This paper is written by Dr Anna Brown who is the same author of the dissertation that received the Best Dissertation of IMPS 2011, and is written based on her dissertation. The research has demonstrated how using multidimensional item response modeling method has solved the problems raised by ipsative data.
This research is important because it has important practical implications. In reality, the findings of this article have been applied robustly in workplace assessment, especially in personality assessment. In commonly used personality inventory, multiple (and sometimes contrasting) traits of personality are assessed. In such a case, the model proposed in this paper is able to provide a more accurate estimate of reliability of the instruments while using a smaller number of ‘item blocks’ because selecting the best items that provide most useful information. Therefore, users are able to spend less time on completing the test while being assessed at higher accuracy and reliability.
The essence of this article that it is able to capture the decision process of a respondent in face of a force-choice items/ questionnaires. There might be in the future further improvement from this version of the model but what was described in the article has made a fundamental innovation from the traditional approach in analyzing this kind of items, which bring the old problem of ipsative data. This article can be seen as a friendly reminder to practitioners in the field of psychometrics, especially working on model development, that any research work towards innovative approaches to educational/ psychological measurement should be initiated from precisely describing the behavior of or the cognition behind the individuals’ response decision and strategies. This is the only way to develop a realistic model that accounts for the true interaction between items and respondent, which is the ultimate goal of the study of Psychometrics.