 |
| a mini-quiz = only 5 questions |
If we shorten the quizzes to mini-quiz length, I've got three quizzes ready to be turned into code and slapped up on the Test Central blog. There'll be an announcement tonight.
Monday, September 08, 2025
if we shorten the quizzes to mini-quiz length...
If we shorten the quizzes to mini-quiz length, I've got three quizzes ready to be turned into code and slapped up on the Test Central blog. There'll be an announcement tonight.
8 comments:
READ THIS BEFORE COMMENTING!
All comments are subject to approval before they are published, so they will not appear immediately. Comments should be civil, relevant, and substantive. Anonymous comments are not allowed and will be unceremoniously deleted. For more on my comments policy, please see this entry on my other blog.
AND A NEW RULE (per this post): comments critical of Trump's lying must include criticism of Biden's or Kamala's or some prominent leftie's lying on a one-for-one basis! Failure to be balanced means your comment will not be published.
Subscribe to:
Post Comments (Atom)
Question about these quizzes: Are you going to give partial credit if someone gets one correct answer but not all of the correct answers in a given question? Likewise, will there be a penalty for selection of multiple incorrect answers? Say, for example, that a test-taker should select B and C for full credit. It feels like they should get something for selecting B even if they miss C. Likewise, negative points seem appropriate for someone who selects A and D.
ReplyDeleteAs things stand now, the intro post at Test Central says "no partial credit." This keeps the exam coding simple. But I imagine that that can be changed if you think partial credit is worthwhile. I could make every correct answer worth a quarter point. More motivating for the grade-grubbers...? Maybe read my rationale in that intro and tell me what you think about how fair the system, as laid out, currently is. Maybe it's too tiger-motherish.
DeleteBut that sort of partial credit is about as far as I'm willing to go. You'll recall that the SAT itself was a weird admixture of whole points, fractional negative points, and "no points given" if a question were skipped entirely. Way too complicated for my pea brain. I think a curve was applied, too—one that changed every year. I don't want to over-complicate things that way.
I can also create a poll on Substack and gather learners' opinions that way. Make the process a little more democratic.
Ah, yes. I read that intro and that completely forgot about the "no partial credit" thing. I will admit that partial credit and negative points would be more complicated, but that's where coding comes in! I'm not sure exactly how you are coding these quizzes, but it would (should?) be a fairly trivial matter to give positive points based on the number of correct answers chosen and negative points for the number of incorrect answers.
ReplyDeleteWhether you would want to do that is another story. In some situations it might not make sense, such as the Eightfold Path question (if a test-taker gets seven of them right and one of them wrong, they would essentially be punished twice for the wrong answer). Then again, as a test-taker, I think I would find it a little odd that a question that required me to choose 8 correct answers out of 17 would be worth the same amount of points (that is, one) as a question that simply had me choose one of four answers. Maybe scrap the negative points and give fractional points for each correct answer so that the Eightfold Path question is worth, say, two or four points? The multiple-choice questions with four choices could then still be worth one point.
I dunno... I'm just spitballing here. I think some subtlety would be good, at least.
You may or may not have seen my latest Substack poll by now. In that entry, I wrote a bit more about the problem and acknowledge there may be a psychological reason for giving partial credit: learner motivation and the minimizing of failure experiences, which can discourage learning. So please, vote in the poll and leave a comment there if the poll's answer selections aren't nuanced enough.
DeleteAs for the issue of whether a physically bigger question should or shouldn't be worth as much as a smaller question (Eightfold Path): from the point of view of "either you know it, or you don't," I'm not at all bothered by assigning the question a single point. I saw it as a simple knowledge question: either you know all the "spokes" of the path, or you don't.
For what it's worth, nothing like that sort of ponderous question will appear on the quizzes I've got planned for Tasty Grammar.
If I were seriously teaching a course on comparative religion, I probably wouldn't use such a quiz, anyway, but let's pretend for a moment that I did, and that I planned on implementing partial credit. I'd probably make every quiz question worth one point (it's a dang quiz, after all), and if a given question had multiple answers—say, 17—I might be persuaded to make every checkbox of that question worth 1/17 of a point, and the number of correctly checked or unchecked boxes would be a fraction out of 1 point for that question. So if a student got 9 out of 17 boxes right, he'd get 9/17 of a point (converted to decimals, of course) for that question.
If I do end up going the partial-credit route with my current quizzes, it'll probably be like that: each multiple-choice question has four answer choices, and each answer choice has a correct answer: it's either checked or unchecked. So the student in that scenario would get 1/4 point for every correctly checked/unchecked box. (I can hear my UK readers going, "You mean ticked or unticked, yeah?" I mean ✓ or not ✓.) So the final score out of 5 points would be increments of 0.25 point. This might be more helpful, psychologically, for students who get a 4.25, 4.50, or 4.75 out of 5. If I didn't use partial credit, the only way to get an "A" would be with a perfect score: a 5/5. A 4/5, on my Fairfax County scale, would qualify as a C+, which wouldn't be very encouraging.
So having partial credit might be better from a psychological perspective.
I did vote in the poll, although I didn't leave a comment because I didn't necessarily have anything to add that I didn't say here. I did notice the one comment that also suggested negative points, though. I'm sort of moving away from that idea at the moment. If you are indeed going to have four choices, each with a correct state, I like the idea of getting incremental points for each choice in its correct state--especially given the harshness of your scale (80% is usually going to be a B in my experience). The fewer questions you have on the quiz, the greater effect a single missed answer will have on the final grade.
ReplyDeleteWow—Fairfax County was harsh! It was something for us students to brag about, though: with everyone else using a "10-point scale" to convert percentages to letter grades, our scale said you really had to be up there to get an A. I ended up being a mediocre student at Mount Vernon High School, mainly because I'd messed up my freshman year. I was a B+ student by the end in terms of cumulative GPA, and I ranked 44 out of 410 people in my class—just outside of the top 10%. Not that any of that matters these days. In fact, one of my grad-school profs, who'd seen how stressed I was about comprehensive exams, reassured me by saying, "In ten years, none of this will matter."
DeleteUPDATE: I did a bit of poking around, and Fairfax County now uses the same 10-point scale as everyone else. This happened around the 2009-2010 school year.
DeleteThere are very few things in our lives that will matter in ten years. Not a bad thing to keep in mind.
Delete