Decoding Grades: An In-Depth Look at Grade Distribution in College Courses

College course grades serve as cornerstones for evaluating student performance, gauging course difficulty, assessing student preparedness, and measuring instructor effectiveness. They also act as critical selection criteria for graduate school admissions. However, concerns often arise among education researchers and practitioners regarding the validity of grades, the impact of grade inflation, and inconsistent grading policies across instructors, academic departments, and universities. This article delves into the complexities surrounding grade distribution in college courses, drawing upon a comprehensive dataset of 33 public universities to analyze grading trends and issues.

The Significance of Grades in Higher Education

Grades have far-reaching implications in the academic landscape. They not only evaluate student performance but also influence decisions regarding scholarship eligibility and serve as a selection criterion for graduate programs and employment. University administrators may also leverage grades to evaluate curricula success and make decisions regarding tenure-track and adjunct instructors' employment.

Variability in Grading: A Cause for Concern

Several studies have revealed inconsistencies in undergraduate grades that cannot be solely attributed to student ability. A 2018 study highlighted significant differences between departmental average grades, even when controlling for student ability and course characteristics, suggesting a department-led policy on grade assignments. Furthermore, research indicates that grades in the sciences and economics tend to be lower than those in the humanities. A study of three public universities revealed that undergraduate K-12 education departments assign significantly higher grades, with one in five classes assigning A's to all students.

The disparity between STEM and non-STEM fields is further emphasized by findings that the average grade difference between STEM and non-STEM course grades and GPAs is approximately four-tenths of a grade point. Students of similar academic caliber, based on ACT scores and other course grades, tend to receive lower grades in STEM courses. Departments with lower student enrollments have also been observed to assign higher grades. Additionally, a correlation has been found between instructors' temporary or part-time employment status and high grades, without a corresponding improvement in learning outcomes.

The Impact of Grades on Students and Faculty

Despite the perceived inconsistencies, grades wield considerable influence in higher education. Students often interpret low grades as an "academic mismatch," leading to higher major withdrawal rates in departments with lower grades, such as Physics. Lower grades in STEM departments have been found to disproportionately affect female students, resulting in a higher major attrition rate compared to their male counterparts earning the same (low) grades.

Read also: Your Guide to Nursing Internships

Grades also play a crucial role in the professor-student relationship. Studies suggest that students often "punish" instructors by assigning lower ratings when they receive lower grades than expected, but rarely "reward" them following a higher grade.

A Comprehensive Dataset for Analyzing Grading Patterns

Prior research on analyzing grade patterns has been limited by the amount and time span of available data. To address this gap, a unified grading dataset was created, encompassing 33 public universities and providing a high-level analysis of the grade data. This dataset is the largest unified grading dataset that is publicly available. The data was collected from original sources published online by institutional research boards or through Freedom of Information Act (FOIA) requests. Each dataset was converted into a unified format to facilitate analysis across institutions.

Data Collection and Manipulation

Grade data was obtained from 33 U.S. public universities through extensive searches for FOIA requests and official publications from institutional research departments. Data obtained through FOIA requests were available via CSV and Excel files, while data in interactive tools from institutional webpages were obtained through direct download or web-scraping using Python scripts. Due to FERPA restrictions, grade data from course sections with fewer than 10 to 15 students (depending on institution policy) were not available. The source of the original data is included in the dataset.

Each dataset was converted into a unified tabular format, where each entry represents the letter grade distributions of a course or section during a particular semester. Institution-specific course names and codes were retained, excluding special characters. The unified dataset format contains 28 columns, including:

Year: The year the course was offered (e.g., 2009, 2022).
Semester: The semester the course was offered (e.g., Spring, Summer, Fall, Winter).
Dept. Name: The name of the department offering the course (e.g., Computer Science, Spanish).
Dept. Code: A 2-5 letter code for the department offering the course (e.g., CS, SPAN).
Course #: A 2-6 digit number representing the course (e.g., 228, 101).
Course Name: The name of the course (e.g., Discrete Structures II, Calculus I).
Section: A numerical or alphabetical index of a section.
CRN: A unique identifier for the course.
Instructor: The name or identification number of the instructor.
Enrollment: The number of individual students/grade instances.
Avg. GPA: The average GPA of a section.
Letter grade %: The percentage of students receiving each letter grade (A+, A, …, F, Pass, Fail, W).
Other: Uncommon fields or notes by researchers.

Depending on the level of detail provided by the institution, data may be provided for only a subset of the columns. If the original dataset explicitly provided section-level data, then "Section" will be filled in. If the dataset did not explicitly provide this information but included unique CRNs for each course section in a given semester, then artificial section numbers were generated.

