The influence of possible selves on global citizenship identification

We examined the influence of a ‘possible self’ activity on antecedents, identification, and outcomes of global citizenship. Participants wrote about either hoped-for selves as active global citizens, feared selves as inactive global citizens, or a typical day (control) and then answered questions to gauge their global citizen identification. Results show that the saliency of a feared self as an inactive global citizen led to greater identification with the global citizen identity. A structural equation model shows that feared self (vs hoped-for self) predicted greater global citizenship identification, through the perception of one’s normative environment as prescribing a global citizen identity and global awareness. Global citizenship identification predicted greater endorsement of prosocial values and behaviours (e.g. intergroup empathy and helping). The results support the use of a ‘feared self’ activity to engender global citizenship identification and prosocial values in students.


The influence of possible selves on global citizenship identification
Human cognitive capability allows people to think about themselves in the present, reflect on who they have been in the past, and imagine who they could become in the future, in a continuous stream of thought (James, 1890).People have different types of selves (e.g. a social self, spiritual self, and ego); these are affected by positive and negative states of mind, feelings about the self, and self-seeking or self-preserving actions.Through a dynamic interplay between intrapersonal and interpersonal interactions, individuals form a variety of self-representations as part of a working self-concept that, when salient, aids in directing thoughts, motivation, and behaviour (Markus and Wurf, 1987).These future self-representations, which can become salient and active in the working self-concept, are categorized as possible selves (Markus and Nurius, 1986).Possible selves (Markus and Nurius, 1986) are influenced by past selves, linked to actual selves, and can represent hoped-for selves (i.e.selves that individuals would like to become) or feared selves (i.e.selves that individuals wish to avoid).

The dynamic self-concept
One area of research interest in the self-system is the self-concept, or mental selfimage, that is created when people think about themselves.In examining the self-concept, researchers have sought to determine whether or not a person's selfconcept remains stable or can change over time.Researchers have long differed over the malleability or durability of the self-concept.There is support for both a durable self-concept that resists change (Greenwald, 1980;Swann, 1983) and a malleable self-concept (Swann and Hill, 1982) that occurs when people do not refute or reject self-disconfirming feedback.Markus and Wurf (1987), contributing to the discussion about the durability or malleability of the self-concept, presented a dynamic model in which the self-concept was portrayed as consisting of various self-representations (e.g.self-schemas, standards, and possible selves) that, along with intrapersonal behaviour, form a part of the affective-cognitive system.This dynamic model of the self-concept operates in the context of the social environment and includes a working self-concept that is composed of active and salient self-representations, which are immediately accessible from the interaction between the affective-cognitive system (that is, the self-concept and intrapersonal behaviour) and interpersonal behaviours.Self-representations may be activated automatically, based on salient stimuli in the social environment, or recruited deliberately, based on motivations such as self-enhancement (development or maintenance of positive affect about the self ), self-consistency (maintenance of a sense of coherence or continuity), and self-actualization (linked to a desire for selffulfilment).The working self-concept is theorized as shaping both intrapersonal processes, such as affect regulating, motivation, and the processing of selfrelevant information, and interpersonal processes such as social perception, social comparisons, and human interaction.The interaction between these intrapersonal and interpersonal processes serves to determine which self-representations will become salient in the working self-concept and influence the person's motivational state in a particular social environment.In this dynamic model, the working selfconcept represents a path through which the possibility for temporary or permanent change can be introduced and realized in the overall self-concept.
To summarize, in the dynamic working self-concept (Markus and Wurf, 1987) possible selves are characterized as active and salient self-representations that form part of the affective-cognitive system.Hoped-for selves are linked to positive affect and feared selves are linked to negative affect (Markus and Nurius, 1986;Markus and Wurf, 1987).These future selves are yet to be realized and therefore are malleable; they may be activated automatically by salient stimuli in the social environment.In addition, possible selves may also be recruited deliberately from intrapersonal action (Markus and Wurf, 1987) through motivations (e.g.self-enhancement or self-actualization).

