Predictors of flow state in performing musicians: an analysis with the logistic regression method.

Introduction: Flow state has been deemed a desirable state for performing musicians given its negative correlations with musical performance anxiety, its relationship to optimal performance, and its possible effect on creativity. In the field of music, there are a few studies that have assessed intervention programmes to promote flow state in performing musicians with varying results in terms of their success. The flow condition-experience model proposes three components that would be the conditions for flow state to occur and six components that describe the experience of being in a flow state. In addition, within the vast academic literature on this experience, other factors that could influence its occurrence have been proposed. The main objective of this research was to detect which are the most suitable predictors from a set of independent variables collected to distinguish performing musicians with a high flow level. Methods: A binary logistic regression analysis was carried out with data from 163 musicians aged between 18 and 65. Independent variables were introduced in the analysis: skill-challenge balance, clear goals and clear feedback (condition-experience model); and also, gender, age, dedication, (musical) style, musical instrument and (performing) situation. Results: The results showed that the three conditions of the condition-experience model and the situation variable had positive associations with flow state. The model explained 78% of the variance of the dependent variable and obtained a 90.8% correct classification rate. Discussion: These variables seem to contribute most to a high flow level, and the importance of keeping in mind the intrinsic reasons why performers dedicate themselves to music is emphasised. The results and their implications for the training of performing musicians are discussed. Future lines of research are proposed, as well as collecting data on personality-related variables to introduce them into the regression model.