Method and system for assessing cognitive potential and development competencies across the life span

Abstract

The application discloses a kind of full age cognitive potential and development competence evaluation method, comprising: constructing multilevel ability model based on CHC theory;According to 9 age stages, weight vector is distributed;Dynamic question selection is carried out according to ability estimation based on IRT CAT engine;Standardized score is adopted and converted into unified comparable NPCAS standard score by equalization;According to the three paths of subject, occupation and interest talent, weighted aggregation adaptation score is obtained;Visual report and top-notch innovative talent archives are generated, and based on threshold, it is automatically identified that extraordinary talent is identified.The application first realizes full age cross-batch comparable standard score, integrates individual cognition, multi-dimensional adaptation and top-notch talent identification, has high scientificity, full-process automation, and fills the gap in the prior art.

Description

Technical Field

[0001] This invention belongs to the fields of computer-aided psychological measurement, artificial intelligence and educational big data technology, and specifically relates to a method and system for evaluating cognitive potential and developmental competence across all age groups. Background Technology

[0002] In the fields of educational planning, career guidance, and talent selection, objectively characterizing individual cognitive characteristics and developmental potential using standardized assessment tools has become a key means to improve the efficiency of educational resource allocation, promote job matching, and identify outstanding innovative talents at an early stage. Currently, widely used cognitive and talent assessment products both domestically and internationally include the Myers-Briggs Type Indicator (MBTI), Enneagram, Multiple Intelligences Test, Dermatoglyphics Test, and DISC behavioral pattern assessment. These tools are extensively applied in college admissions guidance, career planning, and corporate human resource management. However, they generally suffer from weak theoretical foundations, substandard measurement performance, narrow applicability, limited functionality, and insufficient fairness, making it difficult to meet the urgent national need for establishing a scientific, fair, and life-cycle-covering integrated cognitive evaluation system. With the deepening of the concepts of individualized education and lifelong development, developing a cognitive potential and developmental competence evaluation method and system that is applicable to all age groups, relies on nationally unified norms, and possesses automated comprehensive decision-making capabilities is of significant practical importance for improving education quality, optimizing talent allocation, and promoting social equity.

[0003] A published patent application provides a talent assessment and job recommendation system based on MBTI personality types. This system collects participants' responses to a series of forced-choice questions via online questionnaires. Based on MBTI scoring rules, it determines tendencies across four bipolar dimensions: energy tendency, information acquisition, decision-making style, and life attitude, generating 16 personality type labels. These labels are then matched with a job characteristic database to output a list of recommended careers. However, this system has significant technical shortcomings: First, its theoretical foundation is based on Jungian analytical psychology, rather than a testable empirical research framework. Multinomial meta-analysis studies and a 2009 assessment report by the U.S. National Academy of Sciences both indicate that MBTI lacks stable predictive validity for academic achievement and career performance, and its test-retest reliability at intervals of several weeks is typically below the psychometric threshold of 0.8, raising questions about the reliability of the measurement results. Second, this system is designed only for adults; the question content and norms do not cover children and the elderly, failing to provide a continuous assessment throughout the entire lifespan. Regarding the scoring rules, it adopts a simple method of accumulating raw scores and dividing by a median line, lacking a parameter equivalence mechanism based on item response theory and nationally representative norms. This results in scores from different test batches and different combinations of items not being directly comparable, limiting the cross-time and cross-group tracking application of the assessment results. Functionally, the scheme only provides descriptive classification labels and static job matching, lacking in-depth integration functions such as subject-specific adaptation analysis, interest and talent coupling identification, and automatic identification of outstanding innovative talents. Furthermore, the assessment items do not adequately consider the control of social desirability bias, making the evaluation results susceptible to the influence of the test taker's family environment and cultural background, thus compromising fairness.

[0004] Furthermore, the published paper "A Study on Student Ability Assessment Model Based on the Theory of Multiple Intelligences, Educational Measurement and Evaluation, 2020" proposes a student ability assessment scheme based on the theory of multiple intelligences. This scheme collects information on individuals' performance in eight intelligences—linguistic, logical-mathematical, spatial, bodily-kinesthetic, musical, interpersonal, intrapersonal, and naturalistic—through self-assessment questionnaires. It calculates scores for each intelligence using a weighted summation method, creates an individual intelligence profile radar chart, and identifies the test-taker's dominant intelligence areas based on their scores, providing a reference for subject selection and interest development. However, this scheme also has significant limitations: although the theory of multiple intelligences has a wide influence in education, numerous empirical studies and Gardner's own arguments indicate that the various "intelligences" are not independent structures; their scores exhibit a moderate to high positive correlation, making it impossible to obtain mutually independent dimensions through factor analysis. This violates the fundamental requirement of psychometrics for the independence of assessment dimensional structures, resulting in a lack of discriminative validity in explaining individual differences in potential based on this scheme's dominant intelligence combination. This scheme only targets school-aged individuals and lacks assessment modules and corresponding norms for early childhood and old age, failing to meet the needs of early talent identification and cognitive resource assessment in the context of lifelong development. Its scoring relies on raw scores, without incorporating item response theory for parameter calibration and score equivalence, and it lacks nationally unified norms, making it difficult to fairly compare assessment results across different regions and age groups on a single scale. Functionally, the scheme only outputs static descriptions of strengths, lacking the ability to link cognitive potential with specific subject adaptations, career paths, and interest types, and cannot automatically identify the characteristic patterns of outstanding innovative talents from massive amounts of assessment data. Regarding fairness, because the assessment content heavily relies on language comprehension and self-reporting, the educational investment, family cultural capital, and social expectations of test takers significantly interfere with the assessment results, leading to systemic biases in the evaluation and undermining the social fairness of talent assessment.

