Leeds University

Upcoming seminars

Felipe Gonzalez- Identifiability of discrete choice models considering heterogeneous heuristics- 28/06/2017

Felipe González

Identifiability of discrete choice models considering heterogeneous heuristics

Wednesday, 28th June, 11:00-12:00
University of Leeds, Liberty Building SR (1.12)


Random utility maximization -RUM- (McFadden, 1973) is established as the most popular theory of discrete choice, having as the most iconic models the multinomial logit, hierarchical logit, and mixed logit models. Because these models are based on an additive measure of utility, they are compensatory across attributes and can represent a wide variety of behaviours. However, for the same reason, they cannot represent adequately some specific phenomena such as the decoy effect (Guevara and Fukushi, 2016). To address this among other issues, behavioural theories have been developed such as prospect theory (Kahneman and Tversky, 1979), regret theory (Zeelenbeg and Pieters, 2007), and satisficing theory (Simon, 1955). These theories have been materialized in the random regret minimization –RRM– (Chorus, 2008), elimination by aspects –EBA– (Tversky, 1972), and the stochastic satisficing –SS– models (Gonzalez-Valdes and Ortúzar, 2017), among others.

To exploit the advantages of both RUM and non-RUM heuristics, models that incorporate multiple heuristics have been used (Balbontin et al., 2017; Hess et al., 2012). However, combined models of this kind have been found frequently to lead to undistinguishable effects (unidentifiability), which is analysed in this presentation. First, this phenomenon is analysed theoretically to understand how behavioural differences lead to model identifiability. Then, RUM is compared with three choice heuristics (RRM, EBA, and SS) to explain its implications in practice by using a synthetic population together with a real choice set. Finally, these models are analysed to identify the choice mechanism that each individual used using mixed multiple heuristic models. The results of this show that RUM together with EBA leads to an identifiable model, RUM with RRM does not, whilst RUM with satisficing leads to a weakly identified model.

About Felipe: Felipe Gonzalez is an Industrial Engineer from Pontificia Universidad Católica de Chile. He is now a PhD student from that university working with emeritus professor Juan de Dios Ortúzar. He works on discrete choice models analysing new modelling formulations and has developed a model applying the psychological theory of satisficing. Felipe arrived as a visiting postgraduate research student at UCL in September 2017, to work with Benjamin Heydecker to untangle mathematical issues behind models with multiple heuristics.  Some of these results are presented in this seminar.

Prof. Sergio R. Jara-Diaz - Time use models and the values of work, leisure and travel: a personal research perspective - 18/07/2017

Time use models and the values of work, leisure and travel: a personal research perspective

Tuesday, 18th July, 11:00-12:00
University of Leeds, Business School Maurice Keyworth SR (1.15)


Departing from Becker (1965), DeSerpa (1972), Evans (1972), Train and McFadden (1978) and others, the evolution of a series of microeconomic models of time use is presented. The kernel idea was that to obtain the conditional indirect utility function that commands discrete travel choices, conditional goods consumption and time use equations were implicitly required (Jara-Diaz, 1998). This was applied by Jara-Diaz and Guevara (2003) to obtain a labour supply model from which the components of the value of travel time savings could be estimated empirically. This evolved quite naturally into a model of all activities where a proper labour supply equation was pivotal to generate a system of time use – goods consumption set of equations from which the values of work and leisure could be calculated (Jara-Diaz and Guerra, 2003). That framework has been applied by Jara-Diaz et al (2008, 2013) and Munizaga et al (2008).

 The approaches by Konduri et al (2011) using structural equations and Castro et al (2012) using MDCEV models, motivated further developments in time use modelling that allowed to understand better the values of leisure and work by acknowledging explicit relations between time use and goods consumption (Jara-Diaz and Astroza, 2013; Jara-Diaz et al 2016). Recently, new models including domestic work (Rosales-Salas and Jara-Diaz, 2017) and latent classes (Astroza et al, 2017) have been proposed and applied, improving the estimation of the different time values.

The availability of better data sets and the recognition of models and approaches coming from other disciplines should allow for a more comprehensive view of time use in general and for the development of enhanced models to understand specific activities (Jara-Diaz and Rosales-Salas, 2017).

Past seminars (2017)

Dr. Alexander Erath - Bike to the future: using Virtual Reality to study mobility behaviour - 09/05/2017

Principal Investigator of Engaging Mobility & Co-PI of Cognition, Perception, and Behaviour in Urban Environments

Bike to the future: using Virtual Reality to study mobility behaviour

Tuesday 9 May 2017


The opening of new infrastructure can open windows of opportunity for longitudinal studies to assess impact of built environment on mode choice. However, experimental variation is limited to how the new infrastructure has actually been built. Conducting stated preference surveys to understand the perception of various street design options and potential impact on behaviour based on photos or videos from existing infrastructure only can be too restrictive. Not only such an approach would need to restrict to infrastructure options that already exist, but depending on the type of application, it might also not cover important dimensions of perception and hence behavioural reactions.

