This review of Memrise is a long and detailed opinion from someone who loves the science of learning. It has a video, and you can skip to the bottom of this post if you just want the final verdict. Check out the right-side-bar for even more free resources to aid in your studies. Also, you might find that the one month plan to start learning anything is an effective addition to your studying toolkit.
What Is Memrise?
When I first came upon Memrise I was impressed by the clean design of the product, which claimed to provide a fun tool to learn just about anything. As I created an account, I was overwhelmed by the amount of content available. But would it hold up as a learning tool? Could it actually deliver on the promise of making learning fun?
This question was of particular interest to me, as it’s pretty much exactlywhat this blog is about: making learning effective and fun. My background is a combination of video game design and neuroscience, and every time I look at products like Duolingo, Yousician, Rocksmith and Memrise I get excited to see other people interested in building these sorts of games/tools/lessons/products.
So I decided to take Memrise for a spin and see what it might have to offer. I have many good words and bad words about the product throughout this review, including at least one fatal flaw and one redeeming virtue, but I want to be clear up front that I like what the Memrise team is doing and only hope that this will push educational software forward.
Free from Skill Cookbook
If you prefer your reviews in video format, here it is:
What Does It Actually Offer?
There are many scientific claims on the Memrise site. I was excited.
But when I took a first look at the product, I was initially disappointed. It appeared to be little more than vocabulary lists under a thin veneer of gamification. This is all-too-common in the educational software world. I have come to dislike the word “gamification.” When people use this word it usually means they’re trying to slap the science of fun on top of something mundane. They apply badges, stars and points on things that make no sense. No amount of “gamification” can make a boring repetitive task fun. A game must be designed as a game, not built by bolting psychological principles onto a chore.
I started on the mobile phone, because if a company can nail the mobile experience they have crossed the “convenience” barrier for learning. I chose something interesting to me for my tests: ear training. I’m currently attempting to develop music-themed skills, and this seemed like a great opportunity to do so. I have to give the developers credit for doing a great job with a truly versatile platform. The nuts-and-bolts of the learning system that powers Memrise is solid. The music attachments to the cards I studied, for example, were a nice touch.
But in the end it still felt like flashcards. There were other presentations, to be sure. Sometimes I had to type in words. Sometimes I had to do multiple choice. The manner in which the application switched up the way I had to recall the material was impressive. This sort of technique is an essential differentiator which makes Memrise significantly better than plain old flashcards. But it still feels like them.
I had to ask myself, as I was writing this, how this is so terribly different from Duolingo. I reacted very well to the language learning game, yet I felt less positive about Memrise. My conclusion was that Duolingo was used as a supplement for a greater framework of learning, for me. Moreover, Memrise felt like it was trying to do too much.
Check out the above screenshot. It’s funny, to be sure, but it captures my point. When I started trying to learn the composers of classic songs, this was the sort of content I was presented with. If I had written that flashcard myself, this would be great. I did not write that description, though, and it was the first thing I saw when I looked at the card.
The card in the above screenshot manages to say too much and too little all at once. On the one hand, for a student who had been introduced to Beethoven already it would probably be missing any important facts (the biographical history of Beethoven is less relevant here than, say, what is uniquely characteristic about this piece for its time, surroundings, etc.). On the other hand, for a student who had never heard of Beethoven the card would be overwhelming and not particularly useful for recognizing it in the future (the whole point of the game, after all). Still, both types of users are capable of using it to memorize hard facts through drilling and efficiency thanks to Memrise.
So, maybe that’s how Memrise is supposed to be used. As a specific learning supplement for targeted use-cases, like students in a class about classical composers. As it stands, the learning game just drops the content/lesson onto the user and expects her to infer all of the nuance on her own. This “airdropping of content packages” feels a bit sterile… which I’ll return to in the “biggest flaw” section.
Weirdest of all, the creators seem a bit in denial of this…
One Size Does Not Fit All
My biggest complaint with regard to the science of Memrise is that it misrepresents itself as a generic “learning tool.” I wish they would have just said “we’re the best darn rote-memorization tool out there” and stuck with it. They have the opportunity to be the best-in-class at that. By calling themselves a generic “learning tool” and making bold scientific claims they end up over-promising, unfortunately.
Ironically, I think that Memrise is absolutely great for what it actually does: help you to memorize something!
For all that attractive scientific language on the site, I wish they would have just said that. The problem is the task of generic learning and strong transfer outside the digital realm is not a simple one. There’s plenty of research on how to improve effective recall (memorize) and Memrise could substantiate everything on their site with that research. As I’ve said, the basic learning structure of the site appears very technically sound, both on a software and on a scientific level. But the site makes claims that appear to be about transfer into the real world (i.e., making you actually good at using the skills) which I don’t think are substantiated by what I see in their product.
