A method for evaluating computational thinking based on multi-dimensional item response theory

Abstract

The method for evaluating the computational thinking of the subject based on the multidimensional item response theory in the application can analyze and process the answer data based on the multidimensional item response theory, calculate the computational thinking level of the subject, and display the computational thinking ability evaluation result of the subject under the condition of different m-dimensional ability vectors according to a preset form. The method can evaluate the ability level of the individual in the computational thinking, help the individual find the advantages and disadvantages of the individual in the computational thinking, and then formulate appropriate learning and improvement strategies for the computational thinking ability.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of computer data intelligent processing, and particularly relates to a computing thinking evaluation method based on a multidimensional item response theory. BACKGROUND

[0002] Computational thinking (CT) is a problem-solving process, which is a habit and mode of thinking formed by people based on computer science, and includes many characteristics, such as logical sequencing and analyzing data, using a series of ordered steps (or algorithms) to create solutions and handling strategies, such as confidently dealing with complexity and open-ended problems. CT is essential for the development of computer applications, but it can also be used to support problem solving in all disciplines, including mathematics, science and humanities. The cultivation of computational thinking cannot be separated from the research on the evaluation of computational thinking. Since computational thinking is implicit, it cannot be directly measured by related tools like height, weight and temperature, therefore, how to evaluate computational thinking, how to judge whether students have relevant computational thinking literacy and the degree they have reached, and from which aspects and indicators to measure the computational thinking ability of students have become the premise and urgent needs of the research on the cultivation of computational thinking.

[0003] There are various ways to evaluate and approach computational thinking, such as observing the specific performance of students in diversified situations, analyzing and evaluating through student works, and investigating the computational thinking of students through questionnaires or test questions.

[0004] The current process of establishing an evaluation index system of computational thinking is generally to first establish an index system, usually according to the opinions of experts, to select a part of indexes as first-level indexes, and then to further decompose each first-level index into second-level indexes, and then from second-level indexes to third-level indexes, and so on. The second step is to determine the weight of each index.

[0005] The evaluation scheme should not only have a high representativeness, but also be easy to understand and use, and have strong operability. However, in the evaluation index system of computational thinking, the total number of indexes at all levels may be large, and the differences between indexes may be large, resulting in great difficulty in the final selection of indexes.

[0006] Therefore, it is necessary to study an efficient and accurate computational thinking evaluation method. SUMMARY

[0007] The main purpose of the present application is to provide a computational thinking evaluation method based on a multidimensional item response theory, so as to solve the above technical problems.

[0008] To achieve the above purpose, the present application provides a computational thinking evaluation method based on a multidimensional item response theory, comprising the steps of,

[0009] S1, a step of designing an evaluation task, specifically comprising:

[0010] S11, assigning an evaluation task including W questions to N known training testers for testing and obtaining the answering data of the known training testers; wherein the evaluation task includes multiple preset evaluation task question types, and each question type corresponds to a score;

[0011] S12, constructing a 3PL model of the following formula according to the preset multiple ability dimensions of computational thinking and the answering data of the known training testers,

[0012]

[0013] wherein P(X ij | θ i ) represents the probability of the i-th known training tester correctly answering the j-th question; X ij represents the reaction of the i-th known training tester to the j-th question, if the answer is correct, x ij = 1, otherwise x ij = 0; the answering data of the known training tester is X = (x ij ) N×W ; θ i = (θ i1 , θ i2 , … θ im )' represents the m-dimensional ability vector of the i-th known training tester; a j = (a j1 , a j2 , … a jm )' represents the m-dimensional discrimination vector of the j-th question; d j is a question difficulty parameter representing the j-th question; c j is a lower asymptote parameter representing the j-th question; and D is a model adjustment parameter.

[0014] S13, using a quasi-Monte Carlo expectation maximization algorithm on the 3PL model to estimate parameters and obtain the item parameters δ j = (a j , d j , c j ) corresponding to each question;

[0015] S14, combining the W questions according to a j and d j to obtain multiple sets of test papers with discrimination and difficulty meeting preset conditions;

[0016] S2, obtaining the test paper test results of the subjects, and according to the item parameters δ j= (a j ,d j ,c j ) and the 3PL model in step S12, to obtain the m-dimensional ability vector of the subject.

