In this paper we report on the use of a purpose built Computer Support Collaborative learning environment designed to support lab-based CAD teaching through the monitoring of student participation and identified predictors of success. This was carried out by analysing data from the interactive learning system and correlating student behaviour with summative learning outcomes. A total of 331 undergraduate students, from eight independent groups at the University of Surrey took part in this study. The data collected included: time spent on task, class attendance; seating location; and group association. The application of ANOVA and Pearson correlation to quantized data demonstrated that certain student behaviours enhanced their learning performance. The results indicated that student achievement was positively correlated with attendance, social stability in terms of peer grouping, and time spent on task. A negative relationship was shown in student seating distance relative to the lecturer position. Linear regression was used in the final part of this study to explore the potential for embedding predictive analytics within the system to identify students at-risk of failure. The results were encouraging. They suggest that learning analytics can be used to predict student outcomes and can ensure that timely and appropriate teaching interventions can be incorporated by tutors to improve class performance.