Maybe there’s some feature I’m using wrongly or advantage I’m not seeing. But here’s what I read on their website:
In order to learn anything, you first have to connect it to what you already know. Memories aren’t stored nowhere, you know, they’re always made by creating connections to existing memories. Now, the more your brain does to encode a fact or word [‘encode’ is a fancy word for connect or associate with what you already know], the richer and more robust the resultant memory.
This is all gloriously true. But where was that claim in my experience with Ear Training and learning classical composers, above? I’ve used ear training apps before. I’ve also listened to courses on classical composers via the Great Courses. So I had some context coming into both of my test lessons. Yet I felt largely lost when I was dumped into an immediate quiz-based environment.
Maybe they were trying to leverage pretesting. What I felt instead was a lack of instruction on a pretty nuanced subject. It would have been nice if the game had told me a little more about Mr. Beethoven, though. It never really did.
From a gamer’s perspective, this felt like a lack of “onboarding.” As a software developer, onboarding is the process by which we help educate the user about how to use the software. It exists in apps and it exists in games.
But it did not exist inside of each Memrise lesson. If you think of each lesson like a level in a video game, the levels did not have an appropriate “difficulty curve” which matched my frustration threshold. If the designers of the software could nail that difficulty curve so that it introduced composers in logical groups with more context and an escalating difficulty, I think they’d really be on to something.
Sadly, I’ve yet to discuss what I feel is the biggest flaw…
The Biggest Flaw
My post describing 3 common mistakes which most students make with flashcards explains why it is so important to make your own flashcards.
With Memrise, all of the study content is prepackaged for you as a student. The vast array of pre-made content is one of the things which the company seems to be proud of. It’s true: having all of these pre-made lessons probably lowers the barrier for a student to begin studying. Unfortunately, it also sets the student up to fall victim to the dictionary problem.
I totally understand why they did this. It makes perfect sense from the perspective of a software developer. Getting someone to create their own content is hard. And user-generated content is low quality. In education in particular, it’s rather important to have high-quality content. So Memrise could not have launched as a either a DIY study app or a p2p marketplace for content and be still be taken seriously as an educational tool. So they built their own content on top of a backbone built for memorization.
I think they skipped a step between building the lesson content and putting it in front of the user. This step could have been the part that really made the app shine. I can’t tell you exactly what it should have looked like. All I have is the vague notion that it should be an onboarding step that requires the user to choose, organize, create or otherwise engage in the content in a discovery phase.
Think about the classical music composer lesson I described above. The problem was that I was faced with card after card of information. I never got to stop and get my bearings. If the game had first presented me with just one or two questions and then made some linkage between the two, such as describing that both composers lived in the Renaissance, I might be able form bonds between concepts. This is the sort of “discovery phase” I’m talking about: a moment to pause, reflect, and understand what was studied. To metacognate.
I decided to go back and try studying a topic with which I was already very familiar: Chinese. Here’s one of the first cards I was presented with:
There are just so many things wrong with this. First off, I’m not really comfortable with the whole “imagining a bouncy nude woman” as a suggestion in a learning tool (this appears to be user-generated content which is of dubious quality). Second, the card is switching up the way it writes pinyin and even conflating multiple methods in the same sentence and is therefore more likely to confuse than educate a student. More generally, this is totally overwhelming and not an appropriate starting point for a new student of Mandarin Chinese.
The Redeeming Virtue
The next time I was actually impressed by Memrise was when I looked into the premium plan. Something stood out at me: big, beautiful graphs. I’ve said lots of times on this blog that I love using graphs for motivation, and Memrise has them:
Graphs are boring to create by hand. Automatically logging the data points and creating graphs from them is a clear win that Memrise has accomplished over the manual approach I take with most of my studies. Manually recording is an error-prone and tiresome process.
The question is: is this enough of a value to overcome the rest of the deficiencies?
Memrise Review: Final Verdict
Memrise is an excellent piece of software for memorizing content, a comprehensive improvement upon flashcards. But it is not the generic and all-purpose learning tool it claims to be. Some of the very things which make it easy to use (like prepackaged content) are the very things which will hurt a student’s ability to learn well.
Its biggest strengths, like the different forms of testing and the graphs, certainly hold value but could easily be accomplished by a student on his or her own (and done better, in some cases). The attempts at gamification mostly fall flat, feeling more like cheap tricks bolted onto an exercise.
In the end, I will not be using Memrise for continued studies.
The only place I would consider using it is when attempting rote memorization, and in that case I would prefer to create my own content in order to avoid the dictionary problem and other pitfalls common with memorization-based study techniques.