Making Sure You See the Real Me: The Role of Self-Esteem in Spontaneous Self-Expansion

Abstract

People alter their self-concepts by incorporating attributes of close others, such as desired romantic partners, into their own identity. The current research examined self-esteem as a moderator of this spontaneous self-expansion. Two studies tested the hypothesis that when presented with a prospective romantic partner, higher self-esteem people (HSE) would self-expand to adopt positive attributes, while lower self-esteem people (LSE) would self-expand to adopt negative attributes. This tendency corresponds with people’s desire to self-verify and be seen by others in line with their own self-views, which are often negative among those lower in self-esteem. Study 1 (n = 218) and Study 2 (n = 234) confirmed our predictions that when motivated to increase romantic closeness to a prospective partner, self-esteem differentially predicts people’s spontaneous self-expansion to adopt positive versus negative attributes.

Keywords

self-concept self-expansion self-esteem self-verification

Throughout people’s lives, their self-concept, or their sense of who and what they are, fluctuates. Time, life events, and relationships all change who people think they are, for better and for worse (Slotter & Emery, 2018). The present research sought to investigate one way, called self-expansion, in which social others, specifically desired romantic partners, influence people’s selves. Self-expansion occurs when people adopt attributes as part of their identity that they didn’t previously possess (e.g., Aron, 2003). One way that people engage in self-expansion is by integrating new attributes that belong to romantic partners into their identities, thus making their self-concept more similar to their partner’s (e.g., Aron, Aron, Tudor, & Nelson, 1991). Unsurprisingly, this occurs when the attributes in question are positive, such as being athletic/funny, but can also occur when they are negative, such as being disorganized/clumsy (Slotter & Gardner, 2012). The present research investigated self-esteem, or people’s global feelings of positivity about themselves (James, 1890), as a moderator of when people adopt positive versus negative attributes as their own.

The Expanding Self-Concept

The self-concept refers to a person’s sense of identity and includes everything that a person can claim as “me”/“mine” (James, 1890). Although the self-concept is subjectively experienced as consistent over time, research demonstrates that it is continuously constructed through experiences (e.g., McConnell, 2011).

One way of molding the self-concept is via self-expansion. Self-expansion is a motivated process wherein people add novel attributes/information to their self-concept (Aron & Aron, 1997). People are motivated to engage in self-expansion because it leads to a broader set of self-attributes for creating a meaningful sense of identity. People can self-expand individually by engaging in novel activities that allow them to see themselves in new ways; however, one of the primary ways that people self-expand is by adopting attributes of social others as their own (e.g., Aron, 2003).

In romantic relationships, people self-expand by adopting attributes from their partner’s self-concept (e.g., Aron et al., 1991). Self-expansion predicts higher levels of satisfaction in relationships (e.g., Mattingly, Lewandowski, & McIntyre, 2014). This beneficial integration of partners’ selves typically emerges over time from shared experiences (e.g., Aron, 2003). However, people can integrate novel attributes of a potential romantic partner into their own self-concept even in the absence of shared experience (Slotter & Gardner, 2009, 2012).

Spontaneous self-expansion occurs when people are motivated to affiliate with a novel person via a strong romantic interest. In both attraction contexts and ongoing relationships, self-expansion benefits relationships by enhancing perceived similarity between oneself and a partner, which predicts relational longevity and happiness (e.g., Slotter & Gardner, 2009). Thus, if someone is romantically interested in another person, it benefits them to take on that desired partner’s attributes to enhance similarity. Indeed, people who do not report romantic interest in a potential romantic partner do not self-expand, whereas people who express high levels of romantic interest in novel potential romantic partners do (e.g., Slotter & Gardner, 2009, 2012).

The majority of the existing work on spontaneous self-expansion focuses on the self-concept adoption of novel, positive attributes (e.g., Slotter & Gardner, 2009). Indeed, recent theoretical work has limited the use of the term self-expansion to the adoption of positive attributions into the self (Mattingly et al., 2014). However, people will adopt negative attributes as well (Slotter & Gardner, 2012), termed self-adulteration (Mattingly et al., 2014). Specifically, when presented with a hypothetical potential partner who listed moderately negative attributes in a dating profile as a part of his or her identity, participants spontaneously incorporated these attributes as their own if they were romantically interested in the potential partner. In the present work, we refer to the adoption of positive and negative attributes as spontaneous self-expansion in order to be consistent with previous research, despite recently proposed theoretical nuances in terminology (e.g., Slotter & Gardner 2009, 2012).