[0005] In summary, the aforementioned existing technical solutions are either based on theoretical frameworks that lack sufficient empirical support, or are limited by structural defects in psychometrics and crude scoring rules. They also generally suffer from common bottlenecks such as narrow age range coverage, limited functionality, lack of norms, and insufficient fairness. The industry urgently needs a fully automated evaluation method and system that is applicable to all age groups, supported by nationally unified norms and high standards of psychometrics, capable of scientifically and independently characterizing cognitive potential dimensions, and achieving automatic identification of top-tier innovative talents across the entire process, including subject-specific fit, career fit, interest-talent coupling, and automatic recognition. Summary of the Invention

[0006] In view of the above, the present invention provides a method and system for evaluating cognitive potential and developmental competence across all age groups. Based on multidimensional empirical cognitive theory, it covers all age groups, is fully automated, and has dual core functions of individual cognitive ability and accurate identification of outstanding innovative talents. It can solve the technical defects of existing technologies such as insufficient scientific rigor, incomplete age coverage, incomparable scoring, and limited functionality.

[0007] A method for assessing cognitive potential and developmental competence across all age groups includes the following steps: (1) Construct a multi-level ability model based on CHC (Cattell-Horn-Carroll theory) cognitive ability theory to identify the ability dimension to which the question belongs; (2) Configure the weight vectors corresponding to each age group according to the preset age group division; (3) The CAT (Computer Adaptive Assessment) engine, driven by IRT (Item Response Theory), is used to dynamically select and deliver assessment items according to the test taker's real-time ability. (4) According to the question type classification, the corresponding standardized scoring method is used to determine the answer score value of each dimension. Then, through the estimation of IRT ability value and equivalence processing, it is converted into a unified and comparable NPCAS (National Cognitive Potential and Development Competence Assessment System) standard score. (5) Based on the three mapping paths of subject matching, career matching, and interest and talent matching, the NPCAS standard scores of each dimension are weighted and aggregated to generate the matching scores of each path. (6) The NPCAS standard score and path adaptation score of the test takers are stored in a structured manner in the National Education Big Data Platform to generate a visualized personal ability report and an electronic file of outstanding innovative talents.

[0008] Furthermore, in step (2), nine age groups are divided from early childhood to old age: 3-6 years old, 7-12 years old, 13-15 years old, 16-18 years old, 19-22 years old, 23-40 years old, 41-59 years old, 60-74 years old, and 75 years old and above. Different weight vectors are assigned to each age group, and the following conditions are met: in: This represents the weight vector corresponding to the first age group. j The weight value of the dimension (first level).

[0009] Furthermore, the rules for dynamically selecting and delivering assessment questions in step (3) are as follows: in: To select the question number to submit, Q This represents the candidate item set for the test taker. Indicates the first k The estimated ability of the test subjects after rounds of iterations, for the set Q any of the questions i , This indicates the estimated ability of the test taker. Next question i The amount of information, Indicates the topic i The weight of the corresponding dimension (secondary level). Indicates the topic i Content constraint coefficient or exposure control coefficient.

[0010] Furthermore, the capability estimate The iterative update formula is as follows: in: Indicates the first k The estimated ability of the test taker after +1 iteration. Indicates the test taker's response to the question. i The answers are scored. For the capacity estimate The test taker's response to the question i The probability of answering correctly, ε To prevent smooth terms with a denominator of 0, N For set Q The number of questions in; when Less than the threshold δ The iteration terminates at that time.

[0011] Furthermore, the amount of information The expression is as follows: in: for about The first derivative.

[0012] Furthermore, the probability of answering correctly The expression is as follows: in: For the title i Discrimination parameter, For the title i Difficulty parameters, For the title i Guessing parameters DThis is a scaling constant.

[0013] Furthermore, the standardized scoring methods in step (4) include four types: a 0-1 binary scoring method for objective reasoning questions; a combined accuracy-reaction time scoring method for cognitive processing efficiency questions; a norm-referenced Likert 5-level scoring method for situational judgment questions; and a Torrance scoring method covering three dimensions—fluency, originality, and applicability—for creativity questions.

[0014] Furthermore, the transformation expression for the NPCAS standard score in step (4) is as follows: in: For the test subject on the m NPCAS standard score in dimension (second level), For the test subject on the m Capability estimates in dimensional dimensions For the first m Dimensions of national peer norms θ mean For the first m Dimensions of national peer norms θ Standard deviation μ The target mean, σ The target standard deviation is denoted as .

[0015] Furthermore, in step (5), for any adapted path, its adaptation score is... The calculation expression is as follows: in: For the first m Age group weights in dimensions For the first m The adaptation weight of the dimension on this path. For the test subject on the m NPCAS standard score in dimensions M This represents the number of dimensions (secondary) involved in the aggregation along this path.

