Here is an end-of-semester question for you (while we are all buried under a pile of paper). When I look at my grade spreadsheets at the end of the semester, I am struck by the extent to which there seems to be a great deal of consistency in grades across various assignments within individuals. It seems as if it would make little difference if I gave five quizzes or 10 (except for the fact that, since I don't weight assignments, the 10 quizzes will count for twice as much toward the total as the five quizzes would). I have a number of classes with a fairly high load of assignments (written, multiple choice and online) and I am interested in retaining a high reliability for grading purposes (more items generally means greater reliability). However, for my sanity and the sanity of my students and efficiency in grading, I would like to minimize the number of assignments I give without losing reliability. This applies mainly to purely summative assignments. The number of formative assignments (for which summative credit is only given to encourage students to complete them) will need to be determined by the amount of practice required to learn the concept well.
Of course, I have all my graded items on a spreadsheet so I could do item total correlations for each assignment and each type of assignment and a coefficient alpha for all the assignments in the class and various classifications of assignments. Has anyone else done this with your classes and/or does anyone have any suggestions for how best to carry this process out? Thanks, Rick Dr. Rick Froman Professor of Psychology John Brown University 2000 W. University Siloam Springs, AR 72761 [EMAIL PROTECTED] (479) 524-7295 http://www.jbu.edu/academics/sbs/faculty/rfroman.asp --- To make changes to your subscription go to: http://acsun.frostburg.edu/cgi-bin/lyris.pl?enter=tips&text_mode=0&lang=english
