Problems of Japanese Language Testing

Problems of Japanese Language Testing

In my essay I will introduce the way the most widely accepted Japanese language exam, the JLPT works and how the exam is conducted. It is safe to say that the test satisfies every point a proficiency test should, but it lacks some basic points that should definitely be tested in order to accurately measure the language learners’ actual abilities in Japanese.

First, I will introduce just how popular the exam is. This is especially important, as the large amount of examinees strongly influences the way the exam is now conducted and how it could be improved. The second point of my essay is the introduction of question types in the test and how strongly each part is emphasized. This will show the contrast between the importance of each language skill being tested. Lastly, I will talk about what I think is good about the exam and what should be worked on.

The most widely accepted and accredited Japanese as a foreign language exam is the JLPT (Japanese Language Proficiency Test, 日本語能力試験, Nihongo nouryoku shiken), held twice annually, conducted jointly by the Japan Educational Exchanges and Services and the Japan Foundation. It was launched in 1984 and the number of examinees is increasingly growing. When it was launched the number of examinees was around 7000, but in 2011 around 611000 examinees took the exam worldwide. When it was first held, it offered four levels of examination (level 1, 2, 3, 4) but this system was expanded in 2010 to a five-level testing system (N1, N2, N3, N4, N5). In these systems test level 1 and N1 refer to the most advanced level, while 4 and N5 are the beginner level test. The new testing system allows a more detailed and precise way of testing, with a newly established intermediate level test (N3) (JLPT official website). The following chart shows the necessary knowledge for passing each level in the old system, as well as the new. This chart contains the number of words and kanji known, and information about the necessary listening comprehension knowledge. A more detailed chart can be found at the official JLPT website (https://jlpt.jp/e/about/comparison.html).

	Necessary knowledge to pass
N1	2000 kanji 10000 words high level grammar understanding a broad variety of topics both in written and oral communication listening test is at native speech speed
N2	1000 kanji 6000 words understanding a variety of general topics both in written and oral communication listening test is at near natural speed
N3	650 kanji 3750 words understanding written and oral materials about everyday topics understanding slightly difficult texts, such as newspaper headlines written in kanji listening test is at near natural speed
N4	300 kanji 1500 words limited level grammar understanding familiar everyday topics understanding daily life conversations spoken at a slow speed
N5	100 kanji 800 words basic elements of grammar understanding texts written in hiragana and katakana, and basic kanji understanding conversations occurring in everyday and classroom situations, spoken slowly

Even for those who are not familiar with Japanese language it is visible that the difference between each level is never extreme and the test progressively gets more and more difficult from the beginner level N5 till it reaches the proficient level N1.

Each level tests are divided into three main parts: kanji reading and vocabulary knowledge, grammar knowledge and reading comprehension, and finally listening comprehension. Levels N1 and N2 only consist of two parts, as the kanji reading and vocabulary comprehension section, and the section of reading comprehension and grammar knowledge are not separated.

The length of each section varies. At the beginner level (N5) the entire test takes 105 minutes, the most difficult level takes 170 minutes. The following chart shows the length of each test section, and the total length of the test. Between each section, there is an approximately 10–15 minutes long break.

N1	vocabulary+grammar+reading – 110 minutes listening – 60 minutes total: 170 minutes
N2	vocabulary+grammar+reading – 105 minutes listening – 50 minutes total: 155 minutes
N3	vocabulary – 30 minutes grammar+reading – 70 minutes listening – 40 minutes total: 140 minutes
N4	vocabulary – 30 minutes grammar+reading – 60 minutes listening – 35 minutes total: 125 minutes
N5	vocabulary – 25 minutes grammar+reading – 50 minutes listening – 30 minutes total: 105 minutes

These tests are quite long and tiring. Against this, Imai Shingo (2009) however, emphasizes that the J-CAT (Japanese Computer Adaptive Test) saves time and is just as effective in testing as the JLPT, as far as the skills being tested are concerned. Even the shortest JLPT test takes over 100 minutes, while the longest J-CAT test is 90 minutes long.