Self-discrepancy explanation of possible selves
A self-discrepancy explanation of possible selves (Higgins, 1987), which has found support in subsequent research, suggests that individuals are motivated to move away from feared selves, thereby increasing the discrepancy between the current and future feared selves.Discrepancies between actual self and ideal or hoped-for self are associated with dejection-related emotions, like sadness and disappointment, while discrepancies between actual self and 'ought self' (the possible self defined by one's significant others) are associated with agitation-related emotions like fear and restlessness.Carver et al. (1999) found partial support for the self-discrepancy explanation of possible selves, in that discrepancies with distant feared selves predicted agitation-related emotion while discrepancies with close-by feared selves predicted dejection-related emotions.The researchers offered an approach/ avoidance explanation, which suggested that dejection-related emotions arise for close feared selves as avoidance goals become salient.However, agitation-related emotions (including happiness) are more likely to arise when feared selves are distant and approach goals become more dominant.Carver (2006) theorized that human actions are regulated by approach and avoidance tendencies through discrepancy-reducing and discrepancy-enlarging feedback respectively.As a result, thinking of a feared self will generate discrepancy-enlarging feedback; in turn, avoidance action will be taken to create distance between a current self and feared future self.On the other hand, thinking of a hoped-for self will prompt the creation of discrepancy-reducing feedback and approach action is likely to be taken to attempt to realize the desired self.To summarize, the cognitive and affective components of possible selves interact to motivate various types of behaviour associated with approaching hoped-for selves and avoiding feared selves.Hoped-for selves are accompanied by positive emotions and feared selves by negative emotions.

Influence of hoped-for vs feared possible selves on personal identity
Although possible selves are futuristic self-conceptions, their behavioural influence is not deferred to future time periods, but influences current behaviours (Hoyle and Sherrill, 2006).A wealth of research shows that the desire to move towards a hoped-for self or avoid a feared self can influence personal identity.For example, self-regulating possible selves are influential in improving students' academic performance (Oyserman et al., 2004).The inability to identify balanced possible selves (i.e. to recognize expected and feared selves in the same domain) and to identify feasible steps to attaining each of these selves differentiated youths officially termed delinquent from their non-delinquent peers, with the acquisition of these abilities linked to a reduction in youth delinquency (Oyserman and Markus, 1990) and an improvement in academic performance (Oyserman et al., 2006).Negative (feared) possible selves predicted binge versus non-binge drinking (Quinlan et al., 2006), while discrepancies between feared possible selves and both actual selves and hoped-for possible selves were motivational in increasing exercise among people with fitness goals (Murru and Martin Ginis, 2010).Self-regulation for engaging in health-promoting behaviours was far more likely for participants who wrote about health-related feared possible selves than those who wrote about hoped-for selves or were in the control condition (Hoyle and Sherrill, 2006).In essence, hoped-for, feared, and balanced possible selves serve to influence personal identity when persons are motivated to regulate their behaviour to move towards or away from future self-conceptions, and feared possible selves tend to be more effective than hoped-for selves in motivating behaviour change (Hoyle and Sherrill, 2006;Murru and Martin Ginis, 2010;Quinlan et al., 2006).

