Contents lists available at ScienceDirect
Journal of Adolescence
journal homepage: www.elsevier.com/locate/adolescence
Social media social comparison and identity distress at the collegetransition: A dual-path model
Chia-chen Yanga,∗, Sean M. Holdenb, Mollie D.K. Cartera, Jessica J. WebbaaDepartment of Counseling, Educational Psychology and Research, University of Memphis, 100 Ball Hall, Memphis, TN 38152, USAbOffice of Institutional Research, University of Memphis, 211 Administration Building, Memphis, TN 38152, USA
A R T I C L E I N F O
Keywords:Social mediaSocial comparisonRuminationReflectionIdentity
A B S T R A C T
Introduction: Social media provide a convenient platform for social comparison, an activity thatshould play an important role in youth's identity development at the transition to college. Yet, theidentity implications of online social comparison have not been thoroughly explored. Drawing onthe theories of social comparison, introspective processes, and identity distress, we examined adual-path model. The paths from two types of social media social comparison (i.e., comparison ofability and comparison of opinion) to two introspective processes (i.e., rumination and reflection)and finally to identity distress were tested.Methods: Short-term longitudinal survey data were collected from 219 college freshmen at a stateuniversity in the United States of America (Mage= 18.29, S.D.=0.75; 74% female; 41% White,38% Black).Results: Social comparison of ability on social media had a positive association with concurrentrumination, which predicted higher identity distress. In contrast, social comparison of opinion onsocial media had a positive relationship with concurrent reflection, which, however, did notpredict identity distress.Conclusion: Results indicate that different types of online social comparison yield distinct im-plications for young people's identity development. Largely, the study reaffirms the recentlyrising call for distinguishing the competition-based social comparison of ability from the in-formation-based social comparison of opinion. At the same time, the study expands currentknowledge of why these forms of social comparison may lead to differential outcomes, namelythrough the type of introspection they induce.
Identity development is a major developmental task for adolescents and emerging adults (Arnett, 2014; Erikson, 1968). The taskseems particularly challenging in modern society, where major socio-cultural changes make it a more complicated issue (Berman &Weems, 2011). For instance, scholars have discussed how globalization and economic recessions may require more time for identityexploration and delay youth's identity commitment and integration (Arnett, 2002; Sica, Sestito, & Ragozini, 2014). This study turnedto the growth of social media as another contributor to the complexity of identity negotiation in the digital age. Specifically, itexplored the identity implications of a widely experienced behavior on social media—social comparison, the process of comparingoneself with others as a way to understand and evaluate oneself (Festinger, 1954).
Identity development in the social media context is typically studied through the lens of online self-presentation (e.g., Manago,Graham, Greenfield, & Salimkhan, 2008; Michikyan, Dennis, & Subrahmanyam, 2015; Yang & Brown, 2016). Research of online
https://doi.org/10.1016/j.adolescence.2018.09.007Received 22 May 2018; Received in revised form 17 September 2018; Accepted 18 September 2018
∗ Corresponding author.E-mail address: firstname.lastname@example.org (C.-c. Yang).
social comparison, on the other hand, mostly focuses on its relationship with psycho-emotional rather than identity outcomes. Yet,social comparison has been described as an important component in the construction of self-concept and attainment of self-knowledge (Harter, 2012; Rosenberg, 1979), making the lack of attention to the identity implications of social comparison on socialmedia surprising. In particular, social comparison can provide a vital tool for identity development by fostering adolescents'knowledge of themselves, but it can also undermine adolescents' sense of self-adequacy, particularly as they evaluate themselvesrelative to their peers' positive self-presentation on social media.
The act of social comparison can be divided into two forms: the judgmental, competition-based social comparison of ability andthe non-judgmental, information-based social comparison of opinion (Festinger, 1954; Gibbons & Buunk, 1999; Suls, Martin, &Wheeler, 2000, 2002). The two appear to involve different introspective processes, rumination and reflection (Trapnell & Campbell,1999), which may exacerbate or alleviate the distress young people experience in their identity exploration and construction.Drawing on the theories of social comparison (Festinger, 1954), introspective processes (Trapnell & Campbell, 1999), and identitydistress (Berman, Montgomery, & Kurtines, 2004), this study examined a dual-path model, in which the two types of social com-parison (ability and opinion) on social media were hypothesized to predict identity distress through rumination and reflection,respectively.
Based on our short-term longitudinal survey data, the model was tested with a sample of college freshmen, as the transition tocollege presents a heightened need for identity construction (Stephenson-Abetz & Holman, 2012; Thomas, Briggs, Hart, & Kerrigan,2017) and can cause identity distress when college freshmen become too overwhelmed to develop a coherent sense of self. The goal ofthe study was to clarify how the two types of social comparison may relate to young people's identity experience in the digital age at amajor developmental transition.
1.1. Identity distress
Identity distress is defined as “severe subjective distress regarding inability to reconcile aspects of the self into a relativelycoherent and acceptable sense of self” (American Psychiatric Association, 1980, p. 65). Identity distress is associated with low self-esteem, more mental health issues (Samuolis, Barcellos, LaFlam, Belson, & Berard, 2015; Sica et al., 2014), and greater severity ofpsychological symptoms (Berman, Weems, & Petkus, 2009). Originally, this experience was conceptualized as the core feature ofbroader psychological disorders associated with identity, such as the Identity Disorder mentioned in the DSM-III and III-R (AmericanPsychiatric Association, 1980, 1987). In the DSM-V, although identity distress is no longer listed as a distinct disorder, characteristicsassociated with identity distress, such as impairment in self-direction, are recognized as key components across personality anddissociative disorders (American Psychiatric Association, 2013). The change in the DSM-V implies a more prominent position ofidentity problems in the current version of the DSM. In addition to its clinical implications, identity distress can also be experiencedby those without a clinically diagnosable set of symptoms, and the mild to moderate distress surrounding identity formation can stilllead to maladaptive consequences over time (Berman et al., 2004). Given the plethora of pressures and conflicts modern societypresents, identity distress appears to be a common problem for today's youth (Berman et al., 2004). Compared to previous gen-erations, individuals are now faced with an ever-increasing number of choices about what they value and how they want to live(Berman et al., 2009), in part because society has become much more complex, diverse, and interconnected (Arnett, 2002; Bermanet al., 2004).
Identity distress may be especially salient for those in times of transition, when identity is less firmly defined amidst familiarpeople and contexts. The college transition presents one such period. This transition often entails a sense of loss and discontinuity, asstudents leave behind familiar environments and social supports, lose some of their previous sense of belonging, and reconstruct theirknowledge of themselves and their contexts (Chow & Healey, 2008; Scanlon, Rowling, & Weber, 2007). This is mirrored by thegeneral identity trajectories mentioned in Kroger, Martinussen, and Marcia's (2010) meta-analysis, in which the authors found thatfrom late adolescence to young adulthood, there was a growing proportion of youth reaching the identity achievement status. Theinitial lack of identity achievement at the ages of 18 and 19 may coincide with the transition into college, which requires significantindividual effort as freshmen make adjustments based on their new context. In the digital age, the transition to college is furthercomplicated by the rich materials for social comparison available on social media. Depending on how students approach thesematerials, the media may provide opportunities to ameliorate identity distress or present risks that exacerbate it.
