TL;DR
This study uses a novel hybrid approach combining SEM and fsQCA to dissect the complex factors driving toxic behavior in professional esports spectators. It reveals that toxic behavior isn't caused by single traits but by specific combinations of personality, identification, and demographic factors, offering a more nuanced understanding than traditional linear models.
What Problem Does This Paper Address?
The explosive growth of professional esports, particularly games like League of Legends (LoL), has unfortunately been accompanied by a rise in toxic behavior within online spectator communities. This toxicity, manifesting as trolling, flaming, and harassment, negatively impacts the spectating experience and the overall health of the esports ecosystem. While previous research has touched upon factors influencing toxic behavior, there's a significant gap in understanding how individual differences, specifically personality traits (like the Big Five and Dark Triad), fan identification (with games and teams), and demographic variables, collectively contribute to this phenomenon. Existing studies often rely on linear models that may oversimplify the complex interplay of these factors, failing to capture the multifaceted nature of toxic behavior.
Methods
The researchers employed a two-pronged methodological approach:
- Partial Least Squares Structural Equation Modeling (PLS-SEM): This is a standard technique for analyzing complex relationships between latent variables and observed indicators. It was used here to assess the direct effects of individual predictors (personality traits, identification, demographics) on toxic behavior.
- Fuzzy-Set Qualitative Comparative Analysis (fsQCA): This is a more advanced, set-theoretic approach that moves beyond linear relationships. fsQCA is particularly useful for identifying configurations of conditions that lead to specific outcomes. It allows researchers to understand how combinations of factors, rather than isolated variables, contribute to phenomena like high or low levels of toxic behavior. This method addresses the "equifinality" principle, where different pathways can lead to the same outcome.
The integration of these two methods is a key contribution, allowing for both the assessment of individual predictor effects (PLS-SEM) and the exploration of complex causal combinations (fsQCA).
Datasets
The study utilized a survey dataset of 430 League of Legends (LoL) esports audience members. The data was collected in China during the LPL 2022 Summer competitions. The survey included measures for:
- Big Five personality traits (adapted from Chang et al., 2019)
- Dark Triad personality traits (adapted from Jones and Paulhus, 2014)
- Game Identification (adapted from Badrinarayanan et al., 2015 and Qian et al., 2023)
- Team Identification (adapted from Qian et al., 2023)
- Toxic behavior (adapted from Qian et al., 2023)
- Demographic variables (age and gender)
No new benchmarks were introduced; the study relied on existing, validated scales.
Key Insights and Ideas
- Complexity over Simplicity: The most significant insight is that toxic behavior in esports is rarely driven by a single factor. Instead, it emerges from specific combinations of personality traits, fan identification, and demographic characteristics. This challenges the notion that simply identifying "bad apples" is sufficient for understanding or mitigating toxicity.
- The Power of Configurations: fsQCA revealed distinct "solutions" or pathways leading to both high and low levels of toxic behavior. This highlights that different sets of conditions can result in the same outcome, underscoring the need for a configurational approach to analysis.
- Personality's Dual Role: While some personality traits (like conscientiousness and openness) showed a negative correlation with toxic behavior in PLS-SEM, their role in fsQCA was more nuanced, appearing in specific configurations. This suggests that their influence is context-dependent and interacts with other factors.
- Gender Matters, But How? The study found gender to be a significant predictor in PLS-SEM, with males showing a higher likelihood of engaging in toxic behavior. However, the fsQCA solutions further elaborated on this, showing how male gender interacts with other traits in specific configurations leading to toxicity.
- Identification's Double-Edged Sword: Game identification was found to be a positive predictor of toxic behavior in PLS-SEM. This is an interesting finding, as strong identification can also foster positive community engagement, suggesting a complex relationship where strong ties might also lead to more aggressive defense of one's chosen game or team.
Key Results
- PLS-SEM Findings:
- Significant predictors of toxic behavior included: conscientiousness (negative), openness to experience (negative), neuroticism (positive), psychopathy (positive), game identification (positive), and gender (positive, with males being more likely to engage).
- The overall R² for toxic behavior was moderate (.35), indicating that these predictors explained a substantial portion of the variance.
- fsQCA Findings:
- Necessary Conditions: Low conscientiousness, high Machiavellianism, high neuroticism, and high psychopathy were identified as necessary for high toxic behavior. Conversely, high conscientiousness, low Machiavellianism, low neuroticism, and low psychopathy were necessary for low toxic behavior.
- Sufficient Configurations: Two distinct configurations were identified for high toxic behavior (S1 and S2) and two for low toxic behavior (S3 and S4). These configurations involved specific combinations of personality traits, identification levels, and gender. For example, one solution for high toxicity involved young males with high Machiavellianism and psychopathy, coupled with low game/team identification.
Limitations and Future Work
- Self-Reported Data: The study relied on self-reported data, which can be subject to inherent biases.
- Generalizability: The data was collected from Chinese esports fans, limiting the generalizability of findings to other cultural contexts.
- Future Research Directions:
- Cross-cultural comparisons: To develop more robust configurations across diverse cultural settings.
- Qualitative insights: Conducting interviews to delve deeper into the motivations behind toxic behavior.
- Gender interplay: Further exploring the complex relationship between gender and toxic behavior, especially considering the underrepresentation of females in the sample.
- Team performance: Integrating factors like team performance into the analysis.
- Impact on viewing experience: Investigating how toxic behavior affects the overall spectating experience.
- Longitudinal studies: To capture the dynamic nature of toxic behavior and personality traits over time.
Summary updated