The implementation of a bus system is an attractive alternative for decision makers since it has a short lead time and the service can be easily adapted depending on the demand. Due to the flexible nature it allows for the use of fulfilling different magnitudes of demand. Usually bus service provide a lower level quality service, however busses can be configured in such a way that they also provide higher level of service up until a point where they compete with rail systems. In the Netherlands referred to as HOV (“Hoogwaardig Openbaar Vervoer”, high-level public transport). This is a large spectrum with many different configurations. One should still consider the need for a HOV bus service as it comes with extra investment costs compared with the regular bus. Therefore, this research was looking into where there is no difference in valuation between a regular bus line and a higher level of service bus line. Through interviews, analysis of current lines and looking into what decision makers promise as a higher level of service attributes could be defined which could be seen as characteristic for a higher level service bus service. Additionally, the commuters perspective played a important role as most literature mostly considered only the decision makers perspective. A method has been demonstrated which shows how commuter needs can be used as and translated to operational characteristics with which decision makers specify, plan and design current and future higher level of service bus lines. The method includes a MNL model which expresses the valuation of the characteristics attributes of high level bus services which allows for the prediction of the modal split between a bus service and the current commute. This gave insight on which attributes a differently valued between a regular bus and a HOV bus and when a HOV bus service is equally valued as a regular bus service, hence not justifying investments in a higher level service. It was found that in most cases the HOV bus service results in a higher modal split despite of having a lower valuation for most of the characteristic attributes. The commuter seems to have more trust in the reliability promoted by higher level bus services.

The Netherlands suffers from a traffic congestion problem, which is a cause of both economic damages and high NOx and CO2 emissions. Today, the bus accounts for only a small percentage of all trips made in the Netherlands (less than 4%) but the bus plays an important role as an access and egress mode for train travelers. Short access and egress times are considered as crucial criteria for choosing to travel by train. Thus without a proper bus network, traveling by train will become less attractive. Furthermore the Dutch population is both growing and more mobile than ever, but space to construct new roads is limited. These developments mean that the current transport system in the Netherlands will not be sustainable in the future. To prepare it for the future sustainable public transport will be essential. However it is complicated to design a solid bus network as the public transport sector depends heavily on subsidies. For bus operators to allocate their resources efficiently, good information is key. Direct Ridership Models (DRMs) are rising in popularity as they can provide useful information to bus operators. DRMs are statistical models that are able to capture the relationship between surroundings (such as car ownership in the surrounding neighbourhood or the presence of a university) and travel demand. This way first cut predictions can be made for new bus stop locations. Furthermore, the expected number of travelers can be predicted for existing bus stops when there is a change in the surroundings of the bus stop, which allows bus operators to adjust their services accordingly. In this study, DRMs are used to explain and predict bus ridership for the concession area Arnhem-Nijmegen. The independent variables that are included in the models can be categorised into three groups: demographics, built environment and level of service. Three DRMs are developed based on different regression models. One model is the commonly used OLS. The other two models, Spatial Lag X model (SLX) and Spatial Error Model (SEM), are able to take spatial relationships between bus stops into account. The spatial models showed an improvement over the OLS model in terms of explanatory power. With 34 independent variables, the OLS model could explain 71.4% of the bus ridership. The SEM was able to explain 72.4% and the SLX 73.4%. But more importantly, some of the independent variables that were found significant related to ridership in the OLS model turned out to be not significant in the SEM model, of which car ownership was the most surprising one. For Arnhem-Nijmegen, it could be further looked into if the transportation by bus to parks, sport facilities, amusement parks and hospitals could be facilitated more to increase the bus usage as these variables are found to influence bus ridership significantly. Furthermore, it should be examined why P+R's are not found to affect bus ridership significantly. The prediction accuracy of the OLS model and the SLX model was also studied. The predictions of the SLX model turned out to be more accurate, but in general the prediction results were not satisfying. For more than half of the bus stops in Arnhem-Nijmegen, the predicted value of the SLX model was 50% more or 50% less than the actual value. The reason for this is the existence of non-stationary relationships, which means that the relations between ridership and the independent variables can vary based on the locations of the stops. This study contained bus stops in urban areas and in rural areas and it can be expected that for example the influence of a hospital in a village on bus ridership is smaller than the influence of a hospital in a city. In this study, the existence of non-stationary relationships between regions is demonstrated while studying if a DRM based on Arnhem-Nijmegen can be generalised to the region Groningen- Drenthe. It turned out residential land-use has more influence on bus ridership in Groningen- Drenthe while hospitals and universities have a higher impact on ridership in Arnhem-Nijmegen. The non-stationary relationships make it not advisable to generalise the DRM based on Arnhem- Nijmegen to different regions. In scientific literature, nothing has been written about the practical applications of DRMs for policy makers. As it was assumed DRMs can be useful for policy makers, three policy makers were interviewed for their insights on the applications. One of the application mentioned is preventing unsafe traffic situations, as DRMs allow to update ridership predictions when circumstances around the bus stop change. For example, when a bus stops is designed for a low number of travelers and a new neighbourhood is developed close to it, the increase in travelers could result in unsafe situations. With a DRM the increase in travelers can be predicted, which makes it possible to redesign the bus stop. Other applications that were identified in the interviews are: examining if a bus stop is performing below its potential and substantiating plans for new routes, new stop locations and relocating bus stops. Overall, the benefit of DRM is widely acknowledged and in combination with personal expertise, it could lead to optimised decision making. In the end, it is demonstrated how a DRM can be used to identify bus stops which do not fulfil their potential and suggestions are provided how the number of rides can be increased. The stops Europalaan in Renkum, Kerkstraat in Weurt and Beemdstraat, Fransestraat & Hatertseweg in Nijmegen turned out be under performing and several improvement suggestions were made. Although the prediction accuracy of the models were rather poor, future studies can use this study as a basis for models based on a more homogeneous study area, such as only bus stops in cities. Sub-setting the model is expected to have a major positive influence on the prediction accuracy. It would be interesting to see if a DRM based on a city can be generalised to other cities as well. Furthermore, the model can be complemented with more independent variables. Variables that encourage car use over bus use, such as the difference in travel time to the city center or the distance to the closest highway ramp, could be valuable additions.