Abstract
The purpose of this research was to examine the role of self-construal on goal instrumentality in close friendships. Participants (105 female friend dyads, N = 210) were asked to set personal academic goals and were followed up over the course of a semester. Using the actor–partner interdependence model, results revealed that those with a high relational self-construal were more likely to perceive their friends to be instrumental to their goals. This in turn led to greater friendship commitment and closeness as well as greater goal progress over the course of the semester. Actor’s perception of goal instrumentality did not lead to friend’s goal progress but did have positive relational consequences. The results highlight the benefits of relying on others in self-regulation.
Research on the interpersonal influence of self-regulation has increased exponentially over the past decade. As social creatures, we spontaneously and automatically pursue goals associated with others whom we share an identity with and feel connected to. For instance, goal contagion (the unconscious adoption of others’ goals) is more likely to occur amongst members of the same group (Loersch, Aarts, Payne, & Jefferis, 2008). Likewise, increasing belongingness, even when the social connection is minimal, increases the likelihood of goal sharing (Walton, Cohen, Cwir, & Spencer, 2012). However, the inclusion of others in one’s goal pursuit may also happen more intentionally. This process has been termed transactive self-regulation, in which partners develop shared self-regulatory systems and rely on each other to assist them in their own personal self-regulatory tasks (Fitzsimons & Finkel, 2012). This strategy of self-regulation is beneficial to both the relationship (Fitzsimons & Fishbach, 2010; Fitzsimons & Shah, 2008) and the goal pursuit (Fitzsimons & Shah, 2008). This study examines the role of self-construal in transactive self-regulation among friendship dyads.
One of the first studies to examine the schematic integration of goals and significant others focused on the unconscious pursuit of goals when the mental representation of significant others are activated. Participants who shared their academic goal with their mothers put in more effort in pursuing their academic goal when primed with the word ‘mother’ as compared to those whose academic goal was not associated with their mothers (Fitzsimons & Bargh, 2003). Later studies expanded on the dynamics of transactive self-regulation, specifically in the area of goal instrumentality, which is the extent to which significant others are perceived to be helpful and supportive to an individual’s active goal. Fitzsimons and her colleagues found that instrumental others are perceived more positively (Fitzsimons & Shah, 2008), and people draw closer to instrumental others particularly when progress in the focal goal is slow (Fitzsimons & Fishbach, 2010). In addition, individuals who preferred instrumental friends over non-instrumental friends spent more time studying and consequently reported higher grades over the course of the term (Fitzsimons & Shah, 2008). In other words, they capitalized on their friends’ instrumentality and used them to facilitate self-regulatory efforts.
The findings are consistent with the self-expansion model where the other’s resources are included into the self and are used as one’s own resources (Aron et al., 2004). This intrinsic process is motivated by a desire to increase the efficacy of goal pursuit and likelihood of goal achievement. The model highlights the interpersonal aspect of self-regulation, such that it is not merely the individual’s ability to put in effort into goal pursuit but also the extent to which an individual includes others into the self and then utilizes the partner’s resources. Thus far, few studies have examined individual differences in this process of outsourcing self-regulation to close others. Building on the work of Fitzsimons and her colleagues as well as the self-expansion model, we reasoned that individuals who rely on others are likely to be better self-regulators than those who do not. That is, the more individuals defined themselves by their social linkages, the more they would rely and consider others as instrumental to their goals. In line with the findings from past research, we posit that greater reliance on others would in turn facilitate self-regulation efforts and increase relational closeness and commitment.
Relational Interdependent Self-Construal (RISC) refers to the extent to which the self is conceptualized in terms of close relationships (Cross, Hardin, & Gercek-Swing, 2011). As an individual difference measure, its focus is on how people define the self in relation to general close relationships. It differs from other types of relational self-construals that examine the self in relation to one’s social group, such as the collective-interdependent self (Markus & Kitayama, 1991) or the self in relation to specific relationships, such as the Inclusion of Other in Self (IOS; Aron, Aron, Tudor, & Nelson, 1991). Individuals who score high on the RISC scale disclose more personal information in relationships (Gore, Cross, & Morris, 2006), score higher on the IOS scale and consequently have closer and more committed relationships (Cross, Bacon, & Morris, 2000) than individuals who score low on the RISC measure. High RISC individuals also tend to find close relationships a source of motivation for self-regulation (Gore & Cross, 2006; Gore, Cross, & Kanagawa, 2009). It is important to note that the content of their goals is not necessarily more relational in nature or more likely to be shared, but that above and beyond independent personal interest, the interest of others are a motivational source for high RISC individuals (Gore & Cross, 2006). In addition, they are more likely to perceive close others as more supportive and encouraging of their goal pursuit (Gore et al., 2009). Consequently, their tendency to include others in their goals has positive consequences for their goal progress (Cross, 2009). In sum, defining oneself in terms of close relationships is associated with having closer relationships and better self-regulation.
