European and Dutch environmental goals prescribe that national emitted CO2 levels need to be reduced. Therefore, the Dutch government chose to transition the whole Dutch residential sector to gas-free living by 2050, to decrease CO2 levels. Homeowners are responsible themselves for adopting a sustainable heating system, where specifically heat pumps are one of the main proposed solutions. However, renovation rates are not expected to meet their determined goals. Hassle factors are a psychological barrier in homeowners' decision-making process, leading to inaction and delays. Therefore, it is determined what is perceived as a hassle and to what extent it is a barrier in homeowners' decision-making process.

Five main hassles were determined which impact the decision-making: ‘length of disruption’, ‘information gathering’, ‘subsidy and loan applications’, ‘finding a contractor’ and ‘neighbour consultation’. Next, their influence as a barrier has been examined through discrete choice modelling with the help of a stated-preference choice experiment containing in total 155 complete responses. The results of this study, for the first time, empirically validated that all five mentioned hassles negatively influence homeowners' decision-making. However, these five individual influences are best modelled as one combined hassle factor, which is perceived as more important than the financial profit a homeowner receives over the lifetime of a heat pump. Additionally, a latent class analysis was performed, finding significantly different preferences between two homeowners groups within the sample. Looking more closely at what a homeowner perceives as a hassle, the additional factor of ‘profit’ is found to have a de-hassling effect on the overall amount of hassle, indicating that context is important for homeowners. Looking more closely at the perceptions of hassle, differences between conscious and unconscious perceptions can be found.

The findings of this research show that reducing the five main hassles can be used to speed up the adoption process of heat pumps and contribute to reducing the adversities of climate change. Policymakers and businesses can perform interventions to decrease hassle or provide even ‘hassle-free’ alternatives by taking over tasks. The appropriateness of these interventions can be based on the derived value of time indicators created for each individual hassle factor. The use of context differentiations can be used as an additional tool via nudging and sludging to decrease or increase the perceived amount of hassle of choices. This research's overall value provides a better understanding of the psychological barrier hassle, which is rarely considered in homeowners' decision-making.

In 2050, it is expected that 70% of the world’s population will live in cities (Jin et al., 2014), leading to increasing congestion in and surrounding cities. This will raise new challenges, requiring more efficient and interactive cities. A novel paradigm contributing to these so-called smart cities is participatory sensing. Also known as mobile crowdsourcing, this solution enables both public and professional users to actively gather, analyse, and share local data about the urban environment using built-in sensors in smart devices (Truong et al., 2019). Considering that over 94% of the population has access to a mobile network, obtaining real-time data from these already existing sensors can be a low-cost solution for acquiring a huge amount of information (International Telecommunication Union, 2016). These real-time data can be used to analyse and predict mobility flows, and make public and private transport more efficient, safe, and sustainable.

However, a clear benefit is required to motivate smart device users to share data about their activities and their environment. Sharing data comes with the risk of disclosing private information, as location data can lead to the identification of living and work locations, as well as individual habits. Research on motivations of
smart device users to engage in participatory sensing tasks is required in order to be able to design valuesensitive participatory sensing applications. This study aims to identify factors related to incentives and privacy that explain choice behaviour of users in participatory sensing applications. The main research
question being addressed is as follows: “How do factors relating to incentives and privacy affect the willingness among smart device users to contribute to participatory sensing systems for smart mobility?”

A choice modelling approach was taken in order to identify the trade-offs made by users between potential benefits and costs of sharing data. This is an approach not often used before in the field of participatory sensing and provides novel insights in user behaviour in these systems. First, a literature review was
conducted identifying possible factors relating to incentives and privacy, influencing the willingness of people to share data. Five factors were selected: monetary reward, effort, risk of re-identification, types of data, and
data use. These factors were incorporated in a stated choice experiment distributed among smart device users through an online survey. In total, 125 valid responses were collected. The required effort of participating was regarded the most important factor influencing the willingness to share data in sensing
applications for smart mobility. This provides new insights, as previous studies do not include effort in choice experiments regarding data sharing. As expected, the perceived ease of use declines if more inputs by the user are required. Moreover, respondents are reluctant to the collection of contextual and multimedia in
addition to location and motion data, a finding which is confirmed by recent studies. Almost half of the respondents indicated to be highly concerned about their privacy. Therefore, a surprising finding is that the risk of re-identification was regarded the least important factor influencing the willingness to share data.
However, when taking a deeper look at the data, it appears there is a group of people having extreme preferences regarding privacy and trust, who assign a higher importance to privacy related factors (risk of re-identification, types of data, data use).

