How do empirical software engineering researchers assess the credibility of practitioner–generated blog posts?

Background: Blog posts offer potential benefits for research, but also present challenges. The use of blog posts in SE research is contentious for some members of the community. Also, there are no guidelines for evaluating the credibility of blog posts. Objective: To empirically investigate SE researchers’ opinions on the credibility of blog posts, and identify criteria for evaluating blog posts. Method: We conduct an online survey of software engineering researchers (n=43), to gather opinions on blog–post credibility and credibility criteria. Results: There is diversity of opinion. The majority of researchers provide a qualified response to the credibility of blog posts: essentially, it depends. Several credibility criteria are valued by researchers, such as Reasoning, Clarity of writing, Reporting empirical data and Reporting methods of data collection. Approximately 60% of respondents thought the criteria generalised to other practitioner–and researcher–generated content. Conclusion: The survey constitutes the first empirical benchmark of the credibility of blog posts in SE research, and presents an initial set of criteria for evaluating the credibility of blog posts. The study would benefit from independent replication and evaluation.