Past seminars

2016

November 16

Title "External validity: a critical need for DCEs to take a key position in medical decision making"
Presenter

Dr Esther de Bekker-Grob

Affiliation Erasmus University
Abstract

Ideally, decisions in healthcare regarding patient care are based on evidence from randomized controlled trials and formal meta-analyses. These methods have major strengths and impact on clinical practice. However, their focus on single outcomes limits their use in multi-attribute decisions. The discrete choice experiment (DCE) technique has been introduced in healthcare to arrive at optimal decisions, dealing with multiple outcomes and preference heterogeneity among decision-makers. The DCE application in healthcare has grown exponentially since its introduction in the early nineties. Nevertheless, the lack of insight into external validity hampers DCE taking a key position in medical decision making: are stated preferences consistent with actual healthcare utilization? In this seminar consideration is given to two topics:

1) A mixed method approach to empirically and thoroughly test for external validity of DCEs; and

2) Choice task complexity in DCE: the role of 'attribute level overlap' and 'colour coding' to improve choice consistency in DCEs (and possibly external validity of DCEs).

November 3

Title "Modelling housing demand: social influences and policy lessons"
Presenter

Prof Michelle Baddeley

Affiliation University College London
Abstract

Housing markets are driven by many inter-related sources of instability, on both a microeconomic and macroeconomic scale, and their destabilising impacts were demonstrated amply in the sub-prime mortgage and financial crises of 2007/2008. Partly, this instability reflects the fact that housing choices of individual households are interdependent - generating non-linearity, discontinuities and feedback effects. Herding and other social influences will magnify these sources of instability, specifically: informational influences, consistent with Bayesian models of sequential social learning; and normative psycho-sociological influences such as peer pressure and group influence. In this seminar, I will illustrate these economic and psycho-sociological influences using evidence from two econometric models – one based around historical published data and the other from an experimental analysis of housing choices. Based on this evidence, I will explore a range of microeconomic and macroeconomic policy prescriptions for moderating housing market instability in the real world.

October 19

Title "Real world energy transition modelling: Heterogeneous supply and demand decisions"
Presenter

Dr Matt Shahnazari

Affiliation University College London (Australia)
Abstract

A confluence of problems and opportunities affect the energy sector, which has heightened the need for renewed efforts to improve the model-based analysis of energy systems. Decision makers in the energy sector are challenged by security, affordability and resilience of energy supply, as well as environmental concerns. Opportunities exist to bring new technologies to market and provide new sustainable energy production processes to meet increasing demand for energy. Realistic and effective planning or policy processes, require support from the improved modelling of the techno-economic analysis and stakeholder interaction. This talk will briefly address the limitations of existing energy system models, by focusing particularly on the need to include behavioural factors in the modelling process which emphasize heterogeneity in both supply and demand side attributes. In contrast to the conventional cost minimisation approaches, a modelling framework is proposed based on the perspective of energy actors in the liberalised energy markets. The inclusion of social change scenarios and heterogeneous actors in the proposed framework makes it possible to assess the distributive impacts of energy policies.

  

September 8

Title "Incorporating Behavioral Effects from Vehicle Choice Models into Bottom-Up Energy Sector Models"
Presenter

Prof David S. Bunch

Affiliation Graduate School of Management and Institute of Transportation Studies, University of California, Davis
Abstract

Many types of models are used for evaluating climate-change-related programs and policies, ranging from “top down” models of the entire economy (e.g., computable general equilibrium, or CGE models) to specialized models focused exclusively on specific markets (e.g., Oak Ridge’s MA3T model of the personal vehicle market).  Somewhere in the middle are “bottom up” models of the entire energy sector that rely on well-established legacy systems using linear programming formulations.  A major concern is that these models, despite having a high level of technological detail, yield unrealistic consumer market responses, calling into question their value for making policy decisions.  For example, household vehicle choices end up being “all or nothing” adoption of a single energy technology in any given year.  We have developed a microeconomic theory-based framework for modifying these models to produce more realistic market behavior equivalent to those from (nonlinear) discrete choice models. 

