While multi-threading has become commonplace in many application domains (e.g., embedded systems, digital signal processing (DSP), networks, IP services, and graphics), multi-threaded code often requires complex co-ordination of threads. As a result, multi-threaded implementations are prone to subtle bugs that are difficult and time-consuming to locate. Moreover, current testing techniques that address multi-threading are generally costly while their effectiveness is unknown. The development of cost-effective testing plans requires an in-depth study of the nature, frequency, and cost of concurrency errors in the context of real-world applications. The full paper will lay the groundwork for such a study, with the purpose of informing...