As high RISC individuals are in general more likely to integrate representations of significant others into the self (Cross et al., 2011), we reasoned that they would also be more likely to develop a shared self-regulatory system and thereby engage in greater outsourcing of self-regulation. Reliance on others for aid in self-regulation is arguably easier when there is greater inclusion of self as the boundaries between the personal and the other’s (interpersonal) resources are blurred. Our hypothesis that high RISC individuals self-regulate better than low RISC individuals because they are more likely to rely on others and see others as instrumental to their personal goals is supported by past research highlighting the ‘we-ness’ in the cognitive representations of the self for high RISC individuals (Cross, 2009).
There is an important distinction between sharing one’s goals (i.e., pursuing goals together) and pursuing one’s own goals whilst being motivated by others. In the former, both members of the dyads are concerned about the same goal and pursuing the same interest; in the latter, one member of the dyad is pursuing his or her own goal that is not shared by the other member of the dyad. In the pursuit of one’s own personal goal, one can be motivated by thinking of how the success of the goal will benefit the other member of the dyad but the goal itself is not pursued by the other member of the dyad. Nonetheless, the other member of the dyad may play an important role in facilitating goal progress by providing social support and sharing the self-regulatory burden. This study examines the latter, which is consistent with Gore and Cross’ (2006) study, in which they examined reasons for pursuing personal goals.
This article extends past research by looking at the role of self-construal in the transactive self-regulation processes of personal goal pursuit. We expect that people are more likely to share the self-regulatory burden with significant others when there is greater inclusion of the other in the self-construal. Specifically, we expect that relational individuals are more likely to view their friends as being instrumental to their goals. Consequently, we expect that perceived goal instrumentality will lead to greater goal progress, relationship closeness and relationship commitment. In addition, by following female friendship dyads over a college semester, we were able to examine actor–partner processes over time. As stated above, we expected that RISC will predict greater goal instrumentality, which will contribute to one’s own goal progress and relationship outcomes (actor effects). With regard to partner effects, we did not predict direct partner effects on personal goal progress. That is, Sally’s perception of how helpful Marie is to Sally’s goal (goal instrumentality) should affect Sally’s goal progress (actor effect) but will not directly influence Marie’s own goal progress (partner effect). Sally would need to perceive Marie’s resources as her own and instrumental to Sally’s own goals in order to make progress in her own personal goals. Marie’s personal goals may not progress just because Sally sees Marie as instrumental. However, it is possible that there are partner effects on relationship outcomes. Sally’s perception of Marie’s goal instrumentality is expected to lead Sally to draw closer to Marie (actor effect). Consequently, as Sally draws closer to Marie, Marie should likewise report better relationship outcomes (partner effect).
In summary, based on the findings that high RISC individuals make greater goal progress (Cross, 2009) and have closer and more committed relationships (Cross et al., 2011) due to their tendency to include others into their self-construal, we hypothesize that high RISC individuals are more likely to form a shared regulatory system with close others and see close others as instrumental to their goals. We also hypothesized that their perception of others being instrumental to their goals will facilitate their goal progress as well as increase closeness and commitment to the instrumental other over 4 months. Perception of friend’s goal instrumentality is also expected to lead to changes in partner’s relationship commitment and closeness.
Method
Participants and procedure
Participants were 105 pairs of female friends (210 participants) 17 to 32 years old, with a mean age of 20.19 (SD = 1.91). Participants were recruited through advertisements posted online and around Montreal university campuses for a study on ‘friendship and goals’. There were two components of the study. For the first part, participants came into the lab in pairs and each filled out an hour-long survey about their own and their friend’s academic goal. All participants agreed to have their goal disclosed to their friend; each wrote down her goal on a piece of paper that was given to the other friend. Participants were contacted three times at three- to four-week intervals. At each follow-up, they were reminded of their goal and their friend’s goal and were asked to complete a 20-min online survey. Monetary compensation of $40 was provided to each participant.