[0017] Preferably, the types of the questions include single-choice questions, one-point questions, fill-in-the-blank questions, one-point questions, and true-or-false questions, one-point questions; the difficulty levels of the test papers include four levels of entry, advanced, master, and wisdom, and the proportion of difficult questions with a question difficulty parameter greater than a preset value in the questions of the test papers of the four levels is 10%, 20%, 30%, and 40%, respectively; and the number of questions in each test paper is between 20 and 25.

[0018] Preferably, the step S13 comprises:

[0019] S131, estimating initial values of the item parameters δ ij ) N×W of the 3PL model according to the answer data X = (x j ) j ,d j ,c j of the known training testers;

[0020] S132, according to the formula P(θ = μ k |Λ) = P(μ k |λ k ) = λ k and the initial values, recording the m-dimensional ability vectors of all the known training testers as θ = (θ1, …, θ N ), approximating the m-dimensional ability vectors θ of the known training testers in the 3PL model as discrete latent variables, representing the continuous latent variable θ as k known discrete values μ1, …, μ k , and corresponding unknown probabilities λ k , and using Λ = (λ1, λ2, λ3, …, λ k ) to represent the distribution parameters of the latent variables; then estimating the latent variable values of the missing data, and re-estimating the item parameters δ j = (a j ,d j ,c j ) according to the latent variable values and the initial values to obtain iterative values;

[0021] S133, taking the iterative values as new initial values and returning to the step S131 for the next iteration;

[0022] S134, determining whether the iteration reaches a preset condition, and if yes, ending; or if no, continuing the iteration.

[0023] Preferably, the step S132 comprises:

[0024] E step: calculate the conditional probability expectation of the joint distribution, as follows:

[0025]

[0026]

[0027] denotes the sum of the conditional probabilities of all student ability values being μ k denotes the sum of the conditional probabilities of all student ability values being μ k and answering the question, t is the iteration number;

[0028] M step: update the parameter values according to the expected latent ability value using the maximum likelihood estimation method, find the parameter values that maximize the likelihood function using a numerical optimization algorithm, as follows:

[0029]

[0030] Preferably, wherein,

[0031]

[0032]

[0033]

[0034] Preferably, the preset conditions in the step S134 include:

[0035] whether the iteration number reaches a preset iteration number;

[0036] and / or,

[0037] the difference between the log-likelihood function values of two adjacent iterations is less than a set threshold.

[0038] Preferably, the step S134 further comprises a step S135 of model fitting inspection: performing a fitting inspection on the evaluation model to inspect whether the degree of model fitting data is reasonable.

[0039] 8. The multi-dimensional item response theory-based computational thinking evaluation method according to claim 7, wherein the model fitting inspection comprises χ2 inspection, RMSEA, CFI or SRMR.

[0040] Preferably, the d j satisfies the formula for calculating the multi-dimensional difficulty coefficient: wherein, ​MDISC j is a multidimensional discrimination index, MDIFF j is the multidimensional difficulty coefficient of the jth question, MDIFF j represents the transformed multidimensional difficulty coefficient, and f is the scoring standard, f = 1.

[0041] Preferably, the preset multi-capability dimension of computational thinking includes a plurality of first-level dimension indexes, the first-level dimension indexes include a concept knowledge index, a problem inquiry index, and an algorithm thinking index; the concept knowledge index includes two second-level dimension indexes of a definitional knowledge index and an operational knowledge index; the algorithm thinking index includes two second-level indexes of an algorithm understanding index and an algorithm design index; and the problem inquiry index includes three second-level dimension indexes of an abstraction index, a decomposition index, and a transfer index.

