TWI397824B - The system and method of simulation results - Google Patents

Description

System and method for simulation evaluation of test results

The present invention relates to a test system, and more particularly to a system and method for evaluating a test result of a user's learning outcome.

According to the current Taiwanese examination system, it is roughly divided into two types of examinations, that is, one is to pass the examination, such as high, general examination, professional certification examination, etc.; the other is to accept the quota according to the candidate's score ranking. Taking the joint enrollment of high school and university as an example, all the candidates are ranked according to the scores of the candidates' relevant examinations (high school enrollment is the national scholastic test for national middle school students, college enrollment is the subject ability test and the designated subject test). The higher the ranking, the higher the priority of the distribution of the volunteer school or school.

For this large joint enrollment exam, candidates often conduct a number of mock tests before the exam to simulate the types of propositions and exam atmospheres that may be encountered during the exam, and analyze the scores of these mock quizzes to assess possible Which schools or schools are accepted. The general test score analysis is mostly carried out in a fixed way. For example, the total score of the Zhang San simulation test is 450 points, which is regarded as the strength of 450 points in the last year's college entrance examination, and then the evaluation of the possible admission schools. . However, the difficulty of the test for the mock test is high or low. When it is simple, all the test scores of the test will be higher; when it is difficult, all the test scores will be lower. Therefore, using the simulated test scores as a benchmark to compare the results of the joint enrollment test for last year or other years, and then the possible admission of the school assessment is not accurate enough to assess the true scores of the candidates, and will have the actual scores of all candidates. Great gap.

In addition, the number of simulated tests per year is very large. For proposition teachers, in the case that the mock test questions and questions cannot be completely similar to the calendar year, it must be considered that the mock test questions must have a reasonable design and should not be biased. A certain type of question type, so the proposition teacher is often deeply pressured.

As mentioned above, how to allow candidates to conduct a simulation test can accurately assess their true score rankings, which is a practical thinking direction.

Therefore, the main object of the present invention is to provide a system and method for the simulation evaluation of test results, so as to achieve a more accurate evaluation of the candidate's true score ranking effect.

Another object of the present invention is to provide a system and method for the simulation evaluation of test results, thereby achieving the effect of automatically issuing questions to alleviate the pressure of the propositional teacher.

To specifically describe the content of the present invention, a system for simulating evaluation of test results is proposed, which is mainly composed of a test question bank module, a proposition module and a simulation test result module. The test question bank module collects a correct answer of the test question group with the answer reaction statistics and the test question group; the proposition module combines at least one test paper according to the test question bank module; the simulation test result module is utilized The at least one test paper calculates a set of virtual test subject scores.

Another embodiment of the present invention provides a system for simulating evaluation of test results, which is mainly composed of a test question bank module, a proposition module, a simulation test result module, and an evaluation module. The test question bank module collects a correct answer of the test question group with the answer reaction statistics and the test question group; the proposition module combines at least one test paper according to the test question bank module; the simulation test result module is utilized The at least one test paper calculates a set of virtual test subject scores; the evaluation module compares the scores of the at least one test paper test by the user with the set of virtual test subject scores to evaluate the use The learning outcomes of the people.

Another embodiment of the present invention provides a system for simulating evaluation of test results, which mainly includes a test question bank module, a proposition module, an analog test result module, an evaluation module, an input unit, and a The display unit is composed of. The test question bank module collects a correct answer of the test question group with the answer reaction statistics and the test question group; the proposition module combines at least one test paper according to the test question bank module; the simulation test result module is utilized The at least one test volume, and calculating a set of virtual subject scores according to a specific value input from the input unit; the evaluation module accepting the at least one copy of the user via the input unit and the display unit The scores after the test paper are compared with the scores of the virtual test subjects, and the comparison results are displayed on the display unit.

A further embodiment of the present invention, that is, a method for providing a test result simulation evaluation, comprising the steps of: collecting a test question group having a response response statistic and a correct answer of the test question group; and combining at least one copy according to the test question group a test paper; the number of answers to each question of the at least one test paper is proportionally enlarged according to the answer reaction statistics to be equal to the number of virtual subjects; each virtual subject obtains the question in a random and non-repeating manner in each question Each of the multiple answers of the questions in the test paper is enlarged to combine one answer of each virtual subject; after each virtual subject's answer is compared with the correct answer, each virtual subject is calculated. The score of the tester; and the score of a user after receiving the test of the test paper is compared with the score of each virtual testee to evaluate the learning result of the user.

