I survived Gradefest 2011.
Barely.
I clearly need to work on strategies to become more efficient at grading. One (perhaps obvious) solution would be to implement multiple choice... but I am not sure how I feel about multiple choice exams.
I am pretty sure how my students feel about them (we hatesssss it).
Are multiple choice final exams okay? What about if you haven't given multiple choice exams throughout the semester? Are there better ways to evaluate students that won't break the grading bank?
15 comments:
I have never given a multiple-choice test in my life and never will. I believe that the only kind of test that has any educational value is the kind where students have to produce something themselves, be it a paragraph, a sentence, or just even a word or a number. Circling letters on a piece of paper is not academically enriching, in my opinion.
Of course, it's more difficult to grade such tests than multiple choice but I don't think it's fair to make my life easier at the expense of students.
I'm a really big fan of multiple choice exams that are mainly applied questions - I like questions that require them to take a concept we discussed and actually use it and logic to figure out an answer. I think multiple choice exams that are pure regurgitation are stupid and give them a bad name!
I despise them.
I recently took one in an eco field, and it was NOT fun.
Prof has very detailed lectures, no textbook, no lab or discussion sections, and 3-5 papers to read from primary literature each week.
I had NO idea what to study for. Key concepts in field? Figure interpretation? Definitions? Obvious facts? Obscure facts? Applying concept to real situations? Applying concept to hypothetical situations?
Ugh.
The thing about multiple choice exams is that you put all the time in to creating the questions, so you essentially front-load the grading. Writing a good multiple choice question is HARD!!! A good question being one which creates a bell curve (ie: some people actually get them wrong and some get them right) because the mostly likely wrong answers are included as options AND one that doesn't have to be thrown out due to wording issues (happens more often than I like to recall).
The biggest downside is they only assess if the students RECOGNIZE the answer, not if they KNOW it well enough to come up with it on their own. Still, it prevents a lot of beating around the bush where students come close to the answer but don't actually know/name it.
I guess its utility depends on what type of learning you're trying to assess...
Grade 90-110 all-essay exams 4x per semester (just for one course)? More power to you Clarissa, but I just can't do it. Multiple choice questions are part of my arsenal. They can be written so that students must apply what they've learned, rather than as glorified flash cards.
no such thing in history unfortunately. Closest I get is narrative identifications. SIGH too bad. I swear if I were an exploited adjunct though, I'd be using the textbook provided MC with no qualms
I'm with Pharm Sci Grad - writing good MCQs is extremely hard, particularly for calculations where you must guess the main misconceptions and provide answers that address those. I use MCQ tests as formative assessment but probably not as summative, but I know of subjects at my uni that do.
Can somebody tell me what their pedagogical rationale is for providing 3 wrong answers alongside with the 1 right answer? Why not just get them to provide a single date, name, number, etc.? This will be almost as easy to grade as multiple choice.
Besides, if the goal is to save time, then it isn't achieved. In this very thread, people are sharing how much time and effort they spend trying to figure out the "good" wrong answers for the questions.
I never used to give any multiple choice exams for many of the reasons listed here. However, since many standardized exams to get into professional schools are multiple choice, I started including some multiple choice on tests (with essay type problems and mathematical problems). I found that some students can answer multiple choice very well and some can't. I hope that by encountering some multiple choice they will be better prepared when they take entrance exams.
I use multiple choice for a large fraction of my exams (200 students in my course, 1 TA). I started out at 50% multiple choice (with 120ish students, 1 TA), but had to increase the number of multiple choice questions when grading started taking more than 2 full days to complete.
It is hard to write a good multiple choice exam. It really cuts down the exam headache though, and not just the grading. There are many fewer judgement calls, so a lot less grade grubbing and whining. In some ways, I think it is fairer than problem-based exams.
In my case, the mean grade has not budged, but I suspect that different students are now benefiting. I have a "short answer" section, which is mostly conceptual with some simple plug and chug. I also still ask problem solving questions, but don't chain them like I used to, since I can't carry through an incorrect part a anymore.
Multiple choice questions are ideal for quizzes, and quizzes are ideal for keeping track of how core concepts are assimilated on timescales shorter than the spacing between midterms. I often give weekly multiple-choice quizzes, they take about 5 min to do. I usually don't grade them; instead, right after the students do them, we go over the problems and expose common mistakes and misconceptions.
I occasionally have some multiple choice or true/false problems on exams, but they are typically part of a larger problem that requires some thinking and/or some calculation.
(1) You should never use an exam style on the final that has not been used during the semester.
(2) Multiple choice exams that assess what you want are difficult to design, and that is one reason for #1. You want to vet problems on regular exams before using them on a final exam. That said, one way to design a good m.c. problem is to keep "wrong answer" records for a free response problem used one semester. That is a great way to get good wrong answers.
(3) You can design free response exams that are comparatively easy to grade. That is, designed so that you don't have to figure out extra credit from the work except in extreme cases. Specific wrong answers are worth a specific number of points, all figured out when you wrote the key.
You also DON'T have to make any grading notes on a final because it is pure assessment. They aren't going to learn from your comments.
I should have included
(4) Some conceptual questions are best asked as multiple choice or true-false, but in the latter case I always insist that they explain their answer correctly to get any points for the problem.
In my experience, the best exams that I've taken have been a combination of multiple choice, short answer, and essay - kind of similar to the Advanced Placement exams that high school students take. My undergrad chemistry classes had such exams, and they were really effective.
There will always be those students that are good at multiple choice exams (both knowing the answer and being able to choose one out of the lineup of potential answers). I think a nice balance of multiple choice and "other" questions is best.
All of my graduate school exams were short answer or essay, and those were frustrating after a while, and they actually made me miss multiple choice. In my opinion, there's nothing wrong with multiple choice as long as its done right. What "right" means is another discussion, however. :)
I also continue to ask the problem-solving questions, but not chain them like I did because I could not take the wrong side of more.
Post a Comment