Spontaneous Self-Expansion and Self-Esteem

The notion that people alter their self-concept in unflattering ways may appear counterintuitive as people seek to create and maintain generally positive views of themselves (e.g., Alicke & Sedikides, 2009). Although people want others to affirm them as worthy/lovable, they also want others to see them accurately (e.g., Swann, Hixon, & De La Ronde, 1992; Kwang & Swann, 2010), especially in close relationships. Self-verification theory asserts that, when others see us as we see ourselves, there is less risk of social rejection and thus our self-concepts are protected (e.g., Katz & Beach, 2000; Swann, 1983, 1990). Furthermore, in both close relational and group contexts, empirical research has demonstrated that people are willing to take on negative attributes and group memberships as their own to promote social connection and safety (e.g., Mashek, Steuwig, Furukawa, & Tangney, 2006; Slotter & Gardner, 2012)

Dispositional, global self-esteem is thus one factor that may contribute to whether people prefer others to view them more positively versus negatively. Self-esteem refers to the value that people place upon, and positivity they generally ascribe to themselves as a stable tendency (e.g., Baumeister, 2010; James, 1890). On average, people who are low in self-esteem (termed LSE in previous literature; Murray, Griffin, Rose, & Bellavia, 2003) endorse a greater number of negative self-attributes than higher self-esteem people do (HSE). Although both HSE and LSE show automatic tendencies to seek positive information about themselves, when motivated by social safety concerns, both HSE and LSE prefer information that verifies their existing views of themselves. Thus, when social safety matters, such as in romantic attraction contexts, LSEs prefer information and others who view them as they view themselves and thus negatively (e.g., Kwang & Swann, 2010; Stinson et al., 2010).

Studying the influence of self-esteem on people in romantic contexts is important because self-esteem has consequences for relational functioning (Baumeister, Campbell, Krueger, & Vohs, 2003). HSE predicts variety of positive individual and relational outcomes and thus the focus of the present research. LSEs tend to perceive undue amounts of threat/rejection in their romantic relationships and thus engage in self-protective over relationship-promoting behaviors (e.g., Murray, Derrick, Leder, & Holmes, 2008). Ultimately, the relationships of LSEs tend to decline in quality over time, perhaps eroded by their self-protective relational tendencies (e.g., Murray et al., 2003).

Importantly, the more negative self-views that LSEs endorse (e.g., Baumeister, 2010; Forest & Wood, 2012) may be a mechanism behind their sometimes problematic relational behaviors. Therefore, within the framework of self-verification theory (e.g., Swann et al., 1992), LSEs may be especially likely to seek self-verifying feedback that they do, in fact, possess the negative attributes that dominate their self-views, whereas HSEs seek self-verifying feedback for their positive self-views (Stinson et al., 2010).

One way that people might engage in such self-verification efforts is through tailoring their approach to spontaneous self-expansion. Motivated by self-verification needs, people should engage in spontaneous self-expansion that reflects their underlying self-views. Thus, HSEs should adopt more positive attributes from potential romantic partners, whereas LSEs should adopt the more negative attributes.

Current Research and Hypotheses

We posited two primary hypotheses, tested across two preregistered studies (https://osf.io/sk7mp/ (see Slotter, 2016)).¹ We hypothesized that self-esteem would differentially predict people’s spontaneous self-expansion to adopt positive versus negative attributes. Specifically, higher self-esteem would predict greater spontaneous self-expansion for positive attributes (Hypothesis 1a) and lower self-esteem would predict greater spontaneous self-expansion for negative attributes (Hypothesis 1b). However, we predicted that the effects of Hypothesis 1a and b would only be observed if people are motivated to spontaneously self-expand, in this case via romantic interest in a potential partner (Hypothesis 2; e.g., Slotter & Gardner, 2009).

Study 1

Study 1 tested our hypotheses by examining how natural variability in people’s self-esteem would predict their willingness to self-expand to adopt positive versus negative attributes. Participants were shown profiles of a potential dating partner who possessed one attribute that the participants did not. We then predicted whether participants would adopt this attribute as part of their own self-concept from their reported perceptions of the attribute’s positivity, their dispositional self-esteem, and how romantically interested they reported being in the partner.

Method

Participants

Two hundred and thirty-two undergraduates (146 female; M _age = 18.91, SD = 1.95) took part. To be eligible for the study, all participants had to generate a usable “not me” attribute on a self-rating task completed at a prescreening session at least 2 weeks prior to any laboratory session (see below), as well as report that they were not currently involved in a romantic relationship, were generally interested in finding a new romantic partner, and were interested in dating online. Fourteen eligible participants were excluded from the final data set for failing an attention check measure, resulting in a usable sample of 218 participant (136 female; M _age = 18.89, SD = 1.99). See the Online Supplementary Materials, for a discussion of power considerations; however, the sample was adequate to reach .80 power, with an α of .05, estimating a rough effect size of f ² = .15 for our key interaction of interest.

