across the top of the page.
Fold the top of the page along the dotted line. When you
turn the answer sheet back to the front page, a new list of
titles should cover the names of your acquaintances.
The new titles are a list of the various ways you may
have of experiencing yourself. At times you may find
yourself to be "vulnerable" person, at times you may find
yourself to be a "self-critical" person, and so on.
The descriptions you have written under comparisons can
now used as rating scales to describe the various ways you
do experience yourself. Notice that between your
descriptions is a rating scale. You can use the rating
scale to describe moments in which you experience yourself
as silly and moments in which you experience yourself as
serious, for example.
Start with column 1 -the vulnerable me. Try to recall
times in your life when you have experienced yourself as a
vulnerable person. Can you get a picture in your mind of
some of those times when you felt vulnerable? Sometimes,
closing your eyes and making a mental picture can be
helpful. Now try to describe what you are like when you are
a vulnerable person. using each of the 15 rating scales
under comparisons, describe yourself as a vulnerable person.
Ratings should appear in each square under column 1.
Page 73
68
Use a zero (0) rating when neither description fits
your experience of yourself or when you are unsure about
your experience.
Continue on to Column 2 -the spiritual me. Try to
recall times in your life when you have experienced yourself
as a spiritual person. Can you get a picture in your mind
of some of those times when you experiences yourself as a
spiritual person? Again, closing you eyes and making a
mental image can be helpful. Now try to describe what you
are like when you are a spiritual person. Use each of your
15 rating scales under comparisons to describe yourself as a
spiritual person. Ratings should appear in each square
under column 2.
Continue on to each of the other columns. Try to
describe the various ways you may have of experiencing
yourself.
Column sixteen has been left blank for you to fill in
you "usual me", or your predominant self-experience.
Page 74
APPENDIX D
INSTRUCTIONS FOR SCORING VERBAL MEASURES
69
Page 75
70
Time Talking
The number of seconds a subject actually engaged in
talking during the dyadic discussion including overlaps with
other speakers but excluding time spent laughing and time
associated with single-word affirmations of statements.
Silences longer than two seconds were excluded.
Interruptions
The number of separate instances in which a subject
began to speak while another member of the group was still
speaking, excluding single-word affirmations or negations.
Questions
The number of statements that pose a question, whether
actually demanding an answer or rhetorical.
Ah's
Any verbal nonfluency or incoherent sounds that have no
inherent meaning that indicate a break in speech or thought.
Silences
Any pause measured in seconds that lasts longer than
two seconds.
Page 76
APPENDIX E
DOMINANCE AND COMFORTABILITY RANKINGS OF PEERS
71
Page 77
72
ID Code:
Rank each of the people your talked with today as to how
comfortable you felt with each. Put "1" besides the person
with whom you felt the most relaxed, comfortable, or at ease
and put up to "7", "8" or 911, depending upon the number of
persons in your group for the person with whom you felt the
least relaxed, comfortable or at ease.
Rank each of the people you talked with today as to how dominant
they appeared while interacting with you. Put "1" besides the
person you consider to be the most dominant, influential,
assertive, aggressive or strong and put up to "7", "8" or "9",
depending upon the number of people in your group, for the
person you consider to be the most shy, retiring, timid, quiet
or bashful.
Page 78
APPENDIX F
DOMINANCE SCALE OF CALIFORNIA PERSONALITY INVENTORY
73
Page 79
74
ID CODE:
For each item, if you agree with a statement, or feel
that it is true about you, answer TRUE. If you disagree
with a statement, or feel that it is not true about you,
answer FALSE.
1. I think I would enjoy having authority over other
people.
2. I find it hard to keep my mind on a task or job.
3. 1 have sometimes stayed away from another person
because I feared doing or saying something that
I might regret afterwards.
4. When in a group of people I have trouble thinking
of the right things to talk about.
5. School teachers complain a lot about their pay,
but it seems to me that they get as much as they
deserve.
6. I don't blame anyone for trying to grab all he
can get in this world.
