Extensive research has probed the impacts of personality traits on student satisfaction, academic anxiety, and performance, with particular attention paid to their implications during the COVID-19 pandemic. Notably, a conspicuous gap is discernible in the existing literature concerning investigations that scrutinize the influence of personality on group learning among different COVID-19 lockdown phases (e.g., pre, during, and post-lockdowns). This study endeavors to fill this gap by undertaking an examination of the fluctuations in group performance within data science curricula. This inquiry considers student cohorts comprising diverse personality traits across distinct phases of lockdowns. Through the investigation of 84 student groups, our research has unearthed intriguing correlations and patterns linking group performance with four influential factors, including personality traits, the lockdown phases (i.e., pre, during and post-lockdowns), the joint factor of lockdowns and personality traits, and the implemented COVID-19 policies (e.g., mask-wearing and social distancing) during the early-stage of back-to-schools. The interesting findings revealed from our studies have the potential to propel the advancement of our comprehension within the field of educational psychology and may serve as catalysts for heightened scrutiny of student personality traits, thereby motivating educators to refine their pedagogical approaches in collaborative learning experiences.