Multiple choice questions (MCQ) native in Anki

Hello everyone, as long as I began to accompany the forum I observe that the issue always reappears and always the same arguments are placed on the table, in synthesis, that MQCs are ineffective for the study by the ability to “cheat” taking the tip directly from the alternatives . Summarizing the arguments here Multiple choice questions - Frequently Asked Questions.

As much as the general idea is that MQC’s are not good for keeping information in memory because they are too easy, I imagine there are uses for them with very close concepts that instead of leaving the question easier, add a new layer of difficulty that improves information retention.

See that I do not make peremptory affirmations here, I only expose an empirical experience about the utility that I retreat from the approach using MQC (for close and confused concepts in alternatives). Several people who do not use Anki studies by doing past multiple choice tests, learning information and subsequently going well in future tests.

To put in a simple way: we do not know in reality the greater or lesser effectiveness of a basic question-response or an MCQ question-alternative, and what appears the difference rests more in the type of information that is wanted to learn than in the substrate in which it lays. That is, traditional questions-answers or in MCQ format are not better or worse than others, it all depends on the type of information you want to learn.

In this way, without scientific certainty about the effectiveness or ineffectiveness of the best or worse technique (any of them using SRS) so we keep information in memory (also how information is worked before turning into a flashcard), the negative of deployment MCQ natively in Anki bumps more into a technical coding problem than the utility of technique itself.

If such a suggestion is not very complicated, I would very much like to see this in a next Anki update, letting users choose whether or not they want to use the technique instead of discarding it as inappropriate without scientific certainty.

I imagine that several Anki users think in a similar way, because if that were not the case there would be no incredible add-ons that make functionality available (Multiple Choice for Anki - AnkiWeb) or templates that allow it (edit: removed hyperlink), both of which I use.

“Why post this as a suggestion if you already use templates and add-on that solve the problem?” Because they may not be compatible with the next versions of Anki and not all users are able to use them.

Thank you all for the time. Anki is changing my life.

5 Likes

My vote.

1 Like

Definitely. I like to use MC questions for things I tend to get mixed up and it helps a lot, because I’m regularly forced to think about the differences.

2 Likes

I don’t really use MCQ format for my own study, but to as to the discussion, I think is useful for discriminating between closely related concepts tay would be a source of interference. Wozniak would probably say it would cause a convergence of the two offending items until the two concepts are almost back-to-back.

What you are really practicing with the multiple choice is the algorithm that guides you to the correct answer. As long as MCQ doesn’t present you with semantic cues but rather force you to use the algorithm, it should be alright, but that takes a lot of effort in formulating your questions.

2 Likes

MCQs are a format quite common in standardised tests. So that’s reason enough to believe Anki should have built-in support for them. Not all MCQs can just be easily transformed into other formats.

I agree with this. One type of question that gets asked in this format is AR (Assertion-Reason) and there’s just no way to make AR harder by removing the MCQ format.

For anyone unfamiliar with it, AR involves finding out whether or not the assertion and reason express correct statements and whether the reason is a sufficient explanation for the assertion. Then you select between 4 options like “Both A and R are correct but R doesn’t explain A”, etc.


In any case, studying questions that require a MCQ format is essential for standardised tests if not anything else. A built-in feature would be great IMO.

1 Like

(I just now saw your post but wanted to point out a few important things)

MCQ is good for tests because it’s easier and faster to mark. But learning is most effecient when you…

  1. Use active recall (remember / retrieve the answer, not just recognize the answer on a selection of choices).
  2. Use spaced repetition (ideally a SRS that adapts to your memory curve).

That’s the current consensus from psychology and memory sciences as far as I know (I studied psychology).

So while MCQ can work, they also are inferior to normal flashcards.
(You should also consider the fact that re-reading the same paragraph of text in a text book often enough works too, if you want to memorize the contents in said text. It’s very inefficient and far inferior to active recall though).

I’m pretty sure that’s factually incorrect. We do know that MCQ are less effective for learning. This applies to every field were facts should be memorized. In reality, though, MCQ are even worse than in theory – the reason is that MCQ are often too obvious, i.e. the other answer choices are too unplausible.

It’s probably really easy to implement, considering there are a lot of anki templates that already do that.

Doesn’t this just test for comprehension? I imagine after seeing the cards for a few times you wouldn’t really have to read the answer choices at all, because pattern recognition would set in. That’s a common ‘trap’ for people that have / want to learn things: They believe they know something just because they recognize it. But without those triggers most of the knowledge appears to be gone and cannot be recalled. This, too, is an issue that doesn’t happen with active recall, and even those MCQ that you mentioned can be easily converted into the more effective active recall learning.

1 Like

This is a 4 year old post by the way, so OP might not be active.

Let’s see.

Assertion: Water availability can majorly affect the rate of photosynthesis.
Reason: Water supplies the electrons during photosynthesis.

The solution to this is, “A and R are both correct but R doesn’t explain A” because water mostly affects plants in indirect ways. Like changing the shape of leaves. For electron supply not much water is needed.

I just don’t think asking something open-ended like “How water affects photosynthesis?” is more productive than asking the questions I’m more likely to encounter on exams. The AR format is just so much better. It’s more to the point. The open-ended one can be answered in five thousand different ways. Maybe a botanist would know more ways.

As for pattern recognition, I plan to bulk import PYQs and other material and practice them so that’s unlikely to happen.

I’m trying to find a suitable one that I can apply prettify styling to. But almost everything is extremely complicated or is too simple and bare-bones.

True. Questions shouldn’t be open ended and should be as specific as possible.

Obviously studying with MCQ is better than not studying at all. I was just arguing that proper active recall questions are much better for learning things, while proper MCQ are much better at testing what you know or understand about the topic, after the learning phase already happened. The premise of OP regarding the scientific “facts” were just wrong, which I felt like should be pointed out.

So if it would be studying with MCQ vs not studying at all, then MCQ clearly is the better option. Or if studying for exams is the primary goal and long term retention isn’t as important, MCQ works as well.

What do you need help with? Just the design of having a flashcard on a desk? If you open a new topic I’d probably be able to help you.

Just my two cents, but couldn’t you directly say Answer the following question in Assertion and Reason format?

1 Like

Yes, you don’t necessarily have to have the options in AR but my point was more about how multiple choice isn’t a bad format per se.