One of the most important things you need to understand in order the pass the exam is how JLPT scoring works. Without this understanding, how can you devise a good test taking strategy or study plan? For those like me who read the official test website and still couldn’t figure it out, I’ll explain what a passing score is below and give you some tips on where to spend your time.
Minimum Passing Score
Test questions at all levels can be broken into 4 major focus areas: vocabulary, grammar, reading, and listening. (Check out my post about the different question types for more information.) These focus areas are grouped into sections for scoring. The following table shows how the focus areas are grouped per JLPT level.
| Level | Points Per Section | Total Possible Score | Passing Score | |||
|---|---|---|---|---|---|---|
| Vocabulary | Grammar | Reading | Listening | |||
| N1 | 60 | 60 | 60 | 180 | 100 | |
| N2 | 60 | 60 | 60 | 180 | 90 | |
| N3 | 60 | 60 | 60 | 180 | 95 | |
| N4 | 120 | 60 | 180 | 90 | ||
| N5 | 120 | 60 | 180 | 80 | ||
In order to pass, you must meet two criteria: (1) the minimum overall score and (2) the minimum section score.
(1) Minimum Overall Score
The points in each scored section are summed to produce your overall score, which is a number between 0 and 180 The minimum overall score to pass is shown in the right-most column of the table above.
(2) Minimum Section Score
Additionally you must get a score of >30% (at least 19/60 or 38/120) in each section to pass. You can’t miserably fail any one section and pass the test.
Disclaimer
What follows reflects my personal understanding of the JLPT scaled scoring system. I am not an expert in item response theory and you should regard what follows as simply my opinion.
Points Per Question
According to the official website, “Calculation of test scores…is based on a statistical test theory called Item Response Theory (IRT). This is completely different from calculation of raw test scores based on the number of correct answers. ” See https://www.jlpt.jp/e/about/pdf/scaledscore_e.pdf.
This is where I often hear people say that this means we can’t know how much each question is worth and the score you get is random. Yes, we don’t know exactly how much each question is worth. But, no, your score is not random. IRT scores generally show a high (>.95) correlation with raw scores. See, for example, http://journals.sagepub.com/doi/abs/10.1177/0013164498058003001. This means that like raw scores, the more questions you answer correctly, the higher your scaled score will be. Also, the more questions you answer correctly that are of high difficulty, the higher your scaled score will likely be.
Broadly speaking, IRT looks at your overall answering pattern and compares that with the answering patterns of highly/not highly test takers. For example, someone with somewhat low Japanese ability will get easy questions correct and most hard questions wrong. On the other hand, a test taker with high Japanese ability will get many easy and many hard questions correct. Now you come along after studying many long hours (and reading the tips on this website) and get many hard questions correct. You were rushing so you misread a couple of the easy ones and got those wrong. IRT will look at your overall answering pattern and probably group you with the highly skilled test takers, resulting in a high scaled score.
One important thing to remember is that your score is not binary (high or low). It is a continuous scale from 0 to 180–well, it can be any integer, so it is not strictly continuous in a mathematical sense. So, all things being equal, getting a few more questions correct, should boost your score, if only a little. I find myself saying this a lot, but every extra point counts. If you are near the border, a few extra points could mean the difference between passing and having to keep up your studies for another 6-12 months just so you can take the test again.
Summary
Let’s recap. The test is broken into 3 scored sections (2 for N4-N5): vocabulary + grammar, reading, and listening. Each of the 3 sections is equally weighted and scored used a scaling method called information response theory. Although this means that your score isn’t a simple sum of the number of correct answers you made, it should–at least according to IRT research–increase with the number of questions you answer correctly. In order to pass you need to meet two criteria. (1) Your overall score is greater than or equal to the minimum passing score. (2) You must get a score of greater than 30% on each scored section.
So What?
Knowing how the JLPT is scored can help you when creating your study plan and test strategy. Here are some tips to consider.
Spend time studying your weakest area. Remember that you can’t do horribly on any one section if you want to pass the test. For example, if you are borderline or below when it comes to listening, listening needs to be your priority. No matter how high your other sections are, if you bomb the listening section, you fail the test.
Allocate enough time to work on the harder grammar questions. There are two reasons that I say this. One is that vocabulary and easy grammar questions you either know or you don’t. The second reason is that IRT rewards you more for correctly answering more difficult questions. If you spend all your time agonizing over vocabulary you don’t know, you’ll be too rushed to do a good job on the difficult grammar (towards the end of the grammar section).
Allocate enough time to the reading section. Remember that (for the N1-N3) the reading section is worth the same number of points as all the vocabulary + grammar questions. To put things in perspective, you might get 45-50 vocabulary/grammar questions and 20-25 reading questions. From these numbers, you can see that it is easy to get caught up in the vocabulary + grammar section and not have enough time to complete the reading.
Practice your reading comprehension. This is similar to the previous tip, but applies to test preparation. We know that since there are only 20-25 questions and this section is worth 1/3 of the entire test, getting a few more reading questions correct may significantly boost your overall score. Additionally, consider that you could spend hours drilling flashcards to learn vocabulary/kanji that never even appear on the exam. (No, not all the vocabulary for your JLPT level can be used each year.) On the other hand, skills in reading comprehension are useful every year and can be applied to the entire reading section.
Checkout the rest of the JLPT Playbook