An experience sampling study of organizational stress processes and future playing time in professional sport

ABSTRACT

This study examined the relationships between daily cognitive appraisals of organizational events, affective responses, and coping. In addition, a 5-year longitudinal relationship between coping and performance outcomes at the senior professional level was assessed. Using an experience sampling method, professional academy rugby union players (N = 39, M_age = 17.23 years, SD = 0.87) completed daily diary measures of appraisals, affective responses, and coping over 5 weeks of training. Hierarchical linear modeling revealed that daily cognitive appraisals were related to daily affective responses and coping functions enacted by behaviours, after accounting for a series of within (e.g., time, day, week) and between-person (e.g., personality, key decision-makers) differences. Zero-inflated negative binomial regression revealed that coping related to eliciting support was associated with minutes played at the senior professional level five years later. This study extends theoretical knowledge of the within- and between-person relationships that explain organizational stress experiences. The findings suggest that some coping functions enacted by behaviours may be early indicators of future performance outcomes in professional sport.

KEYWORDS:

• Appraisals
• coping
• diary methods
• multilevel
• performance
• personality

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

¹ A copy of the background questionnaire is available from the first author on request.

² By “event”, we refer to characteristics of a specific organizational environment that influence a person’s experience of thoughts, feelings and behaviours (Beal & Weiss, 2013). Although the term may be compared with the terms “demands” and “stressors”, events represent a broader definition by which a range of environmental variables encapsulate daily demands, social constraints and opportunities (Lazarus, 1999, p. 63). Events relevant to the specific organizational environment in this study were identified in consultation between the first author and the academy staff in a staff meeting that occurred prior to the data collection period.

³ An incremental forward stepwise approach was adopted to check for significant variation in regression slopes at level 1. For example, to test hypothesis 1, negative affect was entered as the level 1 outcome, and all appraisals were person mean centered (CWC) as the level 1 independent variables. Within-person control variables (e.g., week, day, time) were all entered prior to the inclusion of independent variables and left in their raw metric form. For each equation, level 1 slopes were initially allowed to vary across individuals (i.e., random slopes). Where slopes had non-significant variance components (p < .10) or low reliabilities (<.05), they were fixed to be invariant across participants (Raudenbush et al., 2011). This step was then repeated to check for further invariance in slopes. This approach was continued until only random slopes were left to vary between participants in the equation at that step. Following this step, between-person control variables were entered incrementally at level 2: (a) person-averaged threat appraisals, (b) person-averaged challenge appraisals, (c) person-averaged harm appraisals, (d) person-averaged negative affect, (e) person-averaged positive affect, (f) neuroticism, (g) extraversion, (h) key decision-makers, and (i) playing position. All control variables were grand mean centered at the overall mean of the participant sample to provide meaning to the intercept.

⁴ To check for the robustness of results, we examined each hypothesis by comparing two regression models. The first model (n = 39, df = 997) included level 2 control variables (i.e., averaged appraisals and affect, neuroticism and extraversion, key decision-makers and playing position). In comparison, the second model (n = 28, df = 698) included level 1 control variables (i.e., week, day, time) in addition to the level 2 controls. In all cases, the hypotheses were supported. However, because some level 1 controls were significantly associated with affect and coping variables, it was decided to accept the hypotheses based on the findings presented from the second model.