Measures
Goal descriptions
At baseline (Time 1), participants were asked to indicate an academic goal they were trying to pursue during the semester. Examples of goals were ‘Obtain a letter grade improvement in my worst class’ and ‘Get into grad school’.
Relational-interdependent self-construal
Relational-interdependent self-construal was assessed at Time 1 using Cross, Bacon and Morris’s (2000) RISC measure. This measure taps the tendency to think of oneself in terms of close relationships. Ratings were made on a 7-point Likert scale, from 1 (strongly disagree) to 7 (strongly agree). Sample items include ‘My close relationships are an important reflection of who I am’ and ‘When I feel very close to someone, it often feels to me like that person is an important part of who I am’. The internal reliability of this 11-item measure was good, α = .89.
Goal instrumentality
Participants rated the extent to which they perceived their friend to be instrumental to their academic goal. Perceived goal instrumentality was assessed at the first (Time 2) and second (Time 3) follow-up using 3 items: ‘My friend plays an important role in my progress and achievement of this goal’, ‘My progress on this goal depends on my friend’s support and encouragement’ and ‘What my friend does is not affecting my goal progress in any way’ (reverse coded). Ratings were made on a 7-point Likert scale, from 1 (strongly disagree) to 7 (strongly agree). A summary of goal instrumentality was calculated by taking the mean of the two follow-up reports. The internal reliability of this 6-item measure was good, α = .88.
Goal progress
Goal progress was assessed at the first (Time 2) and last follow-up (Time 4) using 3 items: ‘I have made a lot of progress toward my goal’, ‘I feel like I am on track with my goal plan’ and ‘I feel like I have achieved my goal’. Ratings were made on a 7-point Likert scale, from 1 (strongly disagree) to 7 (strongly agree). In the present study, the internal reliability of this 3-item measure was adequate, α = .83 and .89 at the first and last follow-up, respectively. The two assessments of goal progress were highly correlated, r = .50, p < .001.
Relationship closeness
Relationship closeness with one’s friend was assessed using Aron, Aron and Smollan’s (1992) IOS measure. This measure taps a person’s sense of interconnectedness with another and consists of a set of seven diagrams representing different degrees of overlap of two circles (one labelled ‘self’ and one labelled ‘other’). Relationship closeness was assessed at Time 1 and at Time 4. The two assessments were highly correlated, r = .62, p < .001.
Relationship commitment
Relationship commitment was assessed twice (at baseline and at the final follow-ups) using the following items: ‘I am very committed to this friendship’ and ‘I consider myself to be very loyal to this friendship’. The internal reliability of this 2-item measure was adequate, α = .77 and .91 at Time 1 and Time 4, respectively. The two assessments of relationship commitment were highly correlated, r = .53, p < .001.
Results
The 105 dyads in this study were pre-existing and self-selected. Dependencies inherent between dyad members were handled by treating the members as interchangeable and using an actor–partner interdependence modelling approach (APIM; Kenny, Kashy, & Cook, 2006; Olsen & Kenny, 2006). The degree of non-independence that is present in dyadic data can be assessed by examining intra-class correlations between dyad members’ scores (see Kashy & Kenny, 2000). Examination of intra-class correlations between friend dyad members’ (105 dyads, 210 individuals) scores for all independent variables and dependent variables revealed non-significant correlations for RISC (r = −.03) and progress at Time 4 (r = .06) and a marginal correlation for progress at Time 2 (r = .15, p = .06). Significant correlations were also found for perceived goal instrumentality (r = .33, p < .001), relationship closeness at Time 1 (r = .21, p < .05) and Time 4 (r = .51, p < .001) and relationship commitment at Time 1 (r = .27, p < .01) and Time 4 (r = .57, p < .001). These correlations indicate a moderate to high tendency for friends to experience similar degrees of perceived goal instrumentality, closeness and commitment in their relationship.
To test the hypothesized predictions, three separate APIM analyses were conducted using structural equation modelling in AMOS. Prior to the analyses, all predictors were centred on the grand sample mean scores calculated for the entire sample. When APIM analyses are conducted with indistinguishable dyads, as is the case in the present research, one intercept is estimated for each dyad member and is constrained to be equal across the two members (see Kenny et al., 2006; Olsen & Kenny, 2006, for more details on conducting APIM with indistinguishable dyads, using SPSS). In line with Cobb, Davila and Bradbury (2001), the model fit was assessed using the χ2 and the comparative fit index (CFI; Bentler, 1992). A non-significant χ 2 indicates that the model and data are consistent. The CFI represents the comparison between the fit of the specified model and a null model. A CFI value of .90 or above indicates a good model fit (Hoyle, 1995).