1.2. Social comparison on social media
In the absence of objective verifications of individual attributes, people seek information about themselves through socialcomparison, particularly by evaluating their abilities and opinions against comparable others' (Festinger, 1954). Social comparison ofability and that of opinion differ in the manner of how the comparison is performed. The competition-based social comparison ofability is inherently judgmental. When people engage in this type of comparison, they view the comparison targets as competitors(Park & Baek, 2018) and assess whether their own performance or achievement is superior or inferior to the targets' (Festinger, 1954;Gibbons & Buunk, 1999). In contrast, during the information-based social comparison of opinion, individuals view comparisontargets as informants, consultants, or role models (Park & Baek, 2018). They attend to the similarities and differences (rather thansuperiority and inferiority) of attitudes, beliefs, and values between themselves and the targets, and determine whether they mightmake adjustments to their own perspectives (Festinger, 1954; Gibbons & Buunk, 1999; Suls et al., 2000).
C.-c. Yang et al.
The affordances of social media create a rich context for positive self-presentation (Yang & Brown, 2016), which nourishesupward social comparison, in which individuals compare themselves with those who seem better off (Vogel, Rose, Roberts, & Eckles,2014). Likely due to the prevalence of upward social comparison on social media, most existing research of online social comparisonfocused on the judgmental form of comparison, such as comparisons of social connectedness, likability, superiority, and physicalattractiveness (e.g., Chua & Chang, 2016; Feinstein et al., 2013; Vogel et al., 2014). Given the competitive nature shared by all thesecomparisons, they can be viewed as examples of social comparison of ability, broadly defined (Yang, Holden, & Carter, 2018). Thesestudies mostly reported negative psycho-emotional implications of such judgmental comparison, including negative self-views, ne-gative emotions, and depressive symptoms (de Vries & Kühne, 2015; Feinstein et al., 2013; Lim & Yang, 2015; Vogel et al., 2014;Weinstein, 2017). Conversely, less emphasis has been given to the non-competitive comparison of opinion, but the few studiesattending to it revealed its positive associations with psychosocial and emotional outcomes. For instance, online social comparison ofopinion is related to better social adjustment among college students (Yang & Robinson, 2018), as well as lower levels of upwardcontrastive emotions (e.g. envy and depression) and thus higher life satisfaction (Park & Baek, 2018).
Whereas scholars have explored the psycho-emotional implications of online social comparison, especially the judgmental type,little research has explicitly measured how online social comparison would associate with young people's identity states and what theunderlying mechanisms would be. It is particularly important to advance this line of research in the context of the transition tocollege, because college freshmen appear to frequently perform both types of social comparison on social media as a way to makesense of the new community and (re)construct an identity. For example, by viewing peers' social media pages, college freshmen get anidea of how well-adjusted they are themselves, but also feel the pressure to present themselves only in a positive light (Thomas et al.,2017), which may hinder full expression and exploration of the self that are essential for identity growth (Gardner & Davis, 2014).Social comparison of ability may be an underlying process; there may be a concern that posting about one's academic struggles,homesickness, and social anxiety (Thomas et al., 2017) would suggest one's poorer academic and social competencies relative topeers. Disconcertingly, such judgmental comparison is associated with avoidance of seeking and processing self-related information,and thus a lower level of identity clarity (Yang et al., 2018). In other words, social comparison of ability can impede identitydevelopment. At the same time, the rich information generated by peers also makes social media a convenient platform for socialcomparison of opinion, an important process for familiarizing oneself with the norms in the new community (Suls et al., 2000), whichshould facilitate adjustment and identity reorganization. Indeed, online social comparison of opinion at the college transition isassociated with active engagement in identity construction (Yang et al., 2018), suggesting the instrumental role of this type ofcomparison in college freshmen's identity development.
It appears that, for college freshmen's identity development, social comparison of ability on social media may be maladaptive,whereas that of opinion may be adaptive. As a self-evaluation process, both types of social comparison necessarily involve in-trospection and attention to the self; however, the two types of comparison may have different identity implications because they areassociated with different manners of introspection. We discuss two related but different types of introspection in the followingsection.
1.3. Rumination and reflection
Rumination and reflection are related yet distinct constructs that involve introspection about oneself and/or one's own experi-ences. Despite their similar self-attentive nature, rumination and reflection are distinguished by their underlying motivations, pro-blem-solving orientation (or lack thereof), and associated psychological states. Rumination is motivated by threats to the self and isassociated with neuroticism (Trapnell & Campbell, 1999). When people ruminate (or “brood”), they are trapped in repetitive andanxious thoughts about themselves or their experiences (Treynor, Gonzalez, & Nolen-Hoeksema, 2003), which contributes to andexacerbates depression (Nolen-Hoeksema, Wisco, & Lyubomirsky, 2008; Takano & Tanno, 2009). In contrast, reflection is motivatedby epistemic curiosity about the self and is associated with openness to experience (Trapnell & Campbell, 1999). When individualsengage in reflection (or “pondering”), they analyze their thoughts and feelings, often with an attempt to overcome problems (Treynoret al., 2003). Reflection is associated with personal growth (Harrington & Loffredo, 2011) and search for meaning in life (Newman &Nezlek, 2017). In short, rumination refers to maladaptive self-attentiveness characterized by passively brooding upon one's thoughtsand emotions, whereas reflection refers to adaptive self-focus characterized by active attempts to understand oneself and one'ssituation, either for the purpose of solving problems or for sheer epistemic pleasure.
In the context of identity distress, rumination and reflection may pave two opposite paths. Rumination is neurotic in nature andfixated on negative thoughts; it inhibits problem solving and instrumental behavior and hinders social support (Nolen-Hoeksemaet al., 2008; Trapnell & Campbell, 1999). Therefore, it would be difficult for a person who ruminates to develop a coherent and self-embracing identity, and thus rumination may increase one's identity distress. Indeed, rumination is negatively associated with self-acceptance (Harrington & Loffredo, 2011), identity integration, and identification with commitment (Luyckx et al., 2007). A relatedconstruct, ruminative exploration, is also related to low identity commitments (Beyers & Luyckx, 2016). On the other hand, re-flection, which correlates with higher identity commitment at the bivariate level (Luyckx et al., 2007), may reduce identity distressbecause it facilitates better self-understanding and enables one to make sense of distressing experiences. In line with this, interventionprograms that effectively reduce identity distress (Berman, Kennerley, & Kennerley, 2008; Meca et al., 2014) seem to tap into thishealthy means of introspection. Further, in a study involving college students' narratives of distressing experiences in relation to bothtypes of self-attentiveness, Marin and Rotondo (2017) found that increases in self-criticism, a form of rumination, predicted higheridentity distress; increases in reflection, in contrast, predicted lower identity distress and higher self-esteem over a one-week span.
