Worked Example: Small Quizzes and a Large Project

Use one unchanged assessment record

Assume two quizzes scored 9/10 and 8/10, followed by a project scored 60/100. Their percentages are 90%, 80% and 60%. The project has ten times the possible points of either quiz.

Calculate a total-points result

Earned points total 77 and possible points total 120. The overall percentage is 77/120 × 100 = 64.1666…%. The project represents 100/120, or five sixths, of the denominator. This result is appropriate if the gradebook sums all those points directly.

Calculate an equal-assessment result

If each assessment has equal explicit weight, the average is (90 + 80 + 60)/3 = 76.6666…%. This is not a more accurate version of the total-points result; it is a different model. Giving each row weight one reproduces the equal-assessment average.

Calculate category weighting

If the two quizzes form a 20% category, their internal total is 17/20 = 85%. A project category worth 80% contributes the 60% project score. The final result is 85 × 0.20 + 60 × 0.80 = 65%. The category denominator must be resolved before applying course weights.

Audit a disagreement with a gradebook

Ask whether the gradebook uses points, equal assessments or categories. Then check missing grades, exclusions and extra credit. Enter the same record under each model to isolate the reason for disagreement. Adjusting source scores to force a preferred output destroys that audit trail.

A second gradebook in which the two methods disagree less

Use a new illustrative record: a lab earns 45/50, a quiz earns 9/10 and a report earns 16/20. The percentages are 90%, 90% and 80%. Total earned points are seventy, total possible points are eighty, and the total-points grade is 70/80 × 100 = 87.5%. An equally weighted percentage average instead gives (90 + 90 + 80)/3 = 86.6666…. Here the gap is less than one percentage point, but that small gap does not make the two policies interchangeable.

In the total-points model the lab supplies 50/80 = 62.5% of the denominator, the quiz supplies 12.5%, and the report supplies 25%. One additional earned lab point changes the overall percentage by 100/80 = 1.25 points. Under an equal-assessment model, one lab point changes the lab's percentage by two points and the overall average by only 2/3 of a point. A question about where a point matters must specify the grading model before assigning influence.

The three rows can be entered unchanged in the total-points calculator. To reproduce an equal-assessment result in weighted-row mode, add a third column with equal positive weights, such as one for each row. The supplied weights total three rather than 100, and this tool normalizes over that supplied total. It reports the mean over those rows, not a claim that only three percent of the course has been completed.

A third scenario: zero and not-yet-graded are different records

Suppose a course has recorded 18/20 and 45/50, while a 30-point assignment is not yet graded. Excluding that future assignment gives 63/70 = 90%. Recording it as a zero gives 63/100 = 63%. These are not two estimates of the same denominator: one describes graded work only and the other describes a full record with a zero assigned. A blank score should not become zero accidentally when a spreadsheet is imported or copied.

If the final assignment later earns 24/30, the completed total becomes 87/100 = 87%. In a model where each of the three assignments has equal weight, the result would instead be (90 + 90 + 80)/3 = 86.6666…. If the first two assignments form a 40% category and the last assignment forms a 60% category, their category percentages are 90% and 80%, giving 90 × 0.40 + 80 × 0.60 = 84%. Keep each result with its method label; reporting a bare 87 hides the assumptions.

For an excluded item, omit its row from a total-points calculation rather than entering a zero denominator. Possible points must remain positive for every supplied row. For a genuine zero score, enter zero earned points with its positive possible points. The distinction can be audited from the total possible-points output: seventy indicates exclusion here, while one hundred indicates the assignment is counted. A calculator cannot infer whether a blank grade means pending work, an excused item or an unsubmitted assignment.

Use score changes and coverage to audit the original record

Return to the original small-quizzes and large-project record. Raising the project from 60/100 to 70/100 adds ten earned points while possible points stay at 120. Total-points grading rises from 64.1666… to 72.5%, an increase of 8.3333… points. Equal-assessment grading rises from 76.6666… to 80%, an increase of 3.3333… points. With quizzes worth 20% and the project worth 80%, the course result rises from 65% to 73%, an eight-point increase. A single changed score therefore provides a distinctive fingerprint for the gradebook's weighting policy.

Partial category coverage introduces another denominator. If only categories with weights twenty and forty have been graded, the weighted calculator divides their contributions by sixty. Its reported average describes that completed sixty-percent slice. It is not the final course grade with the remaining forty percent treated as zero. Preserve the covered-weight output and do not normalize a normalized average a second time.

Keep policy changes outside the arithmetic record: late penalties, dropped lowest scores, extra credit and category caps each need their own treatment. The current points forms accept earned values up to possible points, so an extra-credit system beyond that range requires an explicit adjusted model. These illustrative ledgers establish reproducibility; they do not identify a school's policy from a number alone or suggest that the most favorable method should be selected.