The identified trade-offs were used to evaluate the implications for different use cases in the field of smart mobility. Three interviews were conducted, which each led to the definition of a use case, being crowd management in a city, safety research using car accident information, and real-time travel information in
public transport. By aggregating the quantitative and qualitative parts of the research, it can be concluded that the accuracy of collected information can be improved by collecting more types of data in addition to location data. However, this will lead to a decline in the acceptance rate. A proposed solution is to provide
tailor-made sensing applications, giving the user control to indicate which data they agree to share.

Furthermore, the communication of the purpose of the data collection is important to users. Moreover, being transparent about the risks related to the data collection can help users to make a well-informed decision and
will ensure an ethical design of sensing applications. Finally, increasing the attractiveness of the application is recommended to reduce the perceived effort, which could be done by gamification of sensing tasks.

Besides societal implications, this study provides several recommendations for future research. First, it is recommended to repeat the experiment without the provision of a financial compensation, in order to see if this leads to different choice behaviour and preferences of users. Furthermore, research on the understanding of privacy risks among users is recommended. Finally, this study can be used outside the smart mobility field, in order to analyse the willingness to share data in a broader sense.

In the history of pension funds, the goal of their investment policy was capital growth within a limited investment risk. This goal was part of their fiduciary duty, the legal responsibility to act in the beneficiary's best interest. In recent years, the purpose of pension funds has shifted to aim for long-term social gains in addition to focusing on financial returns. Pension funds have increasingly integrated social, environmental, and ethical considerations in their investment process. This has been attributed to various factors, including the positive correlation between ESG and financial performance but also external pressures from non-governmental groups fuelling public opinion through the media and actions to act more responsibly. The participants must support decisions relating to a responsible investment policy for the pension funds. However, pension fund beneficiaries' preferences for integrating responsible criteria have not been sufficiently investigated.

The aim of this study can be split up into three parts. First, to understand the preferences on the sustainable criteria from the beneficiaries of ADP. Secondly, based on research findings and interviews, find a suitable approach to implementing this in the investment strategy of ADP. Lastly, recommendations will be made on other stakeholders' steps for a successful SRI implementation.

First, this research sets up a choice experiment. It then uses discrete choice modelling to determine participants’ preferences and willingness to trade financial profits against SRI goals using Random utility maximisation (RUM). With this method, we find that participants are willing to contribute part of their pension income to implementing SDGs. Furthermore, we can see that participants think it is especially important that pension funds focus on reducing inequality. There are several motives for integrating ESG into the decision-making process. However, the execution of policy is different per motive. When participants want to integrate ESG to achieve higher returns, the implementation differs from when the motive is to impact the world positively. Therefore, additional analysis has been executed.

To determine whether the participants consider an equal contribution between the step from no attention to do no harm and do no harm and do good. The utility contribution of each attribute level is also measured; we can observe that for most SDGs, there is little willingness to put in extra money to make a positive contribution to the SDGs, except for reducing inequality. The RUM model is very good for calculating the weights given to the attributes that best fit all participants' answers. However, the model does not reflect whether there is heterogeneity in the population. Therefore, interaction effects on individual characteristics are estimated; this method allows heterogeneity between different groups based on their characteristics. However, when using this methodology, an issue could arise. Observable covariates may not produce some heterogeneity. Therefore, latent class analysis was also used.

