Two female friends at the seaside
Relationships and romance

Close relationships - What's in it for us?

Constantine Sedikides delivered the Michael Argyle Lecture at this year’s Annual Conference.

18 August 2005

Humans are social animals. We spend an inordinate amount of time thinking about people, interacting with them, influencing them and being influenced by them. We crave social contact and find our relationships with friends, romantic partners and relatives rewarding for the most part. Why? What do close others do for us? What are the psychological benefits of close relationships? In particular, how does thinking about close relationships boost the self?

Perhaps thinking about close relationships is the equivalent of vitamins. Taking vitamins helps fight off winter illnesses, such as colds and flu. Likewise, thinking about close relationships helps fight off life adversities, such as being confronted with failure or unfavourable evaluations. More succinctly, bringing to mind close relationships helps lessen the blow of stressful life events. Such events now feel more manageable. Close relationships, then, are an important psychological resource. They bolster the self-system to the point where failure is taken more lightly and may even be seen as a challenge rather than a threat.

With negative comes positive

We carried out an experiment to test the above ideas (Kumashiro & Sedikides, in press). All participants took what they thought was a valid and widely used intelligence test. The ominous ‘Allport-Jameson Intelligence Test’ consisted of difficult mathematical, analytical and verbal problems. To make matters worse, participants had only 25–45 seconds to respond to each problem. Due to the nature of the test, participants could not tell how well they did. At the conclusion of the test, participants learned that the experimenter would need around 15 minutes to compile a customised and comprehensive profile of their performance and intelligence. This was a pretext for the crucial manipulation that was about to follow.

Participants were requested to help, during the next 15 minutes, with an ostensibly unrelated short study. They were randomly assigned to three conditions. A third of participants (in the ‘close-positive relationships’ condition) were instructed to bring to mind a ‘warm and positive relationship, that special person with whom you have the best relationship’. Another third (in the ‘close-negative relationships’ conditions) were instructed to bring to mind a ‘cold and negative relationship,
the person with whom you have the worst relationship’. The final third of participants (in the ‘neutral relationship’ condition) were instructed to bring to mind a ‘distant relationship, a person whom you don’t know well and you neither like nor dislike’. Importantly, after participants recorded in each condition the relation’s initials and the nature of the relationship (e.g. friend, co-worker, neighbour), they proceeded to spend three minutes responding in writing to each of five questions, such as ‘What does this relationship personally mean to you?’ and ‘How does this relationship make you feel?’ Thus, participants did not merely bring to mind a relationship, but they also thought in some depth about the personal relevance of the relationship.

The experimental saga was not over yet. The experimenter entered the room, announcing that he had finishing compiling the participants’ individualised profiles. He then dispensed bogus performance feedback, ostensibly based on well-validated norms. Participants learned that their performance on the ‘Allport-Jameson Intelligence Test’ was ‘poor’, as their scores fell only at the 41st percentile of the distribution. The feedback hurt (as our measures indicated). Immediately afterwards, participants were informed that they would have the opportunity to receive additional information about their performance and intelligence. This information was allegedly accurate, pinpointed their weak points and difficulties in each test domain (mathematical, analytical, verbal), and could help them improve their future performance. Participants proceeded to indicate the degree to which they were interested in receiving additional feedback about their weaknesses. This was our main dependent measure.

As a reminder, we hypothesised that close-positive relationships are a self-bolstering resource. This resource armours the self-system to the point where, even following unfavourable performance feedback, individuals become more open to accurate information about their performance limitations and intelligence failings. Despite being hit by a self-threatening evaluation, participants whose minds are now ‘peopled’ by close-positive relationships will be oddly receptive to further self-threatening information. The results confirmed our hypothesis. Participants in the close-positive relationships condition expressed a stronger interest in obtaining additional information about their weaknesses compared with participants in the other two conditions.

Why does this happen?

Why are participants who ponder a close-positive relationship more receptive to unfavourable (albeit potentially improving) feedback? We reasoned that there may be at least three explanations. First, participants are overwhelmed by warm feelings for the person with whom they have a close-positive relationship. Second, participants experience a rise in self-esteem. Third, participants are in a better mood. Warm feelings for the relation, elevated self-esteem, and better mood could each reassure and shield the self to the point where an additional hit can be easily tolerated. We assessed these explanations in a follow-up study (Kumashiro & Sedikides, in press).

Participants completed a difficult version of the Remote Associates Test (RAT: McFarlin & Blascovich, 1984). The RAT was said to be a valid and reliable assessment of an important intellectual ability: integrative orientation. Participants were supplied with 20 sets of three words (e.g. cotton, bathtub, tonic) and were requested to generate fourth words that could go with each word in a set (in this example gin). When participants were finished, they learned that a comprehensive and personalised profile of their integrative orientation would follow soon. In the meantime, their assistance in another, short study would be appreciated.