Influence of possible selves on social identity
The influence of possible selves extends beyond personal identity to social identities; that is, to self-representations developed in a social context.Oyserman and James (2011) proposed that possible identities (i.e.possible selves) evolve from social contexts such as influential connections and relationships, from aspects of social cognition that generate malleable future self-conceptions in working memory, and from future selves that are embedded in social contexts.While significant research attention has been given to possible selves as personal identities (Murru and Martin Ginis, 2010;Oyserman et al., 2004;Quinlan et al., 2006), far less research has examined the influence of possible selves on social identities.One exception, Cameron (1999), found that group membership (i.e.social identity) as a university student was perceived as being helpful to individuals in realizing hoped-for selves and avoiding feared future selves.However, other possible motivations and underlying processes may underlie the link between possible selves and identification with groups.The findings of Fryberg et al. (2008) highlight the potential negative influence of the social environment through a study of the effect of certain social representations of North American Indigenous Peoples, such as in the case of Indian mascots.American Indigenous Peoples students exposed to such representations registered a lower number of academic-achievement-related possible selves than did their counterparts who were not exposed to such images.Packer and Chasteen (2006) found that when youths were asked to think and write about their future selves at age 70, participants who were strongly identified with their own age group demonstrated greater prejudice against older adults than did participants who were weakly identified with their own age group.These findings reflect the potential negative impact of the sociocultural environment on social identity that may be associated with possible selves.
To summarize, people hold self-conceptions that constitute hoped-for (positive) or feared (negative) possible selves, which in turn influence personal and social identity.Future self-representations that influence social identity can result in an approach towards a hoped-for social self or avoidance of a feared social self.In the present study, we examine the utility of a possible self-task to influence individuals' identification with a prosocial superordinate category: the global citizen.Markus and Wurf (1987) posited that the social environment makes certain selfrepresentations salient in the working self-concept, including past, current, and future self-representations (i.e.possible selves).The social environment is becoming increasingly globalized through media, technology, and travel (McLuhan, 1964), and through education and community life (Dower, 2008;Pike, 2008;Schattle, 2008).In this globalized world context, there is increased human interconnectedness across cultures and national borders (Merryfield et al., 2008), a tendency that is posited to affect individuals' social identities (Arnett, 2002).From an 'intentional worlds' perspective (Shweder, 1990), this sociocultural environment is likely to afford individuals the opportunity to identify as global citizens.Using a social identity perspective (Tajfel and Turner, 1979;Turner et al., 1987), when a global citizen identity is salient, greater identification (i.e.greater psychological connection with this social category) should predict greater adherence to the group's prototypical content, such as values, behaviours, and norms (see Reysen andKatzarska-Miller, 2013a, 2013b).In other words, the sociocultural environments in which individuals are embedded can condition individuals, provide them with relevant knowledge and felt connection to others in the world, and afford a global citizen identity, as well as priming this identity in everyday settings like school and through the news media.If individuals accept membership in the group, when a global citizen identity is salient, a greater felt connection should predict greater endorsement of the group's prosocial values and behaviours.

Global citizenship and possible selves
Recent research (Reysen and Katzarska-Miller, 2013a) shows that individuals who perceive valued others (i.e.their normative environment) as prescribing a global citizen identity, and who feel knowledgeable and connected to others in the world (i.e. who demonstrate global awareness), identify more strongly with a global citizen identity.Furthermore, greater identification with global citizens predicts greater endorsement of prosocial values and behaviours representing that group's characteristic content.The outcomes of global citizenship identification include intergroup empathy (a felt connection to individuals outside one's in-group), valuing diversity (a positive view of cultural diversity), social justice (an endorsement of human rights and equality), environmental sustainability (a concern for the natural environment), intergroup helping (a desire to help others outside one's in-group), and a felt responsibility to act for the betterment of the world (for a review of the antecedents and outcomes of global citizenship identification, see Reysen and Katzarska-Miller, 2013b).This model of the antecedents and consequences of global citizenship identification has since received a wealth of empirical support (Gibson and Reysen, 2013;Katzarska-Miller et al., in press;Plante et al., 2014;Reysen and Katzarska-Miller, 2013c;Reysen et al., 2013;Reysen et al., 2012;Reysen et al., in press).The salience of a feared self as an inactive (vs active or hoped-for) global citizen may influence an individual's degree of global citizenship identification.

Present study
The purpose of the present study is to examine the influence of a 'possible self' manipulation on the antecedents and outcomes of global citizenship identification.Participants were randomly assigned to write about themselves as an active global citizen (hoped-for self ), inactive global citizen (feared self ), or about a typical day (control condition), before being rated for affect and measures of global citizenship.We predict that, consistent with a self-discrepancy explanation of possible selves (Carver, 2006;Carver et al., 1999;Higgins, 1987), the saliency of a feared self will result in significantly higher identification as a global citizen than will be seen among those writing about hoped-for selves or in the control condition.Furthermore, we expect to replicate, and find that the saliency of possible selves is a factor in, Reysen andKatzarska-Miller's (2013a, 2013b) model of the antecedents and outcomes of global citizenship identification.