[0042] The method for evaluating computational thinking based on the multidimensional item response theory in the application analyzes and processes the answering data of the testees based on the multidimensional item response theory, calculates the computational thinking level of the testees, and, after the questions of each set of test papers are designed, the answering data of the participants are processed through the 3PL model of the multidimensional item response theory to obtain the parameter information of the test questions through the quasi-Monte Carlo expectation maximization algorithm (QMCEM), the most suitable questions are selected as the questions of the test papers of different levels of difficulty through data index analysis and selection of the better and the worse, and the answering data of the testees are brought into the 3-parameter Logistic model of the multidimensional item response theory, so that the ability parameters of the students can be obtained through a few iterations because the test question parameters have been estimated, that is, the computational thinking level of the students is obtained. The method can evaluate the ability level of the individual in the computational thinking, help the individual find the advantages and disadvantages of the individual in the computational thinking, and then develop appropriate learning and improvement strategies for the computational thinking ability. BRIEF DESCRIPTION OF DRAWINGS

[0043] The accompanying drawings, which are part of the present application, serve to further understand the present application, and the schematic embodiments of the present application and the descriptions thereof serve to explain the present application but do not constitute an improper limitation on the present application. Obviously, the accompanying drawings in the following description are only some embodiments, and other drawings can be obtained by those skilled in the art without any creative effort. In the drawings:

[0044] Figure 1 FIG. 1 is a flowchart of the step of designing the evaluation task of the method for evaluating computational thinking based on the multidimensional item response theory in an embodiment of the present application.

[0045] Figure 2 FIG. 2 is a subflowchart of the step S13 of designing the evaluation task of the method for evaluating computational thinking based on the multidimensional item response theory in an embodiment of the present application.

[0046] Figure 3 Fig. 1 is a schematic diagram of a hardware structure for implementing a method for evaluating computational thinking based on a multi-dimensional item response theory according to an embodiment of the present application.

[0047] The implementation, functional features and advantages of the present application will be further described with reference to the embodiments and the accompanying drawings. DETAILED DESCRIPTION

[0048] The technical problems solved by the embodiments of the present application, the technical solutions adopted and the technical effects achieved will be described clearly and completely below with reference to the accompanying drawings and specific embodiments. Obviously, the described embodiments are only some of the embodiments of the present application, but not all the embodiments. Based on the embodiments in the present application, all other equivalent or modified embodiments obtained by those skilled in the art without creative effort fall within the protection scope of the present application. The embodiments of the present application can be embodied in various different ways as defined and covered by the claims.

[0049] It should be noted that in the following description, many specific details are given to facilitate understanding. However, it is obvious that the implementation of the present application can be realized without these specific details.

[0050] It should be noted that in the absence of explicit limitations or conflicts, the various embodiments of the present application and the technical features therein can be combined with each other to form technical solutions.

[0051] Please refer to the accompanying drawings and the specific embodiments Figures 1-2 The present application provides a method for evaluating computational thinking based on a multi-dimensional item response theory, comprising the steps of,

[0052] S1, a step of designing an evaluation task;

[0053] S2, obtaining a test result of a test paper of a subject to obtain an m-dimensional ability vector of the subject.

[0054] Please refer to the accompanying drawings and the specific embodiments Figure 1 Step S1 specifically comprises S11-S14.

[0055] S11, assigning an evaluation task comprising W items to N known training testers for testing and obtaining the answer data of the known training testers; wherein the evaluation task comprises a plurality of preset evaluation task types, and each type of question corresponds to a score.

[0056] Preferably, the types of the questions include single-choice questions, one-point questions, fill-in-the-blank questions, one-point questions, and true-or-false questions, one-point questions; the difficulty levels of the test papers include entry-level, advanced level, master level, and wisdom level, and the proportion of difficult questions with a difficulty parameter greater than a preset value in the questions of the entry-level, advanced level, master level, and wisdom level test papers is 10%, 20%, 30%, and 40%, respectively; and the number of questions in each test paper is between 20 and 25. The discrimination parameter ensures that each test paper has all the abilities of computational thinking (i.e., the secondary index). For example, when the difficulty coefficient is greater than 1.8, the question is considered too difficult, and when the difficulty coefficient is less than -1.8, the question is considered too easy. The proportion of difficult questions in the entry-level, advanced level, master level, and wisdom level test papers is 10%, 20%, 30%, and 40%, respectively.

