| As a new-generation development in the field of psychological and educational measurement,cognitive diagnostic assessment aims to provide information about students’ cognitive strengths and weaknesses by measuring specific knowledge structures and processing skills,enabling teachers to use this information to provide remedial guidance to students.In order to achieve the goals of assessment,not only high-quality tests and appropriate cognitive diagnostic models are needed to establish the relationship between the subjects’ unobservable knowledge states and observable responses,but also appropriate scoring methods are needed to extract students’ responses.Extracting students’ responses on the test is an important prerequisite for cognitive diagnostic evaluation.Only by analyzing the response data can we know the individual’s knowledge status.The more detailed the analysis of the response data,the more accurate the diagnosis results.Therefore,compiling high-quality cognitive diagnostic tests and selecting appropriate scoring methods and cognitive diagnostic models play a very important role in diagnostic evaluation.Multiple-Answer Multiple-Choice(MAMC)is one of the most commonly used item format in cognitive diagnosis assessment.Because they can avoid guesswork,have high reliability,and measure multiple topics at one time,providing a greater amount of information.However,the existing MAMC cognitive diagnostic scoring methods only use the information of the items and ignore the differences of students who choose different options,which not only causes waste of answering data,but also leads to evaluation biases,which hinders the development of MAMC cognitive diagnostic tests.To address this situation,this article proposes a MAMC cognitive diagnostic scoring method based on option vectors,based on the premise that options can provide richer diagnostic information,starting from the characteristics of multipleanswer multiple-choice questions.The entire study is divided into four parts.:Study 1,analyzes the advantages and feasibility of scoring with options,and proposes a MAMC cognitive diagnostic scoring method based on option information to analyze the responses of different students to options(OVS).Study 2,compares the accuracy of OVS and existing scoring methods under different number of attributes,knowledge state distribution and test length.Study 3,compares the robustness of OVS with existing scoring methods under different item quality.Study 4,to verify the validity of OVS in empirical data.Results demonstrate:(1)OVS retains the original response data,enabling more diagnostic information to be extracted and a greater ability to distinguish between different knowledge states.(2)OVS has more accurate scoring effects in different attributes,items and tests.(3)In order to fit the reality,after restricting the questions of a test to be of different quality,the scoring effect of OVS shows an advantage over other methods.(4)OVS performs equally well in empirical data.OVS can not only effectively distinguish different types of students,but also the data extracted by OVS has the highest consistency of students’ grades,and the discriminant results are closer to the real situation of students. |
