Predictors of Sportspersonship: Contingencies of Self-Esteem, Contesting Orientations, and Moral Disengagement Free

Self-esteem reflects an individual's perception, emotional evaluation, and acceptance of the self. At least in Western culture, most people value feeling good about themselves. In fact, among college students, boosts to self-esteem are valued even more highly than sexual pleasure, being with friends, enjoying food, or receiving a paycheck (Bushman et al., 2011). But people differ in how they seek to enhance and protect their self-esteem. Building on the early theorizing of William James (1890), Crocker and colleagues (Crocker et al., 2006; Crocker et al., 2004; Crocker & Wolfe, 2001) posited that people seek to maintain and buttress their self-esteem through seeking success, not in everything, but in specific domains. Crocker and Wolfe (2001) proposed that state self-esteem increases following success and decreases following failure, fluctuating around a general level of trait self-esteem, when those successes and failures are in domains of contingent self-worth, but not if they are in other domains. Correspondingly, a contingency of self-worth can be defined as “a domain or category of outcome on which self-esteem is staked, so self-esteem depends on perceived successes, failures, or adherence to self-standards in that domain” (Crocker & Park, 2012, p. 310).

Crocker and colleagues (2003), focusing primarily on college students, identified seven specific domains that people use, in varying degrees, as bases for their sense of self-esteem: virtue, God’s love, family support, academic competence, physical attractiveness, competitive success, and gaining approval from others.

Research on contingencies of self-worth has demonstrated that how one seeks to buttress self-worth is predictive not only of fluctuations in self-esteem, but of mental health and depression, in the way people spend time, the kinds of goals they pursue, the situations they put themselves in, relationship choices, mood changes, and activity motivations and performances (see Crocker & Park, 2012, for a review). Of particular relevance to the current investigation, contingencies of self-worth have also been shown to relate to prosocial or ethical behavior. For example, students who base their self-esteem largely on self-perceptions of virtue have been shown to do more volunteering (Crocker et al., 2003). Among college men, but not women, basing self-worth on competitive success predicted more, and basing it on virtue less, cheating in a laboratory setting (Niiya et al., 2008).

If contingencies of self-worth predict prosocial behavior and cheating, they may also predict multiple dimensions of sportspersonship. It seems likely, for example, when athletes seek to boost self-esteem through competitive success, they may be willing to make sportspersonship compromises to achieve that boost (Kohn, 1992). One exception may be in the “respect for full commitment to one’s sport” aspect since such commitment may be seen as necessary to achieve the desired success.

In addition to directly predicting sportspersonship, contingencies of self-worth may influence other variables known to have predictive utility in this area. In particular, we examined whether contesting orientations and moral disengagement mediate relationships between self-worth contingencies and sportspersonship.

Building on research into the cognitive function of conceptual metaphors (Lakoff & Johnson, 1980, 1999; Landau et al., 2010), Shields and Bredemeier (2009, 2011) have identified two contesting orientations, each rooted in one or the other of two root metaphors: contest-is-partnership and contest-is-war. A conceptual metaphor, which typically operates below awareness, works by creating a conceptual mapping between elements of a target concept (e.g., the sport contest) and analogous elements of a source concept (e.g., a partnership or a war). This ubiquitous process, so essential to thinking, begins very early in life as the young child seeks to make sense of experience and language (Kovecses, 2020).

The meaning and purpose of contests are not simply inherent in the structure of sport. Rather, the perceived meaning and purpose of competition are imputed through interpretation. The two root metaphors provide the scaffolding needed to understand and think about the nature of the dynamics, processes, relationships, and aims of sport. While it is likely that athletes have access to both conceptual metaphors, only one can be cognitively active at a time. And research suggests that athletes have dispositional preferences for activating one metaphor more than the other (Funk et al., 2016; Shields et al., 2015a, 2016a, 2018a, 2018b). A person’s relative preference for activating the different metaphors is referred to as their contesting orientation (Shields et al., 2015a, 2016c).

