Collaborative Problem Solving (CPS) is one of the 21st-century 4-C skills that involves complex cognitive and socio-emotional processes, essential for academic and professional success. Understanding individual contributions within these processes is key to effectively assessing CPS skills and predicting outcomes. Traditional methods, such as self-reports and peer assessments, can be biased, highlighting the need for objective evaluation measures. This study examines CPS in computer engineering undergraduates (15 triads, n=45) engaging in data visualization tasks by integrating subjective peer evaluations and objective group speech metrics. Our findings indicate that variations in individual contributions significantly affect CPS outcomes. Furthermore, combining subjective and objective metrics enhances the accuracy of performance predictions, offering a comprehensive understanding of team dynamics. The results underscore the importance of integrating multiple data sources for CPS assessment and suggest a deeper analysis of CPS behaviors to improve future evaluations.