Grading, GPAs, and College Admissions:
It is an open secret in academia that course grading in US secondary schools and colleges is highly idiosyncratic. Published reports by The College Board, by ACT R&D, and by others, prove beyond reasonable debate that course grades vary by year ("grade inflation"), by subject (math and science low, art and music high), and by state (Massachusetts and Connecticut low, Florida and Texas high). Unpublished, but similarly reliable data shows extraordinarily different grading patterns school-to-school, usually with little or no awareness even by otherwise well-informed administrators and boards of education. Teacher-to-teacher variations are well known by administrators, but seldom publicly discussed and virtually never openly documented.
This work began ten years ago as a method to put high school grades on a common basis, but it now also provides a method to directly predict a student's college class rank at almost any college using web-published data on the college and student secondary school data readily generated using a database program like "PowerSchool". This provides precisely the information which admissions officers and scholarship committees need, and which must now be estimated using inconsistent secondary school data, personal experience, qualitative recommendations, unpublished data, and skills amusingly called "gutology" by some, but time-consuming and imprecise for all. The method also avoids the bitterly contentious issue of differing grading "standards".
The method uses a parameter I call "CAPE" -- Course-based Academic Performance Estimation. For a specific course grade,
Course grade CAPE = [(Mean SAT M + Mean SAT CR) of students receiving that grade] x (1/10)
for that year, school, course, track, and teacher. SAT M is the Mathematics SAT. SAT CR is the Critical Reading SAT, called the Verbal SAT in earlier years. It is essential to recognize that these values are for the grade peer group, rather than the scores of the individual student. The individual student may never take the SAT test at all.
Weighted by course credits, an overall GPA-like CAPE follows directly. CAPE values excluding courses like health, music, and art are straightforward. The (1/10) multiplier serves principally to generate a unique range (40 to 160), thus reducing the chance of confusion with other data. Appropriate procedures for the calculation minimize the effect of data anomalies and allow calculation of CAPE values prior to graduation by using comparable recent values at that school. Arbitrary track premiums are neither necessary nor used. Composite ACT scores can be used as-is or converted to SAT values using a well-known chart.
The student's overall secondary school CAPE is a good predictor of the student's first year college CAPE, allowing direct prediction of the student's class rank at any college with published SAT or ACT data (over 4000 colleges on the College Board site alone). Colleges without published data may provide it on request or suggest a comparable college with published data. In a few well-known cases (Juilliard, Gallaudet, etc.), the CAPE value has limited predictive value. CAPE values at typical US colleges are nearly linear with first year class rank, with about a 40 point CAPE range, top to bottom class rank. The calculation uses a linear relationship unless a more accurate correlation is available. Web-published SAT values for 25th and 75th percentile enrolling students are then used to set the CAPE curve for a specific college.
It is important to note that the applicant's CAPE value and the associated prediction are offered data, not demanded conclusions. The college retains full authority to use, modify, or ignore this and any other information. It can (and should) account for planned special academic accommodations like mentoring, tutoring, reduced load, or specific course selection.
The validity of CAPE as a performance predictor for students from a specific secondary school must be confirmed and documented year by year. In most cases, this will most likely be done using first year college class rank values from an e-mail survey of graduates from the prior year, ideally confirmed with their respective colleges using written permission forms requested during secondary school graduation processing. A parallel source is visiting college representatives, who should be asked in advance for the information and provided copies of the signed permission forms during the visit. It is routine for them to obtain such data in preparation for a visit, so no special effort should be required.
Because the CAPE method generates numerical values on a common base, it provides plots which can include all students and all colleges, or any desired subset of either, while preserving the confidentiality of individual students. Use of trendlines makes such plots both easier to understand and highly tolerant of incomplete college class rank data.
CAPE use benefits thus include:
Equitable evaluation of an applicant's grades for college admissions and scholarships, even if from multiple secondary schools, multiple tracks, or secondary schools unfamiliar to the college. This can be critical for success for applicants from secondary schools which grade lower than the national average.
Immediate identification of applicants who would need special academic accommodations at that college, and should other factors (sports skills, racial or geographic balance, etc.) merit acceptance regardless, the opportunity to begin those accommodations immediately.
Greatly improved counseling on selecting colleges for applications.
Unprecedented ability to analyze and graphically present college placement patterns at a secondary school. Similar capabilities to analyze and present performance of its graduates in college. To my knowledge, only the CAPE method allows such analysis on a mathematically sound basis. Encouraging secondary schools to track their graduates' college performance is a major benefit.
An illustrative simple case
CAPE calculations for a secondary school necessarily reflect the specific grading, course weighting, data collection, and other specifics at that school. For illustration, however, consider the following simple case:
The school has only one track, or no premium in GPA calculations for higher track courses.
The student database program (PowerSchool or a competitor) can access the most recent SAT scores for each student, and can identify each student in a given year/course/teacher cohort, together with each student's course grade. Most students in the school take the SAT. The cohort contains at least 30 students with SAT scores.
Calculate the slope of the cohort's (course grade) vs. (SAT M + SAT CR) plot, say 100 SAT points per 1.0 grade point.
Calculate the mean grade in the cohort, say 3.3. Calculate the (mean SAT M + mean SAT CR) in the cohort, say 1050.
A student receiving a 3.3 (B+) in the course has a course CAPE of (1/10) the SAT sum, or 105. A student earning a 4.0 (A) has that CAPE adjusted by [(4.0 - 3.3)/1.0] times the (SAT/CAPE) slope calculated above (100) times (1/10), which is a course CAPE of 112.
105 + [(4.0 - 3.3)/1.0] * 100 * 0.1 = 112
Students' overall CAPE values are found by weighting the course CAPE values by the course credits.