[0016] Furthermore, in step (6), during the process of generating a visualized personal ability report and an electronic file of outstanding innovative talents, the extraordinary talent identification conclusion is automatically determined and generated based on the preset multi-level threshold rules for extraordinary talents according to the NPCAS standard score.

[0017] An electronic system includes a memory and a processor, wherein the memory stores a computer program and the processor executes the computer program to implement the above-described method for evaluating cognitive potential and developmental competence across all age groups.

[0018] A computer-readable storage medium storing a computer program, which, when executed by a processor, implements the aforementioned method for evaluating cognitive potential and developmental competence across all age groups.

[0019] The core inventive aspect of this invention lies in constructing an executable data processing chain: parametric modeling of the question bank → age group weight loading → adaptive question selection → ability value estimation → reaction time correction → abnormal answer identification → standard score conversion → path mapping → rule identification and structured report output. This invention is the first to combine the three-level framework of the CHC cognitive ability theory with the IRT adaptive equivalence scoring mechanism, achieving comparable standard scores across batches for all age groups; the first to integrate a differentiated weight dynamic configuration mechanism for nine age groups from toddlers to the elderly within a single system; and the first to establish a fully functional adaptive mapping and automatic recognition algorithm system covering 15 subjects, 15 occupational categories, 10 talent domains, and 8 types of extraordinary talents. Compared with existing technologies, this invention has the following beneficial technical effects: 1. Significantly enhanced scientific rigor: All dimensions in this invention are based on peer-reviewed top journal research and large-sample meta-analysis, and the reliability of the dimensions reaches or exceeds the Wechsler scale standard (test-retest reliability ≥ 0.90), completely eliminating the defects of pseudo-scientific assessment tools such as MBTI and Enneagram.

[0020] 2. Uniqueness of full life cycle coverage: This invention is the world's first integrated standardized assessment system covering children aged 3 to the elderly aged 75 and above, filling the technological gap in the field of cognitive assessment across all age groups.

[0021] 3. Guarantee of scoring comparability: This invention uses IRT equivalence and the NPCAS standard scoring system to ensure that scores from different assessment batches and different combinations of questions are comparable nationwide, thus solving the problem of cross-batch incomparability in existing questionnaire tools.

[0022] 4. Multifunctional integration: This invention simultaneously realizes five functions within a single assessment framework: individual ability cognition, subject adaptation, career adaptation, talent and interest identification, and automatic identification of outstanding innovative talents, which is significantly better than existing single-function tools.

[0023] 5. Fairness Guarantee: This invention uses a standardized online assessment design that eliminates biases related to family environment and culture. The assessment results reflect only the individual's basic abilities and endowments, ensuring fairness in assessments between urban and rural areas and regions.

[0024] 6. High degree of automation: The entire process of this invention is completed by computer automation, with no manual scoring and no additional financial expenditure, and can be made available to the whole society free of charge as a public welfare project.

[0025] 7. Precise identification of top talents: This invention is based on a multi-level identification rule with strict statistical thresholds (standard score ≥130 / 135 / 140 corresponds to the top 2.28% / 1% / 0.38%), combined with asynchronous development of geniuses and dual special channels for children with special needs, which can accurately identify special talents in various fields that are difficult to discover in the conventional education system. Attached Figure Description

[0026] Figure 1 This is a flowchart illustrating the method for evaluating cognitive potential and developmental competence across all age groups according to the present invention.

[0027] Figure 2 This is a schematic diagram of a three-module, ten-dimensional, and forty-nine-sub-dimensional capability hierarchy model, demonstrating the complete capability dimension hierarchy structure under the CHC theory v2.2 framework.

[0028] Figure 3 A heatmap of the dynamic weight allocation matrix for all age groups is provided, which visualizes the changes in the weights of each primary dimension across the nine age groups.

[0029] Figure 4 This is a schematic diagram of the calculation process of the CAT engine driven by IRT.

[0030] Figure 5 This diagram illustrates the standardized scoring and standard score conversion process for multiple question types, showcasing the complete technical process from raw answer data to NPCAS standard scores via IRT equivalence.

[0031] Figure 6 The diagram illustrates the computational architecture for the three-path adaptation mapping, showcasing the dimensional weight mapping relationship of the three adaptation paths: subject, career, and talent.

[0032] Figure 7 This is a schematic diagram of a rule engine for recognizing extraordinary talents, showing the engine's inputs, outputs, and pseudocode for the recognition rules.

[0033] Figure 8 This is a schematic diagram of the overall architecture of the NPCAS system, showing the overall technical architecture and data flow of user access, CAT engine, six-level scoring process, three-path adaptation mapping, extraordinary talent recognition and data platform.

[0034] Figure 9 A schematic diagram of a JSON format result report file template for the NPCAS system. Detailed Implementation

[0035] To describe the present invention in more detail, the technical solution of the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0036] Example 1 like Figure 1As shown in the figure, this embodiment provides a method for evaluating cognitive potential and developmental competence across all age groups, including the following steps: Step S1: Construct a multi-level ability model based on CHC cognitive ability theory v2.2.

