Final Grades: graduate and undergraduate grades are assigned by separate curves, as announced in the syllabus; this is why some rows with lower overall point totals get higher letter grades.

The columns of the table are: (a) exam 1, (b) exam 2, (c) exam3, (d) quiz 1, (e) quiz 2, (f) quiz 3, (g) homework 1, (h) homework 2, (i) optional homework 1, (j) optional homework 2, (k) optional homework 3, (l) optional homework 4, (m) normalized exam total, (n) normalized quiz total, (o) normalized homework total, (p) normalized project total, (q) weighted maximum of homework and project totals, (r) normalized weighted sum of exams, quiz, and homework/project, (s) the letter grade.

Cumulative, Normalized Scores a b c d e f g h i j k l m n o p q r s --------------------------------------------------------------- 74 73 63 7 _3 _3 10 15 0 0 0 0 71 45 61 100 100 89 A+ 57 68 44 10 6 3 10 12 11 15 45 15 55 76 93 0 66 63 A 57 83 48 _3 3 _3 10 11 0 0 0 0 69 38 59 10 27 36 C 54 61 25 9 6 5 10 15 0 0 0 0 45 78 61 73 78 68 A 70 43 62 7 7 0 10 4 0 0 0 0 57 67 57 62 69 65 A 57 69 96 10 10 0 10 15 6 15 30 15 80 78 87 0 56 65 B 61 63 50 10 0 0 10 15 0 0 0 0 52 60 61 45 55 52 B 42 90 72 9 8 1 10 8 0 0 0 0 83 74 58 19 34 50 B 60 66 50 9 _3 _3 10 15 0 0 0 0 55 48 61 73 78 68 B+ 35 43 71 0 1 0 10 14 15 15 0 0 40 45 72 0 32 30 C 46 50 53 6 _3 _3 10 12 0 0 0 0 35 0 60 83 86 61 A 78 85 0 10 5 1 10 15 0 0 0 0 85 71 61 68 74 78 B+ 96 83 96 10 8 7 10 15 15 15 50 15 100 86 98 0 74 86 A 75 85 100 8 10 12 10 15 15 15 50 0 98 95 92 92 94 100 A 76 53 54 7 5 _3 7 15 15 12 30 10 56 59 86 0 55 53 B 86 80 96 9 9 0 10 15 0 0 0 0 93 74 61 68 74 82 A 85 72 96 8 10 8 5 15 15 0 0 0 92 88 65 97 98 100 A 54 59 57 9 3 7 10 11 0 0 0 0 44 76 59 76 81 70 A 33 60 69 9 5 0 10 14 15 15 50 15 52 67 98 0 74 66 B+ 90 84 66 5 7 _3 10 15 15 15 50 20 93 59 100 0 77 82 A 79 70 100 9 10 0 10 15 0 0 0 0 91 76 61 93 94 95 A 67 78 99 7 10 5 10 10 0 0 0 0 91 81 59 93 94 96 A 45 60 76 10 _3 _3 7 0 15 15 10 0 57 50 69 0 27 33 C 42 75 59 9 _3 1 0 11 5 10 10 10 59 55 69 0 27 34 C 86 64 90 10 8 15 10 15 0 0 0 0 89 100 61 73 78 86 A 74 62 74 5 10 5 10 15 0 0 0 0 69 78 61 97 98 91 A 94 57 70 9 8 15 10 15 0 0 0 0 80 98 61 89 91 92 A

Notes on Normalizations

1. The final exam scores showed considerable statistical deviation between the two exams (exam A and exam B). Therefore it is likely that one of the two exams was more difficult than the other (or that the professor graded harshly for one of the exams). Since this deviation was consistent (for high score, low score, and mean), the following correction was applied: for each of the exam groups, exam scores are linearly scaled so that the maximum is 100 and the mean is 70. This technique will also be applied to the results of the first and second exams, so that all scores apply equally in the final score calculation.
2. To make all exams count equally (in spite of different mean scores), each exam score (first, second, third) was linearly scaled to make the high score 100 and the mean 70. Then all three of the scaled vectors were summed, the lowest score of these three subtracted from the sum, then again normalized, to obtain (m) above.
3. The normalized homework total was obtained by linearly scaling the summed homework scores to maximum 100, mean 70.
4. The project total (p) is derived from the components (see below), which are summed and then normalized, again to maximum 100.
5. Column (q) is obtained from (o) and (p) by a weighted maximum calculation. Here are the considerations for the calculation: first, project effort and accomplishments generally exceeded even the best homeworks; second, the first day of class it was announced that one could possibly get a high grade (A) by doing homeworks instead of a project. Thus, in the weighting, the highest project grade could not count much more than the best homework (however unfair that may seem). Moreover, some project results were minimal, roughly comparable to average homework results. To compare the homework and project, first multiply both scores by 6 (motivated by the 60% weighting in final grade). The formula for calculating (q) was based on equating a project score of 150 with 414 homework and equating 465 with homework 600. Finally, the formula used was

    normalize( max( homework, 
    		  (324 - 0.4 * (100 + 5 * project)) +
   		               (100 + 5 * project)
   	       )
   	  )
   

   where, again, normalization scales linearly with maximum 100 and mean 70.
6. Next, we have project component scores. The raw scores are as follows (again, no ID in table, but the relative row number is the same as in the table above).

   Columns in this table are (a) score on diary entries, (b) score on networking / textbook issues in project, (c) score how competently the chosen topic was pursued, (d) presentation of results, code, and conclusions, and (e) degree of difficulty for this project. After scaling each column linearly (again, max 100 and mean 70), the relative contributions of (a)-(e) in the project total score are (a) 10%, (b) 40%, (c) 20%, (d) 10%, (e) 20%.

   Project Component Scores a b c d e ------------------ 70 100 100 100 100 0 0 0 0 0 20 50 20 20 20 80 85 100 80 70 70 85 70 85 80 0 0 0 0 0 60 80 70 50 70 75 50 60 75 50 80 85 100 80 70 0 0 0 0 0 0 0 0 0 0 80 90 90 95 90 70 80 80 95 90 0 0 0 0 0 50 100 100 90 90 0 0 0 0 0 70 80 80 95 90 90 100 100 80 95 0 0 0 0 0 80 85 100 80 80 0 0 0 0 0 0 0 0 0 0 60 100 100 90 90 60 100 100 90 90 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 80 85 80 100 85 50 100 100 100 100 60 90 100 100 100

7. Finally, column (r) is obtained by the formula

    normalize( (3 * exam) + quiz + (6 * project))
   

   since this gives exams a 30% weight, the quiz total 10% weight, and the project/homework a 60% weight.