Software Review: Anki (Instead of Flashcards)

Guest Blog by Andrew from Match Next

1. How do we get some lower-performing 4th graders to memorize their multiplication tables?

First, let me give the context. 

Match’s first group of 4th graders is 49 total pupils.  They arrived to us as 2nd graders.  Almost all of them are English Language Learners.  

They made what seem to be sizable gains with us their first two years.  On the 3rd grade MCAS, 90% were proficient in math, which may be a record of some sort for English Language Learners.  We think they’ll grow from there, and go on to be unusually good at math thinking, whether that’s measured by Common Core or ultimately by A.P. Calculus or other assessments.  

Nonetheless, we believe that our students could get much better at basic math.

We think there are big advantages to being automatic in “math facts” – able to recall without using a calculator.  See Dan Willingham here (pdf)

Roughly 25/49 of our Match kids are automatic at math facts, with very rapid recall.  Roughly 13/49 kids have made big progress in the past few months.  We’ve used various techniques: paper flashcards, ‘mad minutes’ (sheets where you have to solve 32 problems in 1 minute), around-the-world games, singing the facts.  All of this is a supplement to traditional math learning.  And roughly 11 continue to struggle. 

So how do we get better, even as we continue to work in many ways on “math concepts”?  Improved practice. 

2. Flashcard Software: Anki

As a math supplement, we’ve tried a few math programs on the computer.  We’re able to do this because of Match Next.  Match Next is the name of our blended learning school, and a few of us (like me) have protected time entirely to investigate educational technologies and find out if they work in real life. 

Today I’m writing about one online flashcard program we’ve been working with for a while now called Anki.

If you’re a regular blog reader you may remember guest blogger Ben Rachbach guest-blogged about Anki last February (along with a couple other flashcard programs).

My name is Andrew, I’ve picked up Ben’s work.  I work with Ryan (Math guy), Ray (Match Next Director), and Debby (ELA gal). 

Overall, we’d rate Anki a 6/10 in our particular context.  Better than old-fashioned flash cards?  Absolutely.  (Assuming the kids have access to a computer). 

On the positive side, it’s pretty customizable. What the cards are and how they get displayed is totally up to you. Plus it gets rid of the ‘plastic baggies full of random index cards’ problem.

On the negative side, it’s a needy piece of technology. Takes a lot of time to maintain and keep the flashcard decks up to date for an entire class. Also, while it’s flexible as a program, it took us a long time to figure out how to make the changes to get Anki to do what we wanted.

3. What is Anki?

Anki is a computer-based, downloaded program where you make flashcards and study them. There’s also a website. You can study your cards on the website, but you can only make them on the downloaded program (more on how that happens later).

Making a deck is easy, like paper index cards except on the computer. Front side: question; Back side: answer. Boom, you’re done.

Studying is also just like paper flashcards. You look at the ‘front’ of the card. The screen simply will say "1+1"

Then you think about the answer. Like your old-school index cards, it all happens in your head. When you click ‘show answer’ at the bottom of the screen, the answer is revealed:

At the bottom of the screen, you’re given 3 options to choose from.

1. Again
2. Good
3. Easy

Depending on which option you select, Anki will show you the same card again at ‘just the right time.’ The easier a card is for you, the longer the program will wait to show you the card again.

4. How Anki Works:

That whole idea - easier cards get more time between exposures - is called ‘spaced repetition.’

Ben gave a great description of spaced repetition on this blog a few months ago: 

“The idea is to review the item right when you are about to forget. Spaced repetition software (SRS) neurotically tracks your performance on every single fact that you’re trying to learn and remember. It adjusts those spaced repetition intervals to be exactly right for the difficulty of each fact and your memorizing abilities. That way, you’re not wasting any time reviewing an item that is solid in your memory, but you’re also not forgetting what you’ve learned.”