[0037] This step is used for parametric modeling of the question bank, such as... Figure 2 As shown, the model strictly follows the factor independence requirements of CHC theory v2.2, constructing 3 modules, 10 primary dimensions (broad domain capabilities) and 49 secondary dimensions (narrow domain capabilities). All dimensions meet the psychometric independence dimensional standard of factor loading ≥ 0.7 and cross loading < 0.3; there is no connotation overlap or structural confusion; each dimension has a clear operational definition and academic support.

[0038] The underlying cognitive processing module, namely the brain's basic hardware capabilities, is independent of knowledge, fully reflects innate neural endowment, excludes environmental interference, and is the foundation of all learning, work, and talent development; it includes four primary dimensions: fluid reasoning ability (Gf), sensory processing ability (Gp), memory and learning ability (Gm), and cognitive processing efficiency (Gs).

[0039] The higher-order cognitive regulation module, namely brain software regulation ability, is based on the underlying hardware and determines learning efficiency, problem-solving ability, and developmental ceiling. It is a core predictor of academic and career achievements and includes three primary dimensions: crystallized knowledge ability (Gc), executive function and metacognitive ability (Ge), and creativity and innovation ability (Gr).

[0040] The psychosocial adaptability module, also known as the core dimension of psychology, determines emotion regulation, interpersonal interaction, social adaptation, and career adaptability, and is the core support for an individual's lifelong development. It includes three primary dimensions: emotion regulation and empathy (Gem), social cognition and interpersonal skills (Gsoc), and personality trait dimension (Gper).

[0041] Fluid reasoning ability (Gf) is further divided into five secondary dimensions: inductive reasoning ability, deductive reasoning ability, quantitative reasoning ability, analogical reasoning ability, and causal reasoning ability.

[0042] Perceptual processing ability (Gp) is further subdivided into four secondary dimensions: visual-spatial processing ability, auditory-verbal processing ability, bodily perception and motion perception ability, and detail recognition and pattern discrimination ability.

[0043] Memory and learning ability (Gm) is further divided into five secondary dimensions: working memory capacity, short-term visual memory ability, short-term auditory memory ability, long-term storage and retrieval ability, associative memory and encoding ability.

[0044] Cognitive processing efficiency (Gs) is further divided into five secondary dimensions: information processing speed, perceptual processing speed, numerical calculation speed, attention control and anti-interference ability, cognitive flexibility and task switching ability.

[0045] Crystallographic knowledge (Gc) is subdivided into five secondary dimensions: language comprehension and expression ability, vocabulary breadth, reading comprehension and reasoning ability, general cultural knowledge reserves, and written writing and logical expression ability.

[0046] Executive function and metacognitive ability (Ge) are further subdivided into six secondary dimensions: planning and strategy making ability, inhibition control and impulse management ability, cognitive monitoring and reflection ability, decision-making and risk assessment ability, knowledge transfer and problem-solving ability, and critical thinking ability.

[0047] Creativity and innovation (Gr) is further divided into five secondary dimensions: divergent thinking ability, originality generation ability, association and cross-domain connection ability, problem discovery and reconstruction ability, and creative implementation and realization ability.

[0048] Emotion regulation and empathy (Gem) is further divided into four secondary dimensions: emotion recognition and understanding, self-emotion regulation, empathy and perspective-taking, and emotion expression and management.

[0049] Social cognition and interpersonal skills (Gsoc) are further divided into five secondary dimensions: interpersonal communication and persuasion skills, group collaboration and cooperation skills, social rules and situational understanding skills, leadership and organizational coordination skills, and conflict resolution and negotiation skills.

[0050] The personality trait dimension (Gper) is further divided into 5 secondary dimensions: conscientiousness, openness, extraversion, agreeableness, and neuroticism (reverse scoring).

[0051] Step S2: Generate adaptive weight configuration vectors for each age group according to the preset age group division.

[0052] The system pre-defines multiple age ranges and assigns a dimension weight vector to each age range. The age ranges include at least: 3-6 years old, 7-12 years old, 13-15 years old, 16-18 years old, 19-22 years old, 23-40 years old, 41-59 years old, 60-74 years old, and 75 years and above. Each age range corresponds to a set of weight vectors, satisfying the following: Among them: different age groups Differences are allowed to reflect the computational emphasis when aggregating dimensions at different developmental stages; in this embodiment, the weight of sensory processing ability is ≥25% for the early childhood stage (3-6 years old), while the weight of creativity and innovation ability is 0%; the weight of crystal knowledge ability is ≥20% for the professional maturity stage (41-59 years old); and the weights of memory and learning ability and cognitive processing efficiency are both ≥20% for the elderly stage (75 years and older), as detailed below. Figure 3 As shown: For children aged 3-6: The assessment time is ≤20 minutes. Assessments of crystal knowledge, creativity and innovation, personality traits, and higher-order metacognition are prohibited. The weights for the ten primary dimensions Gf, Gp, Gm, Gs, Gc, Ge, Gr, Gem, Gsoc, and Gper are 8%, 30%, 20%, 25%, 5%, 5%, 0%, 5%, 2%, and 0%, respectively.