[0057] S12, according to the preset multi-ability dimensions of computational thinking and the answer data of the known training test takers, a 3PL model of the following formula is constructed,

[0058]

[0059] wherein P(X ij |θ i ) represents the probability that the ith known training test taker correctly answers the jth question; X ij represents the reaction of the ith known training test taker to the jth question, if the answer is correct, x ij = 1, otherwise x ij = 0; the answer data of the known training test taker is X = (x ij ) N×W ; θ i = (θ i1 , θ i2 , … θ im )' represents the m-dimensional ability vector of the ith known training test taker; a j = (a j1 , a j2 , … a jm )' represents the m-dimensional discrimination vector of the jth question; d j is the difficulty parameter of the jth question; c j is the lower asymptote parameter of the jth question; and D is the model adjustment parameter (generally 1.7).

[0060] Preferably, the d j satisfies the formula for calculating the multi-dimensional difficulty coefficient: wherein, MDISC j is the multi-dimensional discrimination index, MDIFF j is the multi-dimensional difficulty coefficient of the jth question, and MDIFF jThe transformed multi-dimensional difficulty coefficient is represented by f, and f = 1.

[0061] The purpose of the computational thinking ability evaluation is to assess the individual's ability level in computational thinking, help individuals discover their strengths and weaknesses in computational thinking, and then develop appropriate learning and improvement strategies for computational thinking ability.

[0062] The composition of computational thinking elements is generally analyzed by using word frequency statistics method. First, collect text data related to computational thinking, then clean the text to remove useless symbols and stop words, etc. Second, divide the text into individual words. Then count the number of times each word appears in the text to get a word frequency table. Finally, sort the word frequency table according to the frequency, determine the composition of computational thinking according to the frequency, classify the computational thinking elements with similar examination types into one category, and replace them with more accurate words, as shown in Table 1 below:

[0063] Table 1 Computational thinking evaluation index system

[0064]

[0065] According to the characteristics of the first-level indicators in the above computational thinking ability evaluation index system, computational thinking ability evaluation data is generally collected through three ways to achieve multi-source data. Select the type and difficulty of the evaluation task: The types of computational thinking evaluation tasks include multiple-choice questions, fill-in-the-blank questions, and true-or-false questions. The difficulty levels are very easy, easy, average, difficult, and very difficult. The corresponding question quantity ratio can be 2:3:3:1:1.

[0066] (1) Conceptual knowledge has the characteristics of basic knowledge, so the method of test evaluation is used to evaluate students' mastery of definitional knowledge and operational knowledge. This is one of the sources of evaluation data. A question bank is developed to generate a set of random questions that include definitional knowledge and operational knowledge questions, which are used to collect learners' scores on definitional knowledge and operational knowledge.

[0067] (2) Problem exploration and algorithmic thinking have practical properties, so the method of programming questions is used to collect evaluation data. By analyzing the program, the ability level of learners' problem exploration and algorithmic thinking in the second-level indicators is obtained, which is the second source of evaluation data. Specifically, a mapping relationship between programming and computational thinking is established, and based on the mapping relationship, the programming characteristics of learners are mined to obtain learners' scores in abstraction and decomposition, module definition, module design, logic, control, and construction.

[0068] Preferably, the computational thinking ability index includes multiple primary dimension indicators, which include conceptual knowledge indicators, problem exploration indicators, and algorithmic thinking indicators.

[0069] Preferably, the conceptual knowledge indicators include two secondary dimensions: definitional knowledge indicators and operational knowledge indicators; the algorithmic thinking indicators include two secondary dimensions: algorithm understanding indicators and algorithm design indicators; and the problem exploration indicators include three secondary dimensions: abstraction indicators, decomposition indicators, and transfer indicators.

[0070] S13, use the quasi-Monte Carlo expectation-maximization algorithm on the 3PL model to estimate parameters and obtain the item parameter δ for each question. j =(a j ,d j ,c j ).

[0071] For preferred options, please refer to [the following]. Figure 2 Step S13 may include S131 to S135.

[0072] S131, based on the known test subject's answer data X = (x ij ) N×W Estimate the project parameter δ of the 3PL model. j =(a j ,d j ,c j The initial value of ).