A first model was tested to investigate the relationships between RISC, goal instrumentality and goal progress. The model tested (see Figure 1) was composed of four exogenous variables (actor and partner reports of RISC and Time 2 goal progress) and four endogenous variables (actor and partner reports of goal instrumentality and Time 4 goal progress). All variables were centred on the grand mean (Aiken & West, 1991). The four exogenous variables were allowed to co-vary. A covariance was specified between the error terms of actor and partner reports of goal instrumentality and between actor and partner Time 4 progress. For clarity concerns, these covariances do not appear in Figure 1. Actor and partner paths were specified from RISC to goal instrumentality and Time 4 goal progress. Actor and partner paths were also specified between goal instrumentality and Time 4 goal progress. Finally, actor and partner paths from Time 2 goal progress to Time 4 goal progress were also specified. The model presented above was found to fit the data, χ2 = 15.334, ns (CFI = 1.0). Standardized direct effects are reported. As can be seen in Figure 1, actor’s RISC predicted actor’s goal instrumentality (b = .31, p < .01), which predicted actor’s goal progress at Time 4 (b = .16, p < .05), controlling for Time 2 goal progress, which also predicted progress at Time 4 (b = .46, p < .001). All other paths were not significant. Indirect effects were examined to assess whether goal instrumentality mediated the relationship between RISC and changes in goal progress. The total indirect effect of actor’s RISC on actor’s goal progress at Time 4 was found to be significant (bootstrapping point estimate = .050; 95% confidence intervals (CIs) varying from .010 to .112), whereas the total indirect effect of actor’s RISC on partner’s goal progress was not (bootstrapping point estimate = .013; 95% CI varying from −.024 to .073).
The APIM model showing actor and partner effects for goal progress. As the intercepts were constrained to be across the two members, only one intercept is shown for clarity. *p < .05. **p < .01. ***p < .001. APIM: actor–partner interdependence modelling approach.
To distinguish the indirect effect through actor’s goal instrumentality from the indirect effect through partner’s goal instrumentality, the phantom model approach was used (see Macho & Ledermann, 2011). Thus, the specific indirect effect of actor’s RISC on actor’s progress through actor’s perceived goal instrumentality could be examined by constraining to zero the path from actor’s RISC to actor’s progress through partner’s perceived goal instrumentality. The reverse analysis was then conducted; that is, we examined the specific indirect effect of actor’s RISC on actor’s goal instrumentality through partner report of goal instrumentality by constraining to zero the ‘actor’s RISC → actor report of perceived goal instrumentality → actor’s progress’ path. This allowed us to examine if the indirect effects of actor’s RISC to actor’s progress is significant through actor’s and/or partner’s perceived goal instrumentality (see Macho & Ledermann, 2001). The analysis revealed that actor’s goal instrumentality was a significant mediator of the relationship between actor’s RISC and goal progress (bootstrapping point estimate = .045; 95% CI varying from .009 to .100), whereas partner’s goal instrumentality was not (bootstrapping point estimate = −.003; 95% CI varying from −.031 to .004). The results indicate that individuals who are high RISC are more likely to report increases in goal progress as a result of seeing their friends as instrumental to their goals.