[0053] For children aged 7-12: The assessment time is ≤40 minutes. Advanced career suitability assessments and complex critical thinking assessments are prohibited. The weights for the ten primary dimensions Gf, Gp, Gm, Gs, Gc, Ge, Gr, Gem, Gsoc, and Gper are 20%, 16%, 20%, 15%, 10%, 10%, 4%, 3%, 2%, and 0%, respectively.

[0054] For ages 13-15: assessment time ≤ 45 minutes, in-depth assessment of occupational competence and core evaluation of personality traits are prohibited. The weights for the ten primary dimensions Gf, Gp, Gm, Gs, Gc, Ge, Gr, Gem, Gsoc, and Gper are 22%, 12%, 15%, 15%, 12%, 14%, 6%, 2%, 2%, and 0%, respectively.

[0055] For ages 16-18: The assessment time is ≤50 minutes, there are no restrictions, and a full-dimensional basic assessment can be completed. The weights for the ten primary dimensions Gf, Gp, Gm, Gs, Gc, Ge, Gr, Gem, Gsoc, and Gper are 20%, 10%, 12%, 12%, 15%, 16%, 10%, 2%, 2%, and 1%, respectively.

[0056] For ages 19-22: assessment time ≤ 50 minutes, no restrictions, can complete the full-dimensional basic assessment, the weights for the ten primary dimensions Gf, Gp, Gm, Gs, Gc, Ge, Gr, Gem, Gsoc, and Gper are 18%, 8%, 10%, 10%, 18%, 18%, 12%, 3%, 3%, and 0%, respectively.

[0057] For ages 23-40: assessment time ≤ 50 minutes, no restrictions, can complete the full-dimensional basic assessment. The weights for the ten primary dimensions Gf, Gp, Gm, Gs, Gc, Ge, Gr, Gem, Gsoc, and Gper are 15%, 7%, 10%, 10%, 20%, 20%, 12%, 3%, 3%, and 0%, respectively.

[0058] For ages 41-59: assessment time ≤ 45 minutes, no restrictions, can complete the full-dimensional basic assessment. The weights for the ten primary dimensions Gf, Gp, Gm, Gs, Gc, Ge, Gr, Gem, Gsoc, and Gper are 12%, 7%, 10%, 10%, 23%, 21%, 10%, 4%, 3%, and 0%, respectively.

[0059] For ages 60-74: assessment time ≤ 30 minutes, complex advanced cognitive assessments and career suitability assessments are prohibited. The weights for the ten primary dimensions Gf, Gp, Gm, Gs, Gc, Ge, Gr, Gem, Gsoc, and Gper are 10%, 12%, 20%, 20%, 10%, 13%, 3%, 2%, 2%, and 0%, respectively.

[0060] For those aged 75 and above: the assessment time is ≤15 minutes, and all non-essential assessments except for the core cognitive screening are prohibited. The weights for the ten primary dimensions Gf, Gp, Gm, Gs, Gc, Ge, Gr, Gem, Gsoc, and Gper are 15%, 15%, 25%, 25%, 5%, 10%, 0%, 5%, 0%, and 0%, respectively.

[0061] This implementation method will determine the age group of the test subject from at least nine age groups, including infancy (3-6 years old) to old age (75 years and older). When the age of the test subject is received... y Then, the system uses an interval matching function to select and retrieve the corresponding first-level dimension weight configuration vector: in: For the first k For each age range boundary, if the age is near the boundary, a linear interpolation method between adjacent age ranges can be used for a smooth transition.

[0062] Step S3: Using an IRT-driven CAT engine, dynamically select and deliver assessment questions to the test taker based on their real-time capabilities.

[0063] The calculation process of the IRT-driven CAT engine in this embodiment is as follows: Figure 4 As shown: (1) IRT three-parameter model.

[0064] For the question bank Qany of the questions i The test subject's ability parameters are θ The probability of answering the question correctly is defined as: in: For the title i Discrimination parameter, For the title i Difficulty parameters, For the title i Guessing parameters D The scaling constant (preferred in this example) D =1.7).

[0065] (2) Information content function of the question.

[0066] Under the three-parameter model, the problem i The amount of information can be expressed as: in: for about θ The first derivative, Used to measure questions i The degree of contribution of the test subject to the ability estimate near the current ability point.

[0067] (3) Rules for selecting the next question: in: Indicates the first k The estimated ability of the test subjects after rounds of iterations Indicates the topic i The current weight of the dimension to which it belongs. Indicates the topic i The content constraint coefficient or exposure control coefficient; if multiple questions have the same amount of information, the question with the lower exposure rate or lower content coverage shall be given priority.

[0068] (4) θ Value iteration update method: in: This indicates the test taker's understanding of the questions. i The answers are scored. ε To prevent smoothing terms with a denominator of 0; the system in The iteration will terminate when the standard error of the measurement falls below the threshold.

[0069] Step S4: According to the question type classification, adopt the corresponding standardized scoring method to obtain the raw scores of each secondary dimension. Through IRT ability value estimation and equivalence processing, convert the ability values ​​into nationally unified and comparable NPCAS standard scores.