Both rumination and reflection are highly salient processes at the transition to college. During this period, it is common to
C.-c. Yang et al.
experience the tensions between the past and new selves, the desires for both individuality and conformity (Stephenson-Abetz &Holman, 2012; Thomas et al., 2017), and the needs for both autonomy and relational connection (Radmacher & Azmitia, 2016), all ofwhich may prompt college freshmen to engage in either or both types of introspection. Today, as social media become an integral partin students' negotiation of identity and relationships across the transition to college (Stephenson-Abetz & Holman, 2012; Thomaset al., 2017; Yang, 2016), rumination and reflection may be frequently involved in college freshmen's social media use, such as socialcomparison of ability and opinion, and may serve as important mechanisms explaining how online social comparison may influencefreshmen's identity experiences.
1.4. Current study
We proposed a dual-path model, hypothesizing that college freshmen's social comparison of ability and opinion on social mediawould predict identity distress differentially through their respective associations with rumination and reflection (see Fig. 1). Socialcomparison of ability on social media can easily compromise one's self-views (de Vries & Kühne, 2015; Vogel et al., 2014), and suchthreats to the self may motivate one to ruminate (Trapnell & Campbell, 1999), which then predicts higher identity distress (Marin &Rotondo, 2017). Rumination has been identified as a mediator between judgmental online social comparison and depressivesymptoms (Feinstein et al., 2013), and it may play a similar role in the relationship between ability comparison and identity distress.In contrast, online social comparison of opinion is associated with active engagement in identity exploration and construction (Yanget al., 2018), and the attempt to understand and explore oneself happens to be a characteristic of reflection (Trapnell & Campbell,1999). It appears that the non-judgmental comparison of opinion leaves room for individuals to curiously self-reflect rather than toruminate on how to catch up with others. Given that reflection is associated with lower identity distress (Marin & Rotondo, 2017),social comparison of opinion may contribute to lower identity distress if it does associate with higher levels of reflection. In sum, weproposed the following hypotheses:
H1. Social media social comparison of ability would associate with higher levels of rumination, which, in turn, would predict higheridentity distress.
H2. Social media social comparison of opinion would associate with higher levels of reflection, which, in turn, would predict loweridentity distress.
Because the transition to college is an ongoing process that continues after youth enter the new environment, it is a commonpractice to survey or interview freshmen (and sometime sophomores) about their college experiences in the research of collegetransition (e.g., Stephenson-Abetz & Holman, 2012; Thomas et al., 2017; Yang & Brown, 2016). Following this approach, we recruitedfreshmen from a state university during the fall semester (T1). Participants were recruited through both class announcements and e-mails. Interested students followed a link to grant informed consent and complete the initial survey. Students who completed theinitial survey were notified by e-mail of a follow-up survey during the spring semester (T2). On average, participants finished thesecond survey 3.6 months after the completion of the first one. To be eligible to join the study, participants had to use social media atleast once per week. Following the completion of the study, participants received course credit or a small amount of monetarycompensation.
A sample of 219 college freshmen completed the initial survey (Mage= 18.29, S.D.=0.75; 74% female; 41% White, 38% Black);the follow-up survey was completed by 136 participants from the sample (Mage= 18.24, S.D.=0.65; 74% female; 45% White, 37%Black). Females were overrepresented in the study, but ethnic distribution within the sample was representative of the undergraduatepopulation of the university (58% female; 49%White, 36% Black). According to an item assessing participants' preferred social mediaplatform, 35% (T1) and 37% (T2) of the participants selected Instagram as the most frequently used social media site; 27% (T1) and25% (T2) chose Facebook; 20% (T1) and 19% (T2) selected Twitter. Of the participants who responded with “Other,” 78% (or 14% ofthe entire sample at T1) and 69% (or 13% of the entire sample at T2) reported Snapchat as the social media used most often.Participant dropout (83 participants, 38% of the T1 sample) was not related to age, t(217)= 1.44, p= .15, gender, χ2 (1)= 0.04,p= .85, race/ethnicity, χ2 (2)= 2.53, p= .28, or preferred social media platform, χ2 (3)= 2.06, p= .56. There was no mean
Fig. 1. Hypothesized model. SMSC-Ab= Social Media Social Comparison of Ability; SMSC-Op= Social Media Social Comparison of Opinion.
C.-c. Yang et al.
difference in the focal variables between those who participated only at T1 and those who participated at both times: ts= 1.44 to1.74, ps= .08 to .15. As suggested by the results of the Little’s (1988) Missing Completely at Random test, the data were missingcompletely at random, χ2 (8)= 10.47, p= .23.
All the scales were administered at T1. Identity distress was also measured at T2 to allow for the examination of the longitudinalassociations.
2.2.1. Social media social comparison (SMSC)The Social Media Social Comparison Scale (Yang et al., 2018) was adapted from the Iowa-Netherlands Comparison Orientation
Measure (INCOM; Gibbons & Buunk, 1999) to capture social comparison activities on social media. On a 5-point Likert scale, par-ticipants considered the extent to which they compared themselves with others when using social media, and indicated how well eachstatement applied to them (1=Not at All, 5= Very Well). The scale includes an Ability subscale and an Opinion subscale. “On socialmedia, I compare what I have done with others as a way to find out how well I have done something” is a sample item of the Abilitysubscale (5 items; α=0.84). “On social media, I try to know what others in a similar situation would do” is a sample item of theOpinion subscale (4 items; α=0.87). Higher mean scores on either subscale indicated higher engagement in that type of socialcomparison activity.
2.2.2. RuminationSix items of the Rumination subscale of the Rumination-Reflection Questionnaire (RRQ; Trapnell & Campbell, 1999) were used to
measure rumination (α=0.88). Participants indicated how much they agreed with the statements on a 5-point Likert scale(1= Strongly Disagree, 5= Strongly Agree). “I tend to ‘ruminate’ or dwell over things that happen to me for a really long timeafterward” is a sample item. A higher mean score reflected a higher level of rumination.
2.2.3. ReflectionSix items of the Reflection subscale of the RRQ (Trapnell & Campbell, 1999) were used to measure reflection (α=0.82). Par-
ticipants indicated how much they agreed with the statements on a 5-point Likert scale (1= Strongly Disagree, 5= Strongly Agree). “Ioften explore my ‘inner’ self” is a sample item. A higher mean score reflected a higher level of reflection.
2.2.4. Identity distressThree items of the Identity Distress Survey (IDS; Berman et al., 2004) measuring participants' overall level of identity distress were
administered on a 5-point Likert scale (αT1= 0.79, αT2= 0.79). These items are frequently used to measure young people's globalidentity distress (e.g., Gandhi, Luyckx, Maitra, & Claes, 2015). Participants were informed that people can be upset, distressed, orworried over the issues of long-term goals, career choice, friendships, sexual orientation, religion, values or beliefs, and grouployalties. Then they were asked about their overall level of discomfort and distress about these issues (1=Not at All, 5= VerySeverely), the extent to which uncertainty over these issues as a whole had interfered with their life (1=Not at All, 5= Very Severely),and the duration of their distress over these issues as a whole (1=Never or Less Than a Month, 5=More Than 12Months). A highermean score reflected a higher level of identity distress.