[0073] Specifically, this can include the difficulty parameter d of the initial test questions. j Discrimination parameter a j and the guessing parameter c j The estimated value of the ability distribution parameter γ of the students is initialized; the estimated value of the initialization parameter γ of the students is set; the number of iterations is set to t = 0.

[0074] S132, according to the formula P(θ=μ) k |Λ)=P(μ k |λ k )=λ k And the initial value, denoted as θ=(θ1,……,θ) for all the known m-dimensional ability vectors of the trainees. N The m-dimensional ability vector θ of the known trainees in the 3PL model is approximated as a discrete latent variable, and the continuous latent variable θ is represented as k known discrete values ​​μ1, ..., μ k And the corresponding unknown probability is represented by λ. k , use Λ=(λ1, λ2, λ3,...,λ k) represents the distribution parameter of the latent variable; then the latent variable value of the missing data is estimated, and the item parameter δ j = (a j ,d j ,c j ) is estimated to obtain the iterative value.

[0075] Preferably, the step S132 comprises:

[0076] E step: the conditional probability expectation of the joint distribution is calculated, and the formula is as follows:

[0077]

[0078]

[0079] The sum of the conditional probabilities of all student ability values being μ k , The sum of the conditional probabilities of all student ability values being μ k and answering the question, t is the iteration number;

[0080] M step: according to the expected latent ability value, the parameter value is updated using the maximum likelihood estimation method, and a numerical optimization algorithm is used to find the parameter value that maximizes the likelihood function, and the formula is as follows:

[0081]

[0082] The calculation process of the formula in the M step is as follows:

[0083]

[0084]

[0085] Wherein, t is the iteration number.

[0086] Since the item parameter and the student ability distribution parameter are independent of each other, they do not affect each other in the process of maximum likelihood estimation, so the above formula can be split into two parts and maximum likelihood estimation is performed respectively.

[0087] δ (t+1) = argmax φ (δ) + φ (Λ)

[0088] Calculate φ (δ) :

[0089]

[0090] Calculate φ (Λ) :

[0091]

[0092] δ(t+1) δ (t+1) = argmax φ(δ) + φ(Λ) which is equivalent to:

[0093]

[0094] Preferably, wherein,

[0095]

[0096]

[0097]

[0098] S133, taking the iteration value as a new initial value and returning to the step S131 for the next iteration;

[0099] S134, judging whether the iteration reaches a preset condition, if yes, ending; if no, continuing the iteration. Preferably, the preset condition in the step S134 includes: whether the iteration number reaches a preset iteration number; and / or, the difference between the log-likelihood function values of two adjacent iterations is less than a set threshold.

[0100] Preferably, the step S134 further includes a step S135 of model fitting test: performing a fitting test on the evaluation model to test whether the degree of model fitting data is reasonable. The model fitting test includes χ2test, RMSEA, CFI or SRMR.

[0101] The estimated model is subjected to a fitting test to test whether the degree of model fitting data is reasonable. Commonly used test indexes include χ2test, RMSEA, CFI, SRMR, etc. χ2tests the degree of model fitting by comparing the difference between the observed data and the estimated value of the model. The smaller the χ2value is, the better it is, but it is usually impossible to completely deviate from the observed data; RMSEA is an index for measuring the degree of model fitting according to a fitting function. For a model with a good fitting degree, the RMSEA value should be close to 0. Generally, when the RMSEA value is less than 0.05, the fitting degree of the model is considered to be good; CFI is the abbreviation of Comparative Fit Index. It compares the fitting degrees of the hypothetical model and the independent model. The closer the value is to 1, the better the fitting degree of the model is. Generally, when the CFI value is greater than 0.95, the fitting degree is considered to be good; SRMR is the root mean square of the standardized average residual. It is used to measure the fitting degree of the model. The smaller the SRMR is, the better the fitting degree is. Generally, when the SRMR value is less than 0.08, the fitting degree is considered to be good.

[0102] S14, combining the W items to obtain a plurality of test papers with preset conditions of discrimination and difficulty. j and d j Combining the W items to obtain a plurality of test papers with preset conditions of discrimination and difficulty.

