The Calculation Of Information Utility And The Simulation Study On Validity For The Three-Parameter Logistic Model

In educational statistics and psychological measurement,the requirement of precision is very important.Although advances have improved our ability to describe the measurement precision of a test,but summarizing the overall of a test is still challenging.In classical test theory,reliability provides an overall summary of measurement,but it is for specific sample.If the sample is poorly suited for the test's purposes,the reliability might not reflect the potential usefulness of the test.In item response theory,the test information function provides detailed sample-independent information about measurement precision.But it does not provide an overall summary of the test performance.To solve this problem,Kristian E.Markon introduced the concept of information utility suited for two-parameter Logistic model.Information utility is to address the question of how to quantify the overall amount of information provided by a measure.Information utility complements existing notions of psychometric precision,it quantifies the precision of a measure for an observed response or a set of responses,as opposed to latent trait value or set of trait values.Information utility has a number of important applied implications.For example,trait estimation,test selection,hypothesis testing,computerized adaptive testing.Information utility may be especially useful in situations where information about latent trait level is vague or otherwise unclear.And on this basis,the current article is to research information utility of three-parameter Logistic model.And introduce the indices of information utility,for example,reference prior,expected information,the upper bound and lower bound of information utility,and criterion information utility.In addition calculate the information utility of three-parameter Logistic model,and analyze the application.Finally,the main conclusion of this study and the future research problems are prospected.