Procedure

Participants completed the self-rating task, attribute valence rating task, and dispositional self-esteem measures during a pretesting session at least 2 weeks prior to all laboratory sessions. Upon their arrival at the laboratory, participants completed demographic measures before viewing a novel target portrayed as a potential romantic partner (e.g., Slotter & Gardner, 2012). Participants were told the researchers had been asked to develop a new online dating service. Participants saw a profile of a fellow student of the sex they preferred to date described to them as someone looking for a romantic relationship and interested in using the new dating service. They were told that their job was to evaluate this person as if they were evaluating him or her on an online dating site. They were also told that if they joined the service, they would be able to view this same target individual in the future. Information about the target was displayed next to a moderately physically attractive photo. The profile was displayed for 30 s and contained statements about the target’s interests and personality (see Figure 1 for an example). Statements were the same across participants and fairly generic (i.e., “I like spending time with my friends.”) with the exception of the one meant to present the not me attribute generated at pretesting as an attribute that the potential partner possessed. Specifically, this statement started, “One thing you should know about me is…” followed by a description of the not me attribute. After viewing the profile, participants completed filler questions about the profile layout. They then reported their romantic interest in the potential partner and the self-rating task a second time.

Measures

See Table 1, for bivariate correlations.

Table 1.

Study 1: Bivariate Correlations Among Key Variables.

Variable	(1)	(2)	(3)	(4)	(5)	(6)
(1) Pretesting rating
(2) Post-profile rating	.54***
(3) Attribute valence	.17**	.26***
(4) Self-esteem	−.15*	−.13*	.14*
(5) Romantic interest	.10	.19**	.05	−.17
(6) Gender	−.01	.01	−.02	.04	−.07
(7) Age	−.07	.02	.03	−.03	.03	−.07

Note. Gender was coded as follows: Male = 0, Female = 1.

*p < .05. **p < .01. ***p < .001.

Self-rating task

At the pretesting session, participants rated a series of 15 attributes with regard to the degree that each attribute was characteristic of their self-concept (0 = not at all characteristic of me, 6 = extremely characteristic of me; M = 0.98, SD = 0.81; Slotter & Gardner, 2012). The attributes were adventurous, absentminded, agreeable, artistic, athletic, clumsy, creative, disorganized, enthusiastic, irritable, musical, outgoing, sociable, sophisticated, and thoughtful. One idiosyncratic attribute was selected from participants’ self-ratings at pretesting that they had rated as being “not characteristic of me.” Participants who did not generate such an attribute were not eligible to participate in the study.

In order to be considered a not me attribute, participants had to rate the attribute as three or less on the scale provided at the pretesting session. Thus, each participant generated an idiosyncratic not me attribute. If more than one attribute was rated as “not characteristic of me,” the target not me attribute was randomly selected from the eligible set of attributes. This not me attribute was embedded in the novel target’s profile, which participants saw during the laboratory session.

Participants completed the self-rating task a second time after viewing the potential partner’s profile information (M = 2.06, SD = 1.11). We were interested in change in participants’ ratings of their idiosyncratic not me attribute from pretesting to after viewing the profile information.

Attribute valence

Participants rated the same list of 15 attributes with regard to how positively/negatively they viewed each attribute (0 = extremely negative, 6 = extremely positive; M = 3.57, SD = 1.60).

Self-esteem

Also at the pretesting session, participants completed a 10-item measure of their self-esteem (e.g., “I think that I am a person of worth, at least on an equal basis with others”; 1 = disagree strongly, 7 = agree strongly; α = .85; M = 5.59, SD = 0.91; Rosenberg, 1965).

Romantic interest

During their laboratory session, after viewing the potential dating partner’s profile, participants completed a 2-item measure assessing the extent to which they were romantically interested in the potential romantic partner presented to them (“I am romantically interested in the person I just viewed,” “I am attracted to the person I just viewed”; 1 = strongly disagree, 7 = strongly agree; α = .78, M = 4.11, SD = 1.08).

Results and Discussion

We standardized all variables (M = 0, SD = 1). We then conducted a regression to test our central hypothesis that HSE compared to LSE participants would express greater spontaneous self-expansion to endorse more, compared to less, positive attributes when presented with a potential dating partner that they were romantically interested in. To this end, we predicted post-profile viewing ratings of the not me attribute from participants’ romantic interest in the potential partner, self-esteem, pretesting judgments of the not me attribute’s valence, and all possible interactions. We controlled for participants’ pretesting ratings of the not me attribute in our analysis; as such, our result represent participants’ residualized change scores in their ratings of the attribute (Cohen, Cohen, West, & Aiken, 2003). Results are not meaningfully altered by omitting this covariate.²

As predicted, a three-way interaction emerged among romantic interest, self-esteem, and attribute valence, β = .11, t(209) = 2.26, p = .02, 95% confidence interval (CI) [.01, .20], overall R ² = .35 (Table 2). We next decomposed this three-way interaction to examine the predicted two-way interaction between self-esteem and attribute valence across romantic interest. The predicted interaction between self-esteem and attribute valence emerged among participants who expressed higher romantic interest in the potential partner (+1 SD; Figure 2A), β = .21, t(209) = 2.63, p = .009, 95% CI [.05, .37], but not among participants who expressed lower romantic interest, consistent with Hypothesis 2 (−1 SD; Figure 2B), β = −.003, t(209) = −0.06, p = .95, 95% CI [−.12, .11].

Table 2.

Study 1: Parameters From the Full Model Predicting Post-Target Viewing Self-Rating Scores on the Not Me Attributes.