[0103] S2, obtaining a test paper test result of a subject, and obtaining an m-dimensional ability vector of the subject according to the item parameter δ j = (a j , d j , c j ) obtained in step S13 and the 3PL model in step S12. The test paper is one of the plurality of test papers obtained in step S14.

[0104] In step S2, since the test item design module has basically estimated the parameter size of the item, the subject's answer data can be put into the model to further estimate the subject's ability value θ size. We can use Bayesian posterior expectation estimation, without iteration. The formula is as follows:

[0105] Discretization (numerical score calculation)

[0106]

[0107] Therefore, the ability estimate value of the subject can be obtained.

[0108] In step S2, when the result is output, the result format can output the test results in the form of text, graphics, etc., for the user to view and understand. For example, for the ability parameters in the multidimensional item response theory model, they can be output in the form of tables or charts to show the scores of different abilities. In the result explanation, in addition to the results themselves, users need to be provided with corresponding explanations and instructions to help them better understand the meaning of the results. For example, for the score of a certain ability, a corresponding explanation can be given to tell the user the meaning of the ability and the corresponding test tasks, etc. In the result analysis, the results can be further analyzed by data analysis methods to mine the rules and trends therein. For example, for the trend of the score of a certain ability, time series analysis and other methods can be used for analysis to give corresponding conclusions and suggestions. When interacting with the user, the result output module can be interacted with other modules to provide better user experience. For example, the user can obtain more detailed result information or perform specific data processing and analysis through interaction with the result output module.

[0109] The calculation thinking evaluation method based on the multidimensional item response theory in the application analyzes and processes the answering data based on the multidimensional item response theory, calculates the calculation thinking level of the testee, after the questions of each set of test papers are designed and a large number of testers participate, the answering data of the testers is subjected to the 3PL model of the multidimensional item response theory to obtain the parameter information of the test questions through the quasi-Monte Carlo expectation maximization algorithm (QMCEM), the most suitable questions are selected as the questions of the test papers of different levels of difficulty through data index analysis and selection of the best and the worst, and the answering data of the testers is brought into the 3-parameter Logistic model of the multidimensional item response theory, because the test question parameters have been estimated, the ability parameters of the students can be obtained through a few iterations, that is, the calculation thinking level of the students is obtained. The ability level of an individual in the calculation thinking can be evaluated, the individual can find the advantages and disadvantages of the individual in the calculation thinking, and then appropriate learning and calculation thinking ability improving strategies are formulated.

[0110] Figure 3 It is a hardware structure schematic diagram for running the calculation thinking evaluation method based on the multidimensional item response theory provided by an embodiment of the application. As shown in the figure, Figure 3 the embodiment / computer 6 includes a processor 60, a memory 61, and a computer program 62 stored in the memory 61 and executable on the processor 60, for example, a program for running the calculation thinking evaluation method based on the BP neural network and multi-source data fusion. The processor 60 implements the steps in each of the above-mentioned embodiments of the calculation thinking evaluation method based on the BP neural network and multi-source data fusion when executing the computer program 62. Alternatively, the processor 60 implements the functions of each module / unit in each of the above-mentioned device embodiments when executing the computer program 62.

[0111] For example, the computer program 62 can be divided into one or more modules / units, which are stored in the memory 61 and executed by the processor 60 to complete the application. The one or more modules / units can be a series of computer program instruction segments capable of completing a specific function, which are used to describe the execution process of the computer program 62 in the computer 6.

[0112] The computer 6 can be a desktop computer, a notebook computer, a palm computer, a cloud server and other computing devices. The computer 6 device can include, but is not limited to, a processor 60, a memory 61. Those skilled in the art can understand, Figure 3 The computer 6 is only an example and does not constitute a limitation on the computer 6, and can include more or fewer components than the illustration, or combine certain components, or different components, for example, the computer 6 can also include an input / output device, a network access device, a bus, etc.

[0113] The processor 60 can be a central processing unit (CPU), and can also be other general-purpose processors, a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic device, discrete gate or transistor logic device, discrete hardware components, etc. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor.

