ArticlePDF Available

Personality and Roommate Friendship in Chinese Culture


Abstract and Figures

This study examines the relationship of personality traits to mutual friendship among assigned roommates. It was designed to compare actual and perceived personality similarity and of personality desirability on friendship outcomes in a Chinese cultural setting. Hong Kong university students rated their own and their roommate’s personality on a comprehensive, indigenous measure six months after being paired to room together in the campus dormitories. Wright’s (1991) Acquaintance Description Form was completed by each roommate to assess their friendship. Those roommate pairs where each rated the friendship high were compared with those where each rated the friendship low. The decisive personality variable associated with mutual friendship was the roommate’s perceived higher level on four of the eight desirable dimensions of Chinese personality, namely helpfulness, intellect, openness to experience, and extroversion. There were no associations to friendship for the other roommate’s actual self-ratings on any of these eight dimensions; nor was friendship related to roommate similarity in personality, whether actual or perceived. These results for perceived roommate desirability in personality are discussed in light of related Western research and the constraints which apply to the assigned roommate relationship.
Content may be subject to copyright.
Personality and roommate friendship in
Chinese culture*
Royce Yat-Pui Lee and Michael Harris Bond
Chinese University of Hong Kong, China
This study examines the relationship of personality traits to mutual friendship
among assigned roommates. It was designed to compare actual and perceived
personality similarity and of personality desirability on friendship outcomes in a
Chinese cultural setting. Hong Kong university students rated their own and
their roommate’s personality on a comprehensive, indigenous measure six
months after being paired to room together in the campus dormitories. Wright’s
(1991) Acquaintance Description Form was completed by each roommate to
assess their friendship. Those roommate pairs where each rated the friendship
high were compared with those where each rated the friendship low. The
decisive personality variable associated with mutual friendship was the
roommate’s perceived higher level on four of the eight desirable dimensions
of Chinese personality, namely helpfulness, intellect, openness to experience,
and extroversion. There were no associations to friendship for the other
roommate’s actual self-ratings on any of these eight dimensions; nor was
friendship related to roommate similarity in personality, whether actual or
perceived. These results for perceived roommate desirability in personality are
discussed in light of related Western research and the constraints which apply to
the assigned roommate relationship.
There has been little research on friendship in Chinese culture compared to the West
(Goodwin & Tang, 1996), perhaps because of cultural factors in individualistic cultures
which stimulate interest in interpersonal attraction (Hogan & Emler, 1978). This paper
attempts to redress this imbalance by looking at personality variables associated with
friendship in initially unacquainted roommates. It is designed to assess the relative impact of
similarity versus desirability effects, both actual and perceived, in predicting a state of
mutual attraction.
Similarity effects
The issue of similarity–attraction has gained so much attention in the study of interpersonal
relationships that it is regarded by some as the most well-established finding in social
psychology (Berscheid & Walster, 1983; Fishbein & Ajzen, 1972). Despite their renown,
Byrne’s classical studies of attitude similarity–attraction have also attracted criticism
ßAsian Journal of Social Psychology 1998.
Published by Blackwell Publishers, Oxford, UK
Asian Journal of Social Psychology (1998) 1: 179–190
*Authors’ note. Address all communications to Michael Harris Bond, Department of Psychology,
The Chinese University of Hong Kong, Shatin, NT, Hong Kong, China. E-mail:
(Rosenbaum, 1986; Sunnafrank, 1983; Sunnafrank & Miller, 1981). Studies attempting to
validate the notion of similarity–attraction in real-life situations have suggested that the
Byrne results were merely an experimental artifact, lacking external validity (Sunnafrank,
In response to this challenge, Byrne (1992) has maintained that what has been
demonstrated in the laboratory is just a simple relation between similarity and attraction with
all other factors controlled. In a real-life situation, this similarity effect is subjected to the
influence of many other variables. Moreover, Byrne (1992) also asserted that ‘‘Attitudes
were never considered to be the only, the most pervasive, or the most powerful determinant
of attraction’’ (p. 192).
Actual vs. perceived similarity
In explaining the findings of his attitude studies, Byrne (1971) proposed the notion of
consensual validation involving a reinforcement-affect model. According to this reasoning,
it is reinforcing to interact with someone who shares the same attitude, as through this
process a person’s own view is confirmed and validated, thereby boosting self-esteem and
consequently attraction toward the similar other.
Byrne’s major focus in his early studies was attitude similarity (Byrne, 1971; Byrne &
Griffitt, 1973). In most of Byrne’s (1971) studies, subjects were presented with a fictitious
person’s attitude profile which was generated from the subject’s actual attitude ratings. The
subject’s perceptions of the stranger were thereby shaped by the experimenters’ input and
corresponded closely to actual attitude similarity. Perhaps it was for this reason that Byrne
did not distinguish perceived and actual similarity effects in his studies. However, in his
attempt to solve the controversy between the similarity–attraction vs. repulsion hypothesis
(Rosenbaum, 1986), Hoyle (1993) demonstrated that whether ‘‘actual’’ similarity produces
attraction depends upon the subject’s perception of that similarity.
There is typically, however, a considerable discrepancy between self- and peer
perceptions of a target’s attitudes, values, beliefs, and personality (e.g., Newcomb, 1961;
Robbins & John, 1997). Unless actual and perceived ratings can be proved to closely agree,
it becomes necessary to investigate how each of them separately relates to friendship or to
any other interpersonal outcome.
While most researchers have failed to acknowledge the difference between perceived
and actual similarity, those who have done so tend to downplay the importance of perceived
effects (Duck, 1973a). When incompatible results are found, perceived similarity is often
interpreted as the consequence of ‘‘distorted’’ perceptions and so judged as inaccurate.
Byrne’s (1971) portrayal of consensual validation, however, seems to support the
importance of perceived rather than actual similarity. That is, in order to induce
attraction, similarity needs to be reinforcing (Duck & Barnes, 1992), and this outcome
relies on the person’s subjective perception of this similarity. Supporting evidence can be
traced in various studies (Curry & Emerson, 1970; Hill & Stull, 1981; Newcomb, 1961). In
his study of friendship development of previously unacquainted residents in a college
boarding house, Newcomb (1961) showed that only perceived value similarity was related to
reports of liking at the initial stage. Over time, actual value similarity became more
important, although weak, and it was always less important than the effect of perceived
similarity. Parallel findings have been reported by Hill and Stull (1981) in their longitudinal
study of same-sex friendship.