2.3. Data analysis
Using Mplus 7, we started with a confirmatory factor analysis (CFA) to examine the factor structure of the scales, followed by apath analysis testing the hypothesized model. Given the categorical nature of the scale items, we used the mean- and variance-adjusted weighted least squares (WLSMV) estimator for CFA, and then adopted maximum likelihood (ML) for the path analysis, alongwith a bootstrap of 5000 iterations. Missing data were handled by using pairwise deletion in WLSMV and full information maximumlikelihood in ML. Model fit was considered acceptable when most of the following criteria were met: A non-significant χ2 test(Barrett, 2007), the root mean square error of approximation (RMSEA) being smaller than 0.08 (Browne & Cudeck, 1993), and thecomparative fit index (CFI) and the Tucker-Lewis index (TLI) being greater than 0.90. The cut-off value of 0.90 was chosen over a cut-off of 0.95 for CFI and TLI because the 0.95 threshold tends to over-reject models when the sample is smaller than 250 (Hu & Bentler,1999). In addition to the model fit, results of CFA were also assessed by taking into account item factor loadings, scale compositereliability (CR), and average variance extracted (AVE). Item factor loadings were expected to be greater than 0.40. CR and AVE wererecommended to be greater than 0.70 and 0.50, respectively (Fornell & Larcker, 1981; Hair, Ringle, & Sarstedt, 2011).
For the path model, we regressed T2 identity distress on T1 rumination and reflection, which were regressed on T1 social mediasocial comparison of ability (SMSC-Ability) and T1 social media social comparison of opinion (SMSC-Opinion), respectively. Becausegender, race/ethnicity, and amount of time using communication technologies were found to associate with identity distress (Bermanet al., 2004; Cyr, Berman, & Smith, 2015), we controlled for their effects on T2 identity distress. We also controlled for the gendereffects on rumination and reflection (Luyckx et al., 2007; Nolen-Hoeksema et al., 2008; Treynor et al., 2003), and had T1 identitydistress serve as a control variable for the two T1 mediators and T2 identity distress. The error terms of the two mediators wereallowed to covary.
Two alternative models were examined. First, because SMSC and introspection were measured concurrently, we tested a reverse-
C.-c. Yang et al.
order model in which SMSC became the mediators of the relationships between introspection and identity distress. This model wastested against the hypothesized model by comparing their fit indices. Second, due to the conceptual overlap between the two types ofsocial comparison and between the two mediators, we tested an alternative model in which two more paths were included: T1 SMSC-Ability to T1 reflection, and T1 SMSC-Opinion to T1 rumination. A χ2 difference test was performed to determine whether model fitdiffered significantly between this model and the hypothesized model. The Bayesian information criterion (BIC) was also considered;a lower BIC suggested a better model.
Results of CFA suggested that the presumed factor structure was acceptable: χ2 (309)= 619.30, p < .001; RMSEA=0.068, 90%CI [0.060-0.075]; CFI= 0.95; TLI= 0.95. Factor loadings ranged from 0.45 to 0.92, and most loadings were greater than 0.70. ScaleCR ranged from 0.85 to 0.92, and AVE ranged from 0.52 to 0.70. Descriptive statistics and correlations of the major variables areavailable in Table 1.
The hypothesized path model fit well: χ2 (8)= 5.14, p= .74; RMSEA=0.000, 90% CI [0.000-0.057]; CFI= 1.00; TLI= 1.06;BIC=1426.06. The reverse-order alternative model fit poorly: χ2 (8)= 27.23, p < .001; RMSEA=0.105, 90% CI [0.063-0.149];CFI= 0.87; TLI= 0.65. The second, additional-path alternative model fit well: χ2 (6)= 4.75, p= .58; RMSEA=0.000, 90% CI[0.000-0.077]; CFI= 1.00; TLI= 1.03; BIC=1436.44. However, the χ2 difference test was not significant: χ2 difference= 0.39,p= .82, and the hypothesized model had a lower BIC. Based on these results, we concluded that the hypothesized model was the bestone. Below are the results of the hypothesized model (also see Fig. 2).
As hypothesized, T1 SMSC-Ability had a positive association with T1 rumination (b=0.21, p < .001), which in turn predictedhigher identity distress at T2 (b=0.25, p= .010). The overall indirect path was significant (b=0.05, p= .044). The other hy-pothesized route received only partial support: T1 SMSC-Opinion had a positive association with T1 reflection (b=0.13, p= .002),but T1 reflection did not predict T2 identity distress (β=0.08, p= .429), and the overall indirect path was non-significant (β=0.01,p= .470). The controlled paths from T1 identity distress to T1 rumination (b=0.28, p < .001), T1 reflection (b=0.12, p= .008),and T2 identity distress (b=0.55, p < .001) were all significant. The indirect path from T1 identity distress to T2 identity distressvia T1 rumination was significant (b=0.07, p= .017), but the one via T1 reflection was not (b=0.01, p= .474). See Table 2 forstandardized coefficients. No other control variables were associated with T2 identity distress (bfemale=−0.01, p= .937;bWhite= 0.09, p= .543; buse_amount = 0.06, p= .383), T1 rumination (bfemale= 0.09, p= .425), or T1 reflection (bfemale =−0.05,p= .679).
Table 1Descriptive statistics and correlations.
Mean (SD) 1. 2. 3. 4. 5.
1. T1 SMSC-Ab 2.91 (1.05)2. T1 SMSC-Op 2.80 (1.11) .38***3. T1 Rumination 3.59 (0.88) .31*** .114. T1 Reflection 3.60 (0.75) .09 .22** .14*5. T1 Identity Distress 2.60 (1.08) .20** .11 .39*** .20**6. T2 Identity Distress 2.67 (1.10) .27** .13 .39*** .18* .62***
Note. *p < .05, **p < .01, ***p < .001. All scales were 5-piont Likert scales. SMSC-Ab= Social Media Social Comparison of Ability, SMSC-Op=Social Media Social Comparison of Opinion.
Fig. 2. Significant paths. Numbers are standardized coefficients. SMSC-Ab= Social Media Social Comparison of Ability; SMSC-Op= Social MediaSocial Comparison of Opinion. The thin lines represent significant direct paths. The bold lines represent significant indirect paths. For the clarity ofpresentation, control variables other than T1 identity distress are not included in the figure.
C.-c. Yang et al.
4.1. The dual-path model from social media social comparison to identity distress
As hypothesized, the two types of social media social comparison (i.e., social comparison of ability and social comparison ofopinion) had different implications for identity distress through the two introspective processes (i.e., rumination and reflection).Specifically, comparison of ability was positively associated with concurrent rumination, which predicted higher identity distress.Comparison of opinion, on the other hand, was positively associated with concurrent reflection, although it did not predict loweridentity distress.