[0070] like Figure 5 As shown, the standardized scoring methods in this embodiment include the following four: For objective reasoning questions, a 0-1 binary scoring method is used, that is, 1 point is awarded for a correct answer and 0 points are awarded for an incorrect or unanswered question. The original score of this secondary dimension is the sum of the scores of all valid questions in this dimension. This method is applicable to all dimensions of fluid reasoning, all dimensions of crystal knowledge, and all dimensions of memory and learning.

[0071] For questions related to cognitive processing efficiency, a combined scoring method of accuracy and reaction time is adopted, i.e., raw score = (number of correct questions / total number of valid questions) × 100 - reaction time penalty score, where the reaction time penalty score = Max[0, (average reaction time of test takers - average reaction time of norm) / standard deviation of norm], and the penalty score is calculated only when the average reaction time exceeds the 95th percentile of the norm. This method is applicable to questions covering all dimensions of cognitive processing efficiency.

[0072] For situational judgment questions, a norm-referenced Likert 5-level scoring method is used. The optimal option is assigned a value based on the national norm. The option that fully meets the ability standard is worth 5 points, and the option that least meets the standard is worth 1 point. The raw score of this secondary dimension is the sum of the scores of all valid questions in this dimension. It is applicable to all dimensions of emotion regulation, social cognition, and executive function.

[0073] For questions related to creativity, the Torrance scoring method is used, which covers three dimensions: fluency, originality, and applicability. The raw score for this secondary dimension is calculated as follows: Fluency Score + Originality Score + Applicability Score. Fluency Score = Number of valid answers (answers that meet the requirements of the question); Originality Score = 2 points for answers appearing less than 5% of the norm, 1 point for answers appearing between 5% and 10%, and 0 points for answers appearing more than 10%; Applicability Score = 1 point for answers that are feasible / reasonable, and 0 points for answers that are not. This method is applicable to all dimensions of creativity and innovation.

[0074] For questions involving cognitive processing efficiency or those requiring time constraints, a reaction time penalty function is introduced: in: Indicates the topic i In actual response, For the title i When the reference reaction is; The penalty item can be 0; when The penalty value increases with the timeout rate.

[0075] When an abnormally fast response is identified, a rapid guessing metric can be defined: in: For the title i The minimum effective response time threshold, This is an indicator function.

[0076] Criteria for judging abnormal responses, i.e., the system generates abnormal scores for a single dimension or the entire assessment: in: This represents the number of questions missed. To quickly answer the questions, Express the reciprocal of the variance during reaction time. Indicates the coefficient of the repeated response pattern; when When the proportion of missed answers in a certain dimension exceeds the threshold, it is judged as an abnormal response session; If the proportion of unanswered questions in a single dimension exceeds 10%, or if the average response time of test takers is lower than the norm threshold of 1%, the dimension will be marked as invalid, no score will be generated, and a retest prompt will be triggered.

[0077] The formula for converting ability values ​​to NPCAS standardized scores is: in: For the first m The second-level NPCAS standard score For the first m The national peer norm in the second-level dimension θ mean For the first m The national peer norm in the second-level dimension θ Standard deviation μ The target mean (preferred in this example) μ =100), σ The target standard deviation (preferred in this example) σ =15).

[0078] The total score of the first-level dimension is calculated by weighting the NPCAS standard scores of all second-level dimensions under that dimension. The weights are determined by normalizing the factor loadings of the CHC model v2.2 national norm. The higher the factor loading, the greater the weight. The sum of the weights of all second-level dimensions is 100%.

[0079] in: Secondary dimension i The weights in its first-level dimension n The number of secondary dimensions under the primary dimension. For example, the factor loadings of the five secondary dimensions under fluid reasoning ability are 0.82, 0.80, 0.78, 0.76 and 0.75, respectively. After normalization, their weights are 20.5%, 20%, 19.5%, 19% and 18.5%, respectively, with a total of 100%.

[0080] The total ability score of the full scale is calculated by weighting the total scores of all primary dimensions with the corresponding weights for the age group, reflecting the overall cognitive ability level of the test taker.

[0081] in: First-level dimension j Weighting for the corresponding age group.

[0082] Step S5: Based on the three mapping paths of subject matching, career matching, and interest and talent matching, the NPCAS standard scores of each dimension are weighted and aggregated to generate matching scores.

[0083] This invention does not limit the adaptation results to a single conclusion, but rather maps multiple standard scores to multi-path results. Each path is driven by a predefined mapping template, which includes at least a path identifier, a set of participating dimensions, dimension weights, and an output name. The subject adaptation mapping covers at least 15 subjects, including Chinese, mathematics, English, physics, chemistry, biology, history, geography, political science, science, labor education, music, fine arts, physical education and health, and information technology, used to generate mathematics adaptation scores, language adaptation scores, spatial understanding adaptation scores, etc. The occupational adaptation mapping covers at least 15 major occupational categories and over 120 core occupations, including internet / IT / artificial intelligence, finance, healthcare, education and research, and manufacturing, used to generate adaptation scores for job clusters such as algorithm development, clinical medicine, and manufacturing operations. The interest and talent adaptation mapping covers at least 10 talent areas, including language and literature, mathematical and scientific innovation, spatial vision, music performance, dance competition, sports competition, invention, social leadership, nature exploration, and intangible cultural heritage, used to generate tag scores for mathematical and scientific innovation, music performance, spatial vision, and social leadership, etc.