The main difference between the present invention and the ranking evaluation after the general simulation test is that the past statistics for each test question are used as parameters. For the five options of A, B, C, D, E, etc., if there are 30 test questions in the simulated test paper, the proportion of the first question in the past is: A=3%, B= 20%, C=10%, D=25%, E=42%, if you want to virtualize the distribution of the results of 10,000 subjects, then the answer to the question selects the virtual of A, B, C, D, E. The number of subjects tested was 300, 2000, 1000, 2,500, and 4,200, respectively. Each virtual subject then responded to an answer in a random and non-repeating manner, that is, the 10,000 virtual subjects were tested. For the answer to the first question, there will be 300 candidates A, 2000 candidates B, 1000 candidates C, 2500 candidates D and 4200 candidates E; the second question also uses the same way to answer the answer, until the 10,000 virtual After the test papers of the test papers correspond to the answers, the answers are scored, the distribution intervals of the scores are found, and the results of the score simulation evaluation are obtained. Among them, the test questions with the answer response statistics need to consider whether it is enough to represent the test questions of the large group of the target test items. The test source with the answer response statistics can be the answer statistics of each test test of the previous test, for example: national Middle school students' basic scholasticity test, university subject ability test, university designated subject test, four skills and two special entrance examinations, national certificate test, ability test, and so on.

Referring to the first figure, it is a block diagram of the system of the present invention. As shown in the figure, the test result simulation evaluation system 10 of the present invention mainly comprises a test question bank module 11, a proposition module 12 and a simulation test result module. 13 components. The test question bank module 11 collects a correct answer to the test group with the answer response statistics and the test question group. The test group with the answer response statistics uses the past response statistics (proportion) for each test question as parameters. The proposition module 12 combines at least one test paper according to the test question bank module 11.

Since the test question bank module 11 can classify or set parameters according to the test type, difficulty level, scope of the question, proposition origin, test target or previous division ratio, the proposition module 12 can combine various question types and difficulties according to requirements. The test paper of the degree, and the test paper proposition method of the proposition module 12 is based on the test question bank module 11, and the test questions are set by manual selection or computer, and may be according to the principle of proportionality (the test questions are evenly distributed in various ranges) or according to the standardized test theory. Propositional methods (Standardized test theory is a test procedure for the formulation of competency indicators for various disciplines, the design of test questions, the establishment of test procedures, the scoring method, and even the pretest of test questions and the establishment of test reliability and validity. The proposition can be formulated so that the test can have the best measurement effect. The test paper can also be combined with the standard test paper design structure of the test type selected by the proposition teacher (user) or the test design structure of the previous test. The simulation test result module 13 uses the at least one test paper to calculate a set of virtual test subject scores. When the number of virtual subjects in the group is at least two, the scores are statistically significant. The scores of the virtual subjects can be further ranked according to the scores, resulting in a test result distribution. Among them, the virtual number is the number of people who are estimated to participate in the formal test or refer to the number of people who participated in the test.

Among them, the types of questions with the response statistics can be single-choice questions, multiple-choice questions, multiple-choice questions, yes or no questions, filled questions, selective fill questions or short answer questions. When the proposition teacher asks questions in the way of avoiding the archaeological answers, the parameters can be set for the test questions in the test question bank module 11, such as changing the position of the correct answer (from A to C), or changing the test questions. The calculation parameters make the answer change, but it should be noted that the response statistics (proportion) and correct answers need to be changed accordingly.

