Japanese Vocab Quiz Now Tracks Wrong Answers

Yesterday/way early this morning was quite frustrating. It was quite easy to describe what I wanted to build:

Japanese Vocab Quiz should track wrong answers so that users can review at the end of a quiz to see what words gave them the most trouble.

As is, the thought is that users could look at the list and see if the word is still really hard to remember and if so, go look up additional information on it or build flashcards, or just whatever they need to make it sticky for them.

Eventually though, I want to do things like:

• offer printable worksheets at the end of a round that’s just the words you got wrong
• redo the same quiz with only the things you got wrong
• track wrong answers longer than just one quiz so you could come back later that night or the next day and focus on them again
• track wrong answers across everything so that you could try a different quiz type with those same tricky terms as a way to get them to stick better

So yeah. Easy enough to describe, but technically…

tech details

First, I suppose I should say that the vocab quiz was actually tracking wrong answers in a sense.

Currently, you can choose to do a quiz in infinite mode, or to have it only re-use questions until you get them correct once. So in that sense, the non infinite mode tracks what you got wrong. Behind the scenes though, it was actually just starting with X vocab terms in a list, and as you got them correct, it would remove them from the list.

So basically a huge part of building in the answer tracking was that I had to go from a model of just deleting everything as users got the term right, to saving everything. Which I suppose is something that’s also easy to describe but was a massive amount of work to actually build.

As of now, the site is still forgetful. It only tracks your answers inside one “round” of a quiz. If you start a new quiz, it doesn’t have a clue what you got wrong the last time

There are tech limitations as to the why, but the next steps would be:

• tracking answers as long as you remain on the site. Close your browser? Gone. Back out of the site? Gone.
• tracking answers but only for that browser. Were you studying on your laptop but now want to study on your phone? Nope, sorry. No answers for you!
• tracking answers with user accounts. Then it would be saved to a server somewhere, so your right/wrong history would follow you wherever you went. Torment you in your sleep even!

User accounts would be nice, but they’re also waaaaaay down on my list because currently, Japanese Vocab Quiz is a static app with no true server backend. Oh sure, there’s a server in there somewhere that serves the HTML/CSS/JS/JSON files needed to power the site, but past that, the entire thing runs in the browser (if curious, it’s an Elm app).

Someday…

learning new vocab from scratch

I hadn’t previously studied the chapter 3 vocab (other than when I was initially adding it to the site), so this was my first chance to test out what it was like to learn something from scratch using the site. I went through the Japanese -> English flash card quiz for the Chapter 3 vocab and… I heartily would not recommend it.

Though I eventually did brute force most of the terms into my head (not all 73 were brand new to me though), it was a long, tedious, and ultimately demoralizing process.

It was funny/sad to me how often I’d get a card and be like “what was that one again?” and then sometimes I’d see it maybe less than a minute later, and I had already forgotten it again.

This happened several times. There were terms that were just not sticking despite how often I was seeing them.

Some highlights from my own most missed answers:

• のこす - to leave something
• よがす - to make something dirty
• かわる - for something to change
• ぼうえき - trading
• しゅうしょく - getting a job
• さいふ - wallet, purse

Granted, I’m now familiar enough with the vocab, that I don’t get to see it with fresh eyes, but my theory before last night was that the match up mode was the best way to initially train on new vocab. Last night I was most curious to test the wrong/right with the flash card mode though so that’s what I used (we’ll also not mention the fact that I managed to break match up mode in the midst of implementing wrong answer tracking so I couldn’t have used it anyway.

Plus, there are options I want to add to match up mode before I think it’s super suitable for from-scratch learning: specifically, drag and dropping of the answers from the available answer section (quicker than typing/retyping for when you get something wrong and need to correct it) and an optional mode where you briefly get to see all of the answers before you have to fill it in yourself (better than simply guessing if you’re learning the material from scratch). Also… well, there’s no wrong answer tracking on the match up mode at the moment (technical limitation that may or may not get worked around in the future).

I have several theories as to why the match up mode will be better, but the biggest one would just be: quicker wins. If I hadn’t specifically been wanting to see my brand new “here’s your quiz summary” screen, I would’ve given up in the middle of the 73 question flash card quiz. It was far too long, too aggravating, and just seemed like a massive waste of time as a way to brute force new vocabulary into my head.

I would personally argue that optimizing for fun-level (or at least the “I don’t want to throw things” level) is the most important thing to build new learning around. Language learning is hard and it’s easy to get discouraged, so while my thing last night may have theoretically been quicker than if I’d gone through the match up mode several times… it was not pleasant in the least as a way to start from scratch.

Of course, all of these theories can’t be easily tested until the next chapter. Until then…