Parameter	β	t	p	95% CI for Mean [LL, UL]
Intercept	−.12	−2.46	.014	[−.20, −.02]
Pretesting rating	.41	8.32	<.001	[.31, .51]
Romantic interest	.14	2.55	.011	[.03, .24]
Self-esteem	−.02	−0.33	.742	[−.11, .08]
Attribute valence	.14	2.84	.004	[.04, .24]
Interest × Esteem	−.01	−0.28	.776	[−.11, .09]
Interest × Valence	−.01	−0.15	.882	[−.11, .09]
Esteem × Valence	.10	1.98	.049	[.00, .20]
Interest × Esteem × Valence	.11	2.26	.024	[.01, .20]

Note. df = 209. UL = upper limit; LL = lower limit.

Tests of simple effects within high romantic interest revealed that, consistent with Hypothesis 1a, for HSE participants (+1 SD), more positive attribute valence predicted greater endorsement, β = .34, t(209) = 3.17, p = .001, 95% CI [.13, .55]. For LSE participants (−1 SD), more positive attribute valence did not significantly predict endorsement, β = −.08, t(209) = −0.74, p = .46, 95% CI [−.29, .13]; however, descriptively, the association between attribute positivity and endorsement was negative for LSEs.

Auxiliary Analyses

We ran two auxiliary regressions, controlling for the main effect of gender and age, respectively. In both analyses, our key interaction remained robust. We conducted a final auxiliary regression wherein we predicted participants’ romantic interest in the target from their self-esteem, attribute valence rating of the not me attribute, and their interaction. No significant effects emerged. We included this final auxiliary analysis to demonstrate that variations in self-esteem, or attribute positivity did not account for participants’ romantic interest in the target.

Study 2

Study 2 expanded upon the findings of Study 1 by introducing an experimental manipulation of romantic context as well as by examining both positive and negative attribute adoption rather than treating attribute valence as a continuum.

Method

Participants

Two hundred and ninety adult, nonstudent participants qualified for the current study via Amazon’s Mechanical Turk (Buhrmester, Kwang, & Gosling, 2011). To be eligible to participate, all participants had to generate two usable “not me” attributes, one positive and one negative, on a self-rating task completed at the start of the study session (see below), as well as report that they were not currently involved in a romantic relationship, were heterosexual, and were interested in online dating. Of these eligible participants, 27 failed an attention check measure at the conclusion of the study, and 29 viewed the target individual’s profile for less than 10 s, both resulted in exclusion from the data set. After discarding these individuals, 234 usable participants remained in the data set (111 female; M _age = 34.25, SD = 11.61). See the Online Supplementary Materials for further discussion of recruitment as well as a discussion of power considerations; however, the sample was adequate to reach .80 power, with an α of .05, estimating a rough effect size of f ² = .12 for our key interaction of interest.

Procedure

Procedures in Study 2 were similar to Study 1. Participants first completed demographic measures, the self-rating task, and dispositional self-esteem. They later viewed a novel target before answering questions regarding their interest in the target and repeating the self-rating task.

Study 2 procedures differed from Study 1 in three ways:

All measures were collected within a single, ∼20 min online study session. In order to reduce participants’ suspicion regarding repeating our self-rating task, we included filler items (five questions asking participants to allocate hypothetical money) and told participants that they might see the same questions more than once.

Participants saw the same attributes on the self-rating task as in Study 1 embedded in the target profile; however, we identified two idiosyncratic not me attributes for each participant: one positive attribute, an attribute with an average score of 5 or above on the Study 1 attribute valence measure, and one negative attribute, an attribute with an average score of 3 or below on the Study 1 attribute valence measure. Attributes positivity and negativity were based on Study 1 valence ratings since participants completed all measures in Study 2 in a single session.

In Study 2, we introduced an experimental manipulation. Specifically, participants were randomly assigned to view the target individual in one of the three contexts. In two contexts, the target individual was portrayed as a potential romantic partner who participants viewed for an online dating service, and in the third context, the target individual was portrayed as a potential job candidate who participants viewed for an online job-finding service (e.g., Richman, Slotter, Gardner, & DeWall, 2015). The first romantic context, called the public romantic context (n = 79), emphasized that participants’ responses on the subsequent self-rating task would be viewed by the target individual, as part of the system’s “profile sharing network.” Thus, their responses were public and perhaps prone to self-presentation. The second romantic context, the private romantic context (n = 73), emphasized that participants’ responses were entirely private and would not be viewed by the target individual, thus reducing self-presentation. The third context, the nonromantic context (n = 82), portrayed the target individual as someone looking for new employment and thus served as our control condition designed not to elicit the romantic interest necessary to motivate spontaneous self-expansion (e.g., Richman et al., 2015; Slotter & Gardner, 2009). Instructions regarding response privacy in the nonromantic context were the same as in the private romantic context.

