If you are applying this year, then you probably already familiarized yourself with the passing scores of 2018. They were counted according to the old method, and in 2019 the CT scores will be calculated according to the new system. Awareness is never superfluous, he believes and is in a hurry to tell you how the two methods differ and what to expect from CT 2019.
How points were calculated at CT 2018
To calculate the scores at the CT RIKZ used a special methodology that helped to objectively assess the level of knowledge and training of applicants. It was based on a 100-point scale, which allows to take into account the number of correctly completed tasks, as well as their complexity.
The methodology included the calculation of points for centralized testing tasks with one correct answer and several correct answers. Recall: in the Central Television in Russian and belarusian languages in part A there are several correct answers, and in the CT for the rest of the subjects - only one.
By the way, the more applicants coped with any task, the less points were awarded for it.
How scores were calculated for a test item with one correct answer
For the correct answer, 1 was put, for the wrong - 0. The combination of zeros and ones made up the matrix of answers. It also included the number of applicants who completed a certain test variant and the number of tasks in the test.
The final score of the applicant for the test depended on primary score tasks, initial test score of the applicant and weighting coefficient of complexity of each task.
When these elements were multiplied, we gotmodified primary scoreentrant (MPBA). After completing the tasks of one or another test variant, one could get a certain numbermodified primary points (BCH), reflecting the complexity of the tasks in a specific version.
Final test score was equal to the quotient of dividing the MPBA by the MPB, multiplied by 100.
Rounding off the results to an integer followed the rules of mathematics.
How scores were calculated for a test item with multiple correct answers
If you indicated an incorrect answer or did not indicate the correct one, the system considered it an error. For a correctly completed task (all correct answers are indicated), one was charged, for an answer with one error (in addition to the correct options, he indicated an incorrect one) - 0.2, more than one - 0.
Zero points were awarded when the applicant chose more than one incorrect answer or did not choose more than one correct; specified an incorrect answer but missed the correct one. For example, in A12 the correct answers are 2, 3, 4. If the applicant chose 1, 2, 5 or 2, 3, 5, then he received zero points. Choosing 3, 4 or 1, 2, 3, 4, he got 0.2 points.
Taking into account these nuances, the primary assignment score and modified primary score of the applicant... Then everything was calculated, as in the case with one correct answer.
How the points were calculated on the reserve day
You already understood that earlier the CT score depended on the number of applicants who answered the task correctly. Few applicants are taking the CT. It turns out that their scores are higher? Not. The tests of the "reservists" were assessed according to the described method, only the averaged weighting factors, displayed when calculating the results for the main stage of the CT.
For the formulas used to calculate the test score, see.
Director of RIKZ Yuriy Mikyuk called the dependence of the final score on the number of applicants who answered the question correctly as the main drawback of the old method.
Yes she [old methods - approx. ed.]qualitatively ranked well-trained applicants who scored points on difficult problems. However, the error in measuring the level of knowledge of middle and low-level applicants was significant.
Yuriy Mikyuk, director of RIKZ
The new assessment system for CT should not only effectively rank applicants, but also reflect the real level of their knowledge. The fact that the mark in the diary did not match the results of the CT was mentioned by teachers and tutors more than once. That was an additional motivation for applicants to prepare for CT.
How will the points for CT 2019 be calculated?
At the heart of new system calculation is based on the method of normal distribution N (µ, δ).The most important difference of the new method is that now all correct answers, including partially correct ones, are taken into account.
The result is calculated in two steps. First determined primary score - the sum of points for each correctly performed and partially correctly performed test... After that, according to the results of statistical processing, the primary score is assigned test score, which will be the final result.
Another innovation of this technique is maximum primary score - the maximum amount of primary points that an applicant can receive by correctly answering all the test questions. The maximum primary score is known in advance.
How scores are calculated for a test item with one correct answer
The applicant receives 1 point if the task is completed correctly. And zero points if you indicated an incorrect answer, indicated more than one answer, or the answer is not indicated at all.
How scores are calculated for a test item with multiple correct answers
If the applicant has indicated all the correct answers in the form, 2 primary points are credited to him. One mistake made - 1 point, more than one mistake or no answer - zero points.
You can read all the rules for calculating points at.
What is the passing score for applicants in 2019?
It is expected that applicants will receive a real assessment of their knowledge, scores in the area of \u200b\u200blow values \u200b\u200bwill objectively increase. The number of those gaining 75 or more points should remain approximately the same. Consequently, the situation with admission to prestigious specialties, where the passing score already tends to the maximum value, will remain the same. But it is unclear how many points exactly.