Method
Participants and procedure Participants (n = 323, 67.5% women; M age = 30.24,SD = 9.38) included students, mostly from middle-class and working-class families, attending a regional public university in the southern United States.Students were pursuing a variety of majors including business, education, engineering and technology, liberal arts, natural sciences, and social sciences; all participated voluntarily in research for extra credit or course credit.After random assignment to one of two experimental groups or to a control condition, participants were asked to imagine and write about themselves in the future as either an active global citizen (for those in the 'hoped-for self' experimental group), inactive global citizen (for those in the 'feared self' experimental group), or to write about a typical day (control condition).Participants then completed measures regarding their emotional experience, antecedents, identification, and outcomes of global citizenship, and demographic information.All measures used a 7-point Likert-type scale, from 1 = strongly disagree to 7 = strongly agree.

Experimental materials 'Possible self' manipulation
Participants were asked to take a few minutes to think of themselves in 5 to 10 years' time as an active global citizen or an inactive global citizen, and to keep the image in mind while completing a writing exercise.Participants assigned to the control condition were asked to imagine and write about themselves on a typical day (see Table 1 for manipulation instructions).

Global citizenship
To assess antecedents (normative environment, global awareness), identification, and outcomes (prosocial values and behaviour) of global citizenship we adopted measures from prior research (Reysen and Katzarska-Miller, 2013a;Reysen et al., 2012).See Table 2 for number of items, reliability coefficients, and example items.Note: Items rated on a 7-point Likert-type scale, from 1 = strongly disagree to 7 = strongly agree.

Preliminary analysis
To examine whether the manipulation of possible selves influenced participants' emotional experience, we first conducted a MANOVA, with the manipulation as the independent variable and positive and negative affect as dependent variables.

Mean differences
To examine the influence of the manipulation of possible selves on dependent variables, we conducted a MANOVA with the manipulation as the independent variable, global citizen variables as dependent variables, and participants' positive and negative affect as covariates.The omnibus test was significant, Wilks' Λ = .84,F(9, 310) = 3.07, p < .001,η p 2 = .08.As shown in Table 3, participants in the inactive global citizen condition rated their degree of identification with the normative environment and with the notion of global citizenship significantly more highly than participants in the active global citizen and control conditions.No other significant differences were found.

Global citizen structural model
To test the influence of the active versus inactive global citizen 'possible self' manipulation on Reysen and Katzarska-Miller's (2013a) model of the antecedents and outcomes of global citizenship identification, we conducted a structural equation model using Amos 19 (bias-corrected bootstrapping; 5,000 iterations; 95 per cent confidence intervals).Due to the related nature of the prosocial values to one another (and the antecedents to each other), we allowed the disturbance terms for these sets of variables to covary.Two error terms for global awareness items were allowed to covary.Lastly, the researchers allowed positive and negative affect to covary with all of the latent variables, to control for participants' emotional response to the manipulation.The predicted model adequately fit the data, χ 2 (230, N = 323) = 480.74,p < .001;RMSEA = .058,CI{.051; .065},NFI = .918,CFI = .955.
As shown in Figure 1  The indirect effect of global citizen 'possible self' manipulation was reliably carried by normative environment and global awareness on participants' identification with global citizens (see Table 4 for standardized betas of indirect effects and 95% biascorrected confidence intervals; all indirect effects were significant at p < .002twotailed).The global citizen 'possible self' manipulation also significantly predicted greater prosocial values through normative environment, global awareness, and global citizenship identification.The influence of normative environment and global awareness on prosocial values (e.g.valuing diversity) was reliably carried by global citizenship identification.In effect, the global citizen 'possible self' manipulation predicts greater normative environment and global awareness, which then predicts greater identification with global citizens, and then predicts greater endorsement of prosocial values.Note: Standardized betas and 95% confidence intervals; bias-corrected bootstrapping with 5,000 iterations; all indirect effects are significant at p < .002.

