Abstract:

The commitment of large institutions at worldwide level to carry out transitions from paper- to computer-based large-scale assessments of school populations, raises critical questions regarding how much prepared are current students to deal with digital tests and to what extent there is a differential effect by gender. While most of studies about these mode effects are based on data from standardized tests, little attention has been paid to alternative analysis in real school’s settings. In this paper, we use end-of-course school grades awarded in science disciplines to two cohorts of secondary students which were examined either with computers or paper-and-pencil tests. After converting these grades into percentile scores, we studied the effects of the administration mode in student’s ranks, examining through different methods their statistical equivalence from a gender perspective. It was found that, at overall level, females earned higher grades than males in science and these differences were statistically significant. Interestingly, we also found that females and males performed similarly in both testing modes suggesting that the gender gap in science achievement would not be affected by the assessment mode.

Keywords:

CBA, PBA, Testing mode, eAssessments, STEM, Secondary education, Learning Analytics.