Using the tools of screencasting and keylogging for formative and summative assessment of L2 writing

Pedagogical approaches to writing in a second/foreign language (L2) have often focused on the finished product, i.e., the final text as it is submitted by the learner. While many interesting insights can be gained from these texts, they only show the tip of the iceberg. All the different steps that have led to the finished product, including the many revisions that are usually part of the process of writing a text, are indeed kept hidden (at least if the text has been typed, which is now becoming the default case in many pedagogical situations). Digital technologies like screencasting and keylogging, however, by recording the screen and keyboard activity during a writing task, now make it possible to shift the focus from the product to the process and thus reveal the whole iceberg. The value of such process data will be illustrated with examples taken from a digital archive called the Process Corpus of English in Education (PROCEED; Gilquin 2022). This corpus includes screencast videos (recorded by means of OBS Studio) and keystroke logs (recorded by means of Inputlog; Leijten & Van Waes 2013) based on writing tasks carried out by intermediate learners of English as a foreign language (EFL). The presentation will focus on the use of these data for formative and summative assessment. In formative assessment, i.e., assessment for learning, learners can be required to observe their own process data so that they become aware of what their writing process looks like. They can also compare their data with those of their peers, to situate themselves among the group, as well as with the data of model writers, to help them identify successful writing strategies. As for summative assessment, i.e., assessment of learning, it will be argued that the writing process should be taken into account, in the same way as the assessment of speech takes fluency into account. This could involve the frequency and location of pauses or the use of online resources. A corpus such as PROCEED could be exploited to determine the typical writing behaviour of learners at a given proficiency level (for example, in terms of pausing time or number of revisions) and to set this as the expected standard. This corpus-derived standard could then help assess learners’ writing process empirically. The presentation will end with a discussion of how the addition of a writing process dimension could ultimately help improve writing skills in EFL