Best to passing grade 2018 to add points 20-30 and already be guided by this figure. Use all the opportunities to increase your scores: tighten your certificate, prepare more thoroughly for CT in the Belarusian or Russian language, and not only in specialized subjects. It's great if on RT you gain an average of 70 points per test, while not preparing for testing too much. But now there are many such lucky ones.
The teacher of history and social studies at the educational center Adukar Dmitry Zaitsev, having analyzed the new assessment system of the central heating, believes what awaits us. But don't be in a hurry to despair! Yes, it is always harder for pioneers, but the world does not collapse on this.
Good luck and patience in preparing for the CT! Adukar will always support you with advice and will keep you updated on education news.
If the material was useful to you, do not forget to put "I like it" in our social networks
ABOUT CONDUCTING CENTRALIZED TESTING
Is paper provided for working notes (draft)?
During the centralized testing at all exams in educational institutions, paper for work records is issued to applicants in the required amount.
Will a tenth grader be able to pass CT?
Not. A tenth grader is not eligible to participate in centralized testing.
Is it possible to pass the CT without a passport, presenting a notarized copy?
Not. During the test, the applicant must have with him the original identity document (passport, residence permit, refugee certificate).
Can I go to the toilet during a DH?
Yes. You will be escorted to the washroom. However, getting out is a minus of your testing time.
Does the applicant have the right to make comments (during centralized testing) on \u200b\u200bthe content of the tests?
The applicants' comments on the content of the pedagogical tests are entered into the protocol of the centralized testing in the classroom, indicating the number of the pedagogical test variant, the assignment number and the content of the comment (see Chapter 7, Clause 58 of the Regulation on CT).
Applicants' questions on the content of pedagogical tests are not considered in the classroom.
In what language is the centralized testing of the history of Belarus carried out?
Applicants have the right to take specialized entrance examinations in Belarusian or Russian (optional).
Applicants who have not studied the history of Belarus can take the test on the world history of modern times, except for the profile (direction) of the specialty "historical" (paragraph 16, chapter 3 of the Rules for admission to universities).
Are semi-correct answers in the Russian language valid?
Yes. In the CT in Russian, both completely correct and half-correct answers are counted. This applies only (!) To tasks from part "A".
For the correct answer, the applicant receives a 100% point, for a semi-correct one - 20%.
Semi-correct tasks are those where one extra answer is marked or one required one is not marked.
Examples:
The correct answers are 2.4. The entrant noted 1,2,4. This is a semi-correct answer, since it has one mistake: one extra answer is marked (1).
The correct answers are 1,2,4. The entrant noted 1.4. This is a semi-correct answer, since it has one mistake: one required answer is not marked (2).
The correct answers are 1,2,4. The applicant marked 2. This is an incorrect answer, because it contains two errors: two required answers (1 and 4) are not marked.
The correct answers are 1,2,4. The entrant noted 1,3,4. This is the wrong answer, because it contains two errors: the required answer (2) is not marked and an extra answer (3) is marked.
Answers to the tasks from part "B" are only correct or incorrect. If at least one mistake is made in the answer from part "B" (for example, instead of А1Б2В3Г4, А2Б2В3Г4 is written), such an answer is considered completely incorrect. The same goes for misspelled answers (eg wrong suffix instead of correct suffix).
How is a question assessed for which more than one correct answer can be given? (1) if not all correct answers are given; 2) if correct and incorrect answers are given?)
You probably mean the partially correct answers that we use in tests in Russian and Belarusian. Even if you could not decide on all the correct answers for a specific task, computer program takes them into account. But you will get a slightly lower score than if you had chosen all the correct answers. The same is the case if you have chosen correct and incorrect answers.
How are Russian language CT scores calculated?
Each task has a difficulty factor. This coefficient is determined not by the test authors, but by the applicants. If many applicants have coped with the task, it has a low coefficient of complexity, and vice versa, a high coefficient of complexity for tasks that only a small number of applicants managed to do correctly.
Therefore, people with the same number of correctly completed tasks may have different points. If a person made a mistake in simple tasks, but correctly completed complex ones, he gets more points.
If the answer in Part B is misspelled, will it be valid?
Not. The answer (word or phrase) is given in the form (gender, number, case), determined by the condition of the test task, and in the language chosen for passing the exam. Spelling mistakes in the response are invalid.
