Application of item response theory (IRT) for the generation of adaptive assessments in an introductory course on object-oriented programming

Assessment process in one of the most important issues in the learning process and in many cases it is the process that define the sequence of the instruction because it measures the performance of the student in the educational process. In last decades the inclusion of TICs in teaching-learning process has facilitated to address the diversity of students and teachers features. Learning technologies enhance have permitted to adapt the different ways of learning and teaching that coexist in the educational context through the generation of adaptations process as well as user modeling process. This paper describes the item response theory and how it can be applied in a test scenario of an online course for the generation of adaptive assessments within a course of introduction to Object Oriented Programming from items (assessment questions), which are available in online repositories. Being a probabilistic theory the article describes the variables to consider and how to calculate the probability that a student answers correctly a specific item, which is known as the student proficiency level or theta, based on a skill test and their previous answers. In other papers write for the authors it is described the experience of working with fedora commons repositories, which are distributed across a network, in order to have enough questions to be applied properly normalized in the adaptation process. These questions are described under the Dublin Core standard. The main contributions of this project are the implementation of a probabilistic theory in the generation of adaptive assessment and use of distributed repositories that allow the reusability of items properly parameterized. A second phase of the research includes the implementation of the project (APIP) Accessible Profile Portable Item, allowing standardization of assessment items in a suitable format in the LOM.