2011. Vol.2, No.7, 665-668
Copyright © 2011 SciRes. DOI:10.4236/psych.2011.27101
The Analysis of Attachment Styles through Interpersonal
Circumplex Description
Wen Xue, Shouying Zhao
School of Educational, Guizhou Normal University, Guiyang, China.
Received May 18th, 2011; revised July 23rd, 2011; accepted September 9th, 2011.
The study explored how the dyads with different attachment styles behave towards Leary’s circumplex in coop-
erative and competitive tasks. 100 strangers as a sample were surveyed through two experiments on computer,
by using RBQ (Riverside Behav ioral Q-sort) to examine thei r interpersonal traits through rated by experimente rs.
Relation Questionnaire (RQ) and ECR were also used to measure their attachment styles. There are three con-
clusions based on the results of this study that: 1) Individuals of different attachment styles in interpersonal in-
teraction had shown different trend of interpersonal styles; 2) Individuals of different attachment styles take up
different area in interpersonal circumplex, which proves that it is continuous not discrete of interpersonal at-
tachment style; 3) the dyads in different sex of different attachment types showed different satisfaction in these
two tasks.
Keywords: Interpersonal Circumplex, Attachment Style, Interpersonal Relationship, Dyads
It was mentioned in Bowlby and Ainsworth’s attachment
theory that patterns of childhood attachment have examined in
the context of adult’s experience of emotion, especially in their
relationships. This extension is consistent with Bowlby’s asser-
tion that the attachment mechanism is an impact “from the cra-
dle to the grave”. Act ual ly , the continuity of infant’s attachment
style has further effect on individual’s manifestation towards
their relationships or association (such as marriage, romantic
relation, the companionship). Therefore, most attachment re-
searches have been done by the domestic and international
researchers, combined with various composition of character
about cognition, personality or trait. So further exploring will
be continued for the correlation of this individually emotional
link and individual mental state shape, interpersonal behavior,
and social cognition and support, happiness, etc.
With the gradually in-depth social science researches on
adult interpersonal attachment, studies for attachment methods
and means of measurement have become more diversified. In
China, the main measurement of interpersonal attachment is in
two ways. In the first way, individual’s attachment style was
measured to determine the differences between individuals with
different attachment styles; and the Ladder is the determination
of the one’s score on attachment dimensi ons. Individuals’ scores
on different dimension will be used to determine the extent of
their attachment avoidance and anxiety and so on. Question-
naires are highly applied in researches concerning both the
degree in dimension structure and attachment styles. Which is
used the most is the type of Adult Attachment Styles Question-
Early questionnaire survey towards adult attachment was fo-
cused on the discrete types (Hazan & Shaver, 1987), which was
developed by Ainsworth’s three infant attachment types. Since
then, a large number of adult attachment scales were emerging,
greatly promoting the research progress in this field. According
to Bowlby’s theory, the individual was referred to the self-
model and other-model, because individual’s attachment style
depends on their own positive or negative assessment about
themselves and their supporters (especially parents). So the
adult attachment styles based on the integration of these two
models, can be divided in four-categorical scheme, such as
secure, ambivalent, fearful and dismissing. Therefore, the sub-
sequent establishment of attachment scale towards two dimen-
sions—self-model and other-model, became the core content
discussed by the researchers. The individual with secure at-
tachment of cherish both the positive self-evaluation and
other-evaluation, and they want others to response their own
emotional needs; compared with that, ambivalent attachment
style owes negative self-model, whereas of positive other-model;
The individuals of dismissing attachment style possess positive
self-model, but they believe that others are not worth trusting;
The fearful attachment’s self-model and the other-model were
both negative. Based on this theoretical analysis, Bartholomew
and Horowitz developed a self-report scale—Relationship Ques-
Since then, based on the RQ, Brennan developed a standard
adult intimate relationships Attachment Scale—Closed Relation
Experience scale (ECR), a scale of 36 questions, in which, at-
tachment anxiety and attachment avoidance as its the two di-
mensions composition. Currently, Attachment investigation
would be confirmed by of the combination of these two item
polls or they may use the ECR to determine the subjects’ style
while the RQ as the crit er i on for test.