One of the strongest points of the JLPT tests is that it focuses very strongly on testing the examinee’s knowledge of kanji, without which it is impossible to understand any sort of written text in Japanese. For each level, there are multiple books containing the exact number of kanji and their compounds, as well as the necessary vocabulary, and the grammar. According to Cerezo (2008) it is necessary to inform the students, in this case, examinees about what exactly the content of the exam is and how they will proceed on the day of the examination. This is also covered, as all students who sign up for the test are given an information sheet about the examination, in which the necessary information is provided both in English and Japanese. Based on personal experience, compared to other language exams I have taken, the JLPT is the most precisely conducted one I have ever encountered, although it is very strict.

Probably the most obvious downside of the JLPT is the fact that there is no oral examination or composition test, which is necessary for testing the actual knowledge of a student. This is probably one of the points that could and should be improved, even though it would cost a great deal of money. The listening comprehension task concentrates on whether the examinee clearly understands the social position of each speaker (as this is strongly indicated in Japanese speech) but there is no actual test to measure whether the speaker can apply this knowledge directly in their speech. Bachman (1991) considers writing texts and speaking a part of communicative competence and agrees that oral exams are necessary in order to be able to accurately measure an examinee’s knowledge. Bachman also states that language tests that don’t actually test the student’s communicative competence are outdated, and should be replaced with the ones that measure a broader range of language abilities, such as the knowledge of cohesion, or sociolinguistic appropriateness. This factor, however, is only tested in the listening part of the exam. The J-CAT, too, obviously, lacks this part, as it only contains kanji and vocabulary knowledge, grammar, reading comprehension, and listening comprehension sections (Imai, 2009).

What has to be considered in connection with problem of the lack of oral exam is the amount of examinees and the number of exams. Since this exam is only held twice a year, as it was mentioned earlier, and the number of students taking it is extremely high, it would be difficult to test speaking, too. If the number of exams per year was higher, though, maybe the number of examinees taking each exam would be less, and this way it would be easier to have oral exams, too. The number of oral exam takers would decrease by only admitting those to the speaking test who have passed the written examination.

Secondly, the testing of the examinee’s kanji knowledge should and could be improved. The current test contains tasks in which the examinee has to choose the appropriate kanji for a given reading from a group of similar looking kanji. This could be improved by the method Spolsky (1968) offers for vocabulary testing. With this method the examinee would be required to select which one of a group of definitions is appropriate for the meaning of the given kanji. The kanji testing used now is very confusing, as sometimes the difference between two kanji is just one stroke only, and this becomes even more confusing with 4 possible answers and the general stress the examinee feels during the test. This can result in showing an inaccurate picture of the students’ actually kanji knowledge.

Students are given enough information about what to expect in the exam and how to prepare for it, as both the language courses and most Japanese language books are JLPT centered nowadays. As it is an international exam and the tasks are the same in every country, it provides an accurate picture of the knowledge of different students around the world. While JLPT is a highly approved test, and it does give a more or less accurate picture of a student’s Japanese knowledge, it lacks the part that would focus on the student’s actual skill to use the language.

Increasing the number of annual exams could probably improve the test greatly, as the number of students taking each exam might decrease and the local Japanese teachers working at each country’s office of the Japan Foundation could even correct the compositions or test the examinee’s speaking knowledge.

References

Bachman, L. F. (1991). What Does Language Testing Have to Offer? TESOL Quarterly, 25 (4), 671 –704.

Cerezo, L. (2008). Principles of Language Test Design [PDF document]. Retrieved from: https://webs.um.es/lourdesc/miwiki/lib/exe/fetch.php?id=units&cache=cache&media=tecnicas_unit_4.pdf

Imai, S. (2009). J-CAT(Japanese computerized adaptive test)の得点とCan-doスコアの関連づけ. ヨーロッパ日本語教育, 14, 140–147.

Japanese Language Proficiency Test Official Homepage. https://www.jlpt.jp/e/index.html

Spolsky, B. (1968). Language Testing: The Problem of Validation. TESOL Quarterly, 2 (2), 88–94.