Psychological Pathway of Esports Spectators

A quantitative approach comprising self-registered questionnaire surveys was used to examine the proposed psychological pathway of esports spectators.

A convenient sampling method was employed (Saunders et al., 2012). Two rounds of data collection were completed for Phase One (N = 485) and Phase Two (N = 217). The questionnaires for both rounds were prepared using the Chinese online research software called Wenjuanxing. Links to the questionnaires were disseminated in two ways for each round. Firstly, the researchers shared the link across WeChat groups of various offline events, including regional collegiate esports and seasonal league games. Secondly, there were research assistants who helped disseminate the questionnaire links for each round, including one esports lecturer, three esports event organisers, and two esports team staffs. The assistants shared links to WeChat groups and at the event venues to recruit those who went to League of Legends Pro Leagues and Peacekeeper Elite League seasonal games in 2021 for Phase One and in 2022 for Phase Two. For Phase One, the questionnaire was sent to the assistants on the 5th of January 2021. Data collection lasted for three months and finished in late April 2021. For Phase Two, the questionnaire was sent to the assistants on the 7th of January 2022 and the link was closed on the 31st of January 2022.

The two-step approach of structural equation modeling (SEM; Anderson & Gerbing, 1988) was used to analyse the data. Phase One was designed to assess the adequacy (goodness-of-fit, reliability and validity) of the measurement model using a CFA. Phase Two was then followed to test the hypothetical direct and indirect pathways among the constructs using a path analysis. For each phase, preliminary analyses were conducted, including data screening (Hair et al., 2013), elimination of speeders (Smith et al., 2016), and elimination of outliers (Byrne, 2010). Univariate skewness (< |±2.00|) and kurtosis (< |±7|.00) statistics were also calculated for the items to test their univariate normality, while Mahalanobis distances were calculated to identify multivariate outliers (p < .01; Tabachnick & Fidell, 2007). Both phases were run in SPSS and AMOS 26. The maximum likelihood is used as an estimator.

* Please note that files for this dataset are under embargo until 27 January 2025.