Participants were randomly assigned to four conditions. One quarter thought and wrote about a close-positive relationship, another quarter about a distant-positive relationship, still another quarter about a close-negative relationship, and the final quarter about a distant-negative relationship. Those in the former two conditions wrote about how the relation was supportive of them, those in the latter two conditions about how the relation was critical of them. Subsequently, participants received unfavourable performance feedback. They learned that their integrative orientation ability was ‘below average’, as their scores lay on the 41st percentile of the distribution. Not surprisingly, they found the feedback unpleasant.

Next, participants recorded the level of their self-esteem and their mood. Specifically, they filled out the Heatherton and Polivy (1991) state self-esteem scale (example items: ‘Right now, I am worried about whether I am regarded as a success or failure’, and ‘I feel inferior to others at this moment’). Also, participants filled out a mood scale (Martin et al., 1997), which consisted of both negative (e.g. anxious, nervous, down) and positive (e.g. calm, pleased, happy) adjectives. Finally, participants indicated the extent to which they were interested in receiving additional information about their weaknesses in the domain of integrative orientation.

The findings were consistent with those of the previous experiment. Compared with the other three conditions, participants who pondered the personal relevance of a close-positive relationship expressed the strongest interest in obtaining weakness-related information about their integrative orientation ability. Nevertheless, neither self-esteem nor mood accounted for this finding. That is, participants in the close-positive relationship condition neither had higher self-esteem nor were in a better mood than those in the other conditions. Finally, although participants reported relatively warm feelings for the close-positive relation (as manifested by independent codings of their writing protocols), these feelings did not predict higher interest in weakness-related feedback about integrative orientation.

The role of cognitive dissonance

So far, our research has established the resource function of close-positive relationships by demonstrating that participants dare solicit additional feedback about their weaknesses knowing full well that the feedback will be accurate (i.e. self-threatening). We wanted to understand better the consequences of this resource for social behaviour. In what other ways does the self-bolstering function of close relationships manifest itself?

One of the most counterintuitive contributions of social psychological research is the phenomenon of cognitive dissonance (Festinger & Carlsmith, 1959). This phenomenon refers to the discomfort experienced by a perceived inconsistency between one’s attitude and one’s behaviour (e.g. expressed attitude) and to the strategies one uses to re-establish consistency. For example, when becoming aware of an inconsistency between one’s privately held and publicly expressed attitude, a person will strive to resolve this inconsistency by changing their initial attitude. Take the case of a student who privately opposes the policy to combat grade inflation in her university, given that such a policy would increase the rigour of exam-marking standards, thus resulting in substantially fewer first class and 2:1 degrees. Furthermore, assume that the student finds herself in the awkward position of having voluntarily written an essay that advocates such a policy. Cognitive dissonance theory predicts that she will now change her private attitude to be consistent with her public attitude: she will become more favourable towards the policy of curbing grade inflation.

If close-positive relationships function as a self-bolstering resource, they ought to eliminate cognitive dissonance. Pondering a close-positive relationship should affirm and replenish the self to the point where re-establishing cognitive consistency would become irrelevant and unnecessary. No change in private attitudes would occur.

In an experimental test of these ideas (Sedikides & Kumashiro, 2005), participants responded to a 15-item survey about their university. Only one of these items was pertinent to the objectives of our study. The item stated that the university should combat grade inflation by instituting higher marking standards and, by implication, lowering substantially the percentage of students who graduate with firsts or 2:1s. This was a rather covert way of gauging participants’ attitude towards the alleged grade inflation policy. Note that only participants with an unfavourable attitude towards this policy were retained in the data analyses – and a high majority of respondents indeed held such an attitude.
Next, participants were requested to write an essay, under the pretext of assisting a university committee that would investigate the grade inflation issue. Participants in the low-choice condition were instructed to write a strong and forceful essay in favour of combating grade inflation. In contrast, participants in the high-choice condition learned that the committee had enough ‘against’ arguments and needed more ‘for’ arguments. Their decision was voluntary, but nonetheless the committee needed strong, forceful arguments in favour of combating grade inflation. Participants proceeded to sign a form, acknowledging that (1) they were writing the essay voluntarily, and (2) the committee would use their arguments in making a decision.