180 Royce Yat-Pui Lee and Michael Harris Bond
Thus, in order to examine the contribution of similarity effects in real-life friendship
development, both actual and perceived similarity effects will be investigated in this study.
To our knowledge such a comparative study has only been undertaken once for the domain
of personality (Wetzel et al., 1979), who derived a single index for the whole domain of
personality. As we discuss below, the distinguishable domains of personality need to be
distinguished when friendship is to be explained.
Similarity in what respect?
People in their daily lives are exposed to abundant information about others and it is both
impractical and unlikely for friends to be similar in every aspect. Thus, the question of
whether similarity in specific dimensions is associated with relationship development
becomes an important issue in current interpersonal research.
It is argued that only those constructs which are meaningful to the individual have an
impact on relationships. As Duck (1994) put it, ‘‘Partners do not need a full picture of each
other, they only need a workable one’’ (p. 100). In certain relationships, similarities in some
aspects have stronger effects than those in other relationships. Lea & Duck (1982), for
instance, demonstrated that friend pairs, when compared with nominal pairs, were more
similar in important and uncommon values than in rejected, neutral, or common values. It is
argued that these important and uncommon values are more decisive for friendship, as they
provide unique and irreplaceable rewards during the process of social exchange.
Unfortunately, previous studies (e.g., Duck, 1973b) have tended to obtain a global
measure of similarity, with all the dimensions of one type of measure, say personality,
collapsed together (LaPrell et al., (1990) is a welcome exception). In doing so, the
interpretation of the results is simplified but the implications of specific dimensions for
relationship development remain unknown. Therefore, in order to understand how and why
similarity effects influence friendship development, it is instructive to examine the
associations of friendship with specific domains of personality similarity.
Effects of desirable personality characteristics
Would an individual possessing high levels of ‘‘desirable’’ characteristics such as
helpfulness, caring, sincerity, and honesty be more attractive to a given person than an
individual who was merely similar to that given person in these respects? Commonsense
knowledge and academic findings (Fishbein & Ajzen, 1972; Posavac & Pasko, 1974;
Stalling, 1970) both suggest a positive answer. Posavac and Pasko (1974) found that
American college students were not only attracted to those who were similar to themselves
but also reported greater attraction to strangers who possessed socially desirable
characteristics. In their study of ideal intimates, Cheng et al. (1995) found that Chinese
college students prefer their ideal female friends to be high in helpfulness and their ideal
male friends to be high in extroversion and application, regardless of the respondent’s own
level on these traits.
It is argued that people are generally reward-seeking and punishment-avoiding creatures
who try to maximize their rewards and minimize their punishments. Possession of positive
personality traits may make their holders attractive irrespective of similarity. However, this
issue of desirable personality effects has received scant attention in the study of
interpersonal relationships.
Friendship in Chinese culture 181
In this study, the term ‘‘desirability effect’’ refers to the positive association of a
particular personality characteristic with achieved friendship. Like similarity effects,
desirability effects can be separated into ‘‘actual’’ and ‘‘perceived’’ effects based on self-
reported ratings and other-perceived ratings, respectively. It is hoped that by including both
actual and perceived assessments of personality, a more accurate picture of how personality
desirability is associated with friendship will emerge.
The present investigation
A study of friendship in previously unacquainted, same-sex roommates will be conducted.
Subjects’ self-perceived and perceived roommates’ personality will be measured by using a
comprehensive, lexically derived measure of personality: the Sino-American Person
Perception Scale (SAPPS) (Yik & Bond, 1993). Wright’s Acquaintance Description Form
(Wright, 1985) will be used to assess mutual friendship strength.
Self–Self Similarity (S–SS). Although widely used, the term ‘‘actual’’ similarity is
somewhat misleading. When compared with the term ‘‘perceived’’, the term ‘‘actual’
carries the unspoken connotation of real. In fact, in most studies the so-called ‘‘actual
similarity’’ is calculated from the difference between two individuals’ self-ratings (Curry &
Emerson, 1970; Hill & Stull, 1981). However, there is no evidence, either empirical or
theoretical, to suggest that self-reports are more accurate than perceived reports. Therefore,
it is inappropriate to label ‘‘actual similarity’’ as superior to or more real than the perceived
similarity. Following this argument, actual similarity is named Self–Self Similarity(S–SS) so
as to provide a more neutral, less presumptuous label for this measure.
Self-Perceived Other Similarity (S-POS). Along with the above change, ‘‘perceived’’
similarity is labelled Self-Perceived Other Similarity (S-POS) to establish a parallel naming.
It refers to the difference between one’s self-rating of personality and the target’s perceived
personality rated on the same dimension. Although some linkages may exist between the S-
POS and S-SS, it is speculated that they may reflect different psychological processes.
Therefore, separate analyses will be conducted to unveil their different relationship to
Self-Rated Desirability (S-RD) and Perceived Other Desirability (POD). Following the
categorization of similarity effects, desirability is divided into two categories. The Self-
Rated Desirability (S-RD) refers to the target’s self-perception on a particular domain, while
Perceived Other Desirability (POD) indicates the subject’s perception of his or her
roommate’s character.
A comprehensive measure of personality will be used in this study. It taps eight dimensions
of personality variation in Hong Kong Chinese, namely, emotional stability, openness to
experience, extroversion, helpfulness, application, restraint, assertiveness, and intellect (Yik
182 Royce Yat-Pui Lee and Michael Harris Bond
& Bond, 1993). The participants in this study are previously unacquainted college
roommates, so that the personality dimensions relating to subsequent friendship in this study
may be channelled by the nature of the roommate relationship in Hong Kong universities,
specifically that roommates do not choose one another and must occupy the same living
space for a whole academic year regardless of their emerging feelings for one another.
Similarity effects. Extroversion relates to activity preferences (Costa & McCrae, 1992).