A second model was tested to investigate the relationships between RISC, goal instrumentality and relationship commitment. The model tested was composed of four exogenous variables (actor and partner reports of RISC and Time 1 commitment) and four endogenous variables (actor and partner reports of perceived goal instrumentality and Time 4 commitment). The four exogenous variables were allowed to co-vary. A covariance was specified between the error terms of actor and partner reports of goal instrumentality and between actor and partner Time 4 commitment. For clarity concerns, these covariances do not appear in Figure 2. Actor and partner paths were specified from RISC to goal instrumentality and Time 4 commitment. Actor and partner paths were also specified between goal instrumentality and Time 4 commitment. Finally, actor and partner paths from Time 1 commitment to Time 4 commitment were also specified. The model did not yield an adequate fit to the data. Based on the suggested modification indices, a covariance was added between actor’s Time 1 commitment and actor’s goal instrumentality’s error term. The model fit was then found to be adequate, χ2 = 28.74, ns (CFI = .932). As can be seen in Figure 2, the actor paths from RISC to goal instrumentality (b = .30, p < .01) and Time 4 commitment (b = .17, p < .05) were significant, controlling for actor (b = .41, p < .001) and partner (b = .23, p < .001) Time 1 commitment, that both predicted actor’s Time 4 commitment. All other paths were not significant. Indirect effects were examined to assess whether goal instrumentality mediates the relationship between RISC and actor’s and partner’s changes in commitment. The total indirect effect of actor’s RISC on actor’s commitment at Time 4 was found to be significant (bootstrapping point estimate = .054; 95% CI varying from .011 to .120). The phantom model approach (Macho & Ledermann, 2011) was then used to isolate the two indirect effects (i.e., through actor’s and through partner’s goal instrumentality). The analysis revealed that actor’s perceived goal instrumentality significantly mediated between actor’s RISC and actor’s commitment (bootstrapping point estimate = .045; 95% CI varying from .011 to .097), whereas partner’s perceived goal instrumentality did not (bootstrapping point estimate = −.003; 95% CI varying from −.027 to .004). The total indirect effect of actor’s RISC on partner’s commitment was not significant (bootstrapping point estimate = .027; 95% CI varying from −.015 to .095). The results suggest that high RISC individuals became more committed to their friendships if they viewed their friends as instrumental to their goals. Their friends, however, did not report becoming more committed to the relationship.
The APIM model showing actor and partner effects for relationship commitment. As the intercepts were constrained to be across the two members, only one intercept is shown for clarity. † p < .10. *p < .05. **p < .01. ***p < .001. APIM: actor–partner interdependence modelling approach.
The final model investigated the relationships between RISC, goal instrumentality and relationship closeness. The model tested was composed of four exogenous variables (actor and partner reports of RISC and Time 1 closeness) and four endogenous variables (actor and partner goal instrumentality and Time 4 closeness). The four exogenous variables were allowed to co-vary. A covariance was specified between the error terms of actor and partner reports of goal instrumentality and between actor and partner Time 4 closeness. For clarity concerns, these covariances do not appear in Figure 3. Actor and partner paths were specified from RISC to goal instrumentality and Time 4 closeness. Actor and partner paths were also specified between goal instrumentality and Time 4 closeness. Finally, actor and partner paths from Time 1 closeness to Time 4 closeness were also specified. The model was found to fit the data, χ2 = 20.71, ns (CFI = .995). As can be seen in Figure 3, the actor paths from RISC to perceived goal instrumentality (b = .31, p < .01), which then predicted Time 4 closeness at Time 4 (b = .22, p < .01), were significant, controlling for actor (b = .51, p < .001) and partner (b = .15, p < .10) Time 1 closeness, that both predicted actor’s Time 4 closeness. Actor’s perceived goal instrumentality was also found to predict partner’s closeness at Time 4 (b = .16, p < .05). All other paths were not significant. Indirect effects were examined to assess whether goal instrumentality mediates the relationship between RISC and actor’s and partner’s changes in closeness. The total indirect effect of actor’s RISC on actor’s closeness at Time 4 was found to be significant (bootstrapping point estimate = .078; 95% CI varying from .024 to .154). The phantom model approach (Macho & Ledermann, 2011) was then used to isolate the two indirect effects (i.e., through actor’s and through partner’s goal instrumentality). The analysis revealed that actor’s perceived goal instrumentality mediated the relationship between the actor’s RISC and changes in closeness (bootstrapping point estimate = .062; 95% CI varying from .026 to .121), whereas partner’s perceived goal instrumentality did not (bootstrapping point estimate = −.011; 95% CI varying from −.045 to .012). The total indirect effect of actor’s RISC on partner’s closeness was also found to be significant (bootstrapping point estimate = .065; 95% CI varying from .008 to .144). Further analyses revealed that actor’s perceived goal instrumentality significantly mediated the relationship between actor’s RISC and partner’s closeness (bootstrapping point estimate = .048; 95% CI varying from .014 to .106), whereas partner’s perceived goal instrumentality did not (bootstrapping point estimate = −.014; 95% CI varying from −.053 to .015). The results revealed that individuals who are high in RISC drew closer to their friends as a result of seeing their friends as instrumental. Importantly, their friends also reported becoming closer to these high RISC individuals.