The finding of the significant path from social comparison of ability on social media to rumination and finally identity distress isconsistent with existing literature. Heretofore, research has shown the negative associations between judgmental social comparisonand psycho-emotional well-being, such as higher levels of negative emotions (Lim & Yang, 2015; Park & Baek, 2018; Weinstein,2017), more depressive symptoms (Feinstein et al., 2013), and compromised self-views (de Vries & Kühne, 2015; Vogel et al., 2014).It is important to note that many of these studies focus on negative social comparison (e.g., de Vries & Kühne, 2015; Feinstein et al.,2013; Lim & Yang, 2015; Weinstein, 2017), which can be viewed as a specific type of judgmental social comparison. In these studies,social comparison was operationalized as how negative (e.g., unhappy, miserable, undesirable) the participants felt after comparingthemselves with others on social media. In contrast, we did not ask participants about their negative feelings; we simply measured theextent to which they judged themselves against other people when using social media.
The fact that our more neutral measure of social comparison of ability still associated with rumination and thus identity distress isnoteworthy. It suggests that due to the nature of social media, college freshmen may perform more upward than downward socialcomparison when using these platforms (Vogel et al., 2014). The transition to college now involves an effort of presenting a positiveimage on social media (Thomas et al., 2017; Yang & Brown, 2016), which creates ample opportunities for (upward) social comparisonof academic and social achievements. When students engage in this competition-based comparison and observe how well everyoneelse is doing, it may pose threats to their self, which induces rumination (Trapnell & Campbell, 1999). As reported in previousresearch, college students adopt rumination as a strategy to regulate unpleasant emotions derived from online social comparison(Feinstein et al., 2013). Unfortunately, the strategy often relates to poor psycho-emotional (Feinstein et al., 2013; Takano & Tanno,2009) and identity outcomes (Luyckx et al., 2007; Marin & Rotondo, 2017), likely because rumination enhances negative thinking,inhibits problem solving and instrumental behavior, and reduces one's social support (Nolen-Hoeksema et al., 2008). The pathsuggests that when college freshmen perform social comparison of ability on social media, they may ruminate over how inferior theyare to their peers. Given, as theorized by Steele (1988), that individuals strive to maintain the adequacy and integrity of the self, whenpeople dwell on their inferiority, it can hinder identity integration and exacerbate the distress of not being able to develop anacceptable self-image. Although it is not uncommon for students to face tension in their online identity work and to experience mild,short-term identity distress at the transition to college (Sica et al., 2014; Stephenson-Abetz & Holman, 2012; Thomas et al., 2017),those who frequently compare their abilities with others' on social media seem to be at a higher risk of suffering stronger identitydistress due to their engagement in rumination.
Also as expected, social comparison of opinion on social media was associated with a higher level of reflection. The two werepositively associated, perhaps due to the shared characteristic of the desire to learn and be informed: Social comparison of opinionfocuses on collecting information that would help individuals make predictions and assess the need for adjusting their beliefs,opinions, or values (Suls et al., 2000, 2002); reflection involves epistemic curiosity in and exploration of the self (Trapnell &Campbell, 1999). Suls et al. (2000) argue that one major form of social comparison of opinion entails determining whether one's ownreaction to an object or situation is appropriate, given the norms of the affiliated groups. This assessment seems particularly
Table 2Results of path analysis.
β b (SE) p 95% CI
Direct Paths of InterestT1 SMSC-Ab → T1 Rumination .25 .21 (.06) < .001 [.10, .32]T1 SMSC-Op → T1 Reflection .20 .13 (.04) .002 [.04, .22]T1 Rumination → T2 Identity Distress .20 .25 (.10) .010 [.04, .43]T1 Reflection → T2 Identity Distress .05 .08 (.10) .429 [-.12, .28]Indirect Paths of InterestT1 SMSC-Ab → T1 Rumination →
T2 Identity Distress.05 .05 (.03) .044 [.01, .11]
T1 SMSC-Op → T1 Reflection →T2 Identity Distress
.01 .01 (.02) .470 [-.01, .05]
Significant Controlled PathsT1 Identity Distress → T1 Rumination .34 .28 (.05) < .001 [.17, .38]T1 Identity Distress→ T1 Reflection .17 .12 (.05) .008 [.03, .21]T1 Identity Distress → T2 Identity Distress .54 .55 (.08) < .001 [.40, .71]T1 Identity Distress → T1 Rumination → T2 Identity Distress .07 .07 (.03) .017 [.02, .13]
Note. SMSC-Ab=Social Media Social Comparison of Ability; SMSC-Op= Social Media Social Comparison of Opinion.
C.-c. Yang et al.
important as one transitions into a new environment, such as college. Indeed, it is a common practice for today's college students tolearn about peers and peer norms through social media, as a way to gauge their compatibility with peers and adjust their own self-presentation (Thomas et al., 2017). When the online comparison focuses on opinion rather than ability, students report betteradjustment to college (Yang & Robinson, 2018), and findings of this study suggest that reflection, an adaptive mode of self-atten-tiveness, may contribute to the positive association.
Contrary to our hypothesis, however, reflection did not predict lower identity distress. Although reflection and rumination areconceptually distinct enough to be viewed as two types of introspection (Trapnell & Campbell, 1999), the two still had a weak butsignificant correlation (see Table 1). Therefore, when their associations with identity distress were considered simultaneously, theone less predictive of identity distress became a non-significant predictor. This pattern has also been reported in previous research inwhich both reflection and rumination were examined at the same time (e.g., Luyckx et al., 2007; Newman & Nezlek, 2017; Takano &Tanno, 2009). By definition, identity distress (Berman et al., 2004; Berman & Weems, 2011) has a stronger neurotic than epistemicemphasis, and thus it should not be surprising that rumination stood out as the predictor. It is also worth mentioning that, un-expectedly, there was a weak but positive path between T1 identity distress and concurrent reflection, suggesting that collegefreshmen who were distressed with identity issues engaged in not only rumination but also reflection (though to a lesser degree).These students may have performed reflection with the hope of clarifying their identity. However, as Nolen-Hoeksema et al. (2008)concluded about pondering, a process similar to reflection, epistemic introspection can involve distress in the short term, and it takestime for this introspective process to produce positive outcomes. It suggests another possible explanation for the null associationbetween T1 reflection and T2 identity distress in our model: The timespan may have been too short for the positive effect to emerge.We discuss this limitation in the next section and propose a direction for future research.
The additional-path alternative model, in which two more paths were included (comparison of ability→ reflection, comparison ofopinion → rumination), did not improve model fit. It suggests that social comparison of ability was associated with only rumination,whereas that of opinion was related to only reflection. The findings reveal that there may be a maladaptive route from social mediasocial comparison to identity distress, as well as a benign (if not adaptive) route. Whereas there has been rich literature using theadaptive/maladaptive viewpoint to distinguish rumination from reflection (e.g., Marin & Rotondo, 2017; Nolen-Hoeksema et al.,2008; Takano & Tanno, 2009; Trapnell & Campbell, 1999), the difference between social comparison of ability and that of opinionhas just started receiving attention (e.g., Park & Baek, 2018; Yang et al., 2018; Yang & Robinson, 2018).