7. Every citizen should take the time to find out
about national affairs, even if it means giving
up some personal pleasures.
8. I should like to belong to several clubs or
lodges.
9. I am certainly lacking in self-confidence.
10. When I work on a committee I like to take charge
of things.
Page 80
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
75
If given the chance I would make a good leader of
people.
Sometimes at elections I vote for men about whom I
know very little.
I very much like hunting.
A person does not need to worry about other people
if only he looks after himself.
I can honestly say that I do not really mind paying
my taxes because I feel that's one of the things
I can do for what I get from the community.
When prices are high you can't blame a person for
getting all he can while the getting is good.
In school I found it very hard to talk before
the class.
I am a better talker than a listener.
I would be willing to give money myself in order
to right a wrong, even though I was not mixed up
in it in the place.
We should cut down on our use of oil, if necessary,
so that there will be plenty left for the people
fifty or a hundred years from now.
When the community makes a decision, it is up to
a person to help carry it out even if he had been
against it.
I would rather have people dislike me than look
down on me.
22.
Page 81
76
23. I must admit I try to see what others think before
I take a stand.
24. People should not have to pay taxes for the schools
if they do not have any children.
25. In a group, I usually take the responsibility for
getting people introduced.
26. I would be willing to describe myself as a pretty
"strong" personality.
27. There are times when I act like a coward.
28. I must admit I am a pretty fair talker.
29. 1 have strong political opinions.30. I think I am usually a leader in my group.
31. I seem to do things that I regret more often than
other people do.
32. Disobedience to any government is never justified.
33. I would rather not have very much responsibility
for other people.
Page 82
APPENDIX G
TABLES 7-16
77
Page 83
78
Table 7
Regression Results for Total Talk Rate
Beta Change in
Predictor Values Values R2
Total Resistance to Change .01 .00
Differential Resistance to Change .04 .00
Cross-product Resistance to Change -.74 .02
Male/male Dyads .32*** .10
Male/female Dyads -.03 .00
* ** p_< .001.
Table 8
Regression Results for Differential Talk Rate
Beta Change in
Predictor Values Values R2
Total Resistance to Change .10 .01
Differential Resistance to Change .04 .00
Cross-product Resistance to Change -.29 .00
Male/male Dyads .05 .00
Male/female Dyads -.02 .00
Page 84
79
Table 9
Regression Results of Total Interruption Rates
Beta Change in
Predictor Values Values R2
Total Resistance to Change -.26** .07
Difference Resistance to Change -.09 .01
Cross-product Resistance to Change -.43 .01
Male/Male Dyads .23* .05
Male/Female Dyads .15 .01
* <.05; ** 2 <.01.
Table 10
Regression Results for Differential Interruption Rate
Beta Change in
Predictor Values Values R2
Total Resistance to Change -.16+ .03
Differential Resistance to Change .10 .01
Cross-product Resistance to Change -.21 .00
Male/male Dyads .11 .01
Male/female Dyads .01 .00
+ .10.
Page 85
80
Table 11
Regression Results of Total Question Rate
Beta Change in
Predictor Values Values R2
Total Resistance to Change -. 39*** .15
Difference Resistance to Change -.09 .01
Cross-Product Resistance to Change -.28 .00
Male/Male Dyads .10 .01
Male/Female Dyads .02 .00
*** <.001.
Table 12
Regression Results for Differential Question Rate
Beta Change in
Predictor Values Values R2
Total Resistance to Change -.23* .05
Differential Resistance to Change -1.05 .01
Cross-product Resistance to Change .07 .00
Male/male Dyads .07 .01
Male/female Dyads .15 .01
* 2 < .05.