Following the essay writing, participants in the high-choice condition were assigned into four groups. Some of them thought and wrote about a close-positive relationship, some about a close-negative relationship, some about a neutral relationship, and some about no relationship at all. Finally, all participants completed the state self-esteem and mood scales from the previous study and, more importantly, indicated once again their attitude towards the grade inflation policy.
No cognitive dissonance effect was expected in the low-choice condition. Participants would not change their private attitude, because their public attitude (i.e. the essay writing) was imposed by an external factor – the experimenter. However, cognitive dissonance was expected to emerge in most high-choice conditions, given that participants voluntarily elected to write the essay in favour of combating grade inflation and, as such, they should change their private attitudes in order to settle the looming inconsistency. In particular, cognitive dissonance was expected to emerge in three of the four conditions: close-negative, neutral, and no relationship. Importantly, cognitive dissonance was expected to be absent in the close-positive relationships condition. Participants, affirmed by the thoughts of a close other, would not feel the need to re-establish consistency between private and public attitudes.

Our hypotheses were confirmed. Participants in the close-negative relationship, neutral relationship, and no-relationship conditions who freely chose to write the essay manifested cognitive dissonance; that is, their attitudes towards the policy became more positive. In contrast, participants who had pondered a close-positive relationship maintained their anti-policy attitude (as did low-choice participants). Close-positive relationships eliminated cognitive dissonance. Nevertheless, as in our previous study, neither state self-esteem nor mood accounted for this finding.

What is missing?

Our results establish that close-positive relationships function as a self-affirming resource, but also point to several gaps in our knowledge. First and foremost, what is the exact mechanism through which close-positive relationships benefit the self? We ruled out state self-esteem, mood and warm feelings for the relation. There are several other viable candidates. One is unconditional acceptance. Perhaps a reason why individuals are not bothered by weakness-related feedback or by attitude–behaviour inconsistency is that close relationships make them feel secure or provide them with a safe haven. Two other possible mechanisms are generalised positive affect (Tesser, 2000) and self-esteem (Sedikides & Gregg, 2003), with both presumed to occur outside of conscious awareness. If so, measures that are very different from the ones we used would need to be implemented to capture the effect of these mechanisms. Finally, at the physiological level, researchers will do well to zero in on a potential surge in oxytocin levels that might accompany thinking about close-positive relationships (Taylor et al., 2000).

Several other research opportunities present themselves. In our research, we did not distinguish between different types of close-positive relationships. Which type (e.g. romantic partners, friends, family) is most effective as a self-bolstering resource? Also, what are some other behavioural consequences of this resource? For example, does relationship-induced self-bolstering influence performance-related variables such as task persistence, intrinsic motivation, and creativity (Wentzel et al., 2004)? More importantly, does relationship-induced self-bolstering influence prosocial behaviour?

Finally, what are some key individual differences that moderate the effectiveness of close-positive relationships as a self-bolster? We would speculate in favour of attachment style (Mikulincer & Shaver, 2001), self-esteem (Murray et al., 1998), and implicit self-theories (Dweck, 1999) as promising empirical directions. For example, pondering the personal relevance of close-positive relationships may result in more effective self-bolstering among secure (rather than avoidant or dismissive) persons and among high (as opposed to low) self-esteem persons.

Coping with life’s adversities

Close relationships have high personal relevance. Individuals consider close others part of the self (Aron et al., 1991). Also, close relationships are associated with increased self-understanding and self-growth (Sedikides et al., 1994), reduced actual–ideal discrepancies (Campbell et al., 1994) and positive illusions (i.e. positive self-concept, exaggerated perceptions of control, unrealistic optimism: Martz et al., 1998). In addition, close relationships contribute substantially to life satisfaction and subjective well-being (Argyle, 1987). Thus, close relationships constitute a vital resource to the self.Interestingly, thinking of close others changes the self-system in important ways. This process, for example, buffers existential anxiety (Mikulincer et al., 2003) and lowers stress (McGowan, 2002). More importantly, thinking of close others induces exploratory intentions (Green & Campbell, 2000) and a state of cognitive openness (Green-Hennessy & Reis, 1998; Mikulincer & Arad, 1999).

We relied on the above two lines of research to generate our hypotheses. Close relationships have resource potential and contribute to cognitive openness and exploration. Psychological resources (e.g. success experiences, sense of control) increase willingness to obtain accurate but unfavourable feedback (Aspinwall, 1998; Trope & Neter, 1994). Hence, it follows that pondering the personal relevance of close relationships will buffer the self, thus making the individual more receptive to feedback about personal weaknesses and eliminating cognitive dissonance.

