Developing Data Products Course Project

Setting a letter grading scheme is one of the most challenging aspects of designing a course.

Poor grading schemes can have significant consequences:

• Grade inflation
• Student confusion
• Adjustments to avoid undeserved failing grades

R can make determining letter grades simple. In the grading scheme below, grade ranges are set as [90,100] for A, [80,90) for B, and so on. The assignment categories are weighted: homework 30%, midterm exam 1 20%, midterm exam 2 20%, final exam 30%. A test student shows the ease of calculating final grades.

st <- data.frame(hw=98, mid1=84, mid2=88, final=92)
score <- .3*st$hw + .2*st$mid1 + .2*st$mid2 + .3*st$final
labs <- c("F", "D", "C", "B", "A")
bks <- c(-Inf, 60, 70, 80, 90, Inf)
cut(score, breaks=bks, labels=labs, right=FALSE, include.lowest=TRUE)

[1] A
Levels: F D C B A

The Grade Weighter app makes it simple for faculty to test out different grading schemes. Using simulated assignment scores, faculty can customize:

• Letter grade lower bounds
• Assignment weights

The Grade Weighter app has an easy to read visual output of the distribution of letter grades from the simulated scores and your custom grading scheme!