The new scoring methodology for centralized testing is more objective. This was stated on October 31 at a press conference in Minsk by the director of the Republican Institute for Knowledge Control (RIKZ) Yuri Miksyuk, writes BelaPAN .
According to him, the new technique is a mathematical system that completely excludes the human factor. It has been used all over the world for a long time, and from 2019 it will be used in Belarus as well. More than 6 thousand people use the technique. educational institutions in the United States, it is used to evaluate admission tests in Swedish, Finnish, German universities, it is used by world companies in assessing the IQ test.
“It involves more complex processing of responses than simply multiplying by some previously known coefficient, - said Miksyuk. - The methodology first distributes applicants according to the primary scores in percentage terms, and then, in accordance with this percentage distribution, converts the primary results into the final test score ".
For example, he explained, 10% of applicants received 40 primary points. “The statistical distribution is taken into account, how many percent of applicants received 40 points, how many - 50 and so on. And in accordance with this distribution, the final test score is formed. That is, if the given 40 primary points is almost the maximum result, then it will be, for example, 90 test points» , - said the head of RIKZ.
When asked why the new system is better than the traditional one, Miksyuk answered: “This is classic scientific approach... It allows you to determine the level of knowledge objectively, regardless of the complexity of the test or on other points ".
Up to a certain point, the existing method of calculating the CT scores suited everyone, but then questions arose as to why such low scores are received by applicants, Miksyuk continued. “What is this score - 12? And no one could answer that 12 points is not a real assessment of the applicant's knowledge, but rating score in a specific test, - he noted. - If the test is replaced by another, then the results of the applicants will be completely different ".
Miksiuk said why the simple assignment of certain points for tasks of varying difficulty is worse than the new method. “It's all subjective. One expert believes that this task is not very difficult and sets 3 points, while for another expert the same task will be 5 points, for the third - 1 point,- he said. - And all over the world it is accepted that 1 point is given for a test task, and then - how much and how - the method determines. Everything is fair here. Those who know worse will not get more than those who know better ".
According to Miksiuk, in the 2019 inaugural campaign, with the new scoring methodology, the minimum thresholds for points will be higher than in 2018. How much, he found it difficult to answer.
To pass the CT online on the RIKZ website, tests with a new assessment methodology will appear before the end current year, Miksyuk promised.
Applicants who have high scores in the CT certificates for 2018 do not need to worry, the head of RIKZ assured. According to him, the new methodology will in no way affect the results of strong applicants who scored high scores according to the old counting method. “After 70-75 points, both old and new methods give the same results. The rest may need to try to pass the CT again and get higher scores ", - he said.
Most applicants already know the scores on centralized testing (CT). This year, the Republican Institute for Knowledge Control (RIKZ) changed the testing assessment system, which caused many questions among applicants and their parents. Sputnik tried to answer the most frequent ones.
How much did the CT jobs "cost"?
RIKZ announced the "cost" of tasks for testing this year in advance. Thus, the published scheme for the accrual of "primary" points, different for each subject.
For example, in Russian, the correct answers to questions A1-A27, as well as A29 and A30, are "worth" two points, partially correct - one point. The answer to question A28 is also worth two points, and there is no partially correct answer. For correct answers to questions B1-B6, two primary points are given, and for B7-B10 they give two for correct and one for partially correct.
The scoring scheme is different for each subject. More details can be found on the RIKZ website.
What do the primary points mean?
The primary score is the number of conditional points scored during testing. Later, they are converted into a test score based on the correspondence table. It is he, the test score, that is the final result.
How do you know how many primary points correspond to the test?
Why is there such a gap between the points?
Many applicants and teachers paid attention to the discreteness of points: that is, say, the next test score after 100 is 96. A similar situation is at the beginning and in the middle of the assessment table.
As on the eve of the head of RIKZ Yuri Miksyuk, the number of tasks and, accordingly, primary points is limited. Therefore, there is a certain "gap" between the estimates. Otherwise, under the current grading system, it would be necessary to give more assignments in the CU. Therefore, with the new system, discreteness is observed.
Why is it so easy to score 20-40 points?
After the introduction of the new system of assessment of CT, the average scores in the subjects increased. Many have noticed that to get 20-40 points, that is, exceeding the minimum threshold, it is enough to give only a few correct answers.
As the RIKZ management explained, these points correspond to school marks, that is, 40 points - a conditional "four", for which you do not need to know much. At the same time, you need to know much more for 80 points, as well as for the school "eight", therefore, much more should be decided for this mark on the CT.