Our research supported these conjectures. Cognitive activation of close-positive relationships increased willingness to obtain accurate information about one’s weaknesses on a domain, even in the face of prior failure feedback on that domain. Close-positive relationships shield the self to the point where, even following unfavourable feedback, accurate information about personal weaknesses is sought out despite its self-threat potential. In conclusion, our research establishes that close-positive relationships constitute a resource that enables the self to cope more effectively with life’s adversities.

- Constantine Sedikides is Professor of Social and Personality Psychology and Director of the Centre for Research on Self and Identity at the University of Southampton. E-mail: [email protected].


Centre for Research on Self and Identity:
International Association for Relationship Research:
International Society for Self and Identity:


Argyle, M. (1987). The psychology of happiness. London: Methuen.
Aron, A., Aron, E.N., Tudor, M. & Nelson, G. (1991). Close relationships as including other in the self. Journal of Personality and Social Psychology, 60, 241–253.
Aspinwall, L.G. (1998). Rethinking the role of positive affect in self-regulation. Motivation and Emotion, 22, 1–32.
Campbell, W.K., Sedikides, C. & Bosson, J. (1994). Romantic involvement, self-discrepancy, and psychological well-being. Personal Relationships, 1, 399–404.
Dweck, C.S. (1999). Self theories. Philadelphia, PA: Psychology Press/Taylor & Francis.
Festinger, L. & Carlsmith, J.M. (1959). Cognitive consequences of forced compliance. Journal of Abnormal and Social Psychology, 58, 203–210.
Green, J.D. & Campbell, W.K. (2000). Attachment and exploration in adults. Personality and Social Psychology Bulletin, 26, 452–461.
Green-Hennessy, S. & Reis, P.R. (1998). Openness in processing social information among attachment types. Personal Relationships, 5, 449–466.
Heatherton, T.F. & Polivy, J. (1991). Development and validation of a scale of measuring state self-esteem. Journal of Personality and Social Psychology, 60, 895–920.
Kumashiro, M. & Sedikides, C. (in press). Taking on board liability-focused feedback. Psychological Science.
Martin, L., Abend, T., Sedikides, C. & Green, J.D. (1997). How would I feel if…? Mood as input to a role fulfillment evaluation process. Journal of Personality and Social Psychology, 73, 242–253.
Martz, J.M., Verette, J., Arriaga, X.B. et al. (1998). Positive illusions in close relationships. Personal Relationships, 5, 151–181.
McFarlin, D.B. & Blascovich, J. (1984). Remote Associates Test (RAT) as an alternative to illusory performance feedback. Basic and Applied Social Psychology, 5, 223–229.
McGowan, S. (2002). Mental representations in stressful situations: The calming and distressing effects of significant others. Journal of Experimental Social Psychology, 38, 152–161.
Mikulincer, M. & Arad, D. (1999). Attachment working models and cognitive openness in close relationships. Journal of Personality and Social Psychology, 77, 710–725.
Mikulincer, M., Florian, V. & Hirschberger, G. (2003). The existential function of close relationships: Introducing death into the science of love. Personality and Social Psychology Review, 7, 20–40.
Mikulincer, M. & Shaver, P.R. (2001). Attachment theory and intergroup bias: Evidence that priming the secure base schema attenuates negative reactions to out-groups. Journal of Personality and Social Psychology, 81, 97–115.
Murray, S.L., Holmes, J.G., MacDonald, G. & Ellsworth, P.C. (1998). Through the looking glass darkly? When self-doubts turn into relationship insecurities. Journal of Personality and Social Psychology, 75, 1459–1480.
Sedikides, C. & Gregg, A.P. (2003). Portraits of the self. In M.A. Hogg & J. Cooper (Eds.) Sage handbook of social psychology (pp. 110–138). London: Sage.
Sedikides, C. & Kumashiro, M. (2005). On the resource function of relationships: Close-positive relationships as a dissonance reduction mechanism. Unpublished manuscript. University of Southampton.
Sedikides, C., Oliver, M.B. & Campbell, W.K. (1994). Perceived benefits and costs of romantic relationships for women and men: Implications for exchange theory. Personal Relationships, 1, 5–21.
Taylor, S.E., Klein, L.C., Lewis, B.P. et al. (2000). Biobehavioral responses to stress in females. Psychological Review, 107, 411–429.
Tesser, A. (2000). On the confluence of self-esteem maintenance mechanisms. Personality and Social Psychology Review, 4, 290–299.
Trope, Y. & Neter, E. (1994). Reconciling competing motives in self-evaluation. Journal of Personality and Social Psychology, 66, 646–657.
Wentzel, K.R., Barry, C.M. & Caldwell, K.A. (2004.) Friendships in middle school: Influences on motivation and school adjustment. Journal of Educational Psychology, 96, 195–203.