Skip to main content

Table 3 Difference in average number \( \left(\boldsymbol{\Delta} \overline{n}\right) \) of correct answers on subsequent problems per student between D&D and MC format

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

  Sequence 1 Sequence 2 Sequence 3 Sequence 4 Sequence 5
n 3 5 5 5 5
\( \overline{n} \)(MC, 1st attempt) 0.36 0.94 1.94 2.09 1.08
\( \Delta {\overline{n}}_{\mathrm{D}\&\mathrm{D}\hbox{-} \mathrm{M}\mathrm{C},\ 1\ \mathrm{attempt}} \) 0.73a 0.75b 0.64 0.79 0.02
p (U test) 0.007 0.02 0.25 0.07 0.89
\( \overline{n} \)(MC, 2 attempts) 1.68 2.27 3.54 3.74 2.77
\( \Delta {\overline{n}}_{\mathrm{D}\&\mathrm{D}\hbox{-} \mathrm{M}\mathrm{C},\ 2\ \mathrm{attempts}} \) 1.12a −0.15 1.19b 1.1c −0.54
p (U test) <0.001 0.9 0.03 0.02 0.15
  1. n number of problems in each sequence (excluding the 1st one or two problems)
  2. aSignificant at 0.01 level
  3. bSignificant at the 0.05 level
  4. cSignificant at the 0.05 level, but population might be unequal