The Completeness And Accuracy Of Attributes Definition In Cognitive Diagnosis

Cognitive diagnosis tests explore students' mastery of knowledge from the perspective of cognitive psychology and diagnose the source of their errors, the cognitive diagnosis tests can find a more detailed information of individual defects and be able to distinguish the differences of students with the same score, it is seen as the core of psychological and educational measurement for a new generation. Attributes are the specific knowledge structures and processing skills needed to perform a task, an attribute is a description of the declarative knowledge or procedural knowledge needed to perform a task in a specific domain. The completeness and accuracy of attributes have become the basic and key contents in cognitive diagnosis, on this basis , the results of cognitive diagnosis are scientific. The main purpose of this thesis is to provide more accurate indicators for the completeness and accuracy of attributes, in this paper uses the fraction operation of eighth-grade mathematics as example, using experts review and students verbal reports, and using confirmatory factor analysis, regression analysis and other statistical methods to provide more accurate indicators for the completeness and accuracy of attributes in the cognitive diagnosis tests.There are five chapters in this thesis. The opening chapter,generally introduces some background information about the studies,and present research conditions of cognitive diagnosis both domestic and abroad. Chapter two reviews the three major measure theory, summarizes the theory and models of cognitive diagnosis, also summarizes the methods how to certain the attributes and attributes hierarchy. Chapter three proposes the question of this research. Chapter four carries an empirical research for the completeness and accuracy of attributes. Firstly, the study identifies eleven cognition attributes and their attribute structure according to the records of maths teacher's interview and teaching material, creates a test, then please the subject experts carry out assessment and students' verbal reports to validate the attributes and attributes structure. Lastly, according to the dates of 975 examinees' items responses to statistical analysis. Cognitive attributes and item difficulty values are used to establish the regression equation, its determination coefficient( R ) is 0.89 , the R square is 0.792. The variance of item difficulty values, 79.2 percent is caused by the attributes and attributes structure. In order to validate the attributes and attributes structure, combines the parameters s and g of HO-DINA model to interview students again.Using confirmatory factor analysis, more than 90 percent possibility to accept the attributes and attributes structure. Chapter five summarizes the research achievements, research features , the existence of deficiencies and the future outlook.