Language Speed Review Pain Points

Generally speaking, having time constrained exercises are beneficial in that it helps learners approach native fluency of a language. That being said, I believe that the speed review does a poor job of achieving that end.

Below are a few of the pain points of the speed review (in the Korean course) I’ve identified from my usage of the mobile app:

  • Time scaling. It would be one thing for the allotted time to gradually increase with each word/phrase, but the time seems to decrease almost linearly with the number of questions answered. By the time I get to the ~16 question mark, I have < 2 seconds to pick between 4 answers, most of which are 7-10 word phrases.
  • Answer complexity. I like how the course provides phrases to memorize, as it helps to contextualize the vocabulary. But I HATE that these show up in the speed review. All this does is encourages me to do is to find “cheats” in the phrases that I’m matching (ex… for the phrase “I would love…/나는 정말 ~하고싶어요”, I end up just looking for the answer that has the tilde).
  • Audio playback. Speed review becomes nearly impossible with sound on. Why? Because if the answer happens to be a 7 word phrase, then the audio playback starts to read it out, even when the app has moved on and started the timer for the next question. I have a hard enough time answering the questions in a timely manner without some completely unrelated audio clip playing and distracting me. I’ve taken to just muting my audio completely during the speed reviews.

I think Memrise does a lot of aspects of language learning really well, but things like these are broken experiences. If an exercise effectively encourages a user to develop workarounds for it, then it has failed its purpose.

All being said, my experience with Memrise is limited to the Korean course, so other users may have a different opinions on these points.


My idea is that you expect too much from a computerlearning system.
If you want better, you need real teachers , real life situations and real spoken and written texts.
Don’t be naieve in thinking you can learn it all with computers.

Absolutely! I don’t expect any language learning software to take the place of organic usage of a language, and I certainly don’t think that it’s sufficient to completely achieve mastery of a language.

The pain points that I’ve identified were things that I felt detracted specifically from the experience of using the “Speed Review” exercise from a user experience perspective. If you disagree with any of the points that I made, I’d like to hear what you’d have to say.

Another thing to add, which have made me never use speed review personally: the speed review sessions should not affect your normal word algorithm. The fact that these sessions are multiple choice based and often rely on quick guesses would be putting a word I might not have great retention of further back into the review backlog. I’m not sure if it’s still like that but it affected my word algorithm the last time I used it and it’s a huge point I think. If they overdid the whole system to where users could search for, scroll through and select words to review which doesn’t affect the user’s algorithm information, then I could see its usefulness.

Right now I see it as the first attempt of memrise to compete with the games of other websites, but they’ve now taken that direction a bit different. If that direction was fleshed out more, i’m sure people could suggest a slew of other sorts of review besides speed that could really help word comprehension if it wasn’t connected to the algorithm.


I think that the purpose of speed review is to actually review the foreign words. For me, the problem with the fast timer is that often one cannot hear the foreign words, because the app jumps to the next question. So I would like Speed Review to finish pronouncing the foreign words before starting to count time, in order to prevent jumping to the next question before letting the user hear the foreign words.

Whenever the foreign phrase/ string of words is fairly long a similar problem occurs with the writing: Not only is the pronouniation cut off (as indicated above), moreover, the written words are cut off, because the field containing the answer is too small! So the field for text should be bigger and/ or the font should be smaller.

Cutting both written and oral review off = not reviewing!