However, recent research suggests that rather than a three-or
four-categorical, a dimensional, more comprehensive model
may better represent adult attachment (Crowell, Fraley, &
Shaver 1999). Hazan and Shaver (1987) found that “Secure
attachment tend to describe their parents friendly and warm,
whereas Anxiety-Ambivalent attachment of the individuals
described their families aggressive, and avoidant person tend to
report their par ents cold and rejectin g”. Therefore, it is no doubt
the measurement method based on interpersonal relationship
can be applied to the adult attachment study. For example,
Shaver and Brennan (1992) examined the association of at-
tachment with the Big Five personality model (FFM). Accord-
ing to their study, “more securely attached participants reported
higher extroversion and agreeableness. Contrarily, those indi-
viduals with anxious attachment style more likely to report
lower agreeableness, and higher neuroticism”.
Therefore, preliminary research, at least, proofed that adult
attachment style, to some extent, is related to the individuals’
interpersonal traits or personality, interpersonal communication
and social interaction. That is why insecure attachment person
may show a hostile or compliance attitude in their interpersonal
communication with others. The Figure 1 indicates the relations
between four attachment styles and two dimensions—inter-
personal anxiety and avoidance. It can also be observed from
this figure that the continuous space between these four at-
tachment styles is not mutually discrete separate, but some
among them have not been explored. The independence of
these four styles gives us some illusion that they are separate.
Indeed, participates are distributed around all the space, even
the area between two lines of canonical four styles. That is the
reason why some individuals with Preoccupation attachment
style would be changed into Secure after they fell in love or got
marriage, since they may locate originally between these two
In social interaction, a successful communication is always
attributed to the interpersonal behavior of reciprocity or com-
plementarity of two parnters. Established by this theory, Inter-
personal Circumplex can accurately predict the successful
maintaining rate of a relationship between two peoples. The
IPC is divided by two dimensions, control and affiliation, into
eight parts. They are assured-dominant (PA), in the anticlock-
wise order, arrogant-calculating (BC), cold-hearted (DE), aloof-
introverted (FG), unassured-submissive (HI), unassuming-inge-
nuous (JK), and warm-agreeable (LM), and (NO) octants (see
Figure 2). IPC attempts to describe and explain the character of
a person, especially his or her interpersonal style. Carson (1969)
suggested that only the Agreeableness dimension is similar to
the position in the leading position of the interpersonal dimen-
sion of the two sides complement each other to attract each
The two dimensions of The FFM that—extraversion and
agreeableness—are very similar to the affiliation and domnance
Figure 1.
The relation between attachment style and two dimensions—inter-
personal avoidance and anxiety.
Figure 2.
Learys interpersonal circumplex.
of IPC, respectively (McCrae & Costa, 1989). Therefore, utili-
zing the previous theory, we intend to adopt experimental tasks
with the questionnaire method, in order to examine the descrip-
tion of Leary’s interpersonal Circumplex model for interpersonal
attachment of dyads with opposite sex in different tasks. And
explore the association of individual's interpersonal attachment
and interpersonal traits, while analysis the satisfaction of dyads
with different attachment styles matching in different tasks.
Subjects were 50 men and 50 woman students at several
schools in Guizhou, some of whom came from high school, and
others are graduates or undergraduates. They were ranged in
age from 16 ~ 30. They were matched in pairs in random, and
each pair contains a man and a woman. There are two experi-
mental situations, the cooperative situation and competitive
situation, in each of which 25 dyads participated.
Performance task
In cooperative situation, each dyad was surveyed through an
experiment of detective game on computer, in which 13 things
were asked to be found by their cooperation. And in compete-
tive situation, each dyad was surveyed through an experiment
of combat game on computer, in which they should use their
skills or ability to defeat their partner.
After telling the instruction about these two tasks, the experi-
menters ran the detective or combat FLASH game on com-
puter, and asked the opposite-sex dyad to complete it together.
Each interaction lasted approximately 3 min, while the experi-
menters used RBQ (Riverside Behavioral Q-sort) to examine
their interpersonal traits.
