This thesis describes the evaluation of a programming language construct, namely type systems, using methods from social science research. We investigate whether there is a significant difference in development time between developers using either a statically typed API or a dynamically typed API by conducting a controlled empirical experiment. The controlled empirical experiment is conducted using Dart, a language with optional typing, and DartPad, a web-enabled IDE for Dart. The result of the controlled empirical experiment is inconclusive. Additionally we conduct interviews as an inquiry into how developers view types. Interviewees say that static typing improves security, has a documenting effect, and functions as a social contract with other developers. They also say that dynamic typing affords faster development time.
This thesis can be considered a starting point for future research.