When utilizing the contest-is-partnership metaphor (hereafter abbreviated partnership), contests are construed in a collaborative light. While participants who frame the contest as a partnership usually strive hard to win, it is the striving itself that is of most value. Participants who think through the partnership metaphor seek to use the contest’s oppositional tension to promote enjoyment and performance excellence. In contrast, when employing a contest-is-war conceptual metaphor (hereafter abbreviated war), contests are interpreted in a more antagonistic light and the purpose is largely reduced to designating a winner or winners who receive tangible or symbolic rewards, which may include boosts to self-esteem.

It is plausible that contingencies of self-worth are involved in the etiology of contesting orientations. For example, a child who learns to associate competitive success with an augmented sense of self-worth may concurrently come to think of opponents as enemies, an essential element of the war contesting orientation (Shields & Bredemeier, 2011). In contrast, children who learn that their self-esteem is enhanced when they act virtuously may be more inclined to develop a partnership orientation. People who feel good about themselves when they act in a prosocial manner may be prone to construing contests as potential partnerships for a shared good.

As noted, previous research has demonstrated that contesting orientations are good predictors of sportspersonship attitudes (Funk et al., 2016; Shields et al., 2015a). Research has also demonstrated that contingencies of self-worth, at least the virtue and competitive success contingencies, are predictive of prosocial behavior and/or cheating (Crocker, Luhtanen, et al., 2003; Niiya et al., 2008). If contingencies of self-worth are involved in the evolution of contesting orientations as seems theoretically plausible, it may be that contesting orientations mediate the relationship between contingencies of self-worth and sportspersonship.

While people vary in terms of the relative weight they give to virtue as a contingency of self-worth, nearly everyone wants to feel like they are moral (Haan et al., 1985). One prominent way that people protect their sense of moral selfhood, even when acting in ways that violate social norms or ethical principles, is through what Bandura (1991, 2016) calls moral disengagement. Moral disengagement refers to a set of psychological processes that can be used to avoid self-censure for morally problematic behavior. They include such processes as euphemistic labelling, displacement or diffusion of responsibility, and dehumanization. In sport, people’s tendency to morally disengage has been found to be a powerful predictor of morally-related behavior (Boardley & Kavussanu, 2011; Kavussanu et al., 2013; Li et al., 2014; Stanger et al., 2013; Tsai et al., 2014). In a study of predictors of moral disengagement in sport, contesting orientations were found to be significant, with the partnership orientation predicting lower, and the war orientation higher, use of moral disengagement (Shields et al., 2015b).

To our knowledge, there has been no research on the relationship between contingencies of self-worth and moral disengagement. While empirical evidence is lacking, it seems plausible that important relationships exist. For example, it seems likely that when self-esteem is staked heavily on competitive success that moral disengagement processes will be used more frequently. If winning is necessary for self-esteem, then moral disengagement may allow a person to engage in less than desirable behavior without the sting of self-censure.

The current study seeks to broaden our understanding of sportspersonship by examining the relations among contingencies of self-worth, contesting orientations, moral disengagement, and multiple dimensions of sportspersonship. Based on the plausible theoretical assumptions outlined earlier, we believe contingencies of self-worth are likely to predict both contesting orientations and moral disengagement. They may also predict sportspersonship, both directly and as mediated by contesting orientations and moral disengagement.

To test these theoretical assumptions, two research questions guided the analyses. First, do contingencies of self-worth exhibit the hypothesized relations to contesting orientations, moral disengagement, and sportspersonship? Second, do contesting orientation and moral disengagement mediate relations between contingencies of self-worth and sportspersonship?

For manageability and clarity, we focused specific hypotheses on the two contingency scales most relevant to sport and ethical behavior: the virtue and competitive scales. There are good theoretical reasons to hypothesize relationships between these two contingencies and sportspersonship. Lacking equally compelling arguments for the other contingencies, we considered them exploratory and only preliminary analyses were run with the other contingency scales included. Three specific sets of hypotheses were examined:

1. The virtue contingency scale will (a) relate positively with the partnership contesting orientation; (b) relate negatively with moral disengagement; and (c) relate positively with all dimensions of sportspersonship.

2. The competitive success contingency scale will (a) relate positively with the war contesting orientation; (b) relate positively with moral disengagement; and (c) relate negatively with all sportspersonship dimensions except the full commitment subscale for which we had no hypothesis.

