Collaborative Virtual Environments (CVEs) are computer generated worlds, where two or more users can simultaneously interact with synthetic objects to perform a task. User performance is one of the main issues, caused by either weak coordination, network delay, less use of verbal and non-verbal modalities and less awareness or communication among collaborating users etc. The purpose of this paper is to present, analyze and evaluate user performance improvements for collaborative virtual environments (CVEs) in a comprehensive manner. Our approach of analysis is unique from already published work because we have performed detailed investigation on each user improvement factors, and have examined the assessments according to various parameters. We have comprehensively reviewed each factor for user performance improvement in CVEs and described its features, advantages, issues and challenges. We have observed in our survey that every technique has its own strengths to improve user performance. However, due to vast research available in this area, most of the user performance enhancement methods still lack a perfect choice for its implementation. Therefore, most favorable mechanisms, methodologies, algorithms, protocols and techniques are needed to develop in future, which can precisely, efficiently, and automatically enhance user performance in CVEs. This work is expected to provide a deep insight and useful directions for researchers in the field of CVEs, to enhance user performance.