In addition, the method of calculating the scores of a set of virtual subjects in the simulation test result module 13 is to enlarge the number of answers of each test question in the test paper according to the answer reaction statistics, and enlarge the number of answers to be equal to the number of virtual test subjects, each virtual The subject obtains an answer to each test question in the test paper in a random (random number) and non-repeating manner to combine the test papers of one of the virtual test subjects, and compares the correct answers to calculate each virtual test. The score of the tester. After the score distribution of each virtual subject is calculated, the user can compare the scores of the test paper test with the scores of each virtual subject, and judge the ranking of the scores, thereby evaluating the user's score. learning result. Among them, the method of assessing the learning outcome of the user can be further evaluated in combination with other statistical methods according to the evaluation method of the target test. For example, the designated subject test (referred to as the test) conducted by Taiwan in 2002, can be used for each designated test. The cumulative scores of the scores of the subjects were evaluated, and the subject ability test (referred to as the academic test) conducted in 1994 was combined with the cumulative scores of the scores of the test subjects and the total scores to evaluate the learning results.

In this way, the answer to the essay response statistic can be used to evaluate the virtual score distribution of the simulated test, and the parameters are the response statistics (proportion) of the previous test questions, so the virtual score distribution will be very different from the reality. Close proximity, so the simulation of the test results by the system of the present invention is extremely valuable.

In addition, the test result simulation evaluation system 10 of the present invention further includes an evaluation module 14 that automatically compares the test scores of the user with the scores of each virtual subject, and judges the ranking of the scores, thereby evaluating The learning outcome of the user.

Referring to the second figure, it is another system block diagram of the present invention, as shown in the figure: the test result simulation evaluation system 10 includes a test question bank module 11, a proposition module 12, a simulation test result module 13 and an evaluation. In addition to the module 14, there is further included a display unit 15 and an input unit 16. The user can input a specific value, that is, the number of virtual subjects, through the input unit 16, thereby calculating the virtual subject score and the distribution status of the specific number of persons. At least one test paper combined by the proposition module 12 is displayed on the display unit 15, and the user can answer the test paper based on the test unit 15 via the input unit 16; and the evaluation module 14 passes the user through the test unit. The input unit 16 and the display unit 15 accept (perform) the scores of the at least one test paper and compare the scores of the set of virtual subjects, and compare the results. The display unit 15 displays that the user can actually evaluate the distribution and ranking of his or her performance, and can further perform predictive analysis of possible schools or school groups.

Referring to the third figure, it is another system block diagram of the present invention, as shown in the figure: since the simulation test result module 13 obtains an answer to each test question in the test paper in a random (random number) and non-repeating manner, To combine the test papers for each of the virtual subjects. Since the answers are assigned in a random number method, there is a case where the micro-probability has an extreme score distribution, rather than a more normal score distribution. For example, there may be an extreme score distribution of many people's perfect scores or zero points. Therefore, the test result simulation evaluation system 10 of the present invention further includes a verification module 17 for verifying whether the distribution of scores (scores) of all virtual subjects is normal. The verification method is to adopt the concept of statistics. The method is summarized as follows:

1. Define a number of first eigenvalues according to the characteristics of the target test (for example, the test, define the top mark, the front mark, the average mark, the back mark, the bottom mark, etc. according to the definition of the level).

2. Combine a test question into the same group as a random number method (a set of virtual test subject scores, the number of virtual test subjects in this group is determined by the user).

3. Repeat step 2 multiple times to generate a complex sample set.

4. Calculate a number of second eigenvalues of the plurality of sample groups (eg, the test, define the top mark, the front mark, the average mark, the back mark, the bottom mark, etc. according to the definition of the level).

5. Select a sample group corresponding to the second feature value closest to the first feature value, and use the sample group as a control group to virtualize the test subject score.

For example, if 10,000 sample groups are created, the second eigenvalues (top, front, average, post, and bottom) of each sample are taken out, and the sample closest to the first eigenvalue is taken. The group serves as one of the controls for the group of virtual subjects. The first feature value may be obtained from the previous test results of the target test, and the average value of the first feature value obtained by the previous test may also be used as a reference.

The above-mentioned simulation test is based on the joint enrollment examination. The system of the present invention is not limited to the joint enrollment examination, and is applicable to examinations of general schools, cram schools, government agencies, and international organizations, and the test question bank module 11 is The more the collected questions and the correct answer response statistics, the greater the benefit to the accuracy of the present invention.