Discussion
The purpose of the present study was to examine the influence of a 'possible self' manipulation on the antecedents and outcomes of global citizenship identification.
As predicted, the saliency of a feared possible self as an inactive global citizen resulted in significantly greater identification with global citizens compared to the hoped-for possible self (active global citizen) and control conditions.Furthermore, the saliency of the feared self influenced global citizenship identification through participants' perception of their normative environment and global awareness.
The impact of the feared self also influenced the endorsement of prosocial values through the antecedents and identification of global citizenship.Together, the results suggest that thinking of a feared future self as an inactive global citizen was effective at influencing global citizen identification and related prosocial values and behaviors.This outcome is consistent with the findings of previous researchers: that feared selves have been more influential than hoped-for selves (Hoyle and Sherril, 2006;Murru and Martin Ginis, 2010;Quinlan et al., 2006).

Possible selves and global citizenship
In keeping with possible selves theory (Markus and Nurius, 1986), participants were able to develop self-representations of possible selves that were made salient and accessible through the experimental conditions of active or inactive global citizens.
As shown by participants' negative emotional reactions, the thought of becoming an inactive global citizen was indeed one that provoked fear.When the feared identity was salient, participants reacted by endorsing the perception that valued others, like friends and family, provide an injunctive norm to be a global citizen (normative environment), and the manipulation of future selves predicted the antecedents (normative environment and global awareness).These antecedents (global awareness and normative environment) then predicted greater identification as a global citizen, and global citizen identification then predicted endorsement of the prosocial outcomes (e.g.valuing diversity; displaying intergroup empathy) related to identity as a global citizen.
The present results support the self-discrepancy explanation of possible selves (Carver et al., 1999;Higgins, 1987) in which individuals are motivated to move away from a feared self (here, as an inactive global citizen) and increase the discrepancy between an actual and a feared possible self.When a feared possible self was made salientin the inactive global citizen condition, through the 'possible self' manipulation of thinking and writing about this future self -it appears that discrepancy-enlarging feedback served to activate the human tendency towards avoidance (Carver, 2006).Participants in the inactive global citizen condition sought to avoid this feared self and to enlarge their distance from it by endorsing global citizenship and its prosocial attitudes and behaviours.The 'possible self' manipulation had a different effect on participants in the active global citizen (hoped-for self ) condition: the discrepancyreducing feedback that resulted from thinking and writing about a hoped-for self did not create an approach motivation (Carver, 2006).Participants in the 'feared self' condition may have thought about how their current actions are incongruent with the positive identity of global citizen that they wished to realize in the future.This reaction may have been derived from a view that others in their everyday or normative environment endorse the identity of a global citizen, along with a perception of their global awareness.
An alternative explanation would be that simply raising the saliency of global citizenship would result in greater endorsement of this identity.However, if this were the case the degree of identification would not differ between the hoped-for and feared self conditions.The present results also enhance our understanding of prior research examining the link between possible selves and social identities.
Although Cameron (1999) positioned identification with groups (such as university students) as an antecedent to possible selves (e.g.join a group in order to realize a possible self ), the present study suggests that a 'possible self' manipulation involving thinking about a future self may serve to engender a social identity.However, greater identification with global citizens, as found in the present study, may reflect a desire to connect with a social identity in order to realize a future self, as suggested by Cameron (1999).

