Innovation in Language Learning

This study concerns the specification of the language test difficulty level, focusing on the reading comprehension skill.  The main aim of the research is to reveal the moderating factors of the difficulty in a reading comprehension test and to specify the relevant indicators. In this way will possible the arithmetic expression of the test difficulty level before the implementation of the test. The previous research showed that there are many criteria of difficulty both for texts and items of a reading comprehension test. However, these criteria concern specific facets of a relevant test and doesn’t always provide a clear and stable basis for specifying the global difficulty of a test. The need to find out which of them can provide a valid and reliable evidence of the test difficulty, guided to the implementation of an experimental research. The results of this research will be presented and analyzed in order to explore the efficiency of the chosen criteria and their indicators as well as their numerical representation. More precisely, in the aforementioned research, a group of 30 test takers was divided in two groups of 15 and to each group was given a test of different difficulty level. After the correction of the test, was correlated the observed difficulty of the items with the estimated and then they were analyzed with the Correspondence Analysis method. The results showed that there is a highly positive correlation between the estimated difficulty (specified with the chosen indicators) and the observed, according to the item analysis (D.I).