Page 86
81
Table 13
Regression Results for Total Ah Rate
Beta Change in
Predictor Values Values R2
Total Resistance to Change .05 .00
Differential Resistance to Change .10 .01
Cross-Product Resistance to Change 1.21 .01
Male/Male Dyads -.15 .02
Male/Female Dyads -.28 .02
Table 14
Regression Results for Differential Ah Rate
Beta Change in
Predictor Values Values R2
Total Resistance to Change .03 .00
Differential Resistance to Change .09 .01
Cross-product Resistance to Change .32 .00
Male/male Dyads -. 14 .02
Male/female Dyads -.11 .00
Page 87
82
Table 15
Regression Results for Total Silence Rate
Beta Change in
Predictor Values Values R2
Total Resistance To Change -.01 .00
Differential Resistance to Change -.04 .00
Cross-product Resistance to Change .74 .02
Male/male Dyads -. 32*** .10
Male/female Dyads .03 .00
*** 2 < .001.
Table 16
Regression Results for Differential Silence Rate
Beta Change in
Predictor Values Values R2
Total Resistance to Change -.09 .01
Differential Resistance to Change -.07 .01
Cross-product Resistance to Change .94* .04
Male/male Dyads -.17+ .03
Male/female Dyads -.04 .00
+2 < .10; * 2 < .05"
Page 88
83
REFERENCES
Allport, G. W. (1937). Personality, A psychological
interpretation. New York: Holt.
Aries, E. J., Gold, C., & Weigel, R. H. (1983).
Dispositional and situational influences on dominance
behavior in small groups. Journal of Personality and
Social Psychology, 44, 779-786.
Breckler, S. J., & Greenwald, A. G. (1986). Motivational
Facets of the self. in R. M. Sorrentino & E. T. Higgins
(Eds.), Handbook of motivation and cognition (p.
145-164). New York: Guilford Press.
Cantor, N., Hazel, M., Niedenthal, P., & Nurius, P. (1986).
title. in R. M. Sorrentino & E. T. Higgins (Eds.),
Handbook of motivation and cognition (p. 96-121). New
York: Guilford Press.
Danzinger, K. (1976) Interpersonal Communication. New York:
Pergamon Press.
Doster, J. (1975). Individual differences affecting
interviewee expectancies and perceptions of self-
disclosure. Journal of Counseling Psychology, 22, (3),
192-198.
Page 89
84
Doster, J. & Watson, W. (1987, August). The complexity of
Selves. Paper presented at the Seventh International
Congress of Personal Construct Psychology, Memphis, TN.
Duck, S. W. & Spencer, C. (1972). Personal constructs and
friendship formation. Journal of Personality and Social
Psychology, 23, 40-45.
Epstein, S. & O'Brien, E.J. (1985). The person-situation
debate in historical and current perspective.
Psychological Bulletin, 98, 513-537.
Ferguson, N. (1977). Simultaneous speech, interruptions and
dominance. British Journal of Social and Clinical
Psychology, 16, 295-302.
Fruge, E. (1973). A Multimethod investigation of
influential, dominant, or assertive behavior. Unpublished
manuscript.
Gergen, K. J. (1971). The concept of self. New York: Holt,
Rinehart & Winston.
Gough, H. C. (1957). Manual for the California Psychological
Inventory. Palo Alto, California: Consulting
Psychologists Press.
Haley, J. (1972). Strategies of psychotherapy. New York:
Grune and Stratton.
Hinkle, (1965). The change of personal constructs from the
viewpoint of a theory of construct implications.
Unpublished manuscript, Ohio State University.
Page 90
85
Jacob, T. (1975). Family interaction in disturbed and normal
families: A methodological and substantive review.
Psychological Bulletin, 82. 33-65.
Keisler, D. J. & Bernstein, A. J. (1974). A Communications
critique of behavior therapies. Unpublished manuscript.
Kelly, G. (1955). The psychology of personal constructs.
New Your: W. W. Norton.
Landfield, A. W. (1975) The Complaint: a confrontation of
personal urgency and professional construction. In ???,
(Eds.). Issues and Approaches in Psychological
Therapies. London: Wiley.
Landfield, A. W. (1975). Exploring socialization through the
interpersonal transaction group. Unpublished manuscript.