Consequently, we expect that matching on this dimension of personality perception will lead
to greater friendship at the end of the first six months of living together. Openness to
experience relates to political/social ideology along the liberal versus conservative
dimension (Trapnell, 1994), so we likewise expect similarity effects for this dimension.
Desirability effects. We expect that the communal (Bakan, 1966) traits of helpfulness and
restraint should be valued regardless of the subject’s own level of that trait.
Perceived vs. actual effects. We expect that perceived effects whether for similarity or
desirability will be stronger than actual effects, as has been found in the rare, previous
comparisons (e.g., Newcomb, 1961).
All participants were same-sex college roommates sharing a double room in one of the
campus dormitories of either the Chinese University of Hong Kong or the Hong Kong
University. Only pairs who were unacquainted before being assigned to room together in
September were recruited. One hundred and thirty-one pairs of roommates completed the
questionnaires in early March: 46 male pairs, 85 female pairs. Their ages ranged from 18 to
25 years.
All the measures used in this study were in Chinese, the subjects’ native language. The
Chinese version of the Acquaintance Description Form (ADF) (Wright, 1991), and the Sino
American Person Perception Scale (SAPPS) (Yik & Bond, 1993) were adopted. The first
was a translated version of the original English instrument, checked by back translation,
while the latter was an indigenously derived questionnaire.
Participants were required to answer the ADF and the SAPPS corresponding to their
judgment of the relationship and of their own personality traits, respectively. Participants
were also required to rate their perception of their roommates’ personality traits using the
same measure. In both cases, the sequence of the questionnaires was randomized and stapled
together to avoid possible order effects.
Acquaintance Description Form (ADF-F2). The Chinese translated ADF-F2 was
employed as the dependent measure. The translation process involved a bilingual student
translating the original English version of ADF-F2 into Chinese and the translation of the
Chinese version back into English by another bilingual. In order to produce a culturally
Friendship in Chinese culture 183
equivalent instrument, necessary modifications were made by the two authors consulting
together on the discrepancies.
Fourteen scales in the ADF-F2 were included to capture what Wright (1991) regarded as
the most important aspects of friendship. Each scale consists of five items measured on 7-
point Likert-type scales. Although participants had to complete all 14 scales, only ten
appeared to be important to the present investigation. These ten scales, collectively called
‘basic scales’’ by Wright (1991), included two friendship strength scales (Voluntary
Interdependence and Person Qua Person), five scales of different rewards perceived in a
relationship (Affirmation Value, Utility Value, Ego Support Value, Stimulation Value,
Security Value), two maintenance scales measuring tension or strain (Maintenance
Difficulty Personal and Maintenance Difficulty Situational), and a global General
Favorability factor (Wright, 1991).
Sino American Person Perception Scale (SAPPS). The short form of SAPPS, with 32
bipolar, 7-point adjective scales, was adopted. Originally developed from both the Western
Five Factor Model (McCrae & Costa, 1987) and indigenous Chinese adjective checklists
(Yang & Bond, 1990), the SAPPS has demonstrated its reliability and comprehensivness in
tapping Chinese personality perception in various studies (Cheng et al., 1995; Yik & Bond,
1993). This instrument has also been validated for measuring both self- and other perception
in Chinese culture (Bond & Shiu, 1997; Weinreich et al., 1996).
This study was carried out in 13 undergraduate dormitories at the Chinese University of
Hong Kong and in three undergraduate dormitories of the Hong Kong University.
Two weeks before the data collection, participants received a letter describing the study
and were informed that an appointment would be made within one week to confirm their
availability. Participants were contacted after two days and appointments were arranged
within a week once it had been established that the roommates had been unacquainted in
September, six months previously.
Participants rated their own personality and their friendship with their roommate. They
were also required to rate their roommates’ personality, ratings which they were not to
disclose to their roommates. Participants were required to seal their completed
questionnaires with a sticker provided and to sign their names. Souvenirs were presented
to each participant upon collection of their completed questionnaire.
Scale composition
Two categories of indices were composed. The first relates to friendship strength and the
other concerns the two different types of similarity effects and desirability effects.
Identifying Chinese friendship strength. For the sake of simplicity and clarity of the
analyses and latter interpretation, it was necessary to reduce the items from the original ten
sub-scales into a smaller number of factors. Furthermore, as it is possible that the Chinese
184 Royce Yat-Pui Lee and Michael Harris Bond
may have a different understanding of friendship, thereby producing a different grouping of
the ADF items, an exploratory factor analysis was performed. First, all the 70 ADF items
were subjected to Exploratory Factor Analysis (EFA) using the maximum likelihood
estimation, with a varimax rotation followed to identify orthogonal dimensions. A two-
factor model emerged, with 41% of the total variance explained.
Close inspection of the structure matrices revealed that the first factor was defined by
items originally residing in the Relationship Rewards, Person Qua Person, Involuntary
Dependence, and other positive aspect scales of the ADF. Therefore, it was named
Friendship Strength (FS). The second factor consisted of a few items concerning relationship
maintenance problems and relationship differentiation, which do not have a strong
connection to the present investigation. The second factor was therefore eliminated from
further consideration.
One FS index was composed by summing all the items in FS with loadings higher
than .4. Altogether, 41 items were included in FS which showed a Cronbach alpha above
Similarity indices. Similarity indices were generated for the eight dimensions tapped by
the Sino American Person Perception Scale (SAPPS): the Self–Self Similarity (S-SS) indices
and Self-Perceived Other Similarity (S-POS) indices.
S–SS These indices are the absolute difference between one’s self and his or her
roommate’s self-ratings on each personality dimension. Therefore, a lower S–SS
score on a particular dimension indicates a smaller discrepancy between
roommates’ self-perceptions on that dimension, i.e., a more similar profile.
These indices were all characterized by the prefix S–SS denoting Self–Self
Similarity, and followed by the construct (e.g., S–SS Application).
S-POS The other set of similarity indices measures the absolute difference between the
individual’s self-perception and his or her perception of his or her roommate on a
particular dimension. As above, a lower S-POS score indicates a higher perceived
similarity. Each such index begins with S-POS, denoting the Self-Perceived Other
Similarity on a given personality dimension.
Psychometric properties
The means and reliability coefficients of the SAPPS trait dimensions varied between .66 and
.82, consistent with previous studies and acceptable for four-item scales.