As romantic interest should be irrelevant in the nonromantic context, we predicted that spontaneous self-expansion would not occur in this condition. In contrast, and replicating previous research, we predicted the spontaneous self-expansion would occur in the other two contexts, as a function of our key variables (e.g., Richman et al., 2015). It is worth noting the participants who were discarded from the data set as outlined above were roughly evenly distributed across contexts. Irrespective of context, participants were asked to view the target individuals’ profile for as long as they wanted, but at least 30 s (M = 27.57 s, SD = 10.02).

Measures

See Table 3 for bivariate correlations.

Table 3.

Study 2: Bivariate Correlations Among Key Variables.

Variable	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
(1) Preprofile positive rating
(2) Preprofile negative rating	−.19***
(3) Post-profile positive rating	.62***	.17**
(4) Post-profile negative rating	.19**	.70****	.15*
(5) Self-esteem	.06	−.01	.27***	−.14*
(6) Romantic interest	.14*	.17*	.35***	.29***	.01
(7) Gender	.12	.12	.16*	.05	.16*	−.03
(8) Age	−.10	−.03	−.22**	−.01	−.08	−.06	.05
(9) Profile viewing time	−.02	.05	−.05	.03	−.02	.01	−.01	.04

Note. Gender was coded as follows: Male = 0, Female = 1.

*p < .05, **p < .01. ***p < .001.

Self-rating task

The self-rating task was similar to Study 1 (1 = not at all characteristic of me, 7 = extremely characteristic of me; M = 0.80, SD = 0.79). One positive and one negative attribute were selected for each participant (M _pos = 2.29, SD = 1.17; M _neg = 2.18, SD = 1.19) and embedded in the target’s profile that participants subsequently viewed. Participants were only eligible to participate in the study if they generated these two “not me” attributes.

Participants completed the self-rating task a second time after viewing the novel target’s profile information (M _pos = 2.64, SD = 1.58; M _neg = 2.12, SD = 1.37). We were interested in change in participants’ ratings of both idiosyncratic not me attributes from pre- to post-viewing the novel target’s profile.

Self-esteem

At the beginning of the study, participants also completed the 10-item measure of their global self-esteem used in Study 1 (α = .94, M = 5.00, SD = 1.40; Rosenberg, 1965).

Interest

After viewing the novel target’s profile information, participants completed a modified measure of interest. Specifically, since some participants did not see a potential romantic target, we altered the items included to be about motivation for affiliation rather than romantic attraction specifically (“I am very interested in meeting the person I just viewed” and “I like the person I just viewed”; α = .84, M = 3.43, SD = 1.42).

Results and Discussion

We standardized all continuous predictors (M = 0; SD = 1); context was effect coded (public romantic = 1, private romantic = 0, nonromantic = −1). We then conducted two regressions. The first tested our central hypothesis that HSE compared to LSE participants would express greater spontaneous self-expansion to endorse positive attributes when presented with a potential dating partner that they were romantically interested in (Hypotheses 1a and 2). The second tested our hypothesis that LSE compared to HSE participants would express greater spontaneous self-expansion to endorse negative attributes when presented with a potential dating partner they liked (Hypotheses 1b and 2). Thus, we predicted post-profile viewing ratings of the positive and negative not me attributes, respectively, from participants’ interest in the potential partner, self-esteem, randomly assigned context, and all possible interactions. We controlled for participants’ preprofile viewing ratings of both not me attributes, how long participants’ spent viewing the target individual’s profile, and the post-profile viewing score of the negative attribute in the regression predicting endorsement of the positive attribute and vice versa for the regression predicting the endorsement of the negative attribute. Results were not meaningfully altered by omitting these covariates.

Positive Attributes

When predicting endorsement of the positive not me attribute, the predicted interaction emerged among romantic interest, self-esteem, and context, β = .14, t(221) = 3.11, p = .002, 95% CI [.05, .23], overall R ² = .64 (Table 4). We next decomposed this interaction to examine the predicted two-way interaction between self-esteem and context across romantic interest. The predicted significant interaction emerged among participants who expressed higher romantic interest in the target individual (+1 SD; Figure 3A), β = .22, t(221) = 3.11, p = .002, 95% CI [.08, .35], but not among participants who expressed less romantic interest, consistent with Hypothesis 2 (−1 SD; Figure 3B), β = −.07, t(221)=−0.99, p = .32, 95% CI [−.20, .07].

Table 4.

Study 2: Parameters From the Full Model Predicting Post-Target Viewing Self-Rating Scores on the Positive Not Me Attributes.

Parameter	β	t	p	95% CI for Mean
Intercept	−.01	−0.30	.76	[−.09, .06]
Preprofile positive rating	.56	13.11	<.001	[.47, .64]
Interest	.25	5.49	<.001	[.16, .33]
Self-esteem	.22	5.21	<.001	[.13, .30]
Context	.21	4.13	<.001	[.11, .31]
Interest × Esteem	.20	4.94	<.001	[.11, .28]
Interest × Context	.18	3.49	<.001	[.08, .28]
Esteem × Context	.08	1.47	.14	[−.02, .18]
Interest × Esteem × Context	.14	3.11	.002	[.05, .23]
Preprofile negative rating	.01	0.16	.87	[−.10, .13]
Post-profile negative rating	.01	0.21	.83	[−.12, .15]
Profile viewing time	−.03	−0.73	.47	[−.11, .05]

Note. df = 221.

