Other

A Statistical Examination Of Course Scores

The correlation coefficient of exam and homework rankings are calculated below. To calculate the correlation coefficient, the scores of each evaluation is sorted from highest to lowest and the correlation coefficient between two evaluation types is estimated from the class population (49 students). The overall score is a weighted average of all evaluation scores. The weights for this course are: overall_score = 0.10 * sum(hw)/200 + 0.25 * exam1 + 0.30 * exam2 + 0.35 * final Here are the correlation coefficients of homework and exam rankings: Corr. Coef. hw rank ex1 rank ex2 rank final rank overall rank hw rank 1.00 0.41 0.25 0.65 0.61 ex1 rank 0.41 1.00 0.58 0.63 0.77 ex2 rank 0.25 0.58 1.00 0.55 0.81 final rank 0.65 0.63 0.55 1.00 0.89 overall rank 0.61 0.77 0.81 0.89 1.00 Comment : Note that homework assignments are only the 10 % of the final course score. The correlation coefficient of homework rankings and the final rankings (grade distribution) is *the* interesting data of the table. Any value above 0.1 indicates that working out the homework problems helps you in the exams. Remember that we made the solutions available to everyone, so just looking at the solutions and learning how to solve the problems in the set does not help you as much as trying to solve them by yourselves. Conclusion : Spending time on homeworks pays off. You should have asked for more homeworks!