Passionate apologies: understanding apologizing behaviour using the dualistic model of passion

Abstract Research on the dualistic model of passion has shown that people’s responses to adversity while pursuing passionate activities are associated with both harmonious passion (HP) and obsessive passion (OP). In this research we focused on a specific type of adversity, an interpersonal transgression, and tested if HP and OP predict different apologizing tendencies. University students (N = 395) read a hypothetical scenario that described making a mistake that negatively affected a peer. They then reported their apologizing behaviour in response to the scenario. We found that HP predicted a greater likelihood of apologizing to the peer and endorsing more comprehensive apology statements, whereas OP predicted a lower likelihood of apologizing and endorsing more defensive apology statements. These results contribute to our understanding of how and when HP and OP can impact the relationships people have while pursuing activities they love.


Introduction
People often interact with others while pursuing passionate activities in their lives. The dualistic model of passion (DMP) conceptualizes passion as a strong motivational desire to engage in an activity that one loves, values, pursues regularly, and has incorporated into one's identity (Vallerand, 2015). This model therefore considers passion to be a context-specific motivational construct (Vallerand, 1997) that entails a deep connection with a particular activity. Moreover, the DMP distinguishes between two forms of passion. The first, harmonious passion, involves a passionate activity being pursued with a sense of control, flexibility, and personal volition that remains in balance with other life domains. The second, obsessive passion, involves an uncontrollable urge to pursue an activity that occupies an overwhelming part of one's identity and life (Vallerand, 2015). This means that, although both HP and OP involve liking, valuing, and engaging in an activity on a regular basis, they differ in the quality of the relationship one has with an activity; with HP, the person is in control of the activity, whereas with OP, it is as if the activity is in control of the person.
Research relying on the DMP has shown that the quality of the interactions people have with others while pursuing their passions depends on the extent to which a passion is harmonious or obsessive (Vallerand, 2015). HP consistently predicts higher quality interpersonal relationships within the context of a passionate activity Philippe et al., 2010;Utz et al., 2012), whereas OP is unrelated to interpersonal relationship quality and is even associated with greater interpersonal conflict . Efforts to understand the processes through which HP and OP affect relationship outcomes has pointed to the role of emotions; HP predicts greater positive emotional experiences, which in turn predict higher relationship quality, whereas OP predicts greater negative emotional experiences, which in turn predict lower relationship quality (Philippe et al., 2010).
Our focus in this research was on an additional process that may affect the relationships people have while pursuing their passionate activities: apologizing behaviour. People make mistakes, and there are times when these mistakes affect others. Apologizing for transgressions can be helpful because they can resolve conflicts, promote forgiveness, and repair trust in relationships (e.g., Fehr et al., 2010;Kim et al., 2004). However, people differ not only in whether they apologize for a transgression, but also in how they apologize. Schumann (2014) proposed a framework that identifies different elements of comprehensive apologies, such as expression of remorse, acceptance of responsibility, and request for forgiveness. This framework also identifies several elements of ineffective, defensive apologies, such as victim blaming, excuse making, and denial. Apologies can include both comprehensive and defensive elements, but incorporating more comprehensive elements should lead to more positive outcomes for both victims and transgressors (Schumann, 2014).
Giving a comprehensive apology, however, is not always easy. Apologizing is a potentially awkward process that can come at a cost to the apologizer's immediate sense of self-esteem and feelings of power and control (Okimoto et al., 2013). Providing comprehensive apologies may be especially threatening to the apologizer because they highlight the negative implications of the transgression (Schumann, 2014(Schumann, , 2018. Consequently, people often anticipate that giving an apology will be an aversive experience (Leunissen et al., 2014) and, when one is given, some opt to apologize in a more superficial and even defensive way as an act of self-protection (Schumann, 2014).
We predicted that apologizing behaviour within the purview of a passionate activity would show differential associations with both HP and OP. HP involves adopting a non-defensive, mindful engagement in an activity (St-Louis et al., 2018), being resilient in the face of adversity (Paquette et al., 2022), having a secure sense of self-esteem (Mageau et al., 2011), and approaching difficult situations with a tendency toward addressing problems directly (Schellenberg & Bailis, 2016). This means that those with high HP should not find apologizing to be especially threatening, and should be inclined to apologize for transgressions in order to resolve conflicts. Therefore, we hypothesized that HP would facilitate apologizing behaviour by helping people (a) be more willing to apologize for a transgression (hypothesis 1a), find the idea of providing an apology to be less aversive (hypothesis 1b), and (c) provide more comprehensive apologies (hypothesis 1c). On the other hand, OP involves engaging in an activity with a sense of defensiveness (Vallerand, 2015), a fragile sense of self-worth in which one's self-esteem is tied to activity outcomes (Mageau et al., 2011), and with a tendency to rely on avoidance strategies when coping with difficult situations (Schellenberg & Bailis, 2016). This means that people with high levels of OP may find the act of apologizing to be highly threatening, and may self-protect by not offering an apology and, if one is offered, providing an apology that is more defensive. Therefore, we hypothesized that OP would impede apologizing behaviour by causing people to (a) be less willing to apologize for a transgression (hypothesis 2a), find the idea of providing an apology to be more aversive (hypothesis 2b), and provide more defensive apologies (hypothesis 2c).
Our aim was to test if both HP and OP were associated with different aspects of apologizing, including one's willingness to apologize for a transgression, perceived aversiveness of apologizing, and the use of different apologizing elements. Focusing on these different aspects allowed us to examine associations between both forms of passion and different components of the apology process (i.e., whether or not an apology is provided, feelings about apologizing, how an apology is given). We opted to study these relationships with university students because they often report having high levels of passion toward their studies (Schellenberg & Bailis, 2016;Vallerand, 2015), and regularly need to work with others in their programs. Ethics approval was obtained from the University of Manitoba psychology/sociology research ethics board before data collection (approval #P2020:109), and all participants provided informed consent. Anonymous data and syntax files are available on the Open Science Framework at https://osf.io/zbhxu/.

