Morten Ørregaard Nielsen’s research is in econometric theory and methods. Econometrics is the statistical study of data in economics and related disciplines. These data are typically not derived from controlled randomized trials. Consequently, econometric methods of inference need to accommodate features of the data such as complicated structures of dependence across observations.
For example, in educational economics, empirical studies may attempt to ascertain the effects of various policies (smaller class sizes, etc.) on student test scores. Due to peer effects and teacher inputs, for example, individual student scores are not independent, but rather dependent within groups such as classrooms. Empirical analyses need to apply suitable econometric methods to account for such dependence and obtain reliable inferences.
Morten Ørregaard Nielsen’s research program focuses on two main types of statistical dependence. The first is the co-called cluster dependence and associated cluster-robust inference methods. In the test scores example above, it seems natural to assume that observations are only dependent within either classrooms or schools, which would then define the clusters. Ongoing research deals with both the choice of clustering structure as well as valid methods of inference in the presence of dependence within such clusters.
The second strand of the research program deals with dependence in time-series analysis with a particular focus on a type of strong dependence known as long memory. This type of dependence is commonly observed, e.g., in financial data. Developing methods of analysis that are valid in the presence of such dependence is challenging, but necessary to obtain credible inferences.