>>13051762I mean I'm not sure I even understand your question or why it is important to you.
When I first started marking I still felt that it was my duty to pay a lot of attention to detail and provide the fairest mark possible to all students. However, I quickly noticed repeating answers in scripts where teamwork wasn't allowed. Because I just had two large piles (marked and unmarked) I would need to browse the marked pile and open every exam book in a tedious linear fashion until I found the book that had the same answer. I would then record the student numbers in my feedback note to the Professor. My mind would automatically start remembering the surnames as this was easier to spot in the marked pile than remembering the place number or student names. As this process repeats itself, and I note the similarities, these were usually clustered around friend groups of the same ethnicity. Eventually I learned who these groups are in TA sessions. I was also able to use the names to guess gender and ethnicity with remarkable accuracy. Aside from ethnic friend groups, I also figured out Stacey-orbiter relationships and so forth. You can learn quite a lot about the social dynamics of an undergrad year by observing trends in exam answer similarity, it is fascinating observation and could make for an interesting data analytics paper in itself. But I digress. They self-segregate, I have just adapted to their segregation. So that is how "Chen" and "Li" always ended up in the same pile, likewise for "Mohammed" and "Mohammed", and "Brad" and "Jenine".
Now when I am not familiar with a student year group yet I have a system for my first marking. Among my many other tricks and techniques that I've adapted over the years, I organize piles according to ethnicity. This is a mere starting point until I learn the friend circles in more detail and then I start with piles of those circles. It cuts my marking time by more than health without sacrificing marking quality.