Results and Analysis
The Distribution of Individuals with Different
Attachment Types in IPC
By the end of the experiment, subjects should complete the
W. XUE ET AL. 667
self-report scale RQ and ECR also, in order to determine their
attachment styles. There are 43 people of secure attachment, 15
of fearful attachment, whereas 22 subjects of ambivalent at-
tachment, and 20 with dismissing style. Then the subjects’ at-
tachment type was tested with their highest scores in the eight
subscales of RBQ through correlation analysis. Every subscale
would represent a trait in the circumplex, the pie charts are
shown as follow (see Figure 3).
Through the Figure 3, subjects with secure attachment tested
in two tasks mainly reflects a highly-frequency performance in
PA, JK LM, and NO. Among other three insecure attachment
styles, it is easy to find that the interpersonal performance of
ambivalent-attachment individual is similar to the secure one,
but the proportion of each trait is distinguished with it. Fearful
attachment were tested mainly in interpersonal style of BC, DE,
and FG, which is the most distinguished on compared with the
secure style ; while the subjects of dismissing attachment got
high scores in FG and JK.
In order to exclude the possibility that extreme data would be
generated by the highest scores, and to further explore the rela-
tionship between variables about the 4 types of attachment and
the 8 interpersonal traits in IPC octant, we tested the data of
100 individual of these two tasks through Pearson two-tailed
test (Table 1).
It can be well informed in Table 1 that subjects with secure
attachment style (A) was more obvious in the performance of
BC, PA, NO, LM in interpersonal interaction, which shows
significantly positive correlation in PA, but shows a signify-
cantly negative correlation with the HI. This matches the hy-
pothesis that highly positive self-model is the main character of
secure individuals. When it comes to fearful attachment style
(B), the situation may become just opposite, which shows posi-
tive relation towards the traits of BC, DE, FG, HI and a lower
scores on the affiliation dimension of the circumplex, while a
higher score in complaisance, because individuals under this
style take a part in negative self-model and other-model. Sub-
jects with ambivalent attachment take a positive correlation
with BC and DE two traits. Although it may indicate a conflict-
tion against the highest-score pie chart that DE takes no evi-
Figure 3.
The highest scores distribution of RBQ in different attachment style
individuals. (Pies: 1-secure; 2-fearful; 3-ambivalent; 4-dismissingTrait:
from 1-PA to 8-NO).
Table 1.
The correlation coefficent of eight interpersonal traits and attachment
styles in two tasks.
A 0.27** 0.23* –0.07 0.18 0.27** 0.09 0.140.17
B –0.13 0.13 0.013 0.31** 0.03 0.01 0.06 0.09
C 0.13 0.04 0.089 0.03 0.01 0.03 0.070.10
D 0.01 0.025 –0.22* 0.04 0.01 0.09 0.020.01
dence above this attachment style, it can be explained that this
characteristics is due to this type of attachment and subjects’
average score on this trait are not low, but this is not their high-
est score characteristics. What should be a surprise is that the
dismissing attachment shows similarity with the secure on.
Since individuals with this type although evaluate others nega-
tive or not worth trusting, a positive self-model was formalized
in their inner mechanisms of attachment experience in early
stage. So they may also get a not low score on control dimen-
sion, however, they may cherish less ability of an agreeable
nature, which can be indicated on the decreased correlation
coefficient compared with the secure attachment, because their
other-model are negative.
To a contrary, Figure 4 provides a clear distribution of dif-
ferent attachment styles’ location in the circumplex. And each
style is depicted as a consecutive manner in the circumplex,
which proofs the continuum of attachment style shown in hori-
zontal and vertical dimensions. Among them, three types of
attachment are in continuous interpersonal performance, with-
out being interrupted, whereas ambivalent attachment person
take a break in the circular model. In this interpersonal process,
individuals belonged to ambivalent attachment style, tended to
be divided into two forms, one of which relatively came closer
to fearful attachment, and the other form performed relatively
as the secure one, which also well confirmed the assumption
that it is the continuity rather than separation should be the
nature of attachment styles. But whether it is close to the fearful
attachment, or secure attachment, it showed high compliance
more than the dominance on the control dimension. In our daily
life it often seems as manifestation as lacking of confidence
when put it in social relation. This should be attributed to the
negative self-evaluation model of ambivalent attachment.
