On Unobserved Traits and Its Measurements: A Note on Item Response Modelling Approach

Abstract

Often, scientists are interested in studying traits and other unobservable attributes of individuals which can only be measured with instruments such as questionnaire, test let, construct, and so on. These instruments are useful in extracting information on these hidden traits. Traditionally, Karl Pearson's classical approach were in used in modelling these traits with its attendant shortcoming ranges from variant estimates to its challenges of validity and reliability. This study present a better and modern item response modelling that remedy shortcomings of classical approach yielding reliable, consistent, and valid estimate through item response functions.