Friendship reciprocity: dyadic level analysis
Two special types of friendship were investigated: Reciprocated Friendship (RF) and
Reciprocated Non-Friendship (RNF). Using the median split method, participants were first
classified into two types (the high and low FS groups) based on their Friendship Strength.
Roommate pairs both reporting strong friendship strength (high FS) were categorized as RF,
while pairs both reporting a weak friendship strength (low FS) fell into RNF. The
Unreciprocated Friendship pairs were excluded. Fifty-six and 60 participants were included
in RF and RNF respectively.
Friendship in Chinese culture 185
As the main objective in this section focuses on the comparisons between the RF and
RNF, a 2 2 (two levels of friendship and sex) analysis of variance (ANOVA) was
employed for each of the eight trait dimensions. Variables with pvalues above .01 were
reported so as to provide a conservative estimate of different relationships.
Self–Self Similarity (S–SS). There were no differences between the actual similarity scores
of RF and RNF pairs.
Self-Perceived Other Similarity (S-POS). There were no differences between the perceived
similarity scores of RF and RNF pairs.
‘‘Actual’’ desirability effects. There were no differences between the self-ratings of one’s
partner in RF as compared with RNF pairs.
Perceived desirability effects. Of the eight personality traits, four exhibited significant
group differences in the perceived ratings of one’s roommate: openness, F(1, 112) =
7.48, p<.01, extroversion, F(1, 112) = 7.32, p<.01, helpfulness, F(1, 112) = 34.2, p<
.001, and intellect, F(1, 112) = 14.9, p<.001). Members of RF pairs thus perceived
their roommates to be more open, extroverted, helpful, and intellectual than did
members of RNF pairs.
The means of the ratings of the roommates on the eight SAPPS dimensions are
summarized in Table 1.
The results of this study are clear: with respect to personality ratings, rated similarity did
not predict mutual friendship. Instead, it was four of the socially desirable personality
dimensions of the other that were associated with consensual friendship after six months
of acquaintance. These desirable traits, however, are those traits as rated by the
perceiver, not self-ratings of those traits provided by the target. These results are
discussed below.
Table 1 Means of P-OD indices for reciprocated friendship (RF) and reciprocated non
friendship (RNF)
Reciprocated Friendship Reciprocated Non-Friendship
Application 4.64 4.66
Assertiveness 4.60 4.28
Emotional stability 4.42 4.06
Extroversion 4.57 3.77*
Helpfulness 4.97 3.92**
Intellect 4.81 4.22**
Openness 4.31 3.63*
Restraint 4.45 4.04
Note N = 116.
*p<.01 two-tailed; ** p <.001, two-tailed.
186 Royce Yat-Pui Lee and Michael Harris Bond
Similarity vs. desirability effects
In this social context, it was not similarity but rather certain desirable qualities as perceived
in one’s roommate that were related to mutual friendship. Chinese roommates appear to
optimize; they do not match. Of course, other studies have showed similarity effects for
personality in actual friendship formation (Duck, 1973b; Kandel, 1978; Wetzel et al., 1979).
Duck (1973b) found that broad similarity in CPI ratings was associated with initial
friendship in unacquainted pairs. However, these friends were not initially thrown into a
roommate relationship but were free to associate as reinforcements presented themselves.
When one must live with a roommate, the other’s reward value with respect to key socially
desirable traits may make the difference in supporting a friendship.
On the other hand, the present results may be integrated with those from earlier
American work on attraction and similarity of the perceived partner to the personality of
the ideal self. In a series of studies (LaPrelle, et al., 1990; Wetzel et al., 1979; Wetzel &
Insko, 1982), researchers have found that similarity of one’s ideal self to the partner’s
perceived character is a more powerful predictor of liking than is the similarity of the self
to the perceived character of one’s partner. As in this study, similarity of self and
perceived other was of little comparative moment. LaPrelle et al. (1990) concluded:
‘There is no evidence in these studies of any effect for self similarity beyond the
tendency for ideal similarity to relate to liking and for self to resemble ideal’’ (p. 238).
Previous results supporting the similarity of self to other may be explained as special
cases involving these two principles.
The effects of perceived desirability found in this study may derive from the association
of highly desirable ratings of the other with smaller discrepancies between ratings of the
ideal self and the highly regarded, perceived other. If that is the case, attraction to the other
would appear to be based upon some cognitive consistency mechanism (e.g., Rosenberg,
1960). If, on the other hand, desirability alone is the decisive factor, some reinforcement
explanation, based upon the rewards provided by the desirable qualities of the other, would
seem to be indicated.
Obviously, that speculation is an empirical question to be resolved in future research.
Only one study to date (Wetzel et al., 1979) has directly compared desirability effects with
those from ideal similarity. The authors found that ideal similarity was more strongly
associated with roommate liking.
However, given differences in outcome measures used, groups, roommate assignment
procedures, and cultural contexts, any conclusions about the role of desirability effects in
Hong Kong must await a direct, local comparison of these two relationship constructs.
Which dimensions of personality? The present results for desirability do not, of course,
allow conclusions about causality. Perceived qualities in the other do not necessarily lead to
mutual friendship. It is, however, instructive to attend to the particular qualities in the other
which are associated with this shared bond six months after first meeting.
These roommates did not choose one another; they may, however, choose how intensely
and widely to associate once they become roommates. In this sense, some filtering (Duck &
Craig, 1978) may be involved in achieving a reciprocated friendship. Perceptions of one’s
roommate as high in openness, intellect, helpfulness, and extroversion may set in motion
cycles of behavior which promote and sustain friendship. Without these perceptions to guide
behaviors toward the roommate, the relationship is ignored, with affect minimized and
contact limited.
Friendship in Chinese culture 187
A roommate’s perceived helpfulness has obvious implications for building the
interdependencies and mutual support which characterize friendship, especially in a
collective culture (Goodwin & Tang, 1996). Perceived extroversion signals a willingness to
associate with enthusiasm and seems to facilitate friendship development regardless of the
partner’s level on this personality dimension. High levels of intellect and openness to
experience are positively valued cognitive qualities by Chinese students and the results for
these variables suggest that friendships in this environment may be based upon maximizing the
rewards one receives from these qualities in one’s roommate.
