CMPUT 404

Grading System: Percentages and Letters

We do not use pre-defined percentages of your final mark to compute your final grade, but instead use our judgement of how your final percentage indicates the level of mastery in the table below for an A, B, C, D+. This is done after the final exam.

This is the standard system in use by most Computing Science courses at the University of Alberta. However, it may not be what you are used to from other institutions. The department of Computing Science believes this produces a fair evaluation and our extensive experience supports this.

You can find a copy of this table on the Department Course Policies page.

• The table listed above is used to translate the final percentage in the course to a letter grade by using grade breakpoints after the final exam.
  • This is done by using breakpoints: each letter grade is given a minimum percentage required to get that letter grade.
  • The minimum overall percentage demonstrates "Consistently original thinking that extends the material, demonstrated depth and breadth in the material, ability to integrate material with other subjects, ability to analyze and synthesize material at various levels of abstraction." will be the minimum percentage "breakpoint" for an A-.
  • The minimum overall percentage demonstrates "Like an A, but not consistent over time, or weak in a specific area." will be the minimum percentage "breakpoint" for a B-.
  • The minimum overall percentage demonstrates "Understand some of the core material but not its subtleties, can apply it to simple situations but often needs assistance, evidence that the material has had some change on the way of thinking." will be the minimum percentage "breakpoint" for a C-.
  • The minimum overall percentage demonstrates "Poor" will be the minimum percentage "breakpoint" for a D.
  • Anything over 50% is usually at least a D
    • Sometimes 50% might be better than a D, such as a C.
    • This is the only actual percentage used. All other percentages in this document are only examples and do not apply to this semester.
• Depending on the definition of curved, grades are not curved.
  • Some people define curved to mean that grades are adjusted based on some statistics (such as the average) performance of all students in the class. In this sense, grades are not curved.
  • Some people define curved to mean that grades are adjusted based on distributions, e.g. a fixed number or percent of students get an A. In this sense, grades are not curved.
  • Some people define curved to mean that grades may change depending on what happens during the semester, and in this sense graves are "curved" but this is not the University, Faculty, or Department's definition of curved.
  • University Policy: No fixed grade distribution can be used.
    • Class rank and averages are not taken into account when deciding breakpoints.
      • No system which produces a fixed number or fixed percentage of A's of A's is used.
  • No system where the breakpoints are moved based on class averages is used.
  • By this system it is entirely possible for every student to get an "A" in the course if every student demonstrates "A" level mastery of the course material.
  • It is also possible for every student to fail.
• The percentages required to attain a particular letter grade are not decided until after the final exam.
  • This is to ensure that any surprises or differences between semesters can be taken into account.
  • For example, if a new assignment or exam proves to be more difficult than previous semesters, the percentages required for each letter grade (breakpoints) will lower.
  • If a new assignment or exam proves to be less difficult than previous semesters, the percentages required for each letter grade will rise.
• The same breakpoints (final percentage required to get an A, B, C, etc.) apply to everyone.
  • If one student demonstrates performance as listed above then all students with the same or higher final percentage will receive the same letter grade or higher.
  • Example: if one student demonstrates "Consistently original thinking that extends the material, demonstrated depth and breadth in the material, ability to integrate material with other subjects, ability to analyze and synthesize material at various levels of abstraction." and has 81% in the course, than any student with 81% or higher will earn at least an "A-". This is just an example, an A is not 81% this semester.
  • Example: If student X gets an overall 82% and deservers an A, and student Y gets an overall 83%, then student Y will also get an A, regardless of what the instructor thinks of them. This is just an example, an A is not 82% this semester.
• Every semester contains different material since every semester we try to improve the course, and we also need to constantly create new questions and assignments to mitigate cheating.
  • From one semester to the next the breakpoints could be higher or lower.
  • From one exam to the next what percentage demonstrates each of the qualities in the table could be higher or lower.
  • Adjustments to breakpoints do not depend on class averages.
    • The class average can be a D.
    • The class average can be an A.
• Instructors can not decide to make the breakpoints just whatever they want. Breakpoints are validated by at least one of:
  • Associate Chair (Undergraduate Program)
  • Associate Chain (Graduate Program)
  • Department Chair
• Almost nothing in the course outline (syllabus) can be changed after the first day of class. Especially not component weights, like the weights of an exam.
• Everyone in the class can get an A at the same time in the same semester.
  • However, for the whole class to get an A, everyone in the class must demonstrate "Consistently original thinking that extends the material, demonstrated depth and breadth in the material, ability to integrate material with other subjects, ability to analyze and synthesize material at various levels of abstraction."
  • This is highly unlikely in large classes, as a basic fact of probability and statistics.
  • However, University Policy ensures it must remain possible, even if it is unlikely.
• Courses/semesters with open-book exams will usually have lower breakpoints than courses/semesters with closed-book exams.
  • This is because open-book exam questions are usually more difficult, and avoid asking for very much that can be simply looked up.
  • Courses/semesters with open-internet exams will usually have even lower breakpoints, because the exam questions are usually even more difficult.

Breakpoint Calculations and Examples

The numbers in this section examples only and will not be used for this course!

• Example hypothetical scenario:
  • Semester 1: The breakpoint for a C- is determined to be 70%.
  • Semester 2: The semester is exactly the same as Semester 1, except that one exam, worth 13% of the overall grade, has more questions that require "original thinking that extends the material" and fewer questions that require "understanding parts of the material"
  • If the exam were judged by itself, the breakpoint for a C- would drop from 80% in semester 1 to 50% in semester 2, a difference of 30%. But the exam is only worth 13% of the overall grade, so the overall breakpoint would probably be about 4% lower in semester 2 than semester 1. (13% x 30% = 3.9%) because in this hypothetical scenario, none of the other exams or assignments changed.
  • Semester 2: The breakpoint for a C- is determined to be a 66%.
• Breakpoints are not actually decided individually for each assignment, exam, lab, etc. Instead, breakpoints are only decided at the end of the semester of the overall grade. However, the result is the same, since breakpoints can be averaged according to the weights given for each assignment/lab/exam/etc.
• Example hypothetical semester:

Assessment Weight Minimum % required for a C-

Exam 1 20% 40% Exam 2 20% 50% Exam 3 30% 60% Assignment 1 10% 90% Assignment 2 10% 80% Labs 10% 70%

• Calculation for a C- for the above example:
  • 20% * 40% + 20% * 50% + 30% * 60% + 10% * 90% + 10% * 80% + 10% * 70%
  • = 60% overall required for a C-
  • The same could be done for all the letters.
• In practice, it is too time-consuming to determine breakpoints for each component, so the breakpoints are only decided at the end of the semester, since the calculations are mathematically equivalent anyway.
• Consider an exam, "exam 1" on which every question requires: "Original thinking," or "ability to integrate the material with other subjects," or "ability to analyze and synthesize material at various levels of abstraction." Your score on such an exam should be interpreted as the % of the exam that you got an A on. Or the % of the exam that you did better than an F, D, C or B!
• Alternately, consider an exam, "exam 2" where every question only requires "understanding of parts of the material": your % on that exam is the % of the exam that you didn't get an F on. This should very well be a much higher % than "exam 1."