Created by Justin Esarey | based on research by Justin Esarey and N. Valdes
Instructions: Input the characteristics
of the SET, the person and class being evaluated, and the
way that SETs are used administratively to identify
exceptionally strong and weak teaching. A percentile is
the proportion of the population equal to or less than the case
being evaluated; for example, a class at the 50th percentile of
teaching quality is taught as well or better than 50 percent of
all classes at the college/university. Correlation is the
degree of positive linear relationship between two continuous
variables; click this link
for more details. Biases against women or required courses
are subtracted from the value of Teaching Quality used to
calculate SET scores.
The SET Simulator
will then use computational simulation to determine the
likelihood of a class with the specified true teaching quality
being evaluated as exceptionally well or poorly taught. It will
also show the distribution of SET scores given the specified
level of teaching quality, and a sample display of the relationship
between SET scores against teaching quality (with each plotted dot
representing a class or section).
Note: not all combinations of correlation choices are possible. The program will auto-adjust correlation choices if necessary.