The remaining four personality dimensions of application, restraint, assertiveness, and
emotional stability showed no effects on roommate friendship in this study, even though
they are also socially desired qualities (Lai & Bond, 1997). It is likely, however, that they
would form the basis for reciprocated friendship in other interaction contexts. People who
are thrown together in task groups, for example, may well develop friendships on the basis
of the partner’s perceived strength of application and emotional stability. Or, other qualities
may become important if the roommates in this study move into a new type of relationship,
say, socializing after they have left university and are working. A taxonomy of such
relationship types (e.g., Wish et al., 1976) would be useful in generating predictions about
which personality traits become decisive in friendship formation. Results from this study,
however, strongly suggest that it will be the perceived personality of the other that leads to
friendship, not the other’s self-perceived personality.
Perceived vs. ``actual'' effects
It was how roommates rated their partners that related to mutual friendship, not how that
partner rated him- or herself. This result is consistent with those of other studies, where
perceived and so-called actual effects are compared; perceived effects are always much
stronger (e.g., Hill & Stull, 1981; Newcomb, 1961). When actual characteristics of the other
are related to outcomes, it is probably through the mediating agency of perceived
Of course, perceptions of others’ personality and others’ self-perceptions are
moderately related (e.g., Funder & Colvin, 1988; Robins & John, 1997), although even
for married couples these relations are often surprisingly low (McCrae et al., 1996). It
thus becomes a moot epistemological question how the other’s personality may best be
measured. In a social psychology of personality (e.g., Hogan, 1982), the other’s perceived
personality (i.e., his or her reputation) may be the decisive factor in generating and
sustaining the behaviors which nurture the relationship. In this sense, perceptions of the
target’s personality by others may be more ‘‘real’’ than the target’s self-rated personality.
Bakan, D. (1966). The duality of human existence. Chicago, Ill: Rand McNally.
Berscheid, E. & Walster, E. (1983). Interpersonal attraction. Reading, MA: Addison-Wesley.
Bond, M. H. & Shiu, W. Y. F. (1997). The relationship between a group’s personality resources and the
two dimensions of its group process. Small Group Research, 28, 196–219.
Byrne, D. (1971). The attraction paradigm. New York: Academic Press.
Byrne, D. (1992). The transition from controlled laboratory experimentation to less controlled settings:
Surprise! Additional variables are operative. Communication Monograph, 59, 191–198.
188 Royce Yat-Pui Lee and Michael Harris Bond
Byrne, D. & Griffitt, W. (1973). Interpersonal attraction. Annual Review of Psychology, 24, 317–336.
Cheng, C., Bond, M. H., & Chan, S. C. (1995). The perception of ideal best friends by Chinese
adolescents. International Journal of Psychology, 30, 91–108.
Costa, P. T. Jr. & McCrae, R. R. (1992). Revised NEO Personality inventory (NEO PI-R) and NEO
Five-Factor Inventory (NEO-FFI). Odessa, FL: Psychological Assessment Resources.
Curry, T. J. & Emerson, R. M. (1970). Balance theory: A theory of interpersonal attraction?
Sociometry, 33, 216–238.
Duck, S. W. (1973a). Similarity and perceived similarity of personal constructs as influences on
friendship choices. British Journal of Social and Clinical Psychology, 12, 1–6.
Duck, S. W. (1973b). Personality similarity and friendship choice: Similarity of what, when? Journal
of Personality, 41, 543–559.
Duck, S. W. (1994). Meaningful relationships: Talking, sense, and relating. Thousand Oaks, CA: Sage.
Duck, S.W. & Barnes, M. K. (1992). Disagreeing about agreement: Reconciling differences about
similarity. Communication Monographs, 59, 199–208.
Duck, S. W. & Craig, G. (1978). Personality similarity and the development of friendship: A
longitudinal study. British Journal of Social and Clinical Psychology, 17, 237–242.
Fishbein, M. & Ajzen, I. (1972). Attitudes and opinions. Annual Review of Psychology, 23, 487–554.
Funder, D. C. & Colvin, C. R. (1988). Friends and strangers: Acquaintanceship, agreement, and the
accuracy of personality judgement. Journal of Personality and Social Psychology, 55, 149–158.
Goodwin, R., & Tang, C. S. K. (1996). Chinese personal relationship. In M. H. Bond (Ed.), The
handbook of Chinese psychology (pp. 294–308). Hong Kong: Oxford University Press.
Hill, C. T. & Stull, D. E. (1981). Sex differences in effects of social and value similarity in same-sex
friendship. Journal of Personality and Social Psychology, 41, 488–502.
Hogan, R. T. (1982). A socioanalytic theory of personality. In M. Page (Ed.), Nebraska symposium on
motivation (Vol. 30, pp. 56–89). Lincoln, NB: University of Nebraska Press.
Hogan, R. T. & Emler, N. P. (1978). The biases in contemporary social psychology. Social Research,
45, 478–534.
Hoyle, R. H. (1993). Interpersonal attraction in the absence of explicit attitudinal information. Social
Cognition, 11, 309–320.
Kandel, D. B. (1978). Similarity in real life adolescent friendship pairs. Journal of Personality and
Social Psychology, 36, 306–312.
Lai, M. & Bond, M. H. (1997). Gender stereotypes and the self-concept in Hong Kong youth.
Manuscript submitted for publication.
LaPrelle, J., Hoyle, R. H., Insko, C. A. & Beruthal, P. (1990). Interpersonal attraction and descriptions
of the traits of others: Ideal similarity, self similarity, and liking. Journal of Research in
Personality,24, 216–240.
Lea, M. & Duck, S. (1982). A model for the role of similarity of values in friendship development.
British Journal of Social Psychology, 21, 301–310.
McCrae, R. R. & Costa, P. T. (1987). Validation of the five-factor model of personality across
instruments and observers. Journal of Personality and Social Psychology, 52, 81–90.
McCrae, R. R., Stone, S. V., Fagan, P. J. & Costa, P. W. Jr. (1996). Identifying causes of disagreement
between self-reports and spouse ratings of personality. Unpublished manuscript, National Institute
of Health, Bethesda, MD.
