An approach for the determination and correlation of diversity and efficiency of software development teams

Abstract

We examined the impact of diversity on team efficiency. To do so, a model was developed to measure both the efficiency and diversity of the teams. Based on these measures, the correlation between efficiency and diversity was also analysed. In addition, to demonstrate the applicability of the model, it was applied to a real-life problem involving five teams dealing with different software development projects. Firstly, diversity indices were calculated based on age, experience, education and gender information on each member for each team by using Simpson’s Diversity Index. Then, four key performance indicators (KPIs) were defined to measure the success rate of the teams. Depending on these KPIs, efficiencies of the teams were measured through data envelopment analysis (DEA). The correlation between team efficiency and each diversity factor was analysed and all four factors had positive correlation with efficiency. That is, in order to increase efficiency, teams should be composed of members with diverse characteristics. Education was the diversity factor that had the most positive correlation with team efficiency. This result highlights the importance of different educational backgrounds on team efficiency.

Significance: 

• This study represents the first attempt to measure team diversity using Simpson’s Diversity Index.
• A new technique is proposed to measure team efficiency through DEA.
• Team efficiency is positively correlated with diversity, specifically educational level, which is important for many software development teams to consider.