1 Month of Reading Japanese Newspaper: A Detailed Measurement of My Progress

Japanese Language Reading
1

Since I’ve touched on this in another topic, I thought I’d get into detail about it. I’ve been a bit frustrated that it is hard to quantify the progress you make reading native material. You know it gets easier, but it’s a gradual process, so there’s no immediate gratification. I’m very fond of numbers and statistics, so I tried to measure my own progress.

So this is what I did:

  • over the course of 1 month, I read 114 Japanese newspaper articles on NHK’s website (more like 20 days, though, since I had to stop for 9 days in between because of a study trip that left me with little spare time)
  • I tried to cover many different topics, but of course was a bit biased in my selection (didn’t read more than a couple of sports-themed articles)
  • I created a spreadsheet and wrote down (a) the length of each article and (b) the number of words I had to look up
  • (I looked up all words I didn’t know with certainty, including the reading, but didn’t count names of places and people. When I was in doubt, but turned out to be correct, I also didn’t count that.)
  • I then calculated the ratio of unknown words to measure the change.
  • I also re-read most of the articles directly afterwards and counted again (that wasn’t the main purpose, though)
  • I did not do any kind of conscious repeating apart from reading the articles.

Here are the results:

image
image900×558 84.5 KB

Important is the blue stuff here (red is only for 2nd time reading which I recorded more out of curiosity). The horizontal axis show the progress on # of articles, the % numbers on the vertical axis show the percentage of words* I had to look up.

The diagonal blue line shows the trend. As you can see, there are always spikes (often during articles on topics I was unfamiliar with, like politic organs or technology) and lows (I became quite good at reading articles about weather, climate and natural disasters). What’s important, however, is that there was a constant and notable trend downwards, from about 2,4% to 1,2%. So basically I was able to eliminate half of the unknown words in the course of one month or 114 articles.

(*Actually, this is not entirely true: I calculated the value by dividing the # of lookups by the # of characters of the article without taking into account how long the words I looked up were, so the actual percentage should be higher. It just would’ve taken a lot of extra effort otherwise. But what’s important here is the progress, not the exact values.)

To illustrate that further, let’s take a look at these numbers:
image

The table shows (a) the average percentage of lookups in the last 7 days and (b) the total average percentage of lookups of all days. Here, too, you can see both numbers going down steadily.

Conclusion: One month is not a long time, so I was very happy to see that progress was constant and significant. I’m no one special, I don’t think I’m above-average at memorizing stuff, so I think the results can be applied to most everyone.

It is easy to loose motivation if you can’t see or feel your immediate progress, so I hope the results of my little self-experiment will encourage others and reaffirm them that progress is constant and significant if you keep doing something regularly.

23 Likes
2

As someone who also likes to measure progress, and can find it difficult to do so, this is a great post!

Now it has me thinking how this could be applied to measuring progress across different “levels” of texts (simpler vs. more complex readings). Or like if you progressed from reading “NHK Easy News” to reading the main website, how you might quantify that progress. Mmmm now you’ve got me thinking! Love it!

Thanks for sharing!

2 Likes
3

Yeah, it definitely is much easier to measure the progress in a semi-isolated environment. I’m pretty sure if I would do the same thing for reading books – different books, I mean –, the results would be all over the place.

1 Like
4

Yeah it is helpful that you’re measuring progress across one “type” of reading (i.e. news on a specific news site).

The main thing that I’m working on fleshing out is tracking progress across the ACTFL Performance Descriptors for the Interpretive Mode (For more info about the descriptors, see here, and for the actual descriptors for the Interpretive Mode, see here.)

Basically, based on the text level I’m reading, can I say that I fall in the “Novice”, “Intermediate” or “Advanced” range, and then track my progress across that scale. There’s a bit more to take into account, and matching a text against that scale can be a little tricky, but your post got me thinking about that again, so much appreciated :D.

1 Like
5

That sounds very interesting. Didn’t know about ACTFL before, so I’ll definitely look into that. Thanks!

1 Like
6

That’s a really cool idea and approach. I’ve thought about at least reading newspaper articles daily, but haven’t followed through on reading more than one. I also struggled with selecting what to read. My first approach was to focus on the top articles at 朝日新聞, but that often ended up being stuff I didn’t care to read about in detail.

I wonder if there are any other rubrics to use to measure article difficulty. I’ve seen some learning sites that clearly have a scale based upon the JLPT levels, ranking articles from 5.0 to 1.0 to indicate difficulty.

7

As mentioned in my previous post, I’ve been occasionally working on a rubric of sorts, mostly to classify a text based on ACTFL’s performance guidelines scale. It’s a bit of a work in progress, but you can kind of see what I have so far here.

The end goal being that I have a well-defined enough rubric that I can say, “This text falls under the X proficiency level” (replacing X with Novice, Intermediate, Advanced, etc.). This would also assess text difficulty, by proxy, as less challenging texts would be, most likely, labeled as “Novice”, more challenging as “Intermediate” and so on and so forth. I’ve also done some research in what a “Novice” ACTFL proficiency level translates to when it comes to JLPT and CEFR.

Anyway, just thought I’d share! If anyone wants to nerd out more about this kind of thing and/or wants to help me flesh out the rubric, let me know :heart:.

4 Likes
8

This topic was automatically closed 365 days after the last reply. New replies are no longer allowed.