Landfield, A. W. & Barr, M. A. (1976). Ordination: a New
measure of concept organization. Unpublished manuscript.
Landfield, A. W. & Rivers, P. C. (1975). An introduction to
interpersonal transaction and rotating dyads.
Psychotherapy: Theory, Research and Practice, 12,
365-373.
Leary, (1957). Interpersonal diagnosis of personality:_a
Functional theory and methodology for personality
assessment. New York: Ronald Press.
Magnusson, D. & Endler, N. S. (1977). Interactional
Psychology: present status and future prospects. in D.
Magnusson & N. S. Endler, (Ed.), Personality at the
Page 91
86
crossroads: Current issues in interactional Psychology
(p. 3-31). New York: Wiley and Sons.
Mair, (1977a). Psychology and psychotherapy: some common
concerns. British Journal of Medical Psychology, 50,
(1), 21-25.
Mair, J. M. M. (1977b). The community of self. In D.
Bannister (Ed.), New perspectives in personal construct
psychology (p. 125-149). London: Academic Press.
Markus, H. , & Smith, J. (1981). The influence of self-
schemata on the perception of others. In N. Cantor & J.
F. Kihlstrom (Eds.), Personality, cognition, and social
interaction (p. 233-262). Hillsdale, NJ: Lawrence Erlbaum
Associates.
Mischel, W. (1968). Personality and assessment. New York:
Wiley and Sons.
Mischel, W. (1973). Toward a cognitive social learning
reconceptualization of personality. Psychological
Review, 80, 252-283.
Mishler, E. G. & Waxler, N. E. (1968). Interaction in
families: Experimental study of family processes and
schizophrenia. New York: Wiley and Sons.
Mortensen, C. C. (1972). Communication: The study of human
interaction. New York: McGraw-Hill.
Phillips, W. M. (1981). The minimax problem of personal
construct organization and schizophrenic thought
Page 92
87
disorder. Journal of Clinical Psychology, 37, 4,
692-698.
Reusch, J. & Bateson, G. (1951). Communication: the Social
matrix of Psychiatry. New York: Dorton.
Riskin, J., & Faunce, E. E. (1972). An evaluative review of
family interaction research. Family Processes, 11,
365-455.
Roger, D. B., & Schumaker, A. (1983). An evaluative review
of family interaction research. Family Processes, 11,
365-455.
Rogers, W. T., & Jones, S. E. (1975). Effects of dominance
tendencies on floor holding and interruption behavior in
dyadic interaction. Human Communication Research, 1,
113-122.
Scherer, K. R. (1979). Voice and speech correlates of
perceived social influence in simulated juries. In H.
Giles & R. St. Clair (Eds.), Language and social
psychology (]. 255-286). Oxford, England: Blackwell.
Scott, G. (1988). Permeability of selves and compliance
with therapeutic homework tasks. Unpublished manuscript,
University of North Texas.
Siegman, A. W. & Pope, B. (1968). Interviewer warmth in
relation to interviewee verbal behavior. Journal of
Consulting and Clinical Psychology, 32, 588-595.
Watzlawick, P., Beavin, J., & Jackson, D. (1967).
Pragmatics of human communication. New York: Norton.
Page 93
88
Wiens, A., Thompson, S., Matarazzo, J., Matarazzo, R., &
Saslow, G. (1965). Interview interaction behavior of
supervisors, head nurses, and staff nurses. Nursing
Research, 23, 322-329.
Wiggins, G. S. (1979). A psychological taxonomy of trait-
descriptive terms: The interpersonal domain. Journal of
Personality and Social Psychology, 37, 395-412.
Yalom, I. D. (1970). The theory and practice of group
psychotherapy. New York: Basic books.
Zimmerman, D. & West, C. (1975). Sex roles, interruptions
and silences in conversation. In B. Thorne, & N. Henley
(Ed.), Language and sex: Differences and dominance (p.
105-129). Rowley, Mass; Newbury House.