The reason why the two types of social comparison take different routes likely associates with whether value judgement isinvolved in the process and how easy it is to make changes in the given dimension. Social comparison of ability, particularly in thecontext of social media, can easily make people feel inferior (de Vries & Kühne, 2015; Vogel et al., 2014). Yet, improving one's ownabilities takes efforts and time, and there is no guarantee one can achieve the goals set (Festinger, 1954). On social media, whencollege freshmen see peers' posts of colorful social lives or outstanding academic records (selectively showcased by peers) and engagein comparison of ability, they may become trapped in rumination and negative emotions (Feinstein et al., 2013; Park & Baek, 2018;Weinstein, 2017) because they do not think they can catch up anytime soon. In contrast, in the process of social comparison ofopinion, one opinion is not inherently better or worse than another, and it is relatively easy to change opinions (Festinger, 1954).Therefore, in this process, there may be a higher likelihood for freshmen to direct energy to exploring the self and new ideas (i.e.,reflection) as opposed to being fixated on how they are not good enough (i.e., rumination).
The study has theoretical implications at both macro and micro levels. At the macro level, the study identified a new angle to theresearch field of identity development in the digital age. We recognized the increased prevalence of social media as a major socio-cultural change that could influence youth's identity development. Although an increasing number of studies have also explored thisissue, most take the perspective of online self-presentation, discussing the tension of online self-presentation in the collapsed context(Stephenson-Abetz & Holman, 2012; Thomas et al., 2017) and the relationships between different modes of online self-presentationand youth's identity states (Michikyan et al., 2015; Yang & Brown, 2016). This study is among the few shedding light on the topicunder the social comparison framework.
At the micro level, the hypothesized model clarified that there were maladaptive and benign/adaptive paths from social mediasocial comparison to identity distress, suggesting that whether social comparison should be regarded as detrimental depends onwhich types of comparison and self-attentiveness are involved. This is important to delineate because current social media researchhas primarily centered on the judgmental, competitive forms of social comparison that are frequently associated with negativepsycho-emotional outcomes (e.g. de Vries & Kühne, 2015; Feinstein et al., 2013; Vogel et al., 2014). Yet, the current findings suggestthis association may not hold when social comparison is not inherently judgmental, such as during the comparison of opinion. Thedifferential implications of the two types of social comparison have also been reported in a few other studies (Park & Baek, 2018;Yang et al., 2018; Yang & Robinson, 2018). Although the indirect path from social media social comparison of opinion to loweridentity distress via reflection was not significant, the fact that social comparison of opinion was associated with the adaptive type ofself-attentiveness (i.e., reflection) suggests its instrumental nature, which is distinct from that of social comparison of ability. Thetheoretical implications at the macro and micro levels, collectively, indicate that future scholars could explore identity developmentin the digital age by considering young people's engagement in the two types of social comparison. Continued efforts in identifyingmajor mediators and moderators will further expand the area.
At the practical level, the findings provide clues to assisting college freshmen in addressing identity distress at the transition tocollege. At this transition, freshmen face the need to address the sense of anonymity among the large student body (Scanlon et al.,
C.-c. Yang et al.
2007), make themselves known and acceptable to new peers, and negotiate new and old selves (Stephenson-Abetz & Holman, 2012;Thomas et al., 2017). Because identity distress is most salient when identity construction is a focal developmental task (Wiley et al.,2011), professionals are advised to keep track of whether first-year students feel distressed because they are unable to reconciledifferent aspects of the self and/or develop an acceptable self-image. As social media use is almost ubiquitous among college students,it may be necessary to remind students to use social media for opinion assessment but not for ability competition. Specifically, itcould be beneficial to call students' attention to how they approach information about others so that they might use social media as ameans of developing self-knowledge rather than self-judgement. For instance, one has different options for processing a friend'sInstagram post about acceptance into a campus sorority or fraternity. On one hand, one might engage in comparison of ability,considering one's own lack of such group membership to signify that the friend must have better social skills and social competence.On the other hand, one might engage in comparison of opinion, recognizing that the friend is in similar social circles and that onemight enjoy joining a campus organization oneself. The former may lock students into rumination, whereas the latter may propelthem to reflect upon their own values, attitudes, and beliefs to gain self-knowledge. Therefore, being cognizant of the ways in whichone typically approaches information about others is important, particularly as the rise of social media makes such information amore pervasive part of young people's daily life.
4.3. Limitations and future directions
Given our use of a relatively small convenience sample, the results should be interpreted cautiously. In addition, the studyinvolved only two waves of data, and most variables were measured at T1. Specifically, the social comparison and introspection datawere collected contemporaneously, making it difficult to ascertain the directionality. For this reason, we avoided using the term“predict” when reporting results involving these two sets of variables. We also tried to address the issue by testing an alternative,reverse-order model, and the poor fit of this alternative model suggests that our hypothesized sequence was more plausible. Tofurther expand the research and fully clarify the directionality, scholars are encouraged to examine the model with a large, re-presentative sample, and to measure all variables across more time points.
Another limitation, as mentioned in a previous section, is the short interval between the two time points. Related to this lim-itation, it is unclear which one of the two possible reasons (conceptual overlap between rumination and reflection versus timespan)would provide a better explanation for why reflection did not predict identity distress in our study. Future scholars may considerreplicating the study, using a longer time interval. If both reflection and rumination serve as significant predictors of identity distress,then the null result in our study should be interpreted as a result of the limited timespan. On the other hand, if reflection remains non-significant in the long-term study, it would indicate that reflection is indeed less important than rumination in predicting identitydistress.
We focused on college freshmen because we recognized college transition as a crucial time for identity negotiation and (re)construction (Chow & Healey, 2008; Scanlon et al. 2007; Thomas et al. 2017). However, given younger adolescents' avid use of socialmedia (Lenhart, 2015) and adolescence as a prime time for identity development (Erikson, 1968), it would be meaningful to alsoexamine the applicability of the model to middle or high school students. Previous research shows that high school students whospend more time using communication technology report higher identity distress (Cyr et al., 2015); our model could provide morenuanced information by showing how different ways of media use would have differential implications.
Finally, our model can be expanded by including additional or different sets of identity outcomes, depending on which age groupis involved. For instance, although younger adolescents typically do not reach the identity achievement status (Kroger et al., 2010),this phenomenon is normative and should not cause severe identity distress. Thus, when applying the model to younger adolescents,identity distress may be replaced by other identity variables that are more relevant to the age group, such as self-esteem or self-perception. Regarding youth going through the transition to college, previous research suggests that, in addition to identity distress(Sica et al., 2014), place identity, identity integration, and identity clarity (Chow & Healey, 2008; Thomas et al., 2017; Yang & Brown,2016) are also salient constructs to consider. Including these additional identity variables will generate a more comprehensive pictureof how social comparison on social media contributes to youth's identity development.
The present study offers new insights relevant for youth's identity development during what is for many a tumultuous transition.Specifically, the study calls attention to the different ways in which college students might compare themselves with others on socialmedia as they look to reevaluate and rebuild their identity and sense of self after entering the new college context, and whatimplications this may have for their identity. While social comparison of ability on social media was related to higher concurrentrumination and subsequently higher identity distress, online social comparison of opinion was related to greater reflection, whichwas not predictive of identity distress. Largely, the study reaffirms the recently rising call for distinguishing the competition-basedsocial comparison of ability from the information-based social comparison of opinion. At the same time, the study expands currentknowledge of why these forms of social comparison may lead to differential outcomes, namely through the type of introspection theyinduce. Recent research on adolescent development in relation to social media use is moving away from the simplistic narrative thatsocial media use is inherently beneficial or detrimental. Our study joins this effort by showing which specific mode of social mediause would be concerning and why. Our dual-path model may be expanded by including other identity outcomes, and the model maybe applied to other age groups. It is our hope that the model will provide a useful framework to further the understanding of youngpeople's identity development in the digital age.