Referring to the fourth figure, which is a flowchart of the implementation of the present invention, as shown in the figure: the method for simulating the evaluation result of the test result of the present invention comprises the following steps: collecting a test question group having the answer reaction statistics and the correct answer of the test question group ( In step S10), at least one test paper is combined according to the test question group (step S11). The test group with the answer response statistics uses the previous answer statistics (proportion) for each test question as the parameters. The proposition module combines at least one test paper according to the test question bank module. Since the test question bank module can classify or set parameters according to the test type, difficulty level or range, the proposition module can combine various test questions and difficulty test papers according to requirements, and the proposition module test paper proposition method is According to the test question bank module, the test questions are set by manual selection or computer, and according to the principle of proportionality (the test questions are evenly distributed in various ranges) or the proposition method according to the standardized test theory (standardized test theory is the formulation of the ability indicators of various disciplines). , the design of the test questions, the establishment of the test procedures, the scoring method and even the standard test procedures such as test pretests and the establishment of test reliability and validity to test the test papers so that the test can have the best measurement results. . Then, the number of answers of each question of the at least one test paper is proportionally enlarged according to the answer reaction statistics to be equal to the number of virtual subjects (step S12), and each virtual subject is obtained in a random and non-repeating manner. An answer to each test question in the test paper to combine one of the virtual subject's group answers (step S13), and then compare each virtual subject's answer to the correct answer, and calculate each virtual test The score of the person (step S14); finally, the score of the user after receiving the test of the test paper is compared with the score of each virtual subject (step S15), thereby evaluating the learning result of the user.

In step 12, the step of inputting the number of virtual subjects is further included (step S121), and the answer response statistics is multiplied by the number of virtual subjects, and the number of answers to each question of the test paper is obtained.

In many test questions, there is a degree of difficulty. Therefore, when the score distribution of the user's test directly corresponds to the score distribution of the previous joint test, there will be some drop in the score. The present invention provides a system and method for the simulation evaluation of test results, so as to achieve a more accurate assessment of the candidate's true score ranking effect.

Although the present invention has been disclosed in the above embodiments, it is not intended to limit the invention, and any one of ordinary skill in the art without departing from the invention In the spirit and scope, the scope of protection of the present invention is defined by the scope of the appended claims.

10‧‧‧Test results simulation evaluation system

11‧‧‧Test question bank module

12‧‧‧Proposition Module

13‧‧‧Analog test result module

14‧‧‧Evaluation module

15‧‧‧Display unit

16‧‧‧Input unit

17‧‧‧ verification module

S10‧‧·Collect a test group with answer response statistics and the correct answer to the test group

S11‧‧‧ Combine at least one test paper according to the test group

S12‧‧‧Enlarge the number of answers to each question in the at least one test paper according to the response statistics of the questions, and enlarge the number to the number of subjects

S121‧‧‧Enter the number of virtual subjects

S13‧‧‧ Each virtual subject obtains an answer to each question in the test paper in a random and non-repeating manner to combine the answers of one of each virtual subject

S14‧‧‧ Calculate the scores of each virtual subject after comparing the answers of each virtual subject to the correct answer

S15‧‧‧Comparison of the results of a user's acceptance of the test paper with the scores of each virtual subject

The first figure is a block diagram of the system of the present invention.

The second figure is another system block diagram of the present invention.

The third figure is another system block diagram of the present invention.

The fourth figure is a flow chart of the implementation of the present invention.

10‧‧‧Test results simulation evaluation system

11‧‧‧Test question bank module

12‧‧‧Proposition Module

13‧‧‧Analog test result module

14‧‧‧Evaluation module

Claims (16)