Participants
Participants were 395 undergraduates (324 females, 69 males, 1 non-binary, 1 unspecified) at the University of Manitoba (M age = 19.38 years, SD age = 3.29) who participated in the study in exchange for course credit. A power analysis revealed that this sample size would be sufficient to detect a small-to-medium sized effect (f 2 = .085) with .95 power. Three additional students participated in the study but were excluded due to a large amount of missing data (n = 1) or because they indicated at the end of the online survey that they did not respond honestly (n = 2). Most participants identified their ethnic background as White/European (44.1%) or Filipino (12.9%). A full breakdown of participant ethnic/cultural backgrounds is reported in the Supplementary File (see Table S2).

Procedure and measures
Participants completed an online survey about responding to difficult situations in university. First, participants completed the Passion Scale (Marsh et al., 2013;Vallerand, 2015) to measure levels of HP and OP toward their studies. The Passion Scale assesses HP (e.g., "Studying is in harmony with the other activities in my life"; "Studying is well integrated in my life.") and OP (e.g., "I have almost an obsessive feeling for studying"; "I have difficulties controlling my urge to study.") with 6 items each on a scale from 1 (not agree at all) to 7 (totally agree). In line with previous research with students (e.g., Schellenberg & Bailis, 2016), items were altered to refer to studying.
Next, participants read the following scenario that described a difficult situation in university: Imagine that you are taking a university course that requires you to make a class presentation. The presentations are made in pairs, and your professor has selected one of your classmates to be your partner for the presentation. You and your partner are both very busy with school and with work, so you agree to divide the presentation in half -your partner is responsible for preparing and presenting one half of the presentation, and you are responsible for preparing and presenting the other half.
The night before your presentation, your boss asks you to come into work for an extra shift because a few of your coworkers called in sick. You agree because you want to help out your boss and you could also use the extra money. But as a result, you are so tired and exhausted at the end of the day that you sleep through your alarm and are late for your class presentation. Both you and your partner are deducted marks on your presentation because you are late.
Participants then answered questions regarding how they would respond if they were in this situation (Schumann & Orehek, 2019). First, participants answered a single item assessing the likelihood that they would apologize to their partner in this situation. Second, participants answered four questions assessing the extent to which they would find apologizing to their partner aversive ("It would be stressful/hard/unpleasant/humiliating to apologize to my partner"; see, Leunissen et al., 2014). Third, participants read 26 statements that could be made as part of an apology to their partner and indicated the likelihood that they would say each one (e.g., "It was wrong for me to come late to class"). These statements were designed to measure the different apology elements proposed by Schumann (2014). Each element was measured with two items that were created specifically for the scenario (see Table S1 in the Supplementary File). A measure of comprehensive apologizing was created by averaging scores on all eight comprehensive apology elements (α = .78), and a measure of defensive apologizing was created by averaging scores on all five defensive elements (α = .70; see, Schumann & Orehek, 2019). All items assessing responses to the scenario were made on a scale from 1 (not at all/not agree at all) to 7 (totally/totally agree).

