Week 6 Drawing

Pattern of text highlighting while reading Williamson, Bayne & Shay (2020)

This week, I decided to record how I used a highlighting pen to mark texts while I was reading Williamson, Bayne & Shay (2020) paper. where each column represent a page, and each mark represent approximately 3 lines of text. I used a traffic-light system to indicate how much texts were highlighted. This recording was inspired by the text-highlighting functionality of common e-reading software like Kindle. At a functional level, records like this provide a teacher some insights into what their students might think is important message in a prescribed reading. Similar to Kindle, collation of such data from a group of students allows features like “frequently highlighted text” to be highlighted for a reading activity.

However, as suggested by Williamson, Bayne & Shay (2020), datafication increases the risk of pedagogic reductionism as well as limiting the way a teacher see their students. If a teacher uses text highlighting as a parameter to measure students’ intellectual engagement, it also defines such behaviour as the model way a student should interact with a reading material. Teachers may be prompted to reinforce e-reading app as the “official” way or indeed approved way to read one’s course reading. Students might also be prompted to look good in their data by simply highlight most (if not all) of the text, gaming the system to generate an all-green pattern.

Reference:

Williamson, B. Bayne, S. Shay, S. 2020. The datafication of teaching in Higher Education: critical issues and perspectivesTeaching in Higher Education. 25(4), pp. 351-365.

7 thoughts on “Week 6 Drawing

  1. Hi Enoch,
    It’s certainly interesting to think how teachers would react if they had access to such data, and saw that some students were not highlighting at all – would they accept this or be concerned that perhaps they were not engaging wiht the text? This is a problem with having data (as mentioned in this paper), it draws your attention to specifics (e.g. highlighting) that perhaps you didn’t worry about before, but once you have the data, you can’t easily forget about it.

    • Agree – in fact when I was an undergrad (~10 years ago) I had routinely abused highlighters, and my handouts/textbooks then were very heavily (but not purposefully) highlighted. It felt like a “ticking-the-box” behaviour rather than an indicator of my intellectual engagement with the text.

      It seems to indicate there is a “sweet spot” for highlighting behaviour (i.e. neither highlighting too generously nor too sparingly) being able to yield some usable data.

      like you said, “once you have the data, you can’t easily forget about it.” – As a student, I was whelmed seeing the entire page being coloured bright yellow! and it certainly nudged me to modify my highlighting behaviour.

  2. I like the term pedagogic reductionism, it just dawned on me that the quantification of people via dataveillance is not just about observing people, but specifically observing odd disconnected bits of people.

    • I definitely felt that way when dealing with a large cohort – practically can’t really track an individual person and connect the dots without any automation.

  3. ‘At a functional level, records like this provide a teacher some insights into what their students might think is important message in a prescribed reading.’

    Great idea for a visualisation here Enoch, and some good links made to potential use in teaching. I do wonder how teachers might use this data. At a basic level, just seeing that the text is being read, and that particular points are ‘of interest’ to students seems valuable. But without the context of why a particular section has been highlighted, I wonder what kind of intervention or response a teacher might make. Does ‘frequent highlighting’ indicate lack of understanding, or some kind of affinity with the section or phrase?

    ‘it also defines such behaviour as the model way a student should interact with a reading material.’

    Excellent point here, in that students may end up behaving in ways that can be ‘seen’, even if this isn’t their normal way of studying. One would then wonder how ‘authentic’ the data would be, and as you suggest, might invite ‘gaming’ the system.

    • How would a teacher use this data?

      Yes, I agree without the context why a segment is highlighted, it is difficult to have any meaningful deduction about a students’ engagement with the text. I can imagine this highlighting data being overlaid with other data like “notetaking”. So imagine students being able to jot down notes that are associated with a segment of text (somewhat similar functionality to NVivo, but a bit simpler user interface). Forum posts that talk about course readings can also be mined for similar purpose.

      Such data could be processed to highlight parts of a reading in which many students find challenging, and teachers can address it by prompting discussion and/or giving clarification.

      Then again, this subjects the online format to closer monitoring and reinforce e-reading as the endorsed way of learning.

  4. I like the visualization and the discussion here. It made me think of my habits of highlighting text. I’d say that the more I am engaged in reading, the fewer lines I mark with colour. I’m sure it doesn’t work like this with everyone. So as you wisely mention, whose behavior will make a benchmark here?

    We could also sophisticate it a bit more and play with colour. Does the colour used for underlining ideas matter? In automated systems, could red be used for the paragraph you didn’t get, green for something you want to remember, etc.? Some educators might find these data useful for planning their interventions. But still, like you said, datafication opens up more opportunities for students to manipulate the system and therefore the teacher.

Leave a Reply

Your email address will not be published. Required fields are marked *