Using Virtual Environments (VE) is a well-established methodology in the field of cognitive psychology. Although there are several limitations in VE, such as lower resolution, less realism and often no auditory, tactile, proprioceptive and vestibular cues, VE experiments have successfully been conducted in various fields of cognitive studies. Recent advances in computer graphics and lowered barriers to entry into the field of gaming has opened new opportunities to generate realistic 3D scenarios that are suitable for behavioural studies.

In this talk, Alex will provide first hand insights from an ongoing, exploratory research project on using Virtual Reality (VR) to study mobility behaviour using cycling as a case study. The talk will cover the nuts and bolts of creating an immersive Virtual Reality experiment platform by exploring how to combine various software tools to generate VE and to upgrade a simple cycling trainer to a VR simulator using customised sensors and commercially available head mounted displays. In addition, Alex will present the findings from the first experiment which aimed to understand how VR can help participants to state the influence of future infrastructure on their mobility behaviour and to explore the added value of VR as a communication and public engagement tool. The talk will conclude with a critical review of the experiences so far to use VR as a research tool of and an assessment of future research areas that might be relevant to study using VR applications.

Mathieu Plourde - Route choice modelling applications in the city of Québec - 03/05/2017

Mathieu Plourde

Route choice modeling applications in the city of Québec

Wednesday 3 May 2017


Understanding the mechanism behind route choice appears crucial to urban planner and decisions makers. The derivation of the choice set considered by the individual is not straightforward as potential alternatives are almost infinite and the agent is oblivious to most of them. 

The theory of the approach experimented in this study is developed at EPFL by Evanthia Kazagli & Michel Bierlaire and the innovative part lies in the integration of a cognitive representation of the road network. It means the modeling of route as in their conceptual description given by commuters. Instead of looking at the operational decisions, i.e. link to link decision, the goal is to explain the strategic decisions using aggregated spatial cues or mental representation items (MRIs)

The following work presents an application in the city of Québec, Canada. An explanatory approach of the road network structure and the distribution of traffic flows along the main arterials allow the identification of road elements that could serve as MRIs. Using available GPS records from drivers, an operational route choice model integrating latent class in the population will be derived.

Dr. David Palma A. - Modelling wine choices in a realistic way: combining incentive compatibility, sensory science and discrete-continuous modelling - 23/03/2017

Modelling wine choices in a realistic way: combining incentive compatibility, sensory science and discrete-continuous modelling.

Thursday 23 March 2017


Purchasing any food or beverage product is a multi-stage, multi-attribute process. During the first purchase, consumers rely only on extrinsic (i.e. visual) attributes. After purchase consumers can taste the product, perceiving its intrinsic (i.e. sensory) attributes. When re-purchasing the product, consumers have full information about it. Most attempts to model consumers’ choices of food and beverages either focus on only one stage of the process, one kind of attributes, or are limited in terms of their complexity or forecasting possibilities.

We present a novel experimental and modelling framework to study the choice of food and beverages in three stages: (i) purchase, (ii) tasting, and (iii) re-purchase on a controlled yet incentive-compatible setting. Our framework links all stages in a tractable and statistically correct way; it avoids endogeneity issues due to price’s double effect as a cue for quality and as a strain to the budget constrain; it considers the possibility of buying more than one product at a time; and is flexible enough to accommodate several methods of data analysis.

Dr. Edward Webb - Explaining consumer choice anomalies using eye-tracking - 23/02/2017

Explaining consumer choice anomalies using eye-tracking

Thursday 23 February 2017

Consumers often exhibit preference reversals by choosing a low quality, low price good when prices are low, but a high quality, high price good when prices are high, even though the price of the quality premium is held constant. We examine whether such reversals can be predicted from the attention paid to each aspect of the choice set as proposed by salience theory. We conduct an eye-tracking experiment and construct and compare models of consumer choice incorporating visual attention. We find that attention shifts between high and low quality goods reliably predict preference reversals, but inconsistent with salience theory, not shifts between quality and price.

Dr. Tatjana Ibraimovic - Analysing ethnic preferences using a discrete choice modelling framework - 10/02/2017

Dr. Tatjana Ibraimovic

Analysing ethnic preferences using a discrete choice modelling framework

Friday 10 February 2017


Ethnic pluralism and its increasing trend across European countries, has sparked debate on residential segregation, a phenomenon that has various repercussions at economic, social and urban dimensions of modern societies. Social integration and cohesion in residential areas is, indeed, seen as one of the main challenges of urban development today. The question leads to two main issues: on one side, an increasing residential spatial gap between the affluent and less affluent social classes, on the other geographical separation between inhabitants of different origins, cultures and religions.