3. Both contesting orientations and moral disengagement will partially mediate the relationships between the above two contingency scales and sportspersonship.

Participants were National Association of Intercollegiate Athletics (NAIA) first year college student-athletes (N = 1863, 55.7% male) of multiple racial identities (64.6% White/ Caucasian, 12.2% Black/African-American, 11.8% Hispanic/Latino/a, 1% Asian/ Asian-American, 8.5% Multi-racial/Multi-ethnic, 1.7% other, 0.3% not given), ranging in age from 18 to 28 (M = 19.0, SD = 1.4), who had competed for a mean of 10.06 years (SD = 4.04) in their primary sport: football (n = 598), soccer (n = 850), and volleyball (n = 415). Participants were drawn from Fall sports officially-sanctioned by the NAIA, across multiple NAIA member institutions across the United States.

After approval by the Institutional Review Board, participants were recruited through the NAIA’s national office. Specifically, the NAIA had implemented an online educational program for its student-athletes and the registration procedure was modified so that prior to accessing the educational content, athletes were given the option to participate in this study. Interested students were directed to an external site (SurveyMonkey), which housed the consent form and survey; those who declined to participate were taken directly to the NAIA’s educational materials. Approximately 96% of eligible athletes completed the surveys; the high response rate was likely due to the integration of the invitation to participate into the NAIA’s registration process. All instruments were administered in random order, no incentives were provided, and information was collected anonymously. After collection, data were imported into SPSS for screening and analysis.

To control for their potential effects, we collected participants’ age, gender, primary sport, and years of participation in their primary sport. Sport was coded “1” for volleyball (low-contact), “2” for soccer (medium-contact), and “3” for football (high-contact), while gender was coded “1” for males and “2” for females. In addition, to control for the effects of social desirability, a modified version of the Marlowe-Crowne Social Desirability Scale (SDS; Crowne & Marlowe, 1960) was administered. The modified SDS consists of eleven statements, which respondents are asked to rate as either true or false for them (Ballard, 1992). Items included, “When I don’t know something, I don’t mind at all admitting it,” and “I sometimes feel resentful when I don’t get my way.” Responses are coded “1” for each socially-desirable response, and then summed to generate a scale ranging from 0-11, with higher values indicating increased responsivity to social desirability concerns. The instrument has demonstrated a generally acceptable, if occasionally marginal, reliability (Ballard, 1992). Reliability in the current study was also marginal (α = .69).

These variables were assessed with the 35-item Contingencies of Self-Worth Scale (CSWS; Crocker et al., 2003). The CSWS consists of seven 5-item subscales that assess the extent to which self-esteem is based on virtue (e.g., “My self-esteem depends on whether or not I follow my moral/ethical principles”); competitive success (e.g., “I feel worthwhile when I perform better than others on a task or skill”); appearance (e.g., “My sense of self-worth suffers whenever I don’t think I look good”); approval from others (e.g., I can’t respect myself if others don’t respect me”); academic competence (e.g., “My self-esteem is influenced by my academic performance”); family support (e.g., “Knowing that my family members love me makes me feel good about myself”); and God’s love (e.g., “My self-esteem goes up when I feel that God loves me”). Responses were recorded on a 7-point scale from 1 (strongly disagree) to 7 (strongly agree). The CSWS subscales have good test-retest reliability and convergent and divergent validity (Crocker et al, 2003). In the current study, the scales demonstrated acceptable (Virtue a = .75) to excellent (God’s Love α = .93) reliability.

The Contesting Orientations Scale (COS; Shields et al., 2015a) was used to assess the degree to which participants endorsed a partnership and/or war conceptual metaphor for contesting. The COS asks participants to rate their agreement or disagreement with twelve statements, using a Likert scale anchored by 1 (strongly disagree) and 5 (strongly agree). Six items reflected the partnership conceptual metaphor (e.g., “The purpose of competition is to bring out the best in everyone”) and six items reflected the war conceptual metaphor (e.g., “In sports, like in war, opponents stand between you and success”). Reliability in the current study ranged from acceptable (partnership α = .78) to good (war α = .85).

