Career interest tests such as Self-Directed Search instrument, Vocational Preference Inventory, Position Classification Inventory, and so on have been prominent to measure youth people’s vocational interests who often encounter career decision-making before entering the job market. To maximize the efficiency and reliability of the career interest tests, the diagnostic classification models (DCMs) are employed and corresponding new parameter estimation methods are proposed to increase the accuracy of parameter estimates. In this project, three major goals are first to build a new career interest questionnaire for youth people in Taiwan and use diagnostic classification models (DCMs) to analyze the data to yield career interest classification for individuals. Second, new parameter estimation is proposed to deal with the high-attribute issue in DCMs due to a large number of attribute levels. Third, the Q-matrix of DCMs, which describes the relationship between measured attributes and items, is sometimes unknown. A new Q selection Metropolis-Hastings Robbins-Monro Algorithm is proposed to deal with this complicated issue due to the high-dimension attribute space. During the process of this project, a Chinese version of Career Interest Inventory (CII) questionnaire will be developed to measure the six types of career characteristics— Realistic (R), Investigative (I), Artistic (A), Social (S), Enterprising (E), and Conventional (C). Data (item responses to the CII) will be collected from around 50 senior high schools/colleges chosen randomly. The theoretical and statistical methods for the high-attribute DCMs and Q-matrix estimation are developed and validated by a series of simulation studies, followed by a set of empirical data analysis. To maximize the utilities of the new parameter estimation methods, a computer program is developed under the R programming language for free access. Finally, the usage and the result interpretation are summarized in a handbook/manual for career counselors and relevant practitioners.
|Effective start/end date||2020/08/01 → 2021/07/31|
- Interest testing
- item response theory
Explore the research topics touched on by this project. These labels are generated based on the underlying awards/grants. Together they form a unique fingerprint.