In this seminar I will addresses such issues, presenting the analysis of ethnic determinants of residential location choice with the application to the Swiss city of Lugano. I will focus on the use of discrete choice modelling and stated preferences (SP) experimental techniques to assess the importance of voluntary self-segregation (i.e. ethnic preferences) across household with different ethnic and socio-economic background. I will also discuss various choice modelling strategies for answering some of the key questions arising in this domain. Finally, I will discuss the results and policy implications. 

Dr. Marek Giergiczny - Using advanced choice models to study animal behaviour - 08/01/2017

Assistant Professor, Department of Economic Sciences, University of Warsaw

Using advanced choice models to study animal behaviour

Wednesday 18th January 2017

Recent developments in positioning technology have led to new opportunities for investigating resource selection by animals but also new challenges related to the development of proper tools for the analysis of large amounts of information. The two currently most prominent approaches in ecology are Resource Selection Functions (RSF) and Step-Selection Function (SSF).  Resource Selection Functions (RSFs) are used to model habitat selection by animals using data from GPS locations. A RSF is defined as any statistical model deployed to estimate the relative probability of selecting a resource unit versus alternative possible resource units, which in most applications to date has been logistic regression. Another powerful modelling approach in ecology is the Step-Selection Function (SSF), which has been developed to estimate resource selection by animals moving through a landscape (Fortin et al., 2005). The main advantage of using an SSF rather than RSF is that SSFs may better model choices animals make as movement is included and as it constrains selection and availability, which enables association of parameters of movement rules with landscape features.

The main contribution of our paper is to use the state of the art choice modelling approaches in  modelling RSF and SFF. We make use of GPS locations collected within the GLOBE project which aimed to study brown bear behaviour in Poland and Sweden. Within this research project 1.5 million GPS locations for 150 individual bears were collected over a period of 11 years. In our work we have made use of these data and built Multinomial Logit (MNL), Latent Class (LCM) and Mixed Logit (MMNL) at both individual and sample levels. We also incorporate a large amount of interactions with bear-specific characteristics such as age, gender and number of cubs. A variety of different characteristics are used to describe the alternatives, such as road density and building density, land cover area for different types (barren land, forest, shrub land etc.), forest age, terrain ruggedness, and vegetation index.

Our work shows that there is a substantial amount of inter-bear preference heterogeneity among studied animals. Our results clearly show that the current practice in ecological applications assuming that animals have similar behaviour is too restrictive. Our analysis shows that using more advanced discrete choice models gives a much deeper understanding of brown bear behaviour and yields much better predictions of SSF which is a very promising tool in ecology, wildlife management and conservation. We think that our study will propagate the use of more advanced choice models in ecology. 

Past seminars (2016)

Prof. Elisabetta Cherchi - The effect of informational and normative conformity in the preference for electric vehicles - 15/12/2016

Professor of Transport in the School of Civil Engineering and Geosciences, Newcastle University.

The effect of informational and normative conformity in the preference for electric vehicles

Thursday 15th December 2016

According to Crutchfield (1955) individuals consciously or unconsciously tend to “yield to group pressures” and consequently to act in agreement to the majority position. Social conformity has been extensively studied in psychology with also several applications to transport problems. Field experiments are typically used to evaluate the impact of social influence on self-reported changes toward environmentally sustainable transport behaviours. In this seminar, I discuss various aspects of social conformity and the challenge of measuring informational and normative conformity effects in stated preference experiments. Differently from most of the literature in the field, measures of conformity are included as attributes inside a stated choice (SC) experiment, allowing a direct comparison of their effects with typical effects such as purchase price, range and fuel/electricity price. The impact of conformity in terms of policy implication is also discussed. The choice of electric vehicles (EV) is used as an illustrative example. Results from the estimation of hybrid choice models, show that all social conformity effects tested are highly significant and in particular the negative experience of other people has a powerful effect on individual preferences. The relative impact of social conformity effects in the overall utility can be high enough to compensate quite low driving range for EV or significant differences in purchase price (for example 1/3 higher for EV than ICV). Results also confirm that individuals’ behaviour is affected by the image they want other people to have of them, and being watched triggers a propensity to change reducing the inertia to stick with the current type of vehicle.

Dr. Maria Kamargianni - Incorporating Social Influence into Hybrid Choice Models - 24/11/2016

Senior Research Associate in transport and energy, UCL Energy Institute, London.