[0114] The memory 61 can be an internal storage unit of the computer 6, such as a hard disk or a memory of the computer 6. The memory 61 can also be an external storage device of the computer 6, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. Further, the memory 61 can include both the internal storage unit and the external storage device of the computer 6. The memory 61 is used to store the computer program and other programs and data required by the terminal device. The memory 61 can also be used to temporarily store data that has been output or will be output.

[0115] Those skilled in the art can clearly understand that, for the convenience and brevity of description, only the division of the above functional units and modules is exemplified, and in actual application, the above functions can be completed by different functional units and modules according to needs, that is, the internal structure of the apparatus is divided into different functional units or modules to complete all or part of the above described functions. Each functional unit and module in the embodiment can be integrated in one processing unit, or each unit can exist physically, or two or more units can be integrated in one unit, and the integrated unit can be realized in the form of hardware or in the form of software. In addition, the specific names of each functional unit and module are only for easy distinction, and do not limit the protection scope of the present application. The specific working process of the units and modules in the system can refer to the corresponding process in the foregoing method embodiments, which will not be described here.

[0116] In the above embodiments, the description of each embodiment has its own emphasis, and the parts not described or recorded in detail in a certain embodiment can refer to the relevant description of other embodiments.

[0117] Those skilled in the art can appreciate that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be realized in electronic hardware, or in a combination of computer software and electronic hardware. Whether the functions are realized in hardware or software depends on the specific application and design constraints of the technical solution. A person skilled in the art can use different methods to realize the described functions for each specific application, but such implementation should not be considered beyond the scope of the present application.

[0118] In the embodiments provided by the present application, it should be understood that the disclosed apparatus / terminal device and method can be implemented in other ways. For example, the apparatus / terminal device embodiments described above are merely schematic, for example, the division of the modules or units is merely a logical function division, and there can be another division manner in actual implementation, for example, a plurality of units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the displayed or discussed mutual coupling or direct coupling or communication connection between interfaces, devices or units can be indirect coupling or communication connection through some interfaces, devices or units, and can be electrical, mechanical or other forms.

[0119] The units described as separate components can or can not be physically separate, and the components shown as units can or can not be physical units, that is, they can be located in one place, or can be distributed on a plurality of network units. Part or all of the units can be selected according to actual needs to achieve the purpose of the embodiments.

[0120] In addition, each functional unit in each embodiment of the present application can be integrated in one processing unit, or each unit can be physically present separately, or two or more units can be integrated in one unit. The integrated unit can be realized in the form of hardware or in the form of a software functional unit.

[0121] The integrated module / unit, if implemented in the form of a software function unit and sold or used as an independent product, can be stored in a computer readable storage medium. Based on such understanding, all or part of the processes in the above-mentioned embodiment methods can also be completed by a computer program instructing related hardware, and the computer program can be stored in a computer readable storage medium. The computer program can implement the steps of each method embodiment when executed by a processor. The computer program includes computer program code, which can be in the form of source code, object code, executable files or some intermediate forms. The computer readable medium can include any entity or device capable of carrying the computer program code, recording medium, U disk, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal and software distribution medium, etc. It should be noted that the content included in the computer readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction, for example, in some jurisdictions, according to legislation and patent practice, the computer readable medium does not include electrical carrier signals and telecommunication signals.

[0122] The above-described embodiments are only used to illustrate the technical solutions of the present application, rather than limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: it can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement for part of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application, and should be included in the protection scope of the present application.

