The Memrise scheduling algorithm is proprietary. It is unlikely that we have independent and trusted research ))).
I stick to Anki because it is a vendor-independent solution (for example, the AnkiDroid project is an independent implementation of Anki on Android) and it is possible to modify the scheduling schema to some degree.
There are scientific buzzwords in their advertisement page. But consider these info:
From: https://www.reddit.com/r/LearnJapanese/comments/1aelh5/how_good_is_memrise_for_learning_for_vocabulary/
Flash cards should supplement your normal course of study, not replace it. Memrise is no different and will not effectively teach you grammar or composition.
...
If I fail a word it shouldn't take a month for it to show up again.
From: https://community.memrise.com/t/frequency-that-words-are-reviewed/4454/5
For any word that you always get correct, this is the review schedule:
Plant/learn: review again in 4 hours
First review: water/review again in 24 hours
6 days
12 days
24 days
48 days
96 days
180 days
180 days
180 days, etc
If you get a word wrong during the planting session, it will
come due in 4 hours, and then 12 hours. After that, it
follows the normal schedule if you always get it right. (24
hours, 6 days, and so on.)
In some courses, if you plant, and then don't water for 24
hours or more, Memrise adds an extra watering session. (4
hours, 12 hours, 24 hours, 6 days, etc.)
Later, after you've watered an item a few times, if you get
it mostly right (maybe you had a typo), it will come up for
review in 4 hours, and then return to the normal
schedule. (For example: plant, 4 hours, 24 hours, 6 days, got
it mostly right, 4 hours, 12 days, etc)
If you get it completely wrong during a watering session, the
schedule starts over as if you're planting. I'm not sure if
it includes the extra watering at 12 hours or not.
It is somewhat like SRS scheduling, and unlike Supermemo or Anki, this time schema is not adjusted to you and word difficulty (easy words aren't graduated faster).
From the official FAQ What is the science behind Memrise? Why is it effective?:
Memrise is based on several important scientific discoveries about how we learn. First, our system of mems (mnemonics and memory aids) promotes elaborate encoding- encouraging the learning brain to do more by engaging the imagination, and thus helping lay down stronger, more durable memories.
That topic covered by famous: https://www.supermemo.com/en/articles/20rules
But again, cards are made by volunteers; Memrise provides only a rich multimedia platform...
For example, Anki's representation system is built around the Qt WebEngine on desktop and able to show HTML, images, and play sounds, but lacks video support.
Secondly, Memrise makes use of Spaced Repetition, helping you review words at expertly spaced intervals to help you maintain them in memory in the most efficient manner possible. Reminders space out in time as your knowledge for a word gets deeper, meaning you don't forget, but don't waste time reviewing what you already know.
See the above scheduling schema. If they are true (and not adjustable to your performance on individual card basis) - too bad. Supermemo adjust E-factor since 1990 ))
Third, Memrise systematically exploits the Testing Effect, which shows that by actively recalling a memory, you strengthen it. Because the degree of strengthening to a memory correlates with the difficulty of the test, Memrise automatically makes the tests more difficult over time, again helping you boost your learning in the best way possible.
Can someone explain how is that possible (makes the tests more difficult over time) if they do not do cards themselves?
Do they mean making shorter time restriction for an answer? I heard only negative opinions about this feature. For this reason people switched to Anki or other solutions...
I'm wondering if they would provide list of thier:
We've got loads of nifty cognitive science tricks up our sleeves that we translate to the Memrise platform, so that we can adaptively calibrate the tests we give you.
to the public.
