Student Evaluation of Teaching: multilevel IRT model

Romano, Clara (2012) Student Evaluation of Teaching: multilevel IRT model. PhD thesis, Universitò degli Studi di Palermo.


Download (6MB)


The assessment of teaching plays an important role within the Italian university. In 1993, the norm 537 / 93 established a national and local university evaluation system having the duty to monitor the productivity of research and of teaching. In accordance with this norm, Italian universities created theirNV (Nucleo di Valutazione) and began to operate in the wide and complex set of assessment activities. This has contributed to the collection of material relating to di ff erent aspects: the evaluation of university research, Students Evaluation of Teaching (SET), the analysis of student career by the universities. The presence of a rating system has encouraged the formalization of “quality procedures” that are based on an internal and an external activity. The former refers to the process of maintaining and improving the quality,the latter to periodic quality assessment. A meaningful assessment,internal or external, requires the collection of points of view of those who participate in an important way to the training process. So it is necessary to take into account not only the opinions of professional trainers (teachers at various levels and roles) but also students’ opinions. The SET takes into account students’ opinions through an evaluation form (questionnaire): this contains a set of compulsory items for all universities and it is structured in sections concerning several aspects of university courses (teaching,management aspects, class facilities, etc...) The main purpose of this study is to measure the concept of quality of teaching in the opinion of students at the University of Palermo. This concept is considered a latent variable that cannot be directly observable and measurable. In order to measure the latent variable the construct needs to be operationalized in terms of a certain number of dimensions, which are measured through a set of indirect variables or a battery of items. In the psychometrical and psychological literature, multi-item Likert type scales are the main tools for measuring an underlying theoretical concept, which is not directly observable. So we can assess the quality of university courses obtaining an approximation of the true measure by indirect measurements provided by students’ ratings. Rating is an indicator of the level of the specific attribute that it is supposed to be measured by that item. Obviously, it is necessary to emphasize that the result depends on subjective factors, since each student is influenced by his / her own needs and expectations, which on the other hand depend on the di ff erent cultural backgrounds and the di ff erent socio-economic conditions. Apart from the methodological nature of the results, we should not forget that the point of view from which the evaluation originates is the opinion of students. So, due to the presence of heterogeneity of the opinions of students, you cannot expect to arrive at a measure of the quality of teaching based on a system of shared values.

Item Type: Thesis (PhD)
Subjects: 300 Scienze sociali
Depositing User: Clara Romano
Date Deposited: 08 Mar 2013 11:11
Last Modified: 08 Mar 2013 11:11

Actions (login required)

View Item View Item