Abstract:

The entry test to University are a hybrid of level tests, survey tests, diagnostic tests and selective ones. In fact some of them (those relating to planned access degree courses) determines a ranking which will allow, within the limits of available places, registration for the course. The other, even if compulsory, determine scores not affecting the inscription but to highlight the deficiencies to be recovered or, in some cases, give negative opinion about inscription. The "number of pre-registrations" problem, which might be irrelevant for a planned access degree, poses serious problems for an unscheduled access degree. In fact, while the test of degree courses in Primary Education and for Childhood and Preadolescence Training can ensure the objectivity and selectivity of the test, for the degree course in Science of Education there is a vital need for the construction of test, in addition to the required reliability and validity, it can find any deficiencies which, if not remedied, could invalidate future studies (Marlow, 2000; Cheung, Bucat, 2002). In other words, high numbers of registrations to unscheduled access course degree affect the quality of students learning path (Notti, 2010). If entry tests do not find and use any instrument relating to the recovery of any debt, the productivity parameter of the CdL and the faculty will suffer heavily. It is clear that it is necessary and useful to check the instruments built, without any presumptions of infallibility, especially since a bad test provides unreliable results (Steven et al., 1990; 1991). For this reason every effort should be made to build valid and reliable tests.
The objectives of the research have been, in summary, the following: check the validity and the reliability of entry tests of the degree course in Science of Education; show any kind of problem emerged from statistical analysis;suggest, if possible, solutions to be adopted to make tests congruent with the purposes for which they were built. After obtaining the necessary authorizations, the documentation in electronic form concerning tests given to the students and the corresponding tabs of results were acquired. Then, the following statistical processes were made considering the objectives of the research to check if entry tests are able to select students considering their preparation level; to check the ability of the test to measure the skills for which it was constructed and, consequently, its internal coherence.
The test is quite selective, not particularly difficult and with many unreliable items. The study of results examination reveals that the 1133 participant students had more difficulties especially in the two test areas called "Linguistics and literature" and "Geography". In them there is a good level of selectivity. A strong criticism is emerged from the distractors quality. The results we received, show a sufficient quality of test and a capacity to place in a reliable ranking of student results.

Keywords:

Quality, evaluation, test, item analysis.