5. Trying Anki

On a staff computer we created every student a profile on the desktop app and loaded a bunch of multiplication decks we made. For the students, we have a class set of Chromebooks. If you’re not familiar, Chromebooks are $250 small laptops that only do one thing - run the Google Chrome web browser. So no downloaded programs.

For Anki, it meant students could only use the Anki website to study. For us this means that tutors have to sync any new decks from another computer. That’s been a low to medium logistical annoyance, but hasn’t impacted the use from a kid’s perspective.

So why a 6 out of 10 in our context? Three main reasons:

I. Anki’s default spaced-repetition spacing didn’t work for us. If you’re a kid who saw 6 x 8 and you clicked ‘hard’ (ie, telling the program you didn’t know the fact), the default would show you that card again in another 1 minute, then 10 minutes if you ‘sort of’ knew the card.

Maybe this is ideal for a med student learning anatomy, but definitely not good for a 4th grader learning multiplication.

We found that our students needed to see a hard fact again much sooner - more like 5-10 seconds - after seeing it for the first time. Otherwise they would forget the fact and keep getting stuck the every subsequent time they were shown the card.

The way the program is set up, it took us a while to to figure out how to make these adjustments because it we had to comb through this beast of a manual. I think I’ve spent 5+ hours of my life reading that thing and trying to figure out how to tweak the spacing settings.  I swear I wasn’t even drinking.  For the record, I’m a college grad and kind of a techie type.  Still, I may have had more trouble here than you -- often when we test stuff, it's unclear whether the limits come from the product or the humans trying to use the products.

II. We found it super tough to make sure students were using the program with fidelity.

It’s hard enough for adults to truly know how well they’ve learned a fact and when they need to review it. It’s even harder to make an 8 or 9 year old meta-cognate the same way.

The only way to make sure kids were using it absolutely correctly was to have the student tell an adult the answer, then the adult would tell the student what to click based on the response time. Not realistic for a typical classroom with 1 teacher, and even a struggle if you have a bunch of adult tutors in the room like we do at Match.

III. There’s a ton of backend upkeep and maintenance required.

Two main annoyances:

A. Sometimes a student ‘finished’ studying all the cards in a deck, meaning no facts were scheduled to be restudied until the next day. But sometimes we’d have 2 minutes left of Anki studying and just wanted a kid to review a hard deck. Anki wouldn’t let you. Imagine having your flashcards taken from you and being told that you literally couldn’t study them anymore. Not cool. The only way to circumvent this problem was to delete the deck from the user’s profile, then reupload it as if it was a new deck. Very cumbersome.

B. There’s no option to load a deck for multiple profiles. We could upload a set of multiple decks to each account, but we couldn’t tell the program to give all 50 accounts the same decks. We’ve been able to work this in to our tutors’ workflow, but if you were a teacher trying to run this solo it’d be absurdly time consuming.

6. Final Anki thoughts

I. Big Caveat: We’re trying Anki in a particular context.  Your results almost certainly vary. 
II. Anki still beats paper flashcards. Not having to lug around flashcards, risk losing them, or getting disorganized is a huge win.  We rated it lower because we thought the software could be far less clunky. 
III. Students need to metacognate properly in order to get the most out of the program. This is easier said than done.
IV. Upkeep and maintenance can be a huge hassle, especially if a deck needs to be reset.

What’s next for us? 

We think we’ve found a competing program that helps solve a lot of the riddles above: it’s called Reflex Math. One of its shining features → it doesn’t rely on the user to figure out whether she knows the fact. We’re rolling it out with a group of students here at Match Next. Stay tuned.

7. Barely Related Bloggy Bonus: What’s the hardest multiplication fact?

A school in the UK wanted to find out. They tracked the number of correct and incorrect responses of a few hundred students doing thousands and thousands of problems (~60,000).

The result? 

‘6x8’ and ‘8x6’ were answered the most incorrectly. Interestingly, students spent more time on other problems. They just happened to get these two wrong the most often.