All but one school had a letter grading scheme. Letter grades were mapped to their numerical equivalent, with A corresponding to 4.0, F corresponding to 0.0, and with each letter grade transition yielding a 0.33 drop in value (e.g., A- is 3.7). Michigan State University had a unique grading scale that included values 4.0, 3.5, 3.0, 2.5, 2.0, 1.5, 1.0, and 0.0. To maintain consistency within other data, these grades were mapped to A, B+, B, C+, C, C-, D, and F. Several institutions, such as Brooklyn College, assign + and - grades but only released the base grade (i.e. A+, A, and A- are mapped to A). Universities generally used ascending numerical order for Course #, where the first digit represents the course level (1=Freshman and 4=Senior). Exceptions are UC Berkeley, UCLA, and UCSB, where 0-99 are lower-level undergraduate, 100-199 are higher level undergraduate, and >200 are graduate courses. Twenty-six of the datasets (a subset of schools with data at the section level) published instructor names. Although these names are already publicly available online, they were anonymized (e.g., Instructor 1, Instructor 2, …) as the instructors likely did not explicitly agree to the release of this information.

Institutional Information

Each CSV file in the dataset is titled with the institution name, consistent with NCES (National Center for Educational Statistics) naming. The dataset includes information on the period covered, the level of data provided (course or section), whether instructor-level grade data is provided, and the average unweighted and weighted GPAs for the institution.

GPA and Enrollment Computation

Average GPA is used to compare grade levels at the institution, field, and departmental levels. When the average GPAs of various units (e.g., sections, courses) are combined, both an unweighted and weighted average GPA are computed. The average weighted GPA weighs each unit by the number of enrolled students, while the unweighted average GPA weighs each unit equally. Grades corresponding to W ("Withdrawn") and "Pass/Fail" were excluded. The GPA calculations use the following mapping: A+ (4.3), A (4.0), A- (3.67), B+ (3.33), B (3.0), B- (2.67), C+ (2.33), C (2.0), C- (1.67), D+ (1.33), D (1.0), D- (0.67), and F (0.0).

Hierarchy of Academic Disciplines

A classification hierarchy of academic fields is utilized for high-level analysis at the department or major level. The department name associated with each course was identified, and in 2.2% of the cases where the department name could not be located, the field was marked as "NA." The classification hierarchy maps departments into five categories and fourteen subcategories, based on the College Board’s informational guide “College Majors by Academic Area” and the NCES subject codes. The hierarchy includes:

Arts & Humanities: Art, Media, & Architecture; Humanities; Foreign Language
Social Sciences & Public Services: Social Sciences; Public & Social Services
Business: Business
STEM: Science & Math; Engineering; Health & Medicine
Miscellaneous: Trades & Personal Services; Military & ROTCMilitary Science, Air Force ROTC, Aerospace Studies; Interdisciplinary & Individualized Study; Uncategorized

Departments such as ROTC (Reserve Army Training Corps) and interdisciplinary studies (e.g., Agricultural Science and Economics) were placed in the "Miscellaneous" category. The "Uncategorized" subcategory of the "Miscellaneous" category was excluded from analyses. For each institution, the mappings between the department code, department name, and subcategory are stored and included in an excel file included in the dataset. Note that STEM is an acronym for Science, Technology, Engineering, and Math.

Analysis of Grading Data

The compiled dataset is analyzed to identify interesting patterns at the institutional, academic discipline, and department level. All analyses utilize the entire dataset.

Institutional-Level Grading Trends

The distribution of average GPAs across the 33 schools included in the dataset is examined. The unweighted average GPA of each institution and the weighted average GPA of 29 of the 33 schools are compared. The weighted GPA is not available for all institutions.

Tools for Exploring Grade Distributions

Beyond the comprehensive dataset, various tools are available to explore grade distributions at specific institutions. For example, Florida State University (FSU) offers a Business Intelligence dashboard that allows users to track and report the distribution of awarded grades. This dashboard provides the ability to view information sorted by campus, location, college offering the class, academic organization offering the class, course, class section, and primary instructor.

To access the FSU Grade Distribution dashboard:

Navigate to my.fsu.edu and select the "Faculty and Staff" tab.
Under "myFSU Links," click the "BI" icon.
Ensure the "Grade Distribution - Org Offering Class" tab is selected and provide values for the prompts.
Click "Apply."
The dashboard will return a report, which can be viewed by selecting an option from the "View Selector" drop-down box. Selecting the "Table" view results in a report containing values for all parameter values.

The "Grade Distribution - Select Org of Student" tab provides a view of grade distributions based on the student's academic organization.

tags: #grade #distribution #college #courses