On December 3rd, the last JLPT proficiency test in 23 years came to an end.
After the exam, everyone is undoubtedly most concerned about their own results, and the highest voice of the friends in the group is the "scorer", and"Is the scorer accurate?”This question is also the most concerned by everyone.
Regarding the accuracy of the scorer, in fact, the score given by any scorer cannot be 100% consistent with the final score. This has to talk about the scoring principle of "scale score" in the JLPT ability test.
Relative to the fixed score of each question in traditional exams,The JLPT proficiency test adopts the method of "scale and point". The scale score is scored using a method called"Equalization"The method can always be tested with the same scale as the score.
Therefore, the results of candidates will also change according to the examination status of all candidates in the current examination.
If your score band is exactly the same as the score band of the Scorer, it will be very accurate, and conversely, the Scorer will not be as accurate (but the difference will not be very large).
Therefore, for the scorer, you can still use it with confidence, but in the end, it is still subject to the final results given by the official.
Having said all this, what is the "scale point" all about?And how is it scored?Let's take a look
#
What are the benefits of getting a little bit of scale?
Compared with the traditional direct calculation of the paper score, the advantage of the scale score is that it is given based on the candidate's ranking among all the candidates in the current class, and its scoring method excludes the influence of the difficulty of the test paper.
There will be no large fluctuations in the score due to the change in the difficulty of the test questions, and it will also avoid the unfairness caused by some questions with ambiguous answers, and it will test the overall strength of the candidate.
This also allows candidates who take different batches of exams to compare scores side-by-side.
#
How to score the scale points?
Let's start by translating "scale score" into "gear score".
When grading the test paper, JLPT does not directly look at the score of the paper, but counts the correct and false distribution of all candidates' answers to each question in each section (listening, word grammar, reading) (which can be understood as the permutation and combination of answer sheets), and finally classifies candidates with different answer modes into a total of 61 brackets of 0 points and 60 points to obtain the candidate's score.
For example, in a test consisting of 3 questions, there are up to 8 possible response patterns (2 to the 3rd power) as shown in the figure below.
They will be placed in 8 different scoring bands, of course, full marks for all correct candidates, but for candidates who get one question wrong, the first question is wrong, the second question is wrong, and the third question is wrong, they may be placed in different brackets.
It is generally believed that the higher the correct rate of all candidates, the more likely the candidates who answered the question incorrectly will be placed in the lower score bracket.
As shown in the figure, in the actual test, there are 44 questions (N1 word grammar) in the most sections, and theoretically there is a 44-power answer pattern of 2.
To sum up, when we do the questions, we must be careful and careful, and we must not make mistakes in those very simple questions, because the correct rate of all candidates for this kind of questions is often higher, and once we make mistakes, we may be classified at a lower level.
On the other hand, you don't need to pay too much attention to the questions that are too difficult, and even if they are wrong, they may make the question insignificant because of the low accuracy rate of all candidates.