I am a PhD student at the Southampton Education School. My research focuses on the cognitive writing processes used by university students in essay-style writing. As such, a large part of my work uses keystroke logging, a method of collecting real-time information about keyboard and mouse inputs during computer-based writing tasks.
Throughout my PhD, I have learned that whilst keystroke analysis can be a valuable way of gaining insight into the types of cognitive processes that writers may use during writing tasks, the way in which keystrokes are generally analysed lacks transparency. Thus, one of the projects I have been working on is developing a reproducible framework for identifying and calculating different types of writing features (pauses, revisions and bursts). This includes creating clear conceptual definitions for each writing feature and identifying which combinations of keystrokes result in each component. I have also been working toward creating reproducible macro scripts to identify these writing features within a keystroke log.
Finally, I have been using Gaussian Mixture Modelling to analyse pause data from keystroke logging tasks. To make my analysis transparent and reproducible, I have used R to conduct my statistical analysis. This software is open source and code-based, making it easy to follow every step of the analysis. I will also make these scripts openly available via an Open Science Framework repository when I have finished the project.
Through my work with the Southampton ReproducibiliTea journal club, I have become more aware of issues with transparency and reproducibility within scientific research through reading various articles about these topics and engaging in discussions about open science with researchers across disciplines. I am glad that I have been able to apply some of what I have learnt through our meetings in my own research.