The Moral Disengagement in Sport Scale-Short (MDSS-S; Boardley & Kavussanu, 2008) was used to measure moral disengagement. This scale consists of eight items, one for each of the disengagement mechanisms of the longer version of the scale (Boardley & Kavussanu, 2007). Items depict thoughts and feelings that an athlete may have (e.g., “Insults among players do not really hurt anyone”). Respondents indicate their level of agreement with each statement using a Likert scale anchored by 1 (strongly disagree) and 7 (strongly agree). In previous research, the scale has shown good internal consistency, with alphas ranging from .73 to .85. Boardley and Kavussanu (2008) also provide evidence of the scale’s factorial, convergent, and concurrent validity. Reliability was good in the current study (α = .84).

The Multidimensional Sportspersonship Orientations Scale (MSOS; Vallerand,, 1997) was used to assess participants’ sportspersonship attitudes. The original MSOS is comprised of five subscales: 1) respect for rules and officials; 2) respect for social conventions; 3) respect for full commitment to one’s sport; 4) respect and concern for opponents; and, 5) negative approach to sport. Because the fifth subscale—negative approach to sport—has demonstrated problems with factorial validity and internal reliability (e.g., Dunn & Causgrove Dunn, 1999; Lemyre et al., 2002; Vallerand et al., 1997), it was not used in the current study. Participants are presented with a series of statements reflecting the various dimensions of sportspersonship and asked to indicate the degree to which each statement corresponds to them on a five-point scale, anchored by 1 (does not correspond to me at all) and 5 (corresponds to me exactly). Reliability ranged from acceptable to good (rules and officials α = .86, social convention α = .86, full commitment α = .79, respect for opponents α = .73).

Data analysis was conducted via structural equation modeling (SEM) as it represented the most parsimonious method to simultaneously address our three research questions. The analysis proceeded in three major steps. First, we tested the measurement model—a single structural model evaluating all latent variables and their indicators. Adjustments to the measurement model were then made to ensure adequate fit to proceed with subsequent analyses. Second, correlations among the latent variables in the final measurement model were computed using full information maximum likelihood functions afforded by the SEM program being utilized (IBM AMOS). Third, the primary analysis was conducted on a single, unified structural model (Figure 1) that simultaneously analyzed the direct and indirect pathways between two exogenous contingency variables (virtue and competition) and four sportspersonship variables (respect for: opponents, rules and officials, full commitment to sport, and social conventions) via two contesting orientations (partnership, war) and moral disengagement. All four control variables (age, gender, sport contact level, and social desirability) were allowed to predict all four dimensions of sportspersonship. All exogenous variables (contingencies and controls) were allowed to correlate.

For ease of presentation, results were broken into four different diagrams, one for each dimension of sportspersonship (Figures 2a-2d). Three indicators suggested by Kline (2011) and Schweizer (2010) were used to assess model fit: comparative fit index (CFI), standardized root mean square residual (SRMR), and root mean square error of approximation (RMSEA). Cutoff values for good fit are not universally agreed upon; however, models are generally considered well-fitting when CFI ≥ .95, SRMR ≤ .1, and RMSEA ≤ .06 (Kline, 2011; Schweizer, 2010; Tabachnick & Fidell, 2007).

Before conducting the main analyses, data were screened for missing data and outliers (Tabachnick & Fidell, 2007), and the measurement model tested for its fit to the data (Kline, 2011). Missing data analysis revealed approximately 2.12% of data were missing, with no significant patterns. Full-information maximum likelihood (FIML) estimation was used in conducting the analyses (Baraldi & Enders, 2010). The measurement model did not provide a good fit to the data, however (χ²(₂₈₅₃) = 14226.96, p < .001; CFI = .86; RMSEA = .05; SRMR = .06). A series of confirmatory factor analyses (CFA) were performed for each measure, and standardized residual covariances examined for potential sources of misfit. For each scale, items that demonstrated high standardized residual covariances were removed. Specifically, eight items were removed from the CSWS, including two from the competition subscale and one each from the academic, family support, God, virtue, appearance, and approval subscales. Four items were removed from the COS, two each from the partnership and war subscales. Two items were removed from the MDSS. Finally, for each of the four MSOS scales, two items were removed. Skewness and kurtosis for revised scales were ≤ |1|, and alpha levels were acceptable to very good (Table 1 ). Model fit for the revised measurement model was acceptable (χ²(₁₃₉₀) = 5292.81, p < .001; CFI = .93; RMSEA = .04; SRMR = .04). Standardized loadings of indicators on latent variables for this final measurement model are presented in Table 3 .

