30 April 2024
One of my favourite aspects of academia is having the independence to work on topics that intrinsically interest me, and how initially 'innocent' questions can lead me down a path of surprising/unsuspected discoveries. Sharing knowledge is also a part of my job that I really appreciate.
A research project about threshold models. During the last year, we could theoretically extend the methodology to accommodate for serially correlated error terms. By doing this, this approach is now feasible for many more applications. For example, a common problem in testing for the presence of threshold effects (the level or moment when a certain effect is observable) in macroeconometrics is that the errors are by construction serially correlated. These errors are carried over into future periods. Our methodology takes these errors into account.
Having had a few practice lessons in the past, I will be taking my private pilot's licence in the near future.