Tests of simple effects within high romantic interest revealed that, consistent with Hypothesis 1a, interpersonal context was associated with attribute endorsement among HSE participants (+1 SD), β = .60, t(221) = 6.13, p < .001, 95% CI [.41, .80], but not among LSE participants (−1 SD), β = .16, t(221) = 1.58, p = 11, 95% CI [−.04, .38].

More specifically, among participants in both the public, β = .71, t(221) = 6.07, p < .001, 95% CI [.48, .95], and the private romantic context, β = .50, t(221) = 3.63, p < .001, 95% CI [.23, .77], HSE predicted greater attribute endorsement. Also as predicted, among participants in the nonromantic context, self-esteem was not associated with attribute endorsement, β = .09, t(221) = 1.32, p = .19, 95% CI [−.04, .22].

Negative Attributes

When predicting spontaneous self-expansion to endorse the negative not me attribute, the predicted interaction emerged among context, romantic interest, and self-esteem, β = −.16, t(221)=−3.48, p = .001, 95% CI [−.25, −.07], overall R ² = .64 (Table 5). We decomposed this interaction to examine the predicted two-way interaction between self-esteem and context across romantic interest. The predicted interaction between self-esteem and context emerged among participants who expressed higher romantic interest in the target individual (+1 SD; Figure 4A), β = −.38, t(221) = −5.68, p < .001, 95% CI [−.51, −.25], but not among participants who expressed less romantic interest, consistent with Hypothesis 2 (−1 SD; Figure 4B), β = −.06, t(221) = −0.94, p = .35, 95% CI [−.19, .07].

Table 5.

Study 2: Parameters From the Full Model Predicting Post-Target Viewing Self-Rating Scores on the Positive Not Me Attributes.

Parameter	β	t	p	95% CI for Mean
Intercept	.01	0.20	.84	[−.07, .09]
Preprofile negative rating	.64	15.32	<.001	[.55, .72]
Romantic interest	.22	4.83	<.001	[.13, .31]
Self-esteem	−.14	−3.17	.001	[−.22,−.05]
Context	.12	2.45	.02	[.02, .23]
Interest × Esteem	−.15	−3.75	.002	[−.23,−.07]
Interest × Context	.15	3.02	.003	[.05, .25]
Esteem × Context	−.22	−4.48	<.001	[−.31,−.12]
Interest × Esteem × Context	−.16	−3.48	.001	[−.25,−.07]
Preprofile positive rating	.04	0.85	.40	[−.06, .16]
Post-profile positive rating	.01	0.21	.83	[−.12, .14]
Profile viewing time	−.01	−0.15	.88	[−.09, .07]

Tests of simple effects within high romantic interest revealed that, consistent with Hypothesis 1b, context was associated with attribute endorsement among LSE participants (−1 SD), β = .66, t(221) = 6.83, p < .001, 95% CI [.46, .84], but not among HSE participants (+1 SD), β = −.10, t(221) = −0.93, p = .36, 95% CI [−.31, .11].

More specifically, among participants in both the public, β = −.63, t(221) = −5.30, p < .001, 95% CI [−.86, −.39], and private romantic context, β = .43, t(221) = −2.78, p = .007, 95% CI [−.76, −.12], LSE predicted greater attribute endorsement. In contrast, among participants in the nonromantic context, self-esteem was not associated with attribute endorsement, β = .10, t(221) = 1.37, p = .18, 95% CI [−.04, .24].

Auxiliary Analyses

Finally, we conducted the same auxiliary analyses as in Study 1 and found similar, nonsignificant, results.

General Discussion

Past research demonstrates that motivation to be romantically involved with a partner is sufficient to promote spontaneous self-expansion, even when the attributes being adopted are negative (e.g., Slotter & Gardner, 2012). The present research investigated self-esteem as a moderator of people’s spontaneous self-expansion to adopt positive and negative attributes. Drawing from self-verification theory (Swann, 1983), we predicted that HSE would be associated with greater adoption of positive attributes (Hypothesis 1a). Conversely, we predicted that LSE would be associated with greater adoption of negative attributes (Hypothesis 1b). We also predicted that these effects would only be relevant among people who were motivated by romantic interest (Hypothesis 2) in the potential partner. Data from two studies supported these predictions.

The novel contributions of the current research are 3-fold. First, the present work expands on the evidence showing that people will adulterate their self-concepts through the addition of negative attributes (Slotter & Gardner, 2012). Furthermore, the present findings provide additional empirical evidence for a recent theoretical idea that individual differences may predict people self-expanding in different ways (e.g., Slotter & Lucas, 2013). Future research should examine the implications of spontaneously self-expanding in more versus less positive ways. What are the implications of this for their self-views and self-esteem over time or well-being outcomes?