Newcomb, T. M. (1961). The acquaintance process. New York: Holt, Rinehart, & Winston.
Posavac, E. J. & Pasko, S. J. (1974). Attraction, personality similarity, and popularity of the personality
of a stimulus person. Journal of Social Psychology, 92, 269–275.
Robins, R. W. & John, O. P. (1997). The quest for self-insight: Theory and research on accuracy and
bias in self-perception. In R. Hogan, J. A. Johnson & S. R. Briggs (Eds.), Handbook of personality
psychology (pp. 649–679). New York: Academic Press.
Rosenbaum, M. E. (1986). The repulsion hypothesis: On the nondevelopment of relationship. Journal
of Personality and Social Psychology, 51, 1156–1166.
Rosenberg, M. J. (1960). Cognitive reorganization in response to the hypnotic reversal of attitudinal
Friendship in Chinese culture 189
affect. Journal of Personality, 28, 28–63.
Stalling, R. S. (1970). Personality similarity and evaluative meaning as conditions of attraction.
Journal of Personality and Social Psychology, 14, 77–82.
Sunnafrank, M. (1983). Attitude similarity and interpersonal attraction in communication processes: In
pursuit of an ephemeral influence. Communication Monographs, 50, 273–284.
Sunnafrank, M. (1992). On debunking the attitude similarity myth. Communication Monographs, 59,
Sunnafrank, M. & Miller, C. R. (1981). The role of initial conversations in determining attraction to
similar and dissimilar strangers. Human Communication Research, 10, 27–53.
Trapnell, P. D. (1994). Openness versus intellect: A lexical left turn. European Journal of Personality,
8, 273–290.
Weinreich, P., Luk, C. L. & Bond, M. H. (1996). Ethnic stereotyping in a multicultural context:
Acculturation, self-esteem and identity diffusion in Hong Kong Chinese university students.
Psychology and Developing Societies, 8, 107–169.
Wetzel, C. G. & Insko, C. A. (1982). The similarity–attraction relationship: Is there an ideal one?
Journal of Experimental Social Psychology, 18, 253–276.
Wetzel, C. G., Schwartz, D. & Vasu, E. S. (1979). Roommate compatibility: Is there an ideal
relationship? Journal of Applied Social Psychology, 9, 432–445.
Wish, M., Deutsch, M. & Kaplan, S. J. (1976). Percevied dimensions of interpersonal relations.
Journal of Personality and Social Psychology, 33, 409–420.
Wright, P. H. (1985). The acquaintance description form. In S. W. Duck & D. Perlman (Eds.), Sage
series in personal relationships, 1, (pp. 39–62). Beverly Hills, CA: Sage.
Wright, P. H. (1991). The acquaintance description form: What it is and how to use it. Unpublished
manuscript, University of North Dakota.
Yang, K. S. & Bond, M. H. (1990). Exploring implicit personality theories with indigenous or imported
constructs: The Chinese case. Journal of Personality and Social Psychology, 58, 1087–1095.
Yik, M. S. M. & Bond, M. H. (1993). Exploring the dimensions of Chinese person perception with
indigenous and imported constructs: Creating a culturally balanced scale. International Journal of
Psychology, 25, 333–341.
190 Royce Yat-Pui Lee and Michael Harris Bond
... Importantly, perceived self-similarity leads to affiliation and liking (Graves & Powell, 1995;Huston & Levinger, 1978;Lee & Bond, 1998). Perceptions of similarity can come from appearance, as people use appearance-particularly facial appearance-to infer socially relevant information about others, including their personality traits and important preferences (e.g., political affiliation; Penton-Voak et al., 2006;Tskhay & Rule, 2013) in addition to attributes such as race, gender, and age (e.g., Ito & Urland, 2003). ...
Full-text available
People prefer to form relationships with people like themselves—a tendency that extends even to facial appearance, resulting in groups whose members look alike. Here, we investigated the mechanisms underlying homophilic resemblance using facial photos of fraternity/sorority members from two time points: before joining the group and after belonging to the group for three years. Analyses of both subjective trait impressions and objective face-shape measurements revealed that not only did group members look alike, they resembled one another even before joining the group. Moreover, photos of potential fraternity recruits revealed that facial appearance predicted both the group that individuals sought to join and the group’s likelihood of accepting them. Individuals, therefore, seek to join groups consisting of people who look like them, and the groups preferentially accept new members who resemble those already in the group. This bidirectional preference for homophily likely perpetuates intragroup homogeneity, suggesting potential implications beyond appearance.
... Friendships, romantic relationships, and groups form around these similarities, as individuals are happier around others with perceived similar interests and opinions (Caspi & Herbener, 1990;Mackinnon, Jordan, & Wilson, 2011). Critical to the present work, it is the perception of similarity that is crucial in determining liking (Lee & Bond, 1998). Because perceived similarity leads to group formation, facial appearance may play an important and yet relatively overlooked role in determining group membership. ...
Perceivers form strong inferences of disposition from others' facial appearance, and these inferences guide a wide variety of important behaviors. The current research examines the possibility that similar-looking individuals are more likely to form groups with one another. We do so by testing a necessary downstream consequence of this process, examining whether the faces of individuals within groups more physically resemble one another than those in other groups. Across six studies, we demonstrate that individuals' group membership can be accurately classified both from ratings of members' faces, and from direct measurement of members' faces. Results provide insight into how affiliative groups initially form and maintain membership over time, as well as the perception of homogeneity of groups.
... Furthermore, diverse aspects of interpersonal relationships are often correlated with personality traits (for an overview, see: Simpson et al., 2006; Reis et al., 2002) and research of interpersonal relationships could benefit from examining personality constructs (Gaines, 2007). The use of measures that were developed according to the five-factor models of personality are frequent in such research (for instance, see: Asendorpf, 1998; Neyer and Asendorpf, 2001; Asendorpf and van Aken, 2003; Barelds et al., 2007; Donellan et al., 2005; Lee and Bond, 1998; Kurtz and Sherker; 2003; King and Terrance, 2006; Robins et al.;). The research presented below was designed to determine personality traits associated with the dimensions of rules in friendship relationships. ...