The APIM model showing actor and partner effects for relationship closeness. As the intercepts were constrained to be across the two members, only one intercept is shown for clarity. † p < .10. *p < .05. **p < .01. ***p < .001. APIM: actor–partner interdependence modelling approach.
Discussion
Self-regulation research has typically focused on the capacity of the individual to exercise self-control. However, as self-regulation is increasingly acknowledged to be an interpersonal process, the conceptualization of self-regulation must go beyond personal traits (e.g., Tangney, Baumeister, & Boone, 2004) or personal behaviours and attitudes (e.g., Gollwitzer, 1999; Koestner, Lekes, Powers, & Chicoine, 2002) that pertain directly to self-regulation and instead incorporate the non-self-regulatory factors that facilitate the use of the self-regulatory muscle into the understanding of self-regulation. It is not just one’s ability to exercise self-control that makes for effective self-regulation, but also the delegation of self-control to others that contributes to effective self-regulation. Thus far, past research has focused on people picking up the goals of others and the consequence of goal adoption on goal progress and relationship commitment, but few studies have examined the other side of the process, the deliberate delegation of goals to others.
We conducted a longitudinal study with female friendship dyads to examine whether high RISC individuals are better self-regulators through goal instrumentality. As expected, high RISC individuals were more likely to have friends who were instrumental to their goals and consequently they made greater personal goal progress. In addition, the actors reported closer and more committed relationships. On the other hand, actor’s perceived goal instrumentality did not benefit the friend in terms of goal progress or commitment. It is surprising that despite the friend not increasing in relationship commitment, the friend did draw closer to the actor over time. It may be that relationship commitment will increase only if the friend experiences actual goal progress but relationship closeness is fluid and thus more responsive to the actors’ behaviours and perceptions. There is some evidence suggesting relationship commitment (as compared to intimacy/closeness) is more likely to be the result of a conscious decision (Fehr, 1998; Sternberg, 1998). Perhaps the friend responds to the friend’s overtures of closeness without committing to the relationship, if indeed the friend sees the actor as someone who is not instrumental to her own goal. In other words, the friend may like the actor but not invest in the relationship. Further research is necessary to elucidate the distinction between closeness and commitment, particularly in the processes of interpersonal self-regulation.
As this study focused on personal goals, the partner effect on personal goal progress was not significant. Perhaps partners are more likely to influence personal goals if the goals are shared or if there is mutual instrumentality. The more Sally and Marie’s goals are shared and overlap, the more Sally’s perception of Marie’s goal instrumentality may benefit Marie’s own goal progress. As Marie makes progress on her goal, she would increase in commitment and closeness to Sally. This would be consistent with the work by Gere, Schimmack and their colleagues, showing that the more congruence between romantic partners’ goals, the greater the affective well-being in the relationship (Gere, Schimmack, Pinkus, & Lockwood, 2011). Conversely, the more the goals conflict with one another, the lower the relationship quality (Gere & Schimmack, 2013). It is also possible that there are direct partner effects when there is mutual instrumentality that benefits both goals. For example, if Sally and Marie pursue their personal academic goals dyadically and study together, they will experience mutual instrumentality. In this scenario, Sally’s report of Marie’s goal instrumentality may predict Marie’s personal goal progress, as Marie is able to pursue her own goal while being instrumental to Sally’s goal. 1 On the other hand, both members of the dyad may see each other as instrumental to their personal goals but the instrumental behaviours are person and goal specific, such that Sally’s instrumental behaviour for Marie’s goal does not advance Sally’s own goals. For example, if Sally proofreads Marie’s English term paper, it takes time away from her own physics assignment. Marie may be instrumental to Sally’s academic goal by providing her with ample supply of cookies during her late-night study sessions, but this does not advance Marie’s understanding of the Oedipal themes in Macbeth. Thus, while both members of the dyad see each other as instrumental to their personal goal progress, the instrumentality is not transferable but is specific to the needs of the person and the means of attaining the goal. Rather in this type of situation, the more the actor outsources to the partner, the more this reliance may tax the partner’s resources and decrease the partner’s own goal pursuit (Fitzsimons & Finkel, 2012). For instance, recent research found that individuals who are high in agreeableness do not benefit when they are in close proximity with their romantic partners as they are more likely to set aside their personal goal pursuit to focus on their romantic partner (Gore, 2014). Therefore, even if Marie and Sally both see each other as instrumental to their goal pursuit, Sally’s academic goal progress will suffer the more instrumental she is to Marie’s goal due to the constraints of personal resources.
