A randomized controlled trial on the "COVID generation": Were students’assessments biased during the health crisis?

Docimology has documented the biases that impact the assessment of students. Student's ethnicity (Sprietsma, 2013), social background (Rangvid, 2015; Autin et al., 2019), and gender (Lafontaine & Monseur, 2009) may influence their assessment. The SARS-CoV-2 pandemic had significant effects on educational systems: school closures and lockdowns appear to have negatively impacted the academic performance of students (Maldonado & DeWitte, 2020). The result is a generation of students that is qualified as "sacrificed" (Van Nieuwenhuyse, 2021). This experimental research tests the hypothesis that copies of this “COVID generation” are judged more negatively by evaluators. An essay and a mathematical task produced by a grade 6 student were assessed (with a standardized correction grid) by a sample of 297 college students. Based on a random assignment, half of the students received the copy attributed to a student whose class did not experience closure due to COVID, while the other half of the sample received the same copy assigned to a student whose class experienced periods of closure due to this crisis. The magnitude of the differences between the assessments of the copy in each of the experimental conditions (essay: -0.25; math: -0.15) leads to confirmation of our hypothesis. While these effects remain small, their potential reproduction throughout the schooling of this “COVID generation” could have detrimental effects (Autin et al., 2019).