[0084] like Figure 6 As shown, for any adapted path, its adaptation score is defined as: in: For the first m Age group weights in dimensions For the first m The adaptation weight of the dimension on this path. M This represents the number of second-level dimensions involved in the aggregation along this path; The specific weight value is determined by normalizing the standardized regression coefficient (i.e., predictive validity) of this dimension in the academic / career / talent achievement of this field in a large sample study. The higher the predictive validity, the greater the weight.

[0085] Step S6: Based on the preset multi-level threshold rules for extraordinary talents, automatically determine and generate a conclusion on the recognition of extraordinary talents.

[0086] The categories of gifted individuals include at least eight categories: general academic giftedness, mathematical and scientific innovation giftedness, space engineering giftedness, language and humanities giftedness, artistic performance giftedness, athletic giftedness, business leadership giftedness, and creative invention giftedness. This embodiment uses a multi-level threshold system comprising three levels: an entry threshold (NPCAS standard score ≥130), a core recognition threshold (≥135), and a top-level recognition threshold (≥140). It also establishes a domain-specific gifted talent pathway where a standard score ≥140 in any one core secondary dimension is sufficient for recognition, and a special recognition pathway for children diagnosed with learning / developmental disabilities and possessing a score ≥130 in any core secondary dimension. The specific identification rule engine code is as follows: Figure 7 As shown, the rule engine can be configured to be implemented in a threshold table-driven manner, so as to replace the label rules without changing the underlying calculation process.

[0087] Step S7: Store all results in a structured format on the National Education Big Data Platform to generate a visualized personal ability report and electronic files for outstanding innovative talents.

[0088] In this embodiment, the path template can be stored in JSON, a database table, or a rule configuration file, and online updates are supported.

[0089] Example 2 like Figure 8 As shown, this embodiment provides a cognitive potential and developmental competence assessment system for all age groups (i.e., the NPCAS system), including: a user access layer, a CAT adaptive engine, an abnormal answer detection module, an automated scoring engine, an adaptive mapping calculation module, a gifted talent recognition module, a report generation module, and an interface to the National Education Big Data Platform, etc.; all modules are interconnected through a unified data bus, and the entire process is automated without manual scoring; each question is associated with at least the question number and discrimination parameter before entering the calculation process. Difficulty parameters Guessing parameters Dimension, Response Result and during the reaction .

[0090] In addition, the system also includes a top-notch innovative talent tracking and management module. This module establishes electronic files for national top-notch innovative talents for individuals with exceptional talents, provides one free retest opportunity every two years, and supports inclusion in the national talent pool and dynamic exit mechanism after continuous retesting and passing the standards.

[0091] Example 3 This implementation method uses a 13-year-old subject as an example, and the input data is as follows: Age: 13 years old; Question parameters: Question 1 ( a 1, b 1, c 1) = (1.20, 0.30, 0.20), Question 2 ( a 2, b 2, c 2) = (1.05, 0.80, 0.18), Question 3 ( a 3, b 3, c 3) = (1.35, -0.10, 0.22); Answer result: u =[1,0,1]; During the reaction: t =[18s,42s,15s]; Initial ability values: .

[0092] The processing procedure is as follows: Step A: The system identifies that the age of 13 years old falls into the 13-15 age range and loads the corresponding age weight vector.

[0093] Step B: According to Calculate the information content of the candidate questions and select the question with the most information content as the next question.

[0094] Step C: After each response, update based on the 3PL model and the response results. At the same time, the penalty item and the quick guessing index are calculated based on the reaction time.

[0095] Step D: If the score is abnormal A If the result is below the threshold, the evaluation continues; if the result exceeds the threshold, an invalid mark is added to the corresponding dimension.

[0096] Step E: After the assessment, the dimensions will be... Convert to standard score Then, the math fit score is calculated based on the subject path template, and the math and science innovation tag score is calculated based on the interest path template.

[0097] The output results are as follows Figure 9As shown, this JSON is only an example of the output structure. The number of fields, field names, and encoding methods can be adjusted according to the deployment environment.

[0098] Example 4: Using 13-year-old junior high school students as test subjects, the system automatically identifies the age group as "junior high school (13-15 years old)," retrieves the corresponding weight vector (fluid reasoning 22% has the largest weight), runs the CAT engine to complete the assessment in ≤45 minutes, and outputs the NPCAS standard score, the math / physics fit advantage score, and suggestions for choosing subjects for the high school entrance examination.

[0099] Example 5: Using 5-year-old children as test subjects, the system retrieves the child's option weight vector (sensory processing with the largest weight of 30%), uses a gamified online questionnaire without text (with parental assistance), and the assessment time is ≤20 minutes. The system outputs a report on sensory and basic cognitive abilities and early talent and interest screening conclusions.

[0100] Example 6: Using a 28-year-old working professional as the test subject, the system retrieves the career growth stage weight vector, completes a full-dimensional assessment of ≤50 minutes, and outputs a ranking of suitability for multiple career categories such as Internet / Finance / Healthcare, as well as career advancement and development suggestions.