Incorporating Social Influence into Hybrid Choice Models

Thursday 24th November 2016

The aim of this study is to develop a methodological framework for the incorporation of social interaction effects into choice models. The developed method provides insights for modeling the effect of social interaction on the formation of psychological factors (latent variables) and on the decision-making process. The assumption is based on the fact that the way the decision maker anticipates and processes the information regarding the behavior and the choices exhibited in her/his social environment, affects her/his attitudes and perceptions, which in turn affect her/his choices. The proposed method integrates choice models with decision makers’ psychological factors and latent social interaction. The model structure is simultaneously estimated providing an improvement over sequential methods as it provides consistent and efficient estimates of the parameters. The methodology is tested within the context of a household aiming to identify the social interaction effects between teenagers and their parents regarding walking-loving behavior and then the effect of this on mode to school choice behavior. The sample consists of 9,714 participants aged from 12 to 18 years old, representing 21 % of the adolescent population of Cyprus. The findings from the case study indicate that if the teenagers anticipate that their parents are walking lovers, then this increases the probability of teenagers to be walking-lovers too and in turn to choose walking to school. Generally, the findings from the application result in: (a) improvements in the explanatory power of choice models, (b) latent variables that are statistically significant, and (c) a real-world behavioral representation that includes the social interaction effect 

Dr. Amanda Stathopoulos - Empirical analysis of crowd-sourced freight deliveries - 10/11/2017

Assistant Professor of Civil and Environmental Engineering, Northwestern University.

Empirical analysis of crowd-sourced freight deliveries

Thursday 10th November 2016


This seminar presents results from empirical analysis of crowd-sourced freight deliveries in the US. Crowd-sourced deliveries build on the idea that citizens deliver goods, ideally along planned travel routes. Crowdshipping has a potential to match highly fragmented transport capacities with vastly diverse demand for urban freight deliveries, temporally, spatially and in real-time. This is typically achieved through platforms that connect carriers with consumers in need of deliveries. A third-party broker, who operates the platform, provides match-making, analysis and customer services between demand and supply. The main advantage of crowdshipping is the reduced need for fixed facilities, such as cars or warehouses, to run operations. The main obstacles are trust, liability issues, and ensuring a critical mass of couriers and customers. Despite the growth in operations, there is still a poor understanding of the performance, functionality and acceptability of these new delivery methods.
The seminar presents results analyzing the performance in the early stages of operation of crowdshipping. Based on real operational data from 2 years across the US the performance is examined with an emphasis on the specificity of crowdshipping, namely related to delivery variability and the temporal matching dynamics. Based on additional survey experiments the behavior of the main agents in the system is modeled with an emphasis on revealing acceptance and priorities of both occasional drivers and senders.
The research derives from a Partnership-for-Innovation (PFI) project funded by the NSF where a Chicago based research team (NU, UIC) is evaluating the capabilities of CROwd-sourced Urban Delivery (CROUD) in collaboration with a crowd-shipper technology firm.

Dr. Maria Börjesson - The Impact of Accessibility on Labor Earnings - 13/10/2017

Associate Professor, Director of the Centre for Transport Studies,
KTH Royal Institute of Technology

The Impact of Accessibility on Labor Earnings

Thursday 13th October 2016

We estimate to what extent decision makers can induce agglomeration by investing in transport improvements increasing the job accessibility, and how this in turn influences wage growth. We deal with endogeneity by modelling changes over 11 years using micro-level panel data and using the FE estimator, controlling for all time invariant unobserved variables. The endogeneity not controlled for by the FE estimator is dealt with by using an instrument that is also based on temporal changes. We control for both zone-specific and individual-specific fixed effects by separating workers who have changed zone of residence and those who have stayed. The accessibility is derived from a transport model, taking into account consumer behavior and preferences for all travel modes and travel time components.  The elasticity of accessibility defined from the worker’s place of residence is estimated at 0.007. The elasticity of wage earnings with respect to job accessibility at the work location is only significant for workers changing work location and for those estimated at 0.015

Prof. Brett Day - The Value of the English Outdoors: A Cross-Nested Logit Model of Recreation Demand for Greenspaces in England - 23/09/2017

Professor of Environmental Economics, Director of the Land, Environment, Economics and Policy Institute (LEEP), University of Exeter

The Value of the English Outdoors: A Cross-Nested Logit Model of Recreation Demand for Greenspaces in England

Friday 23nd September 2016


This paper reports on the development of a recreation demand model for outdoor greenspace in England. The research is remarkable particularly with regards to the scope of the undertaking, attempting to model recreation behaviour for an entire nation and across all types of greenspace. We report on the development of a detailed spatial dataset describing the location and characteristics of accessible greenspace across England and explain how the estimation dataset was derived by coupling this greenspace map with data from a very large survey of recreation activity amongst English residents. Moreover we describe the estimation of the recreation demand model in the form of a cross-nested logit model and demonstrate a prototype online tool developed from the model intended to aid government, businesses and communities in better understanding the benefits that are derived from accessible greenspace in England.