Assessing students’ progress is particularly important for learning in schools. Teachers need precise information about students’ skills and knowledge to adopt their instruction to individual pupils. The current study is about the project “PIRLS-Transfer”, which aims at providing teachers with valid and reliable standardized tests to assess their students’ abilities. The project is closely related to the Progress in International Reading Literacy Study (PIRLS; Mullis et al., 2003) and makes use of the same reading comprehension model. This model assumes that reading comprehension can be split up into four reading processes (Campbell et al., 2001). Using this model for the development of the “PIRLS-Transfer” tests has important implications because teachers can relate the results from their own pupils to the representative PIRLS sample. In this study, we address the question how to model students’ responses in order to provide teachers with the most useful information about their students’ performance.
    Standard analysis of the “PIRLS-Transfer” study with Rasch-Models provide unidimensional ability values, which most of all allow for a ranking (of abilities) between pupils. Here, more detailed diagnostic information can only be extracted by dividing (discretizing) the ability scale and assuming that the reading processes are ordered hierarchically. In Cognitive Diagnosis Models (CDMs) the central idea is different: Here, mastery of the items is predicted by latent (reading) processes, which are, in addition not assumed to be hierarchically ordered.
    Preliminary analyses of the PIRLS-Transfer response data have been performed by applying the CDM DINA model (Diagnostic input noisy „and“ gate; Junker and Sijstma, 2001) to the PIRLS-Transfer data, containing 175 pupils from second grade (N=116) and third grade (N=59). The test includes two booklets (Nahberger, 2007), containing 16 and 17 multiple choice items, testing for the reading processes from the PIRLS reading comprehension model which are “focus on and retrieve explicitly stated information”, “make straightforward inferences”, “interpret and integrate ideas and information”, and “make complex inferences”.

Campbell, J.R., Kelly, D.L., Mullis, I.V.S., Martin, M.O., & Sainsbury, M. (2001). Framework and specifications for PIRLS assessment 2001 (2nd edition). Chestnut Hill, MA: Boston College. Junker, B., & Sijtsma, K. (2001). Cognitive assessment models with few assumptions, and connections with nonparametric item response theory. Applied Psychological Measurement, 25, 258–272. Mullis, I., Martin, M., Gonzalez, E., & Kennedy, A. (2003). PIRLS 2001 international report: IEA’s study of reading literacy achievement in primary schools. Chestnut Hill, MA: Boston College. G. Nahberger (2007). Lockis Abenteuergeschichten im Urwald/im Wilden Westen. Mit Locki und seinen Freunden spielend Lesekompetenz erlernen. [Lockis adventures in the jungle / in the Wild West. Easily learning reading skills with Locki and his friends.] Baltmannsweiler: Schneider Verlag Hohengehren.