Table 1 presents basic descriptive statistics for all of the study’s variables. Participants displayed a moderate degree of virtue- (M = 4.91) and competition-based (M = 4.96) contingent self-worth, with family- (M = 5.51) and academics-based (M = 5.36) self-worth rated slightly higher. Also, while participants tended to modestly endorse both contesting orientations, they evinced a slight preference for partnership orientation (M = 4.10) over war orientation (M = 3.86). Moral disengagement strategies were found somewhat disagreeable (M = 2.84). Finally, participants tended to see themselves as engaging in positive sportspersonship across multiple dimensions (Rules & Officials, M = 4.22; Social Conventions, M = 4.08; Full Commitment, M = 4.29) though they were noticeably ambivalent about respect and concern for opponents (M = 3.10).

A zero-order correlation analysis was then conducted, which largely exhibited the expected relations. The virtue contingency related positively to partnership orientation (r = 0.48) and all dimensions of sportspersonship (r = 0.31 - 0.46), and negatively to moral disengagement (r = -0.35). The competitive-success contingency related positively to both contesting orientations (partnership r = 0.39, war r = 0.44) and moral disengagement (r = 0.11). As expected, it also correlated negatively with the sportspersonship dimension of respect for opponents (r = -0.06). Contrary to expectations, it also demonstrated small, positive correlations with respect for rules and officials (r = 0.16) and respect for social conventions (r = 0.10). With regard to the full commitment dimension of sportspersonship, for which we had no hypothesis, it evidenced a moderate positive correlation (r = 0.25).

Partnership contesting orientation had substantial positive relations to all dimensions of sportspersonship (r = 0.23 - 0.61), and was negatively related to moral disengagement (r = -0.19). Conversely, war orientation related positively to moral disengagement (r = 0.35), and had largely no or negative relations to sportspersonship, though a positive relation to respect for full commitment was found (r = 0.31).

Structural equation modeling (SEM) was used to address our research hypotheses concerning the direct and indirect (i.e., mediated) relationships between study variables. Specifically, a model in which the virtue and competitive contingencies of self-worth influenced the adoption of contesting orientations, which, in turn influenced moral disengagement and sportspersonship was tested (Figure 1). The model controlled for the effects of athlete age, gender, sport contact level, and social desirability via direct paths between these variables to the four dimensions of sportspersonship (our outcome variables). In addition, the error terms for partnership and war contesting orientations were allowed to correlate, as were the error terms for the four dimensions of sportspersonship. Model fit was reasonably good (χ²₍₅₆₇₎ = 2313.42, p < .001; CFI = .94; RMSEA = .04; SRMR = .05). Examination of standardized residuals suggested that gender and sport contact level exhibited unexplained covariance with both moral disengagement and war contesting orientation. Subsequently, univariate analysis of variances (ANOVAs) were conducted to explore the nature of the relations between these four variables. Results indicated that males and athletes in higher-contact sports were significantly more likely to endorse a war orientation and moral disengagement strategies. Accordingly, the model was re-specified to include direct paths from gender and sport contact level to moral disengagement and war orientation, respectively, to control for these effects upon the study’s mediating variables. This model exhibited a minor improvement in fit (χ²(₅₆₃) = 1838.52, p < .001; CFI = .96; RMSEA = .04; SRMR = .04).

Relations among model variables largely supported the study hypotheses (Figure 2). The virtue contingency was positively related to partnership orientation and negatively related to both war orientation (Hypothesis 1a), as well as moral disengagement (Hypothesis 1b). It was also related positively to all four dimensions of sportspersonship, both directly and indirectly via partnership orientation (Hypothesis 1c). The competitive contingency exhibited a moderately strong, positive relation to war contesting orientation (Hypothesis 2a), as well as a more modest positive relation to moral disengagement (Hypothesis 2b). In addition, its relations to the dimensions of sportspersonship were largely negative, both directly and indirectly, excepting a modest positive relation via partnership orientation (Hypothesis 2c).

