Publication Date

2012

Abstract

Objectives: Relationships between training load, psychobiosocial (PBS) states and performance are dynamic and individual-specific. The nature of these relationships can be investigated using a combination of dynamic linear models (DLMs) and mediating variable analysis, potentially assisting applied sports psychologists in planning and monitoring of individual elite athletes’ intervention programmes. Design: We illustrate this approach by examining the relationships of training loads with a performance-related state (‘self-efficacy’) and the role of potential mediating PBS variables (‘fatigue/lack of energy’ and ‘being in shape’) in explaining these relationships in an elite triathlete across time. Method: Self-reports of PBS states (twice weekly) and training data were collected over 137 days. Using DLMs and mediating variable analysis, direct (unmediated) and indirect (mediated) short-term associations of training load with ‘self-efficacy’ were examined. Results: In this triathlete, we found evidence for positive effects of training on ‘self-efficacy’, which were partly explained by feelings of ‘being in shape’ and suppressed by feelings of ‘fatigue/lack of energy’. Changes in the relationship between lagged training load and ‘fatigue/lack of energy’ were observed across time and were particularly pronounced in temporal proximity of an injury. Conclusion: Strengths of the presented approach are its dynamic nature enabling the observation of changes occurring over time, use of statistical inference rather than visual data interpretation, and quantification of mediating effects to identify potential pathways of intervention. Additionally, the DLM method can identify complex nonlinear associations by examining correspondence between changes in levels of predictors and changes in magnitude and direction of predictor-outcome associations.

School/Institute

Institute for Health and Ageing

Document Type

Journal Article

Access Rights

ERA Access

Access may be restricted.

Share

COinS