GradeMe Analytics

This is what GradeMe data looks like

Project card

On the project card we need to show

• the number of students,

• the number of teams,

• Progress bar: number of evaluations completed so far and the total number of evaluations

Other useful stuff:

• Fraction of teams who gave equal scores to each other.

• A histogram of contribution:

Special analytics

The main goal of special analytics is identifying pairs of students who gave each other a too high or a too low score.

Here is one team:

Below are log-ratios of individual evaluations to true evaluations

##                                      from_1     from_2      from_3       from_4
## CHIA KAI JIE, BENJAMIN        -0.0038022859  0.1297291 -0.07279516 -0.052592450
## CHUA YUNG WEI VINCENT          0.0630581360 -0.0547249  0.14821594 -0.054724900
## NEO WEI QIAN RACHEL            0.0058997221 -0.0547249  0.14821594 -0.054724900
## NICHOLAS CHIN WEI LUN         -0.0007404665  0.1691586 -0.21466736  0.035627178
## PRIYA SHANTHINI D/O ANBAZAKAN -0.0552853157 -0.2271356 -0.02419473  0.091318159
## TAN YAN NING                  -0.0011104943 -0.1188935  0.08404731 -0.001110494
##                                    from_5      from_6
## CHIA KAI JIE, BENJAMIN         0.04271773 -0.07279516
## CHUA YUNG WEI VINCENT         -0.05472490  0.14821594
## NEO WEI QIAN RACHEL           -0.05472490  0.14821594
## NICHOLAS CHIN WEI LUN         -0.11852350 -0.21466736
## PRIYA SHANTHINI D/O ANBAZAKAN  0.09131816 -0.02419473
## TAN YAN NING                   0.10425002  0.08404731

Here, 0 represents accurate evaluations, positive numbers are generous evaluations, negative numbers are evaluations that are too strict. If we show them as a heat map, it’ll give us some info of what is going on

Another way to visually show the same information is some kind of a directed graph. Nodes of the graph represent students and there is an arrow from student A to student B whenever evaluation of B by A is off by more than 10%. Blue arrows are too generous and red arrows are too strict evaluations. Thickness of an arrow is proportional to the deviation of evaluation of the true evaluation.

In this plot, thick red arrows between two nodes represent two mutual evaluations that are too strict (it may indicate a conflict between two students) and thick blue arrows between two students represent mutual evaluations that are too generous.

A perfect situation is the absence of arrows.