The Satisfaction of Dyads with Different Attachment
Styles in Cooperation Task
Calculating through variance analysis, the statistic signify-
cance (Sig) = 0.018 < 0.05 (Table 2), indicated that, between
dyads with the different attachment-style matching, at least one
group showed significant difference towards other groups. As
shown in the satisfaction mean picture (Figure 5), both partners
in the dyads were secure attachment showed the most coopera-
tion satisfaction in this interaction, followed which is the dyads
Figure 4.
The description of four attachment styles in IPC (1-secure; 2-fearful;
3-ambivalent; 4-dismissing).
with only one secure attachment individual, and dyads with
both are insecure attachment took the lowest score compared
with the other two group.
The Satisfaction of Dyads with Different Attachment
Styles in Competition Task
Similar to the cooperative task, the statistic significance (Sig)
in competitive task is 0.018 < 0.05 (see Table 3), which also
indicated at least one group showed significant difference to-
wards other groups. We can be well informed through the linear
figure (Figure 6), that dyads of secure attachment showed sig-
nificantly higher satisfaction than dyads with insecure satisfac-
tion. But compared to cooperative tasks, the scores between the
two groups with insecure attachment style, the difference is not
distinguished. In cooperation task, to some kind of degree, in-
crease peoples’ satisfaction when it comes to exact social asso-
ciation. Figure 6.
The average score of competitive satisfaction degree towards dyads in
different attachment styles. (1-dyads of secure; 2-dyads with only one
insecure partner; 3-dyads of unsecure).
Three conclusions based on the results of this experimental
Table 2.
Variance analysis for cooperative satisfaction degree towards dyads in
different attachment styles.
Sum of squares DfMean of square s F Sig.
Between groups 10.183 2473.895 3.2560.018
In group 23.257 22145.524
Sum 33.440 24
study can be showed that: 1) Individuals of different attach-
ment styles in interpersonal interaction had shown different
trend of interpersonal styles; 2) Individuals of different at-
tachment styles take up different area in interpersonal circum-
plex, which proves that it is continuous not discrete of inter-
personal attachment style; 3) the dyads in different sex of dif-
ferent attachment types showed different satisfaction in these
two tasks. In cooperative task, dyads of secure attachment style
took the highest satisfaction, and of insecure attachment style
with the lowest one, while in competitive task, it is not distin-
guished between dyads of insecure attachment style and dyads
with only one insecure partner.
Gallo, L. C. et al. (2003). An interpersonal analysis of adult attachment
style: Circumplex descriptions, recalled developmental experiences,
self-representations and interpersonal functioning in adulthood.
Journal of Personalit y, 71, 141-182.
Li, T. G. (2005). Attachment and memory for emotional events. Jour-
nal of Peking University, 41 , 965-974.
Ansell, E. B., Kurtz, J. E., & Markey, P. M. (2008). Gender differences
in interpersonal complementarity within roommate dyads. Personality
and Social Psychology Bulletin, 34, 502-512.
Gurtman, M. B., & Pincus, A. L. (2000). Interpersonal adjective scales:
Confirmation of circumplex structure from multiple perspectives.
Personality and Social Psychology Bulletin, 26, 374- 384.
Markey, P. M., Funder, D. C., & Ozer, D. J. (2003). Complementarity
Figure 5.
Average score of cooperative satisfaction degree towards dyads in
different attachment styles.
of interpersonal behaviors in dyadic interactions. Journal of Person-
ality and Social Psychology, 29, 1082-1090.
Mayseless, O., & Scharf, M. (2007). Adolescents’ attachment repre-
sentations and their capacity for intimacy in close relationships.
Journal of research on adolescence, 17, 23-50.
Table 3.
Variance analysis for competitive satisfaction degree towards dyads in
different attachment styles.
Sum of squares DfMean of square s F Sig.
Between groups 11.248 2473.895 4.6670.020
In group 26.512 22145.524
Sum 37.760 24
Sadler, P., & Woody, E. (2003). Is who you are who you’re talking to?
Interpersonal style and complementarity in mixed-sex interactions.
Journal of Personality a nd Social Psychology, 84, 80-96.
Sullivan, H. S. (1953). The interpersonal theory of psychiatry. New
York: Norton.
Zimmermann, P. (2004). Attachment representations and characteristics
of friendship relations during adolescence. Journal of Experimental
Child Psychology, 88, 83-101. doi:10.1016/j.jecp.2004.02.002