Distribution And Absolute Grading System

University grading policy allows instructors to use a distribution system, an absolute measures system, or a combination of the two for calculating final course letter grades. Each grading system has its advantages and disadvantages. Consult with your Subject/Program Area Coordinator and consider the following information before determining a grading system for your course. The grading system used must appear on the course outline. Be aware that upon request, instructors are required to provide the method that was used to translate final and, where appropriate, term marks into grades. Additionally, be aware that instructors need to be able to show the link between the course objectives and the criteria used for assessment. 

Under the distribution (or relative) system, a student's final grade is based on their ranking or relative position in the class upon completion of the course. The General Faculties Council (GFC) has specified that distributions are not mandated by the university. Generally speaking the distribution system is inappropriate unless there is a large class where one would expect a normal distribution to occur.

Advantages:  

• Percentage distribution works well for class sizes greater than 40 and for ‘typical’ classes. 

Disadvantages: 

• Because grades are determined by rank order of students, it may not provide a clear indication of achievement relative to learning objectives. If an instructor cannot show the connection between achievement and learning objectives, they are in contravention of GFC policy.

• Strict adherence to the distribution system is less reliable for smaller classes and atypical classes. For example, a student with a raw score of 90 in a low-achieving section may receive an A, while a student with the same score in a higher-achieving section may receive a B. It also doesn’t account for students who, as a group, have high or low entrance GPAs. 

• Since students are in competition for grades, students may be less willing to collaborate. 

• Students frequently express they find the distribution more stressful because they do not know their standing until the course is over. An additional stressor is the perception that students with high marks will have their grades artificially lowered in order to match a perceived distribution.

• There may be a temptation from instructors to use the distribution system to artificially adjust the grades, particularly at the higher and lower ranks of grades and; thus, try to compensate for a failure to clearly differentiate student achievement.

Procedural Considerations: 

• Once final raw scores are determined, place scores from highest to lowest.

• If the raw scores appear to follow a normal distribution, use the chart on the following page to determine approximately how many students should receive each letter grade. For example, in a typical 300 level class with about 35 students, approximately two students should receive A+, three A, five A-, … one D, zero D+, and zero F. Note: The number of students in a grade band is rounded to the nearest whole number. Moderate variation of the number of students assigned to each grade level is expected. 

• Use natural breaks between raw scores on the Frequency Distribution Sheet for grade divisions when possible. 

• Use the grade descriptors and absolute scale expectations to assist with grade division locations. 

According to University’s Grading Policy, all course outlines must describe the process by which the term marks will be translated into a final letter grade for the course. The process must be consistent with the University of Alberta Assessment Policy and accompanying Grading Procedure, found at the University of Alberta Policies and Principles Online (UAPPOL) website. The following is a suggested statement for explaining how final grades will be determined. In the table below, the first three columns are official university policy. The “Additional Descriptive Information” is merely a suggestion. The information in this column may be deleted or changed, and should be tailored to your own course and then further developed in relation to specific assignments.

According to the University’s Grading Policy, “Grades reflect judgements of student achievement made by instructors and must correspond to the associated descriptor. These judgements are based on a combination of absolute achievement and relative performance in a class.” In this course, your work will be evaluated using the general grading descriptors established by the University, as well as the more detailed assessment criteria that will be provided for specific assignments. Your grade on each assignment will be in one of three formats: a descriptor (excellent, good, satisfactory, etc.), a letter (A, B, C, etc.), or a number. When necessary, descriptor and letter grades will be converted into numerical equivalents in order to weight them properly and average them into a final course grade. Conversions between descriptors, letters, and numbers will be made in accordance with the University’s grading policy. Your final course grade will be reported as a letter grade. It will not be official until it has been approved by the Undergraduate Coordinator and posted on BearTracks.

According to the grading policy, when an instructor gives a grade of A (A+, A, A-), it should mean the following: “In my professional judgement, the quality of this student’s work is excellent (as opposed to merely good).”  When an instructor gives a grade of B (B+, B, B-), it should mean “In my professional judgement, the quality of this student’s work is good (as opposed to merely satisfactory).”

Excellence is not something that is “average” or “typical”. Therefore, measures of central tendency would not be expected to correspond to excellence. In practical terms for instructors and grade approvers, a median of A+, A, or A- would indicate an unusual or anomalous grading pattern. Similarly, a mean higher than 3.3 would indicate an unusual or anomalous grading pattern.

It is important to put a lot of thought into what defines work as excellent, good, satisfactory, etc. and to communicate the definitions to students as clearly and as transparently as possible (preferably in relation to specific assignments).

It is also important to ensure that grading scales and rubrics do not lock you into giving grades that do not reflect your professional judgment. Grading scales and assessment tools should be used to assist the decision-making process rather than substituting for it. In all cases, grades must reflect the level of student achievement, and they must match the descriptor such that work deemed to be “excellent” will receive a grade of A (A+, A, A-), work deemed to be “good” will receive a grade of B (B+, B, B-), work deemed to be “satisfactory” will receive a grade of C (C+, C, C-), etc.

Assessment tools (rubrics, grading scales, etc.) should not be allowed to override your professional judgment. They should be used as tools that assist you in formulating judgments rather than becoming the “decision maker”. It usually takes a lot of experience, calibration, and perfecting over time for assessment tools to produce accurate determinations of the quality of a student’s work. If reporting grades as numbers and applying a grading scale is forcing you to give grades that don’t accurately reflect your professional judgment, then it might be preferable to report grades as descriptors (“excellent work”, “good work”, etc.).

Given that letter grades are assigned at the end of a course after all requisite assignments and assessments have been considered and weighted, instructors are asked to not promise final grades using “bands of percentages” (e.g., 90-100%) to letter grades (e.g., A+) in their course outlines/syllabi. This practice should be avoided as it (a) sets up expectations for student final letter grades without giving the instructor an opportunity to fully consider all requisite assignments and assessments and (b) is vague and possibly misleading as to whether it reflects norm- and/or criterion-referenced assessment for students. 

Working with a fixed grading scale requires instructors to monitor after they grade each assignment to ensure that grades match up with their professional judgment.  When there is not a close match, measures should be taken to bring the grades into alignment with the professional judgment *before* grades are reported to students.  Monitoring of total grades as they develop throughout the term is also a good idea.

Grades in any course, examination or other academic assessment shall not be mandated on the basis of a curve or historic distribution of student grades. The distribution of grades shall not be predetermined by any system of quotas that requires a certain number or percentage of grades at a particular level.