C.-c. Yang et al.
Conflicts of interest
The authors have no conflicts of interest to declare. The research was supported by a fund provided by the first author's in-stitution.
American Psychiatric Association. (1980). Diagnostic and statistical manual of mental disorders (3rd ed.). Washington, DC: American Psychiatric Association.American Psychiatric Association. (1987). Diagnostic and statistical manual of mental disorders (3rd ed., revised). Washington, DC: American Psychiatric Association.American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (5th ed.). Arlington, VA: American Psychiatric Publishing.Arnett, J. J. (2002). The psychology of globalization. American Psychologist, 57, 774–783. https://doi.org/10.1037//0003-066x.57.10.774.Arnett, J. J. (2014). Emerging adulthood: The winding road form the late teens through the twenties (2nd Ed.). New York, NY: Oxford University Press.Barrett, P. (2007). Structural equation modeling: Adjudging model fit. Personality and Individual Differences, 42(5), 815–824. https://doi.org/10.1016/j.paid.2006.09.
018.Berman, S. L., Kennerley, R. J., & Kennerley, M. A. (2008). Promoting adult identity development: A feasibility study of a university-based identity intervention
program. Identity: An International Journal of Theory and Research, 8(2), 139–150. https://doi.org/10.1080/15283480801940024.Berman, S. L., Montgomery, M. J., & Kurtines, W. M. (2004). The development and validation of a measure of identity distress. Identity: An International Journal of
Theory and Research, 4(1), 1–8.Berman, S. L., & Weems, C. F. (2011). Identity distress. In R. J. R. Levesque (Ed.). Encyclopedia of adolescence (pp. 1357–1361). New York, NY: Springer.Berman, S. L., Weems, C. F., & Petkus, V. F. (2009). The prevalence and incremental validity of identity problem symptoms in a high school sample. Child Psychiatry
and Human Development, 40(2), 183–195. https://doi.org/10.1007/s10578-008-0117-6.Beyers, W., & Luyckx, K. (2016). Ruminative exploration and reconsideration of commitment as risk factors for suboptimal identity development in adolescence and
emerging adulthood. Journal of Adolescence, 47, 169–178. https://doi.org/10.1016/j.adolescence.2015.10.018.Browne, M. W., & Cudeck, R. (1993). Alternative ways of assessing model fit. In K. A. Bollen, & J. S. Long (Eds.). Testing structural equation models (pp. 136–162).
Newbury Park, CA: Sage.Chow, K., & Healey, M. (2008). Place attachment and place identity: First-year undergraduates making the transition from home to university. Journal of Environmental
Psychology, 28(4), 362–372. https://doi.org/10.1016/j.jenvp.2008.02.011.Chua, T. H. H., & Chang, L. (2016). Full length article: Follow me and like my beautiful selfies: Singapore teenage girls engagement in self-presentation and peer
comparison on social media. Computers in Human Behavior, 55, 190–197. https://doi.org/10.1016/j.chb.2015.09.011.Cyr, B.-A., Berman, S. L., & Smith, M. L. (2015). The role of communication technology in adolescent relationships and identity development. Child and Youth Care
Forum, 44(1), 79–92. https://doi.org/10.1007/s10566-014-9271-0.Erikson, E. (1968). Identity: Youth and crisis. Oxford, England: Norton & Co.Feinstein, B. A., Hershenberg, R., Bhatia, V., Latack, J. A., Meuwly, N., & Davila, J. (2013). Negative social comparison on Facebook and depressive symptoms:
Rumination as a mechanism. Psychology of Popular Media Culture, 2(3), 161–170. https://doi.org/10.1037/a0033111.Festinger, L. (1954). A theory of social comparison processes. Human Relations, 7(2), 117–140. https://doi.org/10.1177/001872675400700202.Fornell, C., & Larcker, D. F. (1981). Evaluating structural equation models with unobservable variables and measurement error. Journal of Marketing Research, 18,
39–50.Gandhi, A., Lyuckx, K., Maitra, S., & Claes, L. (2015). Non-suicidal self-injury and identity distress in Flemish adolescents: Exploring gender differences and med-
iational pathways. Personality and Individual Differences, 82, 215–220. https://doi.org/10.1016/j.paid.2015.03.031.Gardner, H., & Davis, K. (2014). The App Generation: How today's youth navigate identity, intimacy, and imagination in a digital world. New Haven, CT: Yale University
Press.Gibbons, F. X., & Buunk, B. P. (1999). Individual differences in social comparison: Development of a scale of social comparison orientation. Journal of Personality and
Social Psychology, 76(1), 129–142.Hair, J. F., Ringle, C. M., & Sarstedt, M. (2011). PLS-SEM: Indeed a silver bullet. Journal of Marketing Theory and Practice, 19(2), 139–152. https://doi.org/10.2753/
MTP1069-6679190202.Harrington, R., & Loffredo, D. A. (2011). Insight, rumination, and self-reflection as predictors of well-being. The Journal of Psychology, 145(1), 39–57. https://doi.org/
10.1080/00223980.2010.528072.Harter, S. (2012). The construction of the self: Developmental and sociocultural foundations (2nd ed.). New York, NY: The Guilford Press.Hu, L., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation
Modeling, 6(1), 1–55. https://doi.org/10.1080/10705519909540118.Kroger, J., Martinussen, M., & Marcia, J. (2010). Identity status change during adolescence. Journal of Adolescence, 33(1), 683–698. https://doi.org/10.1016/j-
adolescence.2009.11.002.Lenhart, A. (2015). Teen, social media & technology overview 2015. Retrieved from Pew Research Center website: http://www.pewinternet.org/2015/04/09/teens-
social-media-technology-2015/.Lim, M., & Yang, Y. (2015). Effects of users' envy and shame on social comparison that occurs on social network services. Computers in Human Behavior, 51, 300–311.
https://doi.org/10.1016/j.chb.201.05.013.Little, R. J. A. (1988). A test of missing completely at random for multivariate data with missing values. Journal of the American Statistical Association, 83(404),
1198–1202. https://doi.org/10.1080/01621459.1988.10478722.Luyckx, K., Soenens, B., Berzonsky, M. D., Smits, I., Goossens, L., & Vansteenkiste, M. (2007). Information-oriented identity processing, identity consolidation, and
well-being: The moderating role of autonomy, self-reflection, and self-rumination. Personality and Individual Differences, 43(5), 1099–1111. https://doi.org/10.1016/j.paid.2007.03.003.
Manago, A. M., Graham, M. B., Greenfield, P. M., & Salimkhan, G. (2008). Self-presentation and gender on MySpace. Journal of Applied Developmental Psychology, 29(6),446–458. https://doi.org/10.1016/j.appdev.2008.07.001.