August 31

Title "Environmental Choice Modelling in Action"
Presenter

Prof Jeff Bennett

Affiliation Professor of Environmental Management, Crawford School of Public Policy, Australian National University
Abstract

A number of examples of choice modelling applications in the realm of environmental and natural resource policy making will be presented. The application contexts will range from water supply management in Australia to biodiversity protection in Laos. The examples will be used to illustrate some of the technical challenges involved as well as the difficulties of achieving acceptance of the technique in policy circles.

June 16

Title "Telehealth as a new model of health care service delivery: an investigation of the attitudes and preferences of older Australians"
Presenter

Prof Julie Ratcliffe

Affiliation Professor in Health Economics and Head - Flinders Health Economics Group, Flinders University
Abstract

Telehealth approaches to health care service delivery can potentially improve quality of care and clinical outcomes, reduce mortality and hospital utilisation and complement conventional treatment programs.  However currently substantial research into the potential for integrating telehealth within health care in Australia, particularly in the provision of services relevant to older people including palliative care, aged care and rehabilitation, is lacking. This presentation will report upon the methods and findings from a recent study utilising a discrete choice experiment approach to investigate the preferences of older Australians living in the general community for key attributes of a telehealth service delivery model.  The presentation will illustrate how the results from the discrete choice experiment may be incorporated into an economic evaluation framework to facilitate the future design and delivery of telehealth models of health care service delivery for older people.

 May 2

Title "Using choice experiments to explore the spatial distribution of willingness to pay for proximity to urban amenities"
Presenter

Dr Ali Ardeshiri

Affiliation Post-doctoral Research Fellow, Institute for Choice, University of South Australia
Abstract

This paper reports findings from a choice experiment survey designed to estimate the economic benefits of policy measures to improve the proximity to urban amenities in Shiraz. Using a panel mixed logit specification to account for unobserved taste heterogeneity this paper derived individual-specific willingness-to-pay (WTP) estimates for each respondent in the sample. This study subsequently investigated the spatial dependence of these estimates. Although stated preference studies have been extended to investigate distance-decay effects (see, for example, Bateman et al, 2006; Hanley et al, 2003; Pate and Loomis, 1997), the inherently spatial patterns of WTP are rarely clarified or addressed in stated preference studies (Bateman et al, 2002; Eade and Moran, 1996; Johnston et al, 2002). Aggregate measures of WTP, while useful, can obscure local patterns of heterogeneity (Troy and Wilson, 2006). Exploratory spatial data analysis provides different insights about WTP: its distribution; regional and local outliers; regional trends; and the level of spatial autocorrelation. Furthermore, given that the distribution of benefits are likely to be both spatially and socially uneven (Bateman et al, 2006), evaluating the regional nature of benefits delivers advantages from the political and policy analysis viewpoints.

March 31
Title "Measuring the effect of informational and normative conformity in individuals' preference for electric vehicles"
Presenter

Dr Elisabetta Cherchi

DTU Staff Profile

Affiliation Associate Professor, Department of Transport, Technical University of Denmark
Abstract

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. The impact of conformity in terms of policy implication is also discussed. The choice of electric vehicles (EV) is used as an illustrative example. To measure informational conformity the stated choice experiment was set up such as the same individual answered the choice tasks before and after he/she has received social information on three specific EV features: range, parking spaces reserved for EV and the need to change activity schedule if using an EV. The effect of descriptive norm and other-signalling concern are measured as part of the stated preference experiment, while injunctive norms are measured with typical statements on a 7-point Likert scale. Results from the estimation of hybrid choice models, clearly confirms that the negative experience of other people has a powerful effect on individual preferences and interestingly what people care more about is the fact that with an EV they have to change activity schedule. 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.