[0101] Example 7: Using a 10-year-old primary school student with outstanding performance in mathematics as the test subject, after the system completed the assessment, it detected a standard score of 142 (>140) in the first dimension of fluid reasoning and a standard score of 145 (>140) in the second dimension of quantitative reasoning. The system automatically triggered the "Mathematical and Scientific Talent Certification" process, generated the "National-level Talent Certification Certificate", and started a special re-evaluation.

[0102] The above description of the embodiments is provided to enable those skilled in the art to understand and apply the present invention. Those skilled in the art can readily make various modifications to the above embodiments and apply the general principles described herein to other embodiments without creative effort. Therefore, the present invention is not limited to the above embodiments, and any improvements and modifications made to the present invention by those skilled in the art based on the disclosure thereof should be within the scope of protection of the present invention.

Claims

1. A method for evaluating cognitive potential and developmental competence across all age groups, characterized in that, Includes the following steps: (1) Construct a multi-level ability model based on CHC cognitive ability theory to identify the ability dimension to which the question belongs; (2) Configure the weight vectors corresponding to each age group according to the preset age group division; (3) Using an IRT-driven CAT engine, assessment questions are dynamically selected and delivered to the test taker based on their real-time ability; (4) According to the question type classification, the corresponding standardized scoring method is used to determine the answer score value of each dimension, and then the IRT ability value is estimated and converted into a unified and comparable NPCAS standard score through equivalence processing. (5) Based on the three mapping paths of subject matching, career matching, and interest and talent matching, the NPCAS standard scores of each dimension are weighted and aggregated to generate the matching scores of each path. (6) The NPCAS standard score and path adaptation score of the test takers are stored in a structured manner in the National Education Big Data Platform to generate a visualized personal ability report and an electronic file of outstanding innovative talents.

2. The method for evaluating cognitive potential and developmental competence across all age groups according to claim 1, characterized in that: In step (2), nine age groups were divided from early childhood to old age: 3-6 years old, 7-12 years old, 13-15 years old, 16-18 years old, 19-22 years old, 23-40 years old, 41-59 years old, 60-74 years old, and 75 years old and above. Different weight vectors were assigned to each age group, and the following conditions were met: in: This represents the weight vector corresponding to the first age group. j The weight value of the dimension.

3. The method for evaluating cognitive potential and developmental competence across all age groups according to claim 1, characterized in that, The rules for dynamically selecting assessment questions in step (3) are as follows: in: To select the question number to submit, Q This represents the candidate item set for the test taker. Indicates the first k The estimated ability of the test subjects after rounds of iterations, for the set Q any of the questions i , This indicates the estimated ability of the test taker. Next question i The amount of information, Indicates the topic i The weight of the dimension Indicates the topic i Content constraint coefficient or exposure control coefficient.

4. The method for evaluating cognitive potential and developmental competence across all age groups according to claim 3, characterized in that, The capability estimate The iterative update formula is as follows: in: Indicates the first k The estimated ability of the test taker after +1 iteration. Indicates the test taker's response to the question. i The answers are scored. For the capacity estimate The test taker's response to the question i The probability of answering correctly, ε To prevent smooth terms with a denominator of 0, N For set Q The number of questions in; when Less than the threshold δ The iteration terminates at that time.

5. The method for evaluating cognitive potential and developmental competence across all age groups according to claim 4, characterized in that, The amount of information The expression is as follows: in: for about The first derivative.

6. The method for evaluating cognitive potential and developmental competence across all age groups according to claim 4, characterized in that, The probability of answering correctly The expression is as follows: in: For the title i Discrimination parameter, For the title i Difficulty parameters, For the title i Guessing parameters D This is a scaling constant.

7. The method for evaluating cognitive potential and developmental competence across all age groups according to claim 1, characterized in that, The standardized scoring methods in step (4) include four types: a 0-1 binary scoring method for objective reasoning questions; a combined accuracy-reaction time scoring method for cognitive processing efficiency questions; a norm-referenced Likert 5-level scoring method for situational judgment questions; and a Torrance scoring method covering three dimensions—fluency, originality, and applicability—for creativity questions.

8. The method for evaluating cognitive potential and developmental competence across all age groups according to claim 1, characterized in that, The conversion expression for the NPCAS standard score in step (4) is as follows: in: For the test subject on the m NPCAS standard score in dimensions For the test subject on the m Capability estimates in dimensional dimensions For the first m Dimensions of national peer norms θ mean For the first m Dimensions of national peer norms θ Standard deviation μ The target mean, σ The target standard deviation is denoted as .

9. The method for evaluating cognitive potential and developmental competence across all age groups according to claim 1, characterized in that, In step (5), for any adapted path, its adaptation score is... The calculation expression is as follows: in: For the first m Age group weights in dimensions For the first m The adaptation weight of the dimension on this path. For the test subject on the m NPCAS standard score in dimensions M This represents the number of dimensions involved in the aggregation along this path.

10. The method for evaluating cognitive potential and developmental competence across all age groups according to claim 1, characterized in that: In step (6), during the process of generating a visualized personal ability report and an electronic file of outstanding innovative talents, the system automatically determines and generates a conclusion on the recognition of extraordinary talents based on the preset multi-level threshold rules for extraordinary talents according to the NPCAS standard score.

Images