RAs (Dylan and Scott) are busy transcribing in the lab this week.

In October and November last year, groups of students enrolled in FRAN180 came to the DL2LLab in order to test various online tools and help each other edit their essays.  While students interacted with each other and with the technologies, we used VSC (video screen captures) to record all their interactions with the systems.  The VSC also recorded their voices so that we can analyse the types of interactions that occur between learners. What strategies are they using to help each other?  Are they using a specific metalanguage?  Are they using metacognitive strategies? Are they analyzing their language output in order to enhance its quality?

The first items of analysis are the transcripts of the learner-to-learner interactions. Next we will look at the learner-tool interactions to see whether they are using the tools effectively (see figure 1).  Thirdly we will analyse a questionnaire in which participants had the opportunity to tell us more about the intervention. Last, but not least, we interviewed the instructor (Dr. Catherine Léger) to better correlate the learner activity with the course outcome and get her feedback on the learning task.

Figure 1: image from a VSC session

Transcriptions of audio recording is often tedious.  To facilitate the tasks, we set up a series of abbreviations and coding.  Here are a few examples:

LVH lecture à haute voix (when the learner is reading their own text aloud. No need to transcribe but it is a very important cognitive strategy that we need to be aware of)

DS discours simultané (students talking at the same time)

Some words are used a lot (probably too much) by participants.  The most common are abbreviated when transcribed:  beaucoup (bcp), quelquechose (qch), je (ne) sais pas (jsp), je crois que (jkq), intéressant (int), …

In order to analyse the VSC, we will set up specific markers that we need to look at:

CL every time a user “clicks” on an item or on a page, each click is annotated by a specific action that is also coded, for instance:

B browsing (page)

H help (accessing the help menu/page)

SP search page

Mov mousing over an item

In analyzing the actions within a site, we try to understand the efforts produced by users. We might notice some big discrepancies betweens participants (i.e. some participant will produce a lot more effort to arrive to the same result).  Results will be recycled into various practical outcomes:   (1) train participants on how to use a tool more effectively, or (2) custom a learning task to the tool that will be used, (3) (re)assess the tool in relation to the learning outcome, or (4) find/design a better online tool.

We will discuss the first results of these analysis in a future blog. Stayed tuned and/or add your comments!