RIKZ also assert that no one's scores were "underestimated": for the same amount of knowledge, applicants receive similar marks both on the RT and on the CT, regardless of the complexity of the tasks. Therefore, many children now have the same marks in the rehearsal and centralized testing, as well as in the certificate. The reason is that now the CT does not rank applicants among themselves, but proceeds from the absolute assessment of knowledge.
Why is it still possible to "poke"?
Indeed, you can get 40 points at random on the CT - however, it depends on luck. Given a limited number of questions and answers, the probability of randomly typing a certain mark remains, however, RIKZ emphasizes that they are trying to combat this. Roughly speaking, in any test with closed-ended questions it is possible to randomly get into the correct answers - however, this will not bring high scores.
At the same time, applicants with scores up to forty are considered uncompetitive, they have just as little chances of successful admission as those who in previous years scored 20 points.
Will college passing scores increase?
Due to the increase in the average scores on the CT, first of all, in those specialties where they were not very high, including technical and pedagogical. How much they will grow is not yet known. At the same time, a significant change in the passing scores in prestigious specialties is not expected: the number of applicants with high scores remains the same as in previous years.
It turned out that if you make only three mistakes, you can “slide” to 85, but if you almost accidentally “hit” several tasks in Part A correctly, you will get as much as 40 points. This became possible due to the new system for calculating the results of the DH, which was introduced this year. Chemistry teacher Andrei Kulsha is confident that the innovation will allow weaker applicants to enter universities, but the system is not so bad - you just need to understand how to work with it. Especially for TUT.BY, the teacher explained what's what.
Andrey Kulsha. Photo: from personal archiveGraphs that do not indicate the source are illustrative for understanding the explanations and are not based on RIKZ statistics.
Why the old scoring system (in fact) worked for everyone
Until 2019, the contribution of each task to the test result depended on how well the task was performed on average. Did an easy task - got 1 point, did a hard one - got 3-4. Final result the applicant reflected how he mastered the school curriculum. Got 40 points? This means, roughly speaking, I have mastered about 40% of everything that is taught at school.
The typical distribution of the result looked something like this: a maximum at the beginning, then a sharp decline, a large plateau, and at the end again a decline. And this initial maximum was not at 50 points - usually somewhere around 25-30. This is because very weak applicants always have a chance to guess something. Even if you take and put for the test in the literal sense of the word, monkeys who do not know anything, on average, they "bump" points by 15. In the case of applicants, some knowledge is superimposed on this result. Result: somewhere guessed, remembered somewhere, the result is 30.
The most motivated to study at a university are in the area of \u200b\u200bthe region indicated on the graph: somewhere from 40 points, there is already an adequate ranking of applicants. Those who scored below 40 are weak graduates. According to the old system, the CT perfectly performed the task of weeding out such students and, in fact, did not allow them to compete in universities (except for those with a shortage). But the main thing: people who scored 40 or more ranked well. And so they could honestly compete with each other.
Short:
- The result showed the percentage of knowledge of a particular applicant.
- Weak graduates dropped out well.
- The effect of guessing was negligible for competing people with above average knowledge.
What is the essence of the new system?
Now they decided to use international experience - to bring all scales in different subjects to a single one. The point is now not how much individual person knows the program well, but in how well he knows it in comparison with the average applicant in the country.
In general, everything has remained approximately the same: the one who is stronger gains more - both on the old and on the new scale. But the ranking scale became thicker somewhere, and somewhere less often - a spread of 10 points in different parts scales can be completely unequal.
According to the table, it turns out that random guessing of several tasks in the A part will give a sharp rise in points (up to 50), then there is an almost linear section (from 50 to 80) and from 80 to 100 again there is a strong spread, where you can make a couple of mistakes - and abruptly "move out".
When applicants are assessed on such a scale in practice, the distribution of results turns out to be, as it were, elongated (in comparison with the old one) into an even "bell".
It turns out that the scores do not reflect how well you know the subject by yourself. With mathematics - let's face it - on average in the country, things are categorically bad. Therefore, to get 50 points this year is, by and large, to show "no" result. If we take a subject that is traditionally passed better - for example, English - then here already 50 points according to the new system can speak about something. After all, on average, applicants know this subject not so badly - certainly not to zero.
Last year, it was equally easy to score 60 in Russian and 30 in math. In this, 30 in mathematics will turn into 60, but 60 in Russian will remain 60. The same difficulty in subjects is transformed into approximately the same score. In this sense, there is justice.
