The study highlights the importance of both length of conversation and user longevity as crucial factors for the success of software engineering communities on Discord. Further, the impact of different programming language communities on length of conversations and if user longevity is influenced by user contribution and programming language in software engineering communities on Discord are analysed. The study found both of these relations to be positive, with those in communities dedicated to certain languages having longer conversations compared to others. Furthermore, the research also found that contribution has a positive impact on user longevity on the platform.
The dataset is taken and adapted from the paper by Subash K. et al. titled "DISCO: A Dataset of Discord Chat Conversations for Software Engineering Research"
The data cleaning and processing can be found in the Python notebook
Data analysis was carried out using SPSS, the latest analysis being found in RQ1 and RQ2 files respectively (one is output, one is dataset for SPSS).