As part of the SEM analysis, direct, indirect, and total effects of the two contingencies were calculated for each dimension of sportspersonship, as well as the ratio of indirect to total effects (Pm). Pm is calculated by dividing indirect effects by total effects and is expressed as a percentage - or proportion - of the total effects that come from indirect effects. These results are presented in Table 2 The significant indirect effects of the contingencies on sportspersonship via contesting orientations and moral disengagement support partial mediation (Hypothesis 3). While caution must be used in interpreting Pm when total effects are small, Pm values ranged from 14% to 93% in the current analysis. Pm values for the virtue contingency were particularly substantial, ranging from 41% for respect for opponents to 62% for respect for rules and officials. The competition contingency also demonstrated a substantial proportion of indirect effects for respect for rules and officials (25%), respect for opponents (40%), and full commitment to sport (93%). In total, the final model explained a moderate to fairly large amount of variance in the dependent variables: respect for opponents R² = .24, social conventions R² = .43, full commitment R² = .44, rules and officials R² = .46.

The primary goals of this study were threefold: (1) to determine whether there are significant relationships between an athlete’s tendency to use self-virtue to buttress self-esteem and the multiple dimensions of sportspersonship; (2) to determine if there is a significant relationship between athletes’ proclivity to validate self-esteem through competitive success and sportspersonship; and (3) to determine whether support exists for contesting orientations and/or moral disengagement mediating relations between these self-worth contingencies and the dimensions of sportspersonship. Several specific hypotheses related to these goals were tested.

Consistent with hypotheses, both correlation and SEM results indicated that preference for affirming self-esteem through virtue was significantly related, as expected, to all four measured dimensions of sportspersonship. As expected, it was also related positively to partnership orientation and negatively to moral disengagement.

The structural equation models revealed similar, though not identical, patterns of indirect effects for each of the four dimensions of sportspersonship. Specifically, tethering self-esteem to virtue both directly and indirectly predicted all sportspersonship dimensions. In all cases, the virtue contingency effect on sportspersonship had significant indirect effects through contesting orientations and moral disengagement, providing support for partial mediation.

In brief, the overall pattern of relationships suggests that when athletes tether their self-esteem to being a person of moral character, they are more likely to think of competition through a partnership mindset which, in turn, increases sportspersonship, both directly and by reducing moral disengagement. They are also less likely to activate the war contesting orientation. This appears advantageous for sportspersonship because the war orientation directly and negatively predicts sportspersonship and indirectly through increasing the likelihood of moral disengagement. The only exception to this consistent pattern of results was that there was no direct effect of war orientation on the respect for rules and officials dimension of sportspersonship; all significant indirect effects were via moral disengagement.

There is a thematic congruence to these relationships. It seems plausible that tying self-esteem to virtue occurs when moral concerns are important to the self. This focus on morality carries into the contesting orientation preferences which have implicit moral norms favoring the partnership orientation (Shields & Bredemeier, 2011; Shields et al., 2016b). Similarly, tying self-esteem to perceiving yourself as virtuous may make it more difficult to disable moral considerations or engage in moral excuse-making, as happens with moral disengagement. Finally, while sportspersonship blends both moral and other prosocial values (Bredemeier & Shields, 1998), those who tether self-esteem to virtue are likely predisposed to uphold the prosocial and ethical norms of the contest.

All people who participate in sport want to win. But athletes vary in the extent to which they stake their self-esteem on winning. Examining Table 1 reveals that preference for affirming self-esteem through competitive success was positively related to a war orientation (and, to a lesser extent, the partnership orientation as well), moral disengagement, and all but the respect for opponents dimensions of sportspersonship (which was, as expected, negatively related). While the positive relations between the competitive success contingency scale and two dimensions of sportspersonship (respect for rules and officials and respect for social conventions) were counter to hypotheses, even a cursory examination of the correlations reveals that these positive associations were far less strong than the associations between the sportspersonship scales and the virtue contingency.