Short:
- The "new" scores do not reflect real knowledge: it cannot be treated like that.
- The results for the 2019 system show whether you are better or worse relative to other applicants.
- The new scale will reduce the scatter of scores in different subjects (for example: earlier in physics there were low scores, in Russian - average, now they will become equally average).
Will it really affect something?
From the point of view of an individual applicant, there can be injustice, and very serious. This is especially true for those who are not the strongest, but the average. In the lower part of the scale, the spread due to random guessing (or not guessing) is large, about 20 points - this is the error of the result.
Equally weak guys can have a result with a difference of 20 points. After two or three items, the spread becomes 40. This is a lot, although it is still an error. But when we talk about a level where some kind of preparedness is needed (scores 60-80), the scatter of results here will not turn out to be large, because in this interval a small scatter of the primary score does not turn into a large scatter of the test score. The gap of 20 points between 70 and 90 is no longer an error, but a serious difference in the level of the applicant.
Situation: someone scored 70 in chemistry and 60 in mathematics. In total - 130. Another - 90 in chemistry, and 40 in mathematics. Also 130! But one of the applicants is really better in chemistry, and their results in mathematics differ at the level of a couple of mistakes: both did not prepare much. As a result, the one who is more fortunate may pass, and not the one who is stronger in profile subject... And when you cannot show the difference on the CT, you will fight for a certificate, beating off every point - it can be decisive. The annual assessments are still drawing, but they will be even harder.
The summation remains valid when the passing score is very high - where we cannot drop below 70 in any subject, otherwise we simply will not pass. This applies to specialties that skim all the "cream". Such a minority: after all, more often passages of 250-300 points flash, especially in regional cities... The "stray" people who scored 50 points in mathematics, but do not understand it at all, have a much greater chance of getting into the university.
At this time, university teachers are already climbing the wall: completely unprepared people come to them. In schools, with rare exceptions, they train on tests, instill stereotyped thinking and demotivate the learning process itself. As soon as a person tries to reason, compare and analyze, this is strongly discouraged. Bottom line: we have capable and talented children who are duped. They come to the university and simply reproduce the facts they have learned - it is already difficult to make something out of such people. The teachers have two ways: the first is to look for talent among the masses and work with them, the second is to teach "for show", because they themselves do not care. More often, of course, they choose the latter.
Short:
- The value of the 20-point interval turns out to be too unequal at the “weak” and “strong” levels.
- Summation is warranted in faculties with high passing scores (read: in a minority of cases).
- In most faculties, people with “random” grades will have a better chance of enrolling.
- There will be a decline in the general level of applicants, as a result, the value of education will also fall.
- The "war for a certificate" will intensify.
Can this be "fixed"?
Any task must be comprehensively solved. The problem is not in the scoring system - it is completely normal, it is international practice. The problem is that the admission rules are old and the system is new.
RIKZ went away from the percentage of assignments and issued a normal distribution, reflecting the level of the applicant among others. But at the same time, universities (if they are interested in recruiting strong people) had to react and enter threshold scores according to their profile. For example: if a faculty needs mathematics, then it should enter a threshold score of 70 in this subject. If you do not dial it, then your amount is not even considered. And already if you type, then compete with other worthy ones. This scheme deprives the faculty of "stray" applicants who have random 50 points in mathematics, but trained in 90 languages.
Moreover, such an innovation would, possibly, raise general level students at specific faculties. This would affect the quality of education in general: both teachers are interested in working with strong children, and children are interested when teachers want to interact with them.
Perhaps this will need to be explained to the highest official in Belarus: what has been done does not work; the aggregate works. With us, it is like this: different organizations solve different problems and do not always interact with each other. Surprisingly, in the country, TUT.BY often acts as a buffer between institutions, because otherwise they simply “do not hear” each other: there is not enough capacity of the bureaucratic barrier.
By the way, what about the "partially correct" answers: is this a good change?
Click to find out
More good than bad because it increases the resolution of the test. Previously, if the task was a little unresolved, everything is zero points. Now the test has become more accurate: with the same number of tasks, we can get more points.
This innovation has no effect on "guessing". Earlier we guessed the whole task, now - its parts. It is possible not to guess part of the task in the same way as the whole. It turns out it's like an increase in the number of tasks.
This is good just for strong applicants, for whom 1-2 difficult tasks can "throw off" almost 10 points: from 100 to 91. Not to decide at all - to deprive yourself of as much as 9 points, to solve partially - you can get 97 instead of 91. So in this sense it did not get any worse.