Document Type : Research Paper
Author
Assistant Professor, Department of Educational Sciences, Payame Noor University (PNU), Tehran, Iran
Abstract
The Contingent decision-making styles of sports managers play a vital role in the success of sports organizations. However, predicting these styles based on individual and organizational characteristics has always been a complex challenge. This study employs machine learning algorithms, particularly decision trees and random forests, to develop an accurate model for predicting contingent decision-making styles among sports managers. In this quantitative-analytical study, data were collected from 150 professional sports managers in national leagues. Independent variables included demographic characteristics (age, experience), psychological traits (extraversion, neuroticism), and organizational factors (culture, competitive pressure). The dependent variable was decision-making style (rational, intuitive, avoidant). Decision tree, random forest, and logistic regression algorithms were used for modeling. Model accuracy was evaluated using Accuracy, Precision, Recall, and confusion matrix metrics. The random forest model demonstrated the best performance with 87% accuracy in predicting decision-making styles. Variables of "competitive pressure" and "managerial experience" had the greatest impact on predicting decision styles. Comparison with traditional methods (such as logistic regression) confirmed the significant superiority of machine learning. The findings prove that machine learning is an effective tool for analyzing decision-making behaviors of sports managers. The proposed model can assist sports organizations in selecting and developing effective leaders. This research opens new avenues for integrating artificial intelligence with organizational behavior studies in sports.
Keywords
Main Subjects
)