Claims

1. A method for measuring computational thinking based on multidimensional item response theory, characterized in that, The method comprises the steps of: S1, a step of evaluating a task design, specifically comprising: S11, assigning an evaluation task including W questions to N known trainees for testing and obtaining the trainees' answer data; wherein the evaluation task includes multiple preset evaluation task question types, each type corresponding to a score; S12, constructing a 3PL model of the following formula according to a preset multi-ability dimension of computational thinking and the trainees' answer data; , i e [1, N], j e [1, W]; wherein, represents the probability that the ith known training testee correctly answers the jth question; represents the reaction of the ith known training testee to the jth question, if correct, then , otherwise ; the answer data of the known training testee is X= represents the m-dimensional ability vector of the ith known training testee; represents the m-dimensional discrimination vector of the jth question; is a question difficulty parameter representing the jth question; is a lower asymptote parameter representing the jth question; D is a model adjustment parameter; the preset computational thinking multi-ability dimension includes a plurality of first-level dimension indicators, the first-level dimension indicators include a conceptual knowledge indicator, a problem exploration indicator, and an algorithmic thinking indicator; the conceptual knowledge indicator includes two second-level dimension indicators of a definitional knowledge indicator and an operational knowledge indicator; the algorithmic thinking indicator includes two second-level indicators of an algorithmic understanding indicator and an algorithmic design indicator; the problem exploration indicator includes three second-level dimension indicators of an abstraction indicator, a decomposition indicator, and a transfer indicator; S13, parameter estimation is performed on the 3PL model using a quasi-Monte Carlo expectation-maximization algorithm to obtain item parameters corresponding to each question ; the step S13 comprises: S131, estimating initial values of item parameters of the 3PL model according to the answer data X of the known training testees; ;​ S132, according to the formula and the initial value, record the m-dimensional ability vector of all the known training testers as approximate the m-dimensional ability vector θ of the known training testers in the 3PL model as discrete latent variables, and express the continuous latent variable θ as k known discrete values , , and the corresponding unknown probability is expressed as , the distribution parameter of the latent variable is expressed as ; then estimate the latent variable value of the missing data, and re-estimate the item parameter according to the latent variable value and the initial value, to obtain the iterative value; S133, taking the iteration value as a new initial value and returning to the step S131 for the next iteration; S134, judging whether the iteration reaches a preset condition, if yes, ending; if no, continuing the iteration; S14, according to the and combining the W questions to obtain a plurality of test papers with a preset degree of difficulty and discrimination. S2, obtaining the test results of the subject on the test paper, and obtaining the item parameters in step S13 and the 3PL model in step S12, to obtain the m-dimensional ability vector of the subject. 2.The method of claim 1, wherein, The types of the questions include single-choice questions, one-question-one-point questions, fill-in-the-blank questions, one-question-one-point questions, and judgment questions, one-question-one-point questions; the difficulty of the test papers includes four levels of entry, advanced, master, and wisdom, and the proportion of difficult questions with a question difficulty parameter greater than a preset value in the questions of the four test papers is 10%, 20%, 30%, and 40%, respectively; the number of questions in each test paper is between 20 and 25. 3.The method of claim 1, wherein, The step S132 comprises: E step: calculating the conditional probability expectation of the joint distribution, the formula being as follows: represents the sum of conditional probabilities that all student ability values are represents the sum of conditional probabilities that all student ability values are and answer the questions, and t is the iteration number.​ M step: updating the parameter value using the maximum likelihood estimation method according to the expected latent ability value, and using a numerical optimization algorithm to find the parameter value that maximizes the likelihood function, the formula being as follows: 。 4.The method of assessing computational thinking based on multi-dimensional item response theory of claim 3, wherein, wherein, , , 。 5.The method of assessing computational thinking based on multi-dimensional item response theory of claim 1, wherein, The preset condition in the step S134 includes: whether the number of iterations reaches a preset number of iterations; and / or, the difference between the log-likelihood function values of two adjacent iterations is less than a set threshold. 6.The method of assessing computational thinking based on multi-dimensional item response theory of claim 1, wherein, The step S134 further comprises a step S135 of model fitting inspection: performing fitting inspection on the 3PL model to inspect whether the degree of model fitting data is reasonable.

7. The method for measuring computational thinking based on multi-dimensional item response theory according to claim 6, wherein, The model fitting inspection includes χ² inspection, RMSEA, CFI, or SRMR. 8.The method of assessing computational thinking based on multi-dimensional item response theory of claim 1, wherein, The The formula for calculating the multi-dimensional difficulty coefficient is: Wherein, , is a multi-dimensional discrimination index, represents the transformed multi-dimensional difficulty coefficient, and f is the scoring standard, f=1.

Images