Robotics technology has been widely utilized for multiple applications in recent years. Human comfort has a direct impact on task quality and efficiency during the human-robot collaboration process. As an emerging study, human comfort in human-robot collaboration attracts more and more attention of scholars. In this study, we propose a novel approach to characterize and analyze human comfort based on T-Test to advance the understanding of human factors for the future of smart manufacturing. A factor set of robot performance and work environments that affect human comfort in the human-robot collaboration process is developed. We implement the proposed method to a high-diversity group of participants who are from different educational backgrounds, ages, genders, and countries. Experimental results and analysis indicate that the comfort levels of subjects are effectively evaluated under different/same robot/work environment factors. In addition, the highly statistically significant factors that affect the participants' comfort level positively or negatively are discussed.