Second, the present work has implications for the role of self-verification in close relationships (e.g., Swann, 1983), even in the earliest stages. The current findings are among the first to demonstrate that self-verification consistent social cognition can occur even when people are simply evaluating a potential dating partner. In our studies, relationship initiation hasn’t taken place, and yet, people are adding attributes to their self-concepts in ways that are consistent with their existing self-views. Future research should try to determine the exact mechanism behind this effect. It could be that people are setting themselves up to be viewed by a new dating partner as they view themselves. Alternatively, people may be attempting to keep their self-views consistent due to their individual motivation to seek cognitive consistency.

Third, the current studies shed light on the early relationship processes of people who differ in self-esteem. Much research has been conducted examining how self-esteem is reflected in people’s behavior in ongoing relationships (e.g., Murray, Holmes, & Collins, 2006), but less work examines self-esteem’s role in relationship initiation. It is possible that spontaneous self-expansion for LSEs may confirm, or even exacerbate, their already negative self-views, which could trigger the self-protective tendencies that harm their relationships. Future research should investigate how this process may influence subsequent relational behaviors, as well as how other, related individual differences (i.e., attachment anxiety) might be implicated.

Recent research focused altering LSEs self-views to be more positive (e.g., Stinson et al., 2010). The rationale is that increased positive self-views might help to alleviate some of the negative consequences of LSE, especially in romantic contexts. Unfortunately, LSEs are resistant to receiving positive information about themselves, including compliments from romantic partners (e.g., Marshall et al., 2015). One way to increase LSE individuals’ consideration of positive self-information is through promoting their feelings of self-compassion (Neff, 2012). Future research should examine whether self-compassion is a feasible point of intervention and whether increasing positive self-views increases self-esteem and relationship-promoting behavior.

The current research was limited in that it did not assess any ongoing, behavioral effects. Future research should investigate whether individuals engage in spontaneous self-expansion to promote beneficial similarity with a desired partner (e.g., Aron, 2003) or whether they are simply cognitively picturing themselves as they would be if a relationship began. Similarly, future research should examine the temporal stability of spontaneous self-expansion beyond laboratory settings. Although motivation appears to be sufficient to initiate the spontaneous self-expansion, perhaps subsequent shared experience with the desired romantic partner determines whether self-expansion continues. Another limitation of the current work is that did not fully manipulate our constructs of interest.

In conclusion, the current research illuminates self-esteem as an important moderator of people’s spontaneous self-expansion to include novel positive versus negative attributes into their identities. When motivated by romantic interest in a partner, people self-expand to adopt more positive attributes to the extent they are higher in self-esteem, and more negative attributes to the extent they are lower in self-esteem. This tendency to engage in self-view consistent self-expansion may reflect people’s desires for consistency and self-verification.

Supplemental Material

Supplemental Material, Making_sure_you_see_the_real_me_Supplementary_Materials_2_2019 - Making Sure You See the Real Me: The Role of Self-Esteem in Spontaneous Self-Expansion

Supplemental Material, Making_sure_you_see_the_real_me_Supplementary_Materials_2_2019 for Making Sure You See the Real Me: The Role of Self-Esteem in Spontaneous Self-Expansion by Erica B. Slotter and Lyuboslava Kolarova in Social Psychological and Personality Science

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Supplemental Material

The supplemental material is available in the online version of the article.

Notes

References

Alicke

Sedikides

(2009). Self-enhancement and self-protection: What they are and what they do. European Review of Social Psychology, 20, 1–48.

Aron

(2003). Self and close relationships. In Leary

M. R.

Tangney

J. P.

(Eds.), Handbook of self and identity (pp. 442–461). New York, NY: Guilford Press.

Aron

E. N.

(1997). Self-expansion motivation and including other in the self. In Duck

(Ed.), Handbook of personal relationships: Theory, research, and interventions (2nd ed., pp. 251–270). Hoboken, NJ: John Wiley.

Aron

E. N.

Tudor

Nelson

(1991). Close relationships as including other in the self. Journal of Personality and Social Psychology, 60, 241–253.

Baumeister

R. F.

(2010). The self. In Baumeister

R. F.

Finkel

E. J.

(Eds.), Advances in social psychology: The state of the science (pp. 139–175). London, England: Oxford University Press.

Baumeister

R. F.

Campbell

J. D.

Krueger

J. I.

Vohs

K. D.

(2003). Does high self-esteem cause better performance, interpersonal success, happiness, or healthier lifestyles? Psychological Science in the Public Interest, 4, 1–44.

Buhrmester

Kwang

Gosling

S. D.

(2011). Amazon’s Mechanical Turk a new source of inexpensive, yet high-quality, data? Perspectives on Psychological Science, 6, 3–5.

Campbell

J. D.

(1990). Self-esteem and clarity of the self-concept. Journal of Personality and Social Psychology, 59(3), 538–549.