Through the analysis of interaction effects, we find that there are already many differences in preferences based on individual characteristics. In the latent class analysis, we see three distinct groups. In these classes, we find a large group that gives little utility to pension income and finds all the SDGs important. The 2nd and 3rd groups are smaller, containing about 18% of the population. Group 2 especially finds retirement income very important, and group 3 finds reducing inequality very important. In this analysis, we find that the strong preference of groups 2 and 3 greatly impacts the RUM model. From these analyses, we find that there is heterogeneity in the population. We can use this input when we look at how participants’ preferences can be implemented in the investment policy.

By first analysing the factors and stakeholders involved, we can find which obstacles to overcome when implementing policy. When analysing stakeholders, we find three different groups. First, we find the societal actors; these mainly influence the shaping of policies. Secondly, we find the political actors who influence the rules of the game. Lastly, we find the economic actors who influence the policy's implementation. In this analysis, several challenges have been found.

Lastly, we have looked broadly at options for implementing participants' preferences. In conclusion, the analysis shows that there are different ways in which a pension fund can develop an SRI policy. This can be done differently, depending on the participant's wishes and preferences and the pension fund's contract form. The pension fund first needs to develop a vision of how they want to achieve a sustainable world. After this, they can create an investment strategy and investment mandate. Policymakers also have a role to play in changing the financial sector. Besides implementing policies to improve transparency, they could also implement financial incentives so it will be more financially attractive to adopt ESG practices.

We can conclude that participants are willing to contribute part of their pension income to the implementation of SDGs. Furthermore, we can see that participants think it is especially important that pension funds focus on reducing inequality. Based on the latent class analysis results, we can see that different groups within the population have other preferences. This is important to take into account when formulating policies.

The model results and our analysis are subject to limitations. The model could be improved by allowing participants in a choice experiment to allocate money between different investment options, including policies only focusing on returns. Furthermore, the results could be influenced by self-selection and the population focus. Therefore, we recommend studying a more diverse population with participants from different pension funds. Lastly, we focused on how the results could be implemented and what tools the pension fund could use to accomplish an impact, as most of these recommendations are based on qualitative research. It may be interesting to conduct a quantitative analysis of the established objectives based on implemented policies.

Autonomous vehicles are expected to become an inevitable reality in the coming years, as their uncovered benefits are increasingly proven to be positive, such as increased safety and traffic efficiency. However, their impacts on travel choices have often been represented with an assumed reduction in the travel time penalty, which oversimplified AVs' subsequent effects activity-travel scheduling choices. As such, models and assessment methods do not accurately represent travel behavior implications of autonomous vehicles. Thus, using survey data, this thesis identifies possible rearrangements in activity-travel schedules, classifies the respondents into classes with similar profiles of expected rearrangements, and identifies further classification on the basis of socio-economic, personal, and travel characteristics. The survey to be used is one in which respondents were asked to report a full, regular working day activity schedule using their currently preferred mode of transport, then report the schedule as they expect it to be if they could use an autonomous vehicle. The initial exploration of the data identified the occurrence of activities on-board (work, spare-time, meals, and getting ready), changes to the duration of activities outside travel, as well as travel (delay of work-bound trips, and advancement of home-bond trips). Next, we used latent class models to cluster the responses with respect to on-board activity duration changes, stationary activity duration changes, and travel departure time changes. The clustering uncovered types of classes: no change, single activity on-board (work and spare-time), multiple activities on-board. Interactions between stationary and on-board activities were identified, with some direct activity transfer to travel episodes being common (with work, meals, and getting ready), while other activities were generally not transferred (spare-time). Finally, the addition of personal characteristics and demographics highlighted the limited influence of socio-economic factors, with the exception of education, on activity changes. In contrast, the most significant factors were mostly associated with work (daily time pressure and the ability to do work in the car) and travel time duration. An important insight uncovered was that the travel changes were limited and not as dramatic as expected, highlighting that the value of time impacts alone are not as representative as expected.