 

March 16
Title "Scores versus Models" - "Simplicity versus Complexity" or "Poor versus Good Data"?
Presenter

Anthony Marley, PhD

http://www.unisanet.unisa.edu.au/staff/homepage.asp?Name=Anthony.Marley

Affiliation Institute for Choice, UniSA
Abstract

Best-worst scaling (BWS) is a method that asks individuals to choose the most and the least preferred option from a set of available options. There has been extensive discussion and evaluation of i. scores (data summaries) in exploratory analysis of such data and ii. models where the worst choices are, or are not, assumed to be "mirror images" of the best choices. We evaluate three models and two score measures on 17 BWS data sets involving data aggregated across individuals, and three models on one data set with extensive choice and response time data for each individual. The scores and the models give somewhat different perspectives on the aggregate data, so we are now exploring approaches that include heterogeneity in decision strategies and preferences when a limited amount of data is available for each individual.

2015 

November 25
Title Specification and Estimation of Behavioural Models for the 2015 UK Value of Time Study
Presenter

Stephane Hess, PhD

http://www.its.leeds.ac.uk/people/s.hess

Affiliation Institute for Transport Studies, University of Leeds
Abstract

In early 2014, the UK Department for Transport (DfT) commissioned the first national value of travel time (VTT) study since the mid 1990s. This presentation looks the major methodological innovations of this study, both in terms of survey design and modelling, introducing flexible new ways for modelling reference dependence within a multiplicative model structure. Our findings show a rich pattern of heterogeneity across the travelling public, in terms of an impact on the VTT by both person and trip characteristics, as well as a major role for a number of characteristics that relate to the specific choices faced in a hypothetical stated choice setting, including reference dependence and non-linearities in sensitivities.

November 5
Title The Composition of Optimally Wise Crowds
Presenter

Clintin Davis-Stober, PhD

https://psychology.missouri.edu/stoberc

Affiliation University of Missouri
Abstract

We investigate optimal group member configurations for producing a maximally accurate group forecast. Our approach accounts for group members that may be biased in their forecasts and/or have errors that correlate with the criterion values being forecast. We show that for large forecasting groups, the diversity of individual forecasts linearly trades off with forecaster accuracy when determining optimal group composition.

May 28
Title The Statistical and Policy Benefits of Integrated Choice and Latent Variable Models
Presenter

Akshay Vij, PhD

http://people.unisa.edu.au/Akshay.Vij

Affiliation University of South Australia
Abstract

Integrated Choice and Latent Variable (ICLV) models are an increasingly popular extension to discrete choice models that attempt explicitly to model the cognitive process underlying the formation of any choice. Though the ICLV model has been employed extensively by studies across a wide spectrum of disciplines, the value of the framework has remained unclear. On one hand, ICLV models allow for the proper integration of psychometric data and provide a framework with which to test the influence of latent variables, such as attitudes and perceptions, on observable behavior. On the other, questions have been raised regarding their value to econometricians, practitioners and policy-makers. This talk undertakes a systematic comparison of the statistical and policy benefits offered by the ICLV framework over reduced form choice models without latent variables. The talk will present easily generalizable analytical proofs regarding the benefits, or lack thereof, of ICLV models and use synthetic datasets to validate any conclusions drawn from the analytical proofs.

March 26
Title Dominancy in stated choice surveys and its impact on scale in discrete choice models
Presenter

Michiel Bliemer, PhD

http://sydney.edu.au/business/itls/staff/michielb

Affiliation University of Sydney
Abstract

Stated choice surveys have been used for several decades to estimate preferences of agents. Typically orthogonal or efficient experimental designs underlie such surveys. These experimental designs may suffer from choice tasks containing a dominant alternative, which we show is problematic because it affects scale and therefore may bias parameter estimates. We propose a new measure to calculate dominancy and automatically detect such problematic choice tasks in an experimental design. Further