One reason why athletes who rely on competitive success to buttress self-esteem may still endorse some aspects of sportspersonship could be that they may feel a boost to self-esteem only if they also feel they won the contest fairly. The fact that the same positive relation between the competitive success contingency scale and respect for opponents did not hold, and in fact was negative, is consistent with this explanation. One can be fair, follow the conventions of one’s sport, and still disrespect those who stand in the way of your self-esteem affirming victory. Further research will be needed to confirm, refine, or refute these conjectures.

The mediation models shed further light on these relationships. Examining the models reveals that the positive correlations between the competition contingency and three dimensions of sportspersonship were not based on direct effects. Except for respect for opponents, where a direct negative prediction was demonstrated, the ability of the competition contingency to predict sportspersonship was entirely indirect, through contesting orientations and moral disengagement. The competition contingency most strongly predicted a war contesting orientation which, in turn, negatively predicted (directly or partially through the mediation of moral disengagement) sportspersonship. However, to a lesser extent the competition contingency also positively predicted partnership orientation which, in turn, positively predicted all aspects of sportspersonship. Thus, the small positive correlations between several dimensions of sportspersonship and the competitive success contingency appears to be a result of the competitive contingency predicting a partnership contesting orientation, which it does to a lesser extent than its prediction of a war orientation.

In brief, the overall pattern of relationships suggests that when athletes tether their self-esteem to competitive success, they are more likely to think of competition through a war mindset which, in turn, lowers sportspersonship, both directly and/or by increasing moral disengagement. There may be multiple reasons for these relationships. Unlike tethering self-esteem to virtue, tying self-esteem to competitive success may diminish the salience of moral concerns which, in turn, may decrease a focus on sportspersonship, especially if it is perceived to interfere with competitive success. Similarly, since competitive success is more outside the control of the athlete than is their own moral commitment, the competition contingency may increase a sense of opponents as enemies because they could, by emerging victorious, block the desired affirmation of self-esteem. This would likely activate the war contesting orientation which has been consistently found to relate to higher moral disengagement (Shields et al., 2015a, 2015b) and lower sportspersonship (Funk et al., 2016; Shields et al., 2015a, 2016c).

While the competition contingency most strongly predicts the war orientation, to a lesser extent it also predicts the partnership orientation. This suggests that the negative impact that tying self-esteem to competitive success has on sportspersonship may be offset, even reversed, when athletes adhere to a partnership mindset. This may occur when the tendency of the competition contingency to encourage a war orientation is overshadowed by environmental factors, such as coaching messages or team culture, that encourage a partnership approach.

Youth sports are a complex field comprised of many different organizations, sport areas, competitive levels, demographic characteristics, and interested constituents. More than 60 million youth take part in agency-sponsored youth sports each year (National Council of Youth Sports, n.d.). The coaches involved in youth sports range from occasional volunteers to highly trained professionals. When evaluating the potential usefulness of an investigation, this complexity needs to be kept in mind.

The current study was conducted with participants at the upper end of the youth sport continuum: first-year collegiate student-athletes. It is reasonable to assume that the patterns found in the present investigation developed through many years of participation in lower level youth sport programs. Are there implications, then, for coaches and other leaders involved in youth sports? Can the current findings inform an approach to youth sport-based character development?

Examination of both the correlation table and the four structural equation models reveals that of all the predictors of sportspersonship, partnership orientation appears to be the strongest. The only exception was respect for opponents, where the virtue contingency (.22) was marginally stronger than partnership orientation (.21). In turn, the virtue contingency (.40) was a much stronger predictor of a partnership orientation than was the competition contingency (.22). From these findings, two important implications for youth sport leaders can be drawn.