This study also highlights the difference in short-term versus long-term consequences of goal instrumentality. In the short term, there is some evidence that participants who outsource to their instrumental others put in less effort in their personal goal pursuit (Fitzsimons & Finkel, 2012). However, this study shows that in the long run, this strategy ultimately benefits the outsourcer. In particular, it seems that high RISC individuals are very effective in utilizing interpersonal resources to enhance personal self-control, rather than weakening their self-regulatory capacity. It is clear from the results of this study and past studies (Fitzsimons & Finkel, 2011, 2012) that individuals who rely on others also draw closer to and are more committed to instrumental others. This suggests that high RISC individuals deliberately position themselves in interpersonal situations that will increase the chances of successful self-regulation.
One of the limitations of this study is that we examined only Sally’s view of Marie’s instrumentality (perceived instrumentality), but it is possible that Marie’s perception of how instrumental (felt instrumentality) she is to Sally’s goal will lead to her drawing closer to Sally. 2 Our results showed that the actor’s perception of friend’s goal instrumentality predicts the actor’s self-regulation effort and actor and friend’s relational outcomes. However, it is also possible that the friend’s felt goal instrumentality will also lead to her own report of closeness. It is reasonable to expect that not only do people draw closer to those who are instrumental to their goals, but people are more likely to be instrumental to those who are close to them, as would be predicted by the self-expansion model (Aron et al., 2004). Such that, both members of the dyad would mutually make their personal resources available to each other to enable each other to pursue and attain their goals (as the significant intra-class correlation of perceived goal instrumentality in this study suggests). Future research can distinguish between perceived and felt instrumentality and the unique outcomes associated with both. Nonetheless, our findings do highlight the interactive and dynamic nature of interpersonal self-regulation such that the actor’s perceived goal instrumentality predicts the friend’s report of closeness.
Another limitation of this study is that it did not differentiate between actual instrumental behaviours and the cognitive instrumentality of the close other. When the close other is not available to give tangible support, is the cognitive representation of close instrumental others enough for outsourcing self-regulation processes to take place? Indeed this is implied by the experimental manipulations used by Fitzsimons and her colleagues such that thinking about an instrumental other is sufficient for participants to engage in outsourcing self-regulatory processes (Fitzsimons & Finkel, 2011, 2012; Fitzsimons & Fishbach, 2010; Fitzsimons & Shah, 2008). This aspect of transactive self-regulation may be particularly useful in preventing the depletion of the significant others’ resources. If high RISC individuals can draw support from instrumental others schematically, this will free the partner to pursue his/her own goals. This allows the actor to remain instrumental to the partner’s goals even in times when no tangible helpful behaviours can be engaged in. If the actor regularly invests time and energy into the shared self-regulatory storage by doing concrete instrumental behaviours, the partner will benefit not just in the moment but also during times when the actor has to invest time and energy into his/her own goal.
It is becoming increasingly clear that self-regulation is not just a personal skill to strengthen, but also includes the efficiency with which one can gather and use interpersonal resources to facilitate one’s own efforts. Nonetheless, more research is necessary to elucidate the specific self-regulatory processes. For instance, is there a difference between instrumental behaviours that are directly linked to the partner’s goal (e.g., proofreading a term paper) and those that are indirect but enable the partner to pursue the goal (e.g., baking cookies for late night study sessions)? Future research can also examine the dynamic interplay between being instrumental to one’s partner while receiving support from one’s partner to pursue one’s own goals. What happens when one member of the dyad invests more than the other? Does being a self-regulatory leech whose withdrawal from the shared self-regulatory storage outweighs his/her investment create a deficit in the system that will be detrimental to personal self-regulation and the interpersonal relationship? We suspect that as long as both members of the dyad are investing into the system, the reliance is mutual, the system is in equilibrium and personal self-regulation is enhanced. When one member of the dyad withdraws more than he/she is investing in, this puts the system in jeopardy and the other member’s personal self-regulation may suffer.
Footnotes
Funding
This research was supported by grants to Richard Koestner from the Social Sciences and Humanities Research Council of Canada (SSHRC) and the Fonds Québecois de Recherche Sur la Société et la Culture, Quebec (FQRSC).