Dwell-time station-service analysis using a Rasch analysis technique

In order to ensure punctual and robust service, it is vital to have a good understanding of the current performance of a railway network. Several approaches to doing so exist but lack the ability to compare both service and station performance in a single dimension. The study presented here proposes the use of the Rasch analysis technique within an operational context to compare the relative dwell time performance of stations and services. To do so, we make use of data from commuter trains in Sweden and the UK. The results from the study suggest that the method can be used to study dwell times on a line level and can capture the variability in dwell times. Assessing the model output also shows that the approach adequately reflects the expected variability in both service performance and station difficulty. Comparing the model output to more commonly used indicators for dwell time, we find that the Rasch analysis allows us to better identify cases where planners can make adjustments to reduce the likelihood of dwell time delays. In addition to this, we highlight that a common assumption that more passengers lead to worse dwell times does not hold. Having more in-depth insights into where dwell time performance is troublesome can help planners to make more informed decisions which helps towards improving overall dwell time performance, reducing delays, and improving the attractiveness of trains as a mode of transport.