Results
Descriptive statistics, correlations, and internal consistencies are reported in Table 1. These preliminary analyses showed that HP and OP were strongly correlated with one another (r = .60, p < .001). Correlations between HP and OP of this magnitude are not uncommon in the passion literature given that they are both dimensions of the passion construct (Marsh et al., 2013;Vallerand, 2015). Mean levels of HP were also greater than means levels of OP (t [394] = 24.23, p < .001, d = 1.07), which is also a regular finding in passion research (Vallerand, 2015).
We tested the relationships between both passion types and apologizing responses using multiple regression as this allowed us to control for the other passion type in all analyses (Vallerand, 2015). Separate analyses were conducted for each dependent variable (i.e., likelihood of apologizing, apology aversiveness, levels of comprehensive apologizing, levels of defensive apologizing). In each regression analysis, both HP and OP were entered simultaneously as independent variables. Results are displayed in Table 2. HP predicted a greater likelihood of apologizing (b = 0.10, p = .002), and greater levels of comprehensive apologizing (b = 0.10, p = .032). OP, however, predicted a lower likelihood of apologizing (b = −0.19, p < .001), and greater levels of defensive apologizing (b = 0.27, p < .001). Both HP and OP did not predict feelings of apology aversiveness.
When analyzing apologizing likelihood, we found that most participants (91.4%) indicated that they would definitely apologize in the scenario. We re-analyzed the results by grouping participants into those who would and would not definitely apologize using logistic regression, and found that the likelihood of definitely apologizing was positively associated with HP (B = 0.667, SE = .224, p = .003, OR = 1.949), but negatively associated with OP (B = −0.992, SE = .202, p < .001, OR = 0.371).

Discussion
This research showed that apologizing behaviour within the context of a passionate activity is associated with the quality of one's passion. After imagining themselves committing an interpersonal transgression, students' levels of HP predicted a greater likelihood of apologizing to the victim in the scenario and greater endorsement of comprehensive apology statements, whereas OP predicted a lower likelihood of apologizing and greater endorsement of defensive statements. This research contributes to our understanding of how HP and OP impact people's interactions with others while pursuing activities that they love.
Apologizing has the potential to repair damaged relationships and rebuild trust (Kim et al., 2004). By being less likely to apologize and by offering more defensive apologies, those with Note. Cronbach's alpha values are reported on the diagonal. Gender was coded as 0 (female) and 1 (male).
Participants who did not identify as either male or female or who did not specify a gender were excluded from the analyses with gender. * p < .05. ** p < .01 high OP may be at risk of allowing mistakes, missteps, and misunderstandings to negatively impact their relationships. Apologizing may thus be a reason why OP predicts greater relationship conflict , and could be an additional mechanism that explains why HP, but not OP, predicts higher quality relationships (see, Philippe et al., 2010). More broadly, these findings are in line with the tenets of the DMP (Vallerand, 2015) and with previous research studying how passionate people respond in challenging situations (e.g., Schellenberg & Bailis, 2016); the defensive, self-protective engagement that is characteristic of OP prevents people from addressing challenges directly by being vulnerable, accepting responsibility, and asking for forgiveness for transgressions, whereas the flexible and non-defensive engagement that is characteristic of HP allows people to tackle problems head-on and provide more comprehensive apologies.
These findings also contribute to the literature on the types of people who are more likely to apologize, and who are more likely to apologize effectively. Having a general tendency to apologize for transgressions has been associated with various person factors such as having high levels of honesty-humility and conscientiousness (Dunlop et al., 2015), being prone to experiencing guilt and feeling empathy toward others (Howell et al., 2012), and having high self-compassion, neuroticism, and low narcissism (Howell et al., 2011). The current research shows that HP and OP, both context-specific motivational factors (Vallerand, 1997), also play a role in predicting apologizing behaviour. Studying apologizing within the context of passionate motivation may be especially valuable given that passion entails devoting a great deal of time and energy toward an activity that one likes, values, and has incorporated into one's identity (Vallerand, 2015). People Note. Confidence intervals (CIs) are bias-corrected and accelerated (BCa) and based on 5000 samples. All items were assessed on a Likert scale ranging from 1 to 7. * p < .05. ** p < .01 with such a high level of engagement in an activity are bound to make mistakes that affect others. Offering effective apologies may therefore be critical in order to maintain social harmony while pursuing the activities that are meaningful and central in people's lives.
We consider this research to be an initial yet revealing step forward in understanding how passionate people respond to interpersonal transgressions. Additional research is needed to address several limitations of the current investigation, including its reliance on self-report assessments, focus on undergraduates (some of whom had low levels of both HP and OP toward their studies), and use of a hypothetical scenario. Our sample also consisted mainly of female participants, which may have skewed overall apologizing tendencies (see, Schumann & Ross, 2010). We should note that a disproportionate number of students (91.4%) indicated that they would definitely apologize if they were faced with the situation described in the scenario. This is almost certainly due to the scenario that was presented to participants, which described a situation that very clearly warranted an apology. But this overrepresentation of apologizers in our sample may have introduced bias in our estimates (see, Salas-Eljatib et al., 2018). We therefore believe the results concerning apologizing likelihood should be interpreted with caution. Future research should focus more directly on the connection between passion and one's likelihood of apologizing for a transgression by focusing on situations or scenarios in which the expectation of an apology is less clear. Additional research is also needed to test the long-term consequences of different apologizing responses within passionate activities for both victims and transgressors. More research in this area will continue to reveal how passion affects people's interactions with others, especially when these interactions do not go so well.