Conference Paper
Full-text available
This paper presents research conducted on the factor structure of rules in friendship relationships. Based on the previous research of other authors and the process of collecting the rules that young adults use in their friendship relationships, an IR-1 self-assessment measure was developed that consisted of 141 friendship-related rules which was applied on a large set of subjects (N=501) in the age of early adulthood. A factor analysis of the IR-1 items revealed five dimensions of rules in friendship relationships that were labeled benevolence, sociability, appropriateness, self-control, and self-directedness. The NEO PI-R inventory, which measures the personality dimensions according to the five-factor model of personality, was applied on a smaller subset of subjects (N=114) together with the IR-1 measure. Many statistically significant correlations were uncovered between the dimensions of interpersonal rules on the one hand, and the personality dimensions and their facets on the other hand, most of which were related to extraversion and agreeableness. The five uncovered factors of interpersonal rules were used for the development of a self-assessment instrument IRF-2 with 60 items and five subscales designed to measure the related dimensions of rules in friendship relationships. In addition, a brief measure of positive and negative aspects of friendship quality was developed and a correlation analysis was performed on a small convenience sample (N=61) between variables related to the dimensions of rules in friendships and the variables related to the quality of friendship with a best friend. The results indicated that the use of rules in friendship associated with the dimensions benevolence and sociability had the largest positive influence on the quality of friendship with a best friend.
... Another factor that has been shown to increase liking is similarity. That is, we are more inclined to have increased attraction to and stronger relationships with those who are similar to us (Lee and Bond, 1988). Researchers have found that individuals comply significantly more with a request from individuals who are more similar than dissimilar (see Burger et al., 2001). ...
Purpose – A recent Supreme Court of Canada (SCC) ruling resulted in stricter rules being placed on how police organizations can obtain confessions through a controversial undercover operation, known as the Mr. Big technique. The SCC placed the onus on prosecutors to demonstrate that the probative value of any Mr. Big derived confession outweighs its prejudicial effect, and that the police must refrain from an abuse of process (i.e. avoid overcoming the will of the accused to obtain a confession). The purpose of this paper is to determine whether a consideration of the social influence tactics present in the Mr. Big technique would deem Mr. Big confessions inadmissible. Design/methodology/approach – The social psychological literature related to the compliance and the six main principles of social influence (i.e. reciprocity, consistency, liking, social proof, authority, scarcity) was reviewed. The extent to which these social influence principles are arguably present in Mr. Big operations are discussed. Findings – Mr. Big operations, by their very nature, create unfavourable circumstances for the accused that are rife with psychological pressure to comply and ultimately confess. A consideration by the SCC of the social influence tactics used to elicit confessions – because such tactics sully the circumstances preceding confessions and verge on abuse of process – should lead to all Mr. Big operations being prohibited. Practical implications – Concerns regarding the level of compliance in the Mr. Big technique call into question how Mr. Big operations violate the guidelines set out by the SCC ruling. The findings from the current paper could have a potential impact of the admissibility of Mr. Big confessions, along with continued use of this controversial technique. Originality/value – The current paper represents the first in-depth analysis of the Mr. Big technique through a social psychological lens.
... The Openness domain, however, has not been consistently extracted in crosscultural studies of the Big Five (Bond, 1994). It was found in a number of studies (Yik & Bond, 1993) and was shown to predict mutual friendship in a Chinese sample (Lee & Bond, 1998). Openness to Experience is conceptualized as a broad construct that encompasses tender-mindedness, imaginativeness, liberal thinking, receptivity to many varieties of experience, and a fluid and permeable structure of consciousness. ...
Full-text available
One of the early cross-cultural studies on the Minnesota Multiphasic Personality Inventory (MMPI) (S. Sue & D. W. Sue, 1974) showed that Asian American students using a psychiatric clinic had higher scale elevations on the MMPI than non-Asian students. Twenty years later, similar findings were obtained. The MMPI- 2 profiles of Asian American students showed more somatic complaints, depression, anxiety, and isolation than their Caucasian counterparts (S. Sue, K. Keefe, Enomoto, Durvasula, & Chao, 1996). There may be different interpretations of this finding. First, Asian American students were more prone to psychopathology. Alternatively, the elevated scores on the MMPI scales may be less a reflection of the problems of the students than those of the MMPI itself.
... The association between similarity and attraction is most commonly evaluated in terms of liking (Byrne 1971), which has been shown to increase trust (Doney and Cannon 1997). The reality of the similarity matters, but the perception of similarity, the degree to which an individual believes another's distinguishing characteristics are similar to their own, is often more influential when it comes to attraction (Hoyle 1993;Klohnen and Luo 2003;Lee and Bond 1998). However, individuals do not just like those who are similar in personality or thought, but also in behavior. ...
Conference Paper
Full-text available
The increasing use of robots and their role in society has important implications for research and development in human robot interactions (HRIs). The purpose of the present study was to develop a new measure to assess attitudes toward robots in HRIs. Measures of attitudes humans have " about " robots are relatively uncommon. Those that exist have potential problems limiting utilization in research evaluating the human element in HRIs. The Robot Perception Scale (RPS) was developed to redress this gap by examining a new set of factors in unique ways. The RPS consists of two subscales in which participants rate their agreement with statements concerning general attitudes toward robots and attitudes toward human-robot similarity and attractiveness. Findings provide preliminary support for a ro-botic perception scale that can be used to further our understanding of robots engaged in a variety of HRI settings.
... A personality-based social rule that governs interpersonal interaction, the similarity attraction rule holds that people favour the company of those that are most similar to themselves (Infante, Rancer, & Womack, 1990). The reality of the similarity matters, but the perception of similarity – the degree to which an individual believes another's distinguishing characteristics are similar to their own – was often more influential when it came to attraction. Lee and Bond (1998) discovered that the friendships between roommates thrived when there was common ground in personality traits and shared values. However, the friendships flourished even more when roommates also perceived one another as similar. ...