First, helping young athletes, through both modeling and open discussion, to embrace a partnership approach to sports should be an important goal of youth sport programs. Educating coaches on the distinctions between the two contesting orientations and how to recognize and resist triggers that may activate the war orientation should be critical components of any training program for coaches (Shields & Bredemeier, 2010, 2015). This does not mean that coaches, particularly at the younger end of the continuum, need to educate their young charges on the vocabulary of the contesting orientations. Rather, they can build the component parts of the partnership approach by emphasizing themes central to that orientation, such as a mastery goal orientation (Elliot, 2005), intrinsic reasons for sport participation (Hagger & Chatzisarantis, 2007), and a deep respect for rules, officials, teammates, opponents, and the sport itself.

Findings from the current investigation also suggest that to promote sportspersonship, coaches and other athletic leaders should seek to disconnect athletes’ sense of self-esteem from winning and losing, a finding consistent with other research on youth sport coaching effectiveness (Smith & Smoll, 2002; Smoll & Smith, 1997, 1981, 2002). Coaches should repeatedly share the perspective that people do not need to be scoreboard “winners” to be worthy human beings. This does not mean, however, that striving to win is unimportant. In fact, it is essential to sports. But an over-emphasis on winning, making it a source of pride, can be detrimental to developing positive attitudes toward sportspersonship, particularly respect for opponents. Young athletes should be encouraged to recognize that winning appropriately produces positive emotions, but self-worth should be tethered to the more enduring and intrinsic values of seeking to live a good and moral life. Losing should also not be equated with failure; in fact, accepting loss with grace and respect is an important part of a moral life. Respecting and valuing opponents is one of the most salient areas within the sport domain to put this into practice.

Some may wonder whether disconnecting competitive success from self-esteem will undermine motivation and performance. Based on substantial research, Crocker and Park (2012) suggest that the answer to that question is a definitive “no.” Rather than optimizing performance, making self-esteem contingent on competitive success tends to create stress and distraction that lead to suboptimal outcomes. Clearly, anchoring self-esteem to good character, which is a more internal, stable, and controllable contingency, is preferable.

Promoting both a partnership approach to contesting and a virtue anchor to self-esteem will also reduce the use of moral excuse-making. Moral disengagement is a constant threat to ethical behavior both in and out of sport and can be minimized by these two strategies.

Future researchers may wish to examine some of the relationships among the other sources of self-esteem and sportspersonship that were only addressed tangentially in this investigation. There are a number of potentially interesting relationships reflected in Table 1 that were not examined in depth in the current study. Anchoring self-esteem to academic achievement, a religious belief in God’s acceptance, or family support, for example, also appear to have positive associations with sportspersonship; conversely, tying self-esteem to physical appearance or approval from others have either a negative relationship or a nonsignificant relationship to sportspersonship. Future research could explore these relationships in more depth.

Another fruitful area for future research stems from the limited age range of the current sample. All participants were first year collegiate student-athletes with a mean age of 19. Certainly, this population differs in many ways from high school students who, in turn, differ from younger ages. A fruitful avenue for future research would be an investigation of how the relationships observed in the current study might vary across development.

Despite the large, multi-sport sample, our results are limited by several factors, including participant characteristics. All participants were U.S. intercollegiate athletes from three sports and results cannot be generalized to other countries, ages, competitive levels, or sports without due caution. It is important to note, for example, that a very small portion of respondents (0.7%) were above the age of 24, and thus fell well outside the typical ages of athletes found in colleges/universities in the U.S. It is also important to recognize that this study was cross-sectional in design and consequently no causal relationships were able to be directly examined. The SEM models were consistent with the hypothesis that contingencies of self-esteem influence the adoption of contesting orientations, and that contesting orientations influence both moral disengagement and sportspersonship. But these hypothesized pathways will need to be tested with longitudinal investigations. Finally, modifications were made to the original measurement model in order to achieve an acceptable fit, with deletion of multiple indicator items for each latent variable. Thus, caution is specifically warranted when comparing the current results to those from prior studies using the same variables. Even with these limitations, the current study provides compelling evidence of the theoretical significance of contingencies of self-worth, contesting orientations and, to a lesser extent, moral disengagement for predicting sportspersonship.

David Light Shields is now professor emeritus, St. Louis Community College. We have no conflicts of interest to disclose.

The authors wish to extend appreciation to Kristin Gillette and the National Association of Intercollegiate Athletics (NAIA) for their invaluable assistance with data collection.