Grading, GPAs, and College Admissions:
The SASS (Standardized Academic Scoring System) Plot

 There are no recognized national or (to my knowledge) state grading standards.  Connecticut definitely has no state grading standards.  Indeed, it does not even collect grading data directly:  it simply receives the self-reported GPA data from the College Board.  To the best of my knowledge, no other state has meaningful grading standards either.

Grades are well known to vary with time (“inflation”) and by course type (math and science low, music and art high); published College Board data show that they vary by state.  It seems to be an open secret in the field that grades also vary by school, but there is relatively little research on the matter, and specific comparative data by school is not easily available.  Similarly, it is common knowledge that some teachers grade significantly higher or lower than their school averages, but again, there is typically little to no information easily available.

Despite this variability, American high schools solemnly issue millions of transcripts each year, usually with school seals or other indications of official status.  GPAs are calculated using a variety of systems, and then typically reported to three or four significant figures.  The accompanying information (often called the “school profile”) may or may not contain information on grade distributions at the school in one form or another.  Even with the most detailed profiles, it is at best a guess to compare grades from one school to another.

But despite that uncertainty, every college admissions officer I know carefully examines the transcript and the GPA, and tries to appropriately include the information in the admissions decision.  Members of scholarship committees with merit criteria do likewise.  In many cases, however, no adjustment whatever is made for variability high school to high school.  Even when some adjustment is made, it is virtually always a subjective estimate rather than an analytical process.

In short, grades and GPAs are the most widely reported undefined values in the nation.  In my sometimes humble opinion, that situation cries out for an understandable, mathematically sound method for comparing academic performance from one school to another, or better, from course to course, both within the school and between schools.  The “GPA Plot” allows school to national average comparisons.  The “SASS Plot” allows comparisons right down to individual classes, and scaled against students at the specific college of interest.

The grading cohort considered in SASS is students with the same school, course, track, teacher, and school year, e.g., all Ms. Smith’s Honors Physics students at Anywhere High School for 06-07.  SASS reports the mean SAT I, Math + Reading (or other standard measure like ACT), of those students as a measure of the competition in that cohort.  SASS then reports the individual student’s performance relative to that cohort mean as a calculated score from 400 to 1600.

The two values – the cohort mean and the student’s relative performance – are then plotted as paired horizontal bars on the SASS Plot:  one pair for each graded course taken by the student.  A vertical stripe on the SASS Plot compares these values to published averages at the college of interest, thus providing immediate visualization of how the student is likely to perform at that college.