Global education, global citizenship, and possible selves
In developed nations such as the United States (Gollnick and Chinn, 2013), the UK (Dower, 2008), and Canada (Pike, 2008), education typically occurs in multicultural settings, a fact that creates opportunities for cross-cultural interaction and the development of a global citizen identity (Schattle, 2008).In addition, as students prepare for life in an increasingly globally interconnected world (Merryfield et al., 2008) it is proposed that having a more inclusive social identity (such as global citizenship) will make for easier adjustment to life in school, the workplace, and the community.The model of the antecedents and outcomes of global citizenship identification shown in the present study, and supporting prior tests of the model (e.g.Gibson and Reysen, 2013;Katzarska-Miller et al., in press;Reysen andKatzarska-Miller, 2013a, 2013c), suggests that such diverse educational settings are apt environments in which to engender global citizen identity.Furthermore, the outcomes of global citizenship identification (e.g.valuing diversity and environmental sustainability) are values and behaviours that schools wish to engender in their students (see Reysen and Katzarska-Miller, 2013b).A practical implication of the present findings is that teachers might offer an educational task in which students are asked to imagine themselves as inactive global citizens.Participants in the present research appeared to have sufficient understanding of this identity that they feared and wished to avoid it.Although the present research utilized university students, further research may show the activity is applicable to younger students.

Limitations and future research
Despite the novel findings presented in the present study, there are limitations to consider when interpreting the results.First, this research was conducted in a single university in the United States with mostly undergraduate students.Henrich et al. (2010) have identified significant differences across populations and cultures on a variety of psychological and behavioural measurements, including selfconcept and self-view.These indicate the need for caution in the interpretation and generalization of research findings relating to psychological constructs such as possible selves, and in the use of participant samples from a university population.
The results may differ for individuals embedded in different cultural settings, or for younger or older participants.Second, although the saliency of the feared possible self predicted greater global citizenship identification through the antecedents (i.e.normative environment and global awareness), other possible mediators that were not examined may also provide a link between possible selves and global citizenship identification.Further research utilizing different participant samples and additional mediators is needed.In addition, future research could include an examination of the duration and behavioural influence of the saliency of a feared possible self as an inactive global citizen.

Conclusion
The present study examined the influence of a 'possible self' manipulation on the antecedents and outcomes of global citizenship identification.The saliency of a feared self as an inactive global citizen (vs that of a hoped-for self as an active global citizen) served to increase global citizenship identification, through the perception that others in one's normative environment prescribed the identity of a global citizen and through greater global awareness.Global citizenship identification, in turn, predicted greater endorsement of prosocial values and behaviours.As educators continue to seek methods to increase global citizenship identification (Reysen and Katzarska-Miller, 2013b), the present results show that a 'feared self' activity can directly influence global citizenship identification and indirectly influence related prosocial outcomes.

Figure 1 .
Figure 1.Influence of 'possible self' manipulation on antecedents, identification, and outcomes of global citizenship (controlling for positive and negative affect; standardized betas are significant at p < .05)

Table 1 : 'Possible self' manipulation instructions Experimental group Instructions given
Active global citizen conditionWe are interested in your impression of yourself 5 to 10 years from now.More specifically, we would like you to think about yourself in the future as a person who is an active global citizen.In other words, think of yourself 5 to 10 years from now as a person who is globally aware, caring for other people, interested in other cultures, standing up for the rights of others, living an environmentally friendly life, and feeling responsible to help when you see other people being oppressed.When you think of yourself 5 to 10 years from now as an active global citizen, what images come to mind?

Table 3 : Means (standard deviation) by condition
Means with different subscripts are significantly different (p < .05).Results control for participants' positive and negative affect.Post-hoc comparisons conducted with Sidak correction for multiple comparisons.

Table 4 : Indirect effects of global citizen 'possible self' manipulation, normative environment, and global awareness
Texas A&M University-Kingsville.She teaches classes related to social psychology, organizational psychology, and cognitive psychology.Her research interests include possible selves, global citizenship, and mathematics self-efficacy.Contact: Marion Blake, Department of Psychology, Texas A&M University-Kingsville, Kingsville, TX, 78363, Email: Marion.Blake@tamuk.eduStephen Reysen is an assistant professor of psychology at Texas A&M University-Commerce.He teaches classes related to social psychology, intergroup relations, and multicultural diversity.His research interests include topics related to personal (e.g.fanship) and social identity (e.g.global citizenship).