Full-text available
It has been speculated that robots will become a normal feature of everyday life in the future and effective human-robotic partnerships will be important. The purpose of the present study was to identify the impact certain social design elements have on trust in human-robot interactions (HRIs) and address the question as to whether personality traits and trust in humans can predict trust in robot types. A robot’s appearance, voice, and personality all function as social cues that allow an individual to decide how to react (Powers & Kiesler, 2006). Eyssel and Hegel (2012) found that male robots were perceived as more masculine and agentic whereas female robots were perceived as more feminine and communal, indicating that the gender stereotype process affected participants’ perception of the robots’ capability to perform sex-typed tasks. The current study examined college students (N = 50, Mage = 19.64, SD = 5.10) at a small Southeastern university. An ANOVA comparing mean ratings from the four groups revealed that robot appearance did influence likability, with the human-like gender-neutral robot liked most (F (3,46) = 8.77, p = .0001, ηp2= .36). Robot appearance also played a role in shaping evaluations of fear, with the human-like feminine robot feared most (F (3,46) = 11.96, p = .0001, ηp2= .44).
Positive friendship relationships are connected to the psychological well-being and social adjustment of an individual. Relationship quality in friendships is influenced by the rules that participants use in the interaction with their partners. This paper presents the research of the factor structure of rules in friendship relationships. Based on the previous research of other authors and the process of collecting rules that young adults use in their friendship relationships, the IR-1 questionnaire with 141 statements/rules was developed and then applied on subjects (N=501) in the age of early adulthood. Factor analysis revealed five dimensions of rules in friendship relationships that were labeled benevolence, sociability, appropriateness, self-control, and self-directedness. These dimensions of interpersonal rules were compared with the results of research of other authors in the field of friendship relationships and social competence. The NEO PI-R inventory that measures the personality dimensions according to the five-factor model of personality was also applied on a smaller subset of subjects (N=114). Many statistically significant correlations were uncovered between the dimensions of interpersonal rules, on the one side, and the personality dimensions and their facets, on the other side, and most of them were related to extraversion and agreeableness.
Since 1970, mainland China has been undergoing social and economic reforms that have brought about astonishing economic success. This article reveals that these social and economic changes have led to many changes in Chinese interpersonal relationships, and family structures and processes. They are not limited to mainland China but can also be seen in other Chinese habitats, such as Hong Kong, Singapore, and Taiwan. The article focuses on Chinese relationship research conducted in the last decade and examines how the rapid social and economic changes in Chinese societies have exerted an impact on the various types of interpersonal relationship. Furthermore, it looks at how traditional Chinese values continue to shape interpersonal relationships amidst the increasing influence of modernization and globalization. In terms of interpersonal relationships, it focuses on four broad domains, namely friendship, romantic relationships, marital relationships, and family relationships, providing a review of the relevant studies in each domain.
The present study examines the productivity of task groups in relation to the personality resources of its members and its two dimensions of group process: task focus and shared exchange. It is hypothesized that the depth of a group's personality resources impacts upon productivity both directly through the application of its member's personality resources to successful group outputs, and indirectly through the mediating agency of group process variables that also contribute separately to successful group outputs. To test these hypotheses, the performance of 43 groups in a 3-month social psychology class was measured across three course assignments. This averaged performance score was related to the groups' process scores and to their total levels of personality resources measured by a comprehensive personality inventory administered at the beginning of the groups' life. Using multiple regression, we found that a group's performance was predicted by total member intellect, openness (negatively), and emotional stability (negatively). Blocked regression revealed that group intellect exercised a direct effect upon group performance, but that the effects of group openness and group emotional stability on performance were mediated through the group's two group processes of task focus and shared exchange. It is hoped that this demonstration of a two-step approach to studying the impact of group member's personality through its direct effects on group performance and its indirect effect on performance-linked aspects of group process will be extended to other types of personality measures, and to other types of groups addressing different tasks in other cultural settings.
Previous research has shown that similarity of values can significantly enhance interpersonal attraction. The present study sought to refine and clarify this finding within the context of friendship development. The relationship between value similarity and friendship choice for subjects was examined cross-sectionally after three periods of acquaintance: 1–2 months, 4–6 months, and 12+ months. Two factors, namely the importance of the values to the subjects and the uncommonness of the similarity for them, were hypothesized to affect this relationship. The hypothesis was supported. A further hypothesis was that the salience of these factors for subjects was dependent upon factors that related to the length of their friendships. The results offered only marginal support for this hypothesis. Additional data on unreciprocated friendship choices are presented. The results are discussed in relation to the hypothesized predictive function (to reduce interpersonal risk or uncertainty) as well as the supportive function of value similarity, and the wider implications of the results are considered.
This study tests two theoretical approaches to interpersonal attraction. While the main effort is a replication of Newcomb's Acquaintance Process, portions of Chambliss' theory of attraction are tested also. We replicate Newcomb's study in a residence-hall setting that featurs eight-person natural living groups, and employ nine groups of initial strangers, six male and three female, over an eight-week period. Using nine groups allows us to expand the replication by examining inter-group differences, sex differences, and instrument effects. Although the data for some of our groups parallels Newcomb's, the data also reveal considerable variation between groups unexplainable by the AB-X formulation. Because of this variation, we explore alternative interpretations of the data, and conclude that group substructuring and perceptual-judgmental interpretations of ambiguous stimuli appear to account for relations previously attributed to the AB-X processes. The variable most strongly associated with attraction at all times is Chambliss' variable of "success," based upon an interactional approach. Interpretations of this variable via social exchange theory parsimoniously achieves explanation of the data gathered on the longitudinal aspects of reciprocity in social relations, and further questions the advisability of a cognitive balance approach to long-term attraction.
The present study focused on identifying the independent and conjoint influence of attitude similarity and initial interaction on interpersonal attraction to relative strangers. Participants were informed they would be working on a project with either an attitudinally similar or an attitudinally dissimilar stranger. Half of the participants next engaged in an initial interaction with their partner and the other half did not. All participants then filled out a scale that included a measure of interpersonal attraction. Results indicated that the conjoint, nonadditive effects of attitude similarity and initial interaction overrode the significant main effects of these variables. Although attitudinally similar noninteractants were more attracted to their partners than dissimilar noninteractants, no differences in attraction were observed among similar and dissimilar interactants. Dissimilar interactants were more attracted to their partners than dissimilar noninteractants, but no differences in attraction were observed between similar interactanls and noninteractants.