Cohen

West

S. G.

Aiken

L. S.

(2003). Applied multiple regression/correlation analysis in the behavioral sciences (3rd ed., p. 76). Mahwah, NJ: Lawrence Erlbaum.

10.

Forest

A. L.

Wood

J. V.

(2012). When social networking is not working: Individuals with low self-esteem recognize but do not reap the benefits of self-disclosure on Facebook. Psychological Science, 23, 295–302.

11.

James

(1890). The principles of psychology. Cambridge, MA: Harvard University Press.

12.

Katz

Beach

S. R.

(2000). Looking for love? Self-verification and self-enhancement effects on initial romantic attraction. Personality and Social Psychology, 26, 1526–1539.

13.

Kwang

Swann

W. B.

Jr (2010). Do people embrace praise even when they feel unworthy? A review of critical tests of self-enhancement versus self-verification. Personality and Social Psychology Review, 14, 263–280.

14.

Marshall

Parker

P. D.

Ciarrochi

Sahdra

B. K.

Jackson

C. J.

Heaven

P. C.

(2015). Self-compassion protects against the negative effects of low self-esteem: a longitudinal study in a large adolescent sample. Personality and Individual Differences, 74, 116–121.

15.

Mashek

Stuewig

Furukawa

Tangney

(2006). Psychological and behavioral implications of connectedness to communities with opposing values and beliefs. Journal of Social and Clinical Psychology, 25, 404–428.

16.

Mattingly

B. A.

Lewandowski

G. W.

McIntyre

K. P.

(2014). “You make me a better/worse person”: A two-dimensional model of relationship self-change. Personal Relationships, 21, 176–190.

17.

McConnell

A. R.

(2011). The multiple self-aspects framework: Self-concept representation and its implications. Personality and Social Psychology Review, 15, 3–27.

18.

Murray

S. L.

Derrick

J. L.

Leder

Holmes

J. G.

(2008). Balancing connectedness and self-protection goals in close relationships: A levels-of-processing perspective on risk regulation. Journal of Personality and Social Psychology, 94, 429–459.

19.

Murray

S. L.

Griffin

D. W.

Rose

Bellavia

(2003). Calibrating the sociometer: The relational contingencies of self-esteem. Journal of Personality and Social Psychology, 85, 63–84.

20.

Murray

S. L.

Holmes

J. G.

Collins

N. L.

(2006). Optimizing assurance: The risk regulation system in relationships. Psychological Bulletin, 132, 641–666.

21.

Neff

K. D.

(2012). The science of self-compassion. In Germer

C. K.

Siegel

R. D.

(Eds.), Compassion and wisdom in psychotherapy (pp. 79–92). New York, NY: Guilford Press.

22.

Richman

Slotter

E. B.

Gardner

W. L.

DeWall

C. N.

(2015). Reaching out by changing what’s within: Social exclusion predicts self-concept malleability. Journal of Experimental Social Psychology, 57, 64–77.

23.

Rosenberg

(1965). Society and the adolescent self-image. Princeton, NJ: Princeton University Press.

24.

Slotter

E. B.

(2016, December 31). Pre-registration of self-esteem and spontaneous self-expansion. Retrieved from osf.io/678hp

25.

Slotter

E. B.

Emery

L. F.

(2018). On the rebound: Self-concept malleability and the tendency to engage in rebound relationships predict self-concept clarity. Personal Relationships.

26.

Slotter

E. B.

Gardner

W. L.

(2009). Where do you end and I begin? Evidence for anticipatory, motivated self-other integration between relationship partners. Journal of Personality and Social Psychology, 96, 1137–1151.

27.

Slotter

E. B.

Gardner

W. L.

(2012). The dangers of dating the “bad boy” (or girl): Romantic desire encourages the adoption of even negative qualities of potential partners. Journal of Experimental Social Psychology, 48, 1173–1178.

28.

Slotter

E. B.

Lucas

G. M.

(2013). Validating a scale of self and partner change in romantic relationships: The Perceived Change in Relationships Scale. Self and Identity, 12, 177–185.

29.

Stinson

D. A.

Logel

Holmes

J. G.

Wood

J. V.

Forest

Gaucher

… Kath

(2010). The regulatory function of self-esteem: Testing the acceptance and epistemic signaling systems. Journal of Personality and Social Psychology, 99, 993–1013.

30.

Swann

W. B.

Jr (1983). Self-verification: Bringing social reality into harmony with the self. In Suls

Greenwald

A. G.

(Eds.), Social psychological perspectives on the self (Vol. 2, pp. 33–66). Hillsdale, NJ: Lawrence Erlbaum.

31.

Swann

W. B.

Jr (1990). To be adored or to be known: The interplay of self-enhancement and self-verification. In Sorrentino

R. M.

Higgins

E. T.

(Eds.), Foundations of social behavior (Vol. 2, pp. 408–448). New York, NY: Guilford.

32.

Swann

W. B.

Jr Hixon

J. G.

De La Ronde

(1992). Embracing the bitter truth: Negative self-concepts and marital commitment. Psychological Science, 3, 118–121.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.02 MB