Gender Disparities in Tech : Evidence and Insights from Sentiment and Psycholinguistic Analysis

Gender disparities and biases constitute one of the biggest problems facing the technology industry in recent times. Using an anonymized dataset of employee peer reviews and managerial performance evaluations from a large technology corporation, we study the nature of gender disparities and detrimental stereotypes that persist in the industry. We find preliminary evidence of a statistical performance ceiling whereby men are awarded a disproportionate share of the top performance outcomes compared to women. Sentiment analysis of the textual feedback provided in employee peer reviews finds weak evidence that reviews of female employees tend to be more positive than those of male employees. A multi-dimensional psycholinguistic analysis of peer reviews further reveals many of the commonly ingrained workplace stereotypes that can be detrimental to organizational culture, productivity, and equity. Our study serves to promote the strategic analysis of large-scale human resource data in technology organizations to detect and correct gender disparities and prevent such disparities from coloring the development of technologies designed for general widespread use. ACM Reference format: Abhinav Maurya. 2019. Gender Disparities in Tech: Evidence and Insights from Sentiment and Psycholinguistic Analysis. In Proceedings of Anon, Anon,