We present the design of a test for measuring students' abilities concerning the application of control structures. Validated test instruments are a valuable tool for the evaluation of teaching both in a research setting as well as in a classroom setting. Our test is based on item-response-theory, in particular the Rasch model, and comprises a set of items all following the same format and using a simple, artificial programming language. We field-tested and modified the instrument in four iterations using only small samples and special statistical methods instead of the large samples usually required for IRT models. After the fourth iteration, the test has now reached a usable state. Based on the results, we were able to identify two misconceptions that are occurring very frequently in our test population - students of grade 7 to 10 in secondary schools.

The "Bebras" is an international large scale contest that aims to motivate children for engaging in Computer Science. In this paper we present an investigation of the psychometric factors that determine the success in this contest, e.g. Abilities in Computational Thinking or intelligence factors. For this purpose, we looked for subsets of Bebras questions that are requiring joint psychometric constructs, according to the responses of the test persons, applying latent trait analysis on all combinations of questions. After identifying suitable sets of questions, we tested the fit of the mono-parametric Rasch Model and evaluated the distribution of person parameters. Additionally, the identified sets of questions were analyzed by qualitative methods to find out the nature of these common psychometric constructs. As a test bed for feasibility studies, we have chosen the Bebras Contests of Germany 2009. We have found and analyzed several sets of questions that met our requirements, analyzed the questions, and identified differences in the performance of boys and girls in these sets.