A system for testing test results includes: a test question bank module, which collects a test answer group with a question response statistic and a correct answer of the test question group; a proposition module is assembled according to the test question bank module At least one test paper; and a simulation test result module, according to the answer number statistics of each test question in the at least one test paper, the number of answers is proportionally enlarged to be equal to the number of virtual test subjects, each virtual test Obtaining an answer to each test question in the at least one test paper in a random and non-repeating manner to combine one of the virtual test subject's test paper answers, and comparing the correct answer, calculating each virtual test subject's Grades, and combinations get a set of virtual subjects' scores. For example, the system for testing the test results described in the first paragraph of the patent application, wherein the number of virtual subjects in the group is at least two or more. For example, the system for testing the test results described in claim 1 of the patent scope, wherein the scores of the virtual subjects can be further ranked according to the scores, and a test result distribution is generated. For example, the system for testing the test results described in claim 1 of the patent scope, wherein the test questions having the answer response statistics may be a single multiple choice question, multiple multiple choice questions, multiple choice questions, yes and no questions, filled questions, and selective fill questions. Or a short answer question. For example, the system for testing the test result described in claim 1 of the patent scope, wherein the test question bank module can classify or set parameters according to the type of test questions, difficulty level, range of questions, proposition origin, test target or previous distribution rate. A system for simulating test results as described in claim 1, wherein the proposition module performs a combination of the at least one test paper with reference to a standard test paper design of a test type selected by a user or a test design of a previous test. . A system for simulating evaluation of test results includes: a test question bank module, which collects a test question group having a response to a question and a correct answer of the test question group; a proposition module is combined according to the test question bank module At least one test paper; A simulation test result module is to count the number of answers of each test question in the at least one test paper according to the answer response statistics, and enlarge the number of answers to be equal to the number of virtual test subjects, and each virtual test subject is random and non-repeated The method obtains an answer of each test question in the at least one test paper, and combines one of the test questions of each virtual test subject, and compares the correct answer, calculates the score of each virtual test subject, and obtains one combination in combination. The group of virtual subject scores; and an evaluation module that compares the scores of a user after receiving the at least one test paper with the scores of the set of virtual subjects. For example, the system for simulating and evaluating the test results described in claim 7 of the patent scope, wherein the scores of the virtual test subjects can be further ranked according to the scores, and a test result distribution is generated. For example, the system for simulating evaluation of test results described in claim 7 of the patent scope, wherein the test questions having the answer response statistics may be single multiple choice questions, multiple multiple choice questions, multiple choice questions, yes and no questions, filled questions, and selective fills. Question or short answer question. For example, the system for simulating and evaluating the test results described in claim 7 of the patent scope, wherein the test question bank module can classify or set parameters according to the type of the test question, the difficulty level, the range of the question, the source of the test, the test target or the previous matching rate. . A system for simulating evaluation of test results includes: a test question bank module, which collects a test question group having a response to a question and a correct answer of the test question group; a proposition module is combined according to the test question bank module At least one test paper; a simulation test result module, the number of answers of each test question in the at least one test paper is based on the answer response statistics, and the number of answers is proportionally enlarged to be equal to the number of virtual test subjects input by one input unit. Each virtual subject obtains an answer of each test question in the at least one test paper in a random and non-repeating manner to combine one of the virtual test subject's test paper answers, and compares the correct answer to calculate each a score of the virtual subject and a combination of obtaining a set of virtual subject scores; and an evaluation module for accepting the score of the at least one test paper and the group by the user via the input unit and a display unit Virtual subject score Right, and the comparison result is displayed on the display unit. For example, the system for simulating and evaluating the test results described in claim 11 of the patent scope, wherein the scores of the virtual test subjects can be further ranked according to the scores, and a test result distribution is generated. For example, the system for simulating and evaluating the test results described in claim 11 of the patent scope, wherein the test questions having the answer response statistics may be a single multiple choice question, multiple multiple choice questions, multiple choice questions, yes and no questions, filled questions, and selective fills. Question or short answer question. For example, the system for simulating and evaluating the test results described in claim 11 of the patent scope, wherein the test question bank module can be classified or set according to the test type, difficulty level or range. A method for simulating evaluation of a test result comprises the steps of: collecting a correct answer to a test group having a response to a question and a test question group; and combining at least one test paper according to the test question group; answering each question of the at least one test paper The quantity is proportionally enlarged according to the answer response statistics to be equal to the number of virtual subjects; each virtual subject obtains an answer to each question in the test paper in a random and non-repeating manner to combine each virtual subject. One set of answers of the tester; after comparing the answers of each virtual subject to the correct answer, calculate the score of each virtual subject; and the score of each user after accepting the test of the test paper and each virtual subject The scores of the testers are compared to evaluate the learning outcomes of the user. The method for simulating the test result as described in claim 15 of the patent application further includes the step of inputting the number of the virtual subject.