Agent-based modelling (ABM) is a widespread and commonly used technique for the understanding of social system behaviour. Despite ABM being a commonly used method for the analysis of complex social systems, simulated agents still can best be described as autistic before a specific set of social rules is ascribed to them. This discrepancy leads to model outcomes that can differ from reality. For improving agent-based models on this aspect, a social practice theory-based approach can be used as a theoretical framework. The main question to be researched is: How does enabling agents to reason about others' beliefs on social practices affect dynamics of belief and behaviour formation as opposed to action adaption models, with respect to transport mode choices? Two agent-based models have been created within this study to provide an answer to this research question. Through including a case study within the transport domain, model results can be hold against macro-behaviour within the state of the art of models on transport mode choices. The case study depicts a group of students making a decision for a transport mode to move from their university faculty to a compulsory seminar. The students influence each other, leading to belief adaption. Within the transport domain, usually this social influence is translated to compared and adapted actions instead. This study differs itself by letting beliefs be the reference point for social comparison. Therefore, belief dynamics and agent behaviour for the models with belief adaption are compared to the state of the art action adaption models. Lastly, the effect of limitations towards having insight to other agents' personal characteristics is analysed. The analysis provides the following conclusions. Within the field of transport mode choices, enabling agents to reason about others' beliefs on social practices results in the same opinion dynamics and behaviour as is presented by action adaption models. However, an agent-based model that allows agents to reason about others' beliefs on social practices with limited insight in others' private characteristics through a theory of mind capacity, does provide a different reaction to an increase in the social capital of agents. More specifically, enhancing social network connectivity through increasing the group size or similarity threshold does not lead to more consensus within the theory of mind model, which is the case in the other models presented within this study. This stresses the importance of considering the extent to which are capable of forming a just conjection on others' private characteristics when designing an agent-based model with social interaction.

This research examines the incorporation of morality in discrete choice modelling, applied to the specific case of the choice for ADAS in lease cars. Implicitly, the decision to include or exclude ADAS in a car affects the safety of the driver himself and other road users. Therefore this decision has a moral dimension; a moral dilemma arises during the trade-off between safety, comfort, and investments in ADAS. This research describes a stated preference experiment in which respondents choose between different option packages for lease cars. The target respondents need to make trade-offs between costs, comfort, and safety provided by ADAS. The results show that an increased perceived safety and comfort for the driver/passengers are both more preferred than an increased perceived safety for other road users. Also, the results show that innate morality plays an important role in these preferences, which is tested regarding five moral foundations. Furthermore, a booster that aims at increasing moral choice behaviour is tested, which results to have the opposite effect.

Capturing the individual’s behavioral responses to the actions driven by market and government will always be an interesting activity towards a broad spectrum of society. It could not be neglected that an individual may have different views in responding towards a new policy intervention or a new product market. However, references to use of the consumer–citizen concept in transport research studies are still scarce. Knowing this condition, this research is trying to close the gap by conducting an extensive analysis of the consumer–citizen duality concept. This study intends to test to what extent the consumer and citizen preferences may differ when the users make tradeoff between safety and travel time, in a case study of on-demand motorbike taxi services in Jakarta using stated-choice experiment. By elaborating on the role of an individual as a citizen, the research proposed in this study expect could contribute to an empirical evaluation of the operation of on-demand motorbike taxis, not only from a consumer perspective but also from the perspective of citizens. Subsequently, it also expects to gives nuances to the research field with citizen-stated choice experiments.

In this study, a context-dependent discrete choice model to investigate cyclist’s preferences on link-level cycling infrastructure design features is to investigate the extent the preferences on cycling infrastructure design is affected by the
characteristics of a road segment. Stated preference experiment utilizing visual representation of the alternatives were conducted. The model estimation indicates that cyclists attach different values on certain design features of a cycling infrastructure under different urban contexts. Nevertheless, there are significant heterogeneity on the preferences across respondents which still cannot be explained by the variation of contexts. Segmenting the data based on personal characteristic attributes, we found that differences in traveler’s personal characteristics are associated with certain differences in valuation of design features and also in sensitivity toward context effects.