Cookies and Privacy

The University uses cookies on this website to provide the best experience possible including delivering personalised content on this website, other websites and social media. By continuing to use the site you agree to this, or your can go to our cookie policy to learn more and manage your settings.

Psychology Research Seminars

Centre for Applied Psychology - Our Research Seminars 350x263 - People at a desk Throughout the academic year, the Centre for Applied Psychological research run open seminars for visiting lecturers to showcase their research.

Below, you'll find information on our next seminar, as well as information about upcoming and past events.

Check back regularly to find out more details about what the Psychology Research Seminar Series has in store.

Next Seminar

6 February 2019
3pm, C285

Pelham Carter (Birmingham City University)
Applying large-scale corpus analysis methods to online forum content: a study of a community of sex workers and ‘punters’ 

Research on online forum content and communities often adopts a traditional thematic analysis approach (Attard & Coulson, 2012; Carter et al., Submitted). While this allows detailed qualitative analyses, it also introduces issues with the subjective selection of the themes and the practicality of dealing with the volume of written data within a sample. In contrast, Corpus Linguistics offers a data-driven approach that can extract similar information from very large text databases.

In this paper, we compare the two approaches by drawing on data from an online sex work forum. A sample of posts were collected for the thematic analysis (containing 2631 posts across 60 different threads covering one week). Nvivo was used to conduct a thematic analysis (Braun & Clarke, 2014), which revealed five themes: preferences, personal narrative sharing, advice, philosophical issues, and community maintenance.

The second approach involved the construction of a corpus containing the text of the general discussion forum (comprising 16.5 million words from 255,891 posts covering six years) using web-scraping methods. Topic Modelling (Blei et al. 2003) was used to automatically identify themes within this larger corpus. While this approach revealed some of the same themes as the thematic analysis, it also highlighted additional ones such as the community’s concerns with health, cleanliness and legal issues.

We argue that such methods provide a more rigorous and objective framework for the analysis of online forum content. We will also present ideas for further analysis using Corpus Linguistic methods that are enabled by studies at this scale.

Upcoming Seminars
Upcoming Seminars
Coming soon!