Skip to main content

Table 4 Comparison of selected population size and IRT ability between D&D and MC groups

From: Researching for better instructional methods using AB experiments in MOOCs: results and challenges

  Sequence 1 Sequence 2 Sequence 3 Sequence 4 Sequence 5 Average
% pop. (D&D) 22 % 26 % 26 % 36 % 30 % 28 %
% pop. (MC) 25 % 35 % 18 % 24 % 41 % 29 %
p (Fisher) 0.87 0.18 0.23 0.07 0.1 0.8
Mean IRT (D&D) −0.65 −0.33 −0.54 −0.43 −0.60 −0.45
Mean IRT (MC) −0.69 −0.66 −0.64 −0.65 −0.18 −0.54
p (t test)a 0.89 0.13 0.82 0.55 0.1 0.43
df 37 53 26 43 46 269
t 0.13 1.54 0.23 0.76 −1.44 0.78
  1. aSince the distribution is also non-normal, Mann-Whitney U test was also performed, with very similar results