Marin, K. A., & Rotondo, E. K. (2017). Rumination and self-reflection in stress narratives and relations to psychological functioning.Memory, 25(1), 44–56. https://doi.org/10.1080/09658211.2015.1124122.
Meca, A., Eichas, K., Quintana, S., Maximin, B. M., Ritchie, R. A., Madrazo, V. L., et al. (2014). Reducing identity distress: Results of an identity intervention foremerging adults. Identity: An International Journal of Theory and Research, 14(4), 312–331. https://doi.org/10.1080/15283488.2014.944696.
Michikyan, M., Dennis, J., & Subrahmanyam, K. (2015). Can you guess who I am? Real, ideal, and false self-presentation on Facebook among emerging adults. EmergingAdulthood, 3(1), 55–64. https://doi.org/10.1177/2167696814532442.
Newman, D. B., & Nezlek, J. B. (2017). Private self-consciousness in daily life: Relationships between rumination and reflection and well-being, and meaning in dailylife. Personality and individual differenceshttps://doi.org/10.1016/j.paid.2017.06.039 Advance online publication.
Nolen-Hoeksema, S., Wisco, B. E., & Lyubomirsky, S. (2008). Rethinking rumination. Perspectives on Psychological Science, 3(5), 400–424. https://doi.org/10.1111/j.1745-6924.2008.00088.x.
Park, S. Y., & Baek, Y. M. (2018). Two faces of social comparison on Facebook: The interplay between social comparison orientation, emotions, and psychological well-being. Computers in Human Behavior, 79, 83–93. https://doi.org/10.1016/j.chb.2017.10.028.
Radmacher, K. A., & Azmitia, M. (2016). Family emotional support and the individuation process among Asian- and Latino-heritage college-going emerging adults.Journal of Research on Adolescence, 26(4), 979–990. https://doi.org/10.1111/jora.12251.
C.-c. Yang et al.
Rosenberg, M. (1979). Conceiving the self. New York, NY: Basic Books.Samuolis, J., Barcellos, M., LaFlam, J., Belson, D., & Berard, J. (2015). Mental health issues and their relation to identity distress in college students. Identity: An
International Journal of Theory and Research, 15(1), 66–73. https://doi.org/10.1080/15283488.2014.989443.Scanlon, L., Rowling, L., & Weber, Z. (2007). ‘You don't have like an identity…you are just lost in a crowd’: Forming a student identity in the first-year transition to
university. Journal of Youth Studies, 10(2), 223–241. https://doi.org/10.1080/13676260600983684.Sica, L. S., Sestito, L. A., & Ragozini, G. (2014). Identity coping in the first years of university: Identity diffusion, adjustment and identity distress. Journal of Adult
Development, 21(3), 159–172. https://doi.org/10.1007/s10804-014-9188-8.Steele, C. M. (1988). The psychology of self-affirmation: Sustaining the integrity of the self. In L. Berkowitz (Vol. Ed.), Advances in experimental social psychology: Vol.
21, (pp. 261–302). San Diego, CA: Academic Press.Stephenson-Abetz, J., & Holman, A. (2012). Home is where the heart is: Facebook and the negotiation of “old” and “new” during the transition to college. Western
Journal of Communication, 76(2), 175–193. https://doi.org/10.1080/10570314.2011.654309.Suls, J., Martin, R., & Wheeler, L. (2000). Three kinds of opinion comparison: The triadic model. Personality and Social Psychology Review, 4(3), 219–237. https://doi.
org/10.1207/S15327957PSPR0403_2.Suls, J., Martin, R., & Wheeler, L. (2002). Social comparison: Why, with whom, and with what effect? Current Directions in Psychological Science, 11(5), 159–163.
https://doi.org/10.1111/1467-8721.00191.Takano, K., & Tanno, Y. (2009). Self-rumination, self-reflection, and depression: Self-rumination counteracts the adaptive effect of self-reflection. Behavior Research and
Therapy, 47(3), 260–264. https://doi.org/10.1016/j.brat.2008.12.008.Thomas, L., Briggs, P., Hart, A., & Kerrigan, F. (2017). Understanding social media and identity work in young people transitioning to university. Computers in Human
Behavior, 76, 541–553. https://doi.org/10.1016/j.chb.2017.08.021.Trapnell, P. D., & Campbell, J. D. (1999). Private self-consciousness and the five-factor model of personality: Distinguishing rumination from reflection. Journal of
Personality and School Psychology, 76(2), 284–304.Treynor, W., Gonzalez, R., & Nolen-Hoeksema, S. (2003). Rumination reconsidered: A psychometric analysis. Cognitive Therapy and Research, 27(3), 247–259.Vogel, E. A., Rose, J. P., Roberts, L. R., & Eckles, K. (2014). Social comparison, social media, and self-esteem. Psychology of Popular Media Culture, 3(4), 206–222.
https://doi.org/10.1037/ppm0000047.de Vries, D. A., & Kühne, R. (2015). Facebook and self-perception: Individual susceptibility to negative social comparison on Facebook. Personality and Individual
Differences, 85, 217–221. https://doi.org/10.1016/j.paid.2015.05.029.Weinstein, E. (2017). Adolescents' differential responses to social media browsing: Exploring causes and consequences for intervention. Computers in Human Behavior,
76, 396–405. https://doi.org/10.1016/j.chb.2017.07.038.Wiley, R. E., Berman, S. L., Marsee, M. A., Taylor, L. K., Cannon, M. F., & Weems, C. F. (2011). Age differences and similarities in identity distress following the Katrina
disaster: Theoretical and applied implications of Erikson's theory. Journal of Adult Development, 18(4), 184–191. https://doi.org/10.1007/s10804-011-9130-2.Yang, C.-c. (2016). Social media as more than a peer space: College freshmen encountering parents on facebook. Journal of Adolescent Research. https://doi.org/10.
1177/0743558416659750 Advance online publication.Yang, C-c., & Brown, B. B. (2016). Online self-presentation on Facebook and self development during the college transition. Journal of Youth and Adolescence, 45(2),
402–416. https://doi.org/10.1007/s10964-015-0385-y.Yang, C-c., Holden, S. M., & Carter, M. D. K. (2018). Social media social comparison of ability (but not opinion) predicts lower identity clarity: Identity processing style
as a mediator. Journal of Youth and Adolescence. https://doi.org/10.1007/s10964-017-0801-6 Advance online publication.Yang, C-c., & Robinson, A. (2018). Not necessarily detrimental: Two social comparison orientations and their associations with social media use and college social
adjustment. Computers in Human Behavior, 84, 49–57. https://doi.org/10.1016/j.chb.2018.02.020.
C.-c. Yang et al.
- Social media social comparison and identity distress at the college transition: A dual-path model
- Identity distress
- Social comparison on social media
- Rumination and reflection
- Current study
- Social media social comparison (SMSC)
- Identity distress
- Data analysis
- The dual-path model from social media social comparison to identity distress
- Limitations and future directions
- Conflicts of interest