Many-Facet Rasch Model with Minifac

In this example, we have a simple test with 4 short-answer questions, 6 students (ratees) and 4 lecturers (raters) who mark all of the answer scripts. Using Minifac, I can highlight the severity/leniency of the raters in giving marks to the students' answers. In the Wright Map below, under the column -Rater,  we can see that the raters (A, B and C) have similar levels of leniency (their measures are below 0 and close to one another). This graphical information is more accurately represented by the table that follows it: The measures range from -.20 to -.44.

To demonstrate severe raters, I have reduced the scores from rater C by 1 unit. In other words, for rater C, I deduct 1 point across all questions for all students (except when the score is 1 - I did not make it into a 0). So, what do we get? Now, the Wright Map shows that rater C is high up the ruler and far away from raters A and B. The exact values of the severity measures are in the table that follows the Wright Map.