Large-scale assessments administer items mostly arranging them in booklet designs. In this case the same item often has various positions in different booklets which can cause a bias of item difficulty depending on the position of the item. In further consequence the ability of the test person would be biased. Another important aspect of test administration is the choice of a particular response format with different kinds of multiple choice formats or constructed response formats being available. It is conceivable that using a different response format also has an impact on the difficulty of the item. The LLTM is a proper IRT approach for examining effects of item position as well as response format on item solving probability (Kubinger, 2008). Using this method the impact of item position was analyzed in a mathematic competence test of the national standard tests in Austria for 4th grade. The test was administered as a typical large-scale assessment with various test forms with items on varying positions in different booklets. The data set consisted of nearly 1800 students. Results indicate that a small fatigue effect take place in this test (Hohensinn et al., 2008). The impact of the response format was examined in a mathematic competence test of the national standard tests in Austria for 8th grade which was administered to aroung 3.000 students. For this analyses it is expected that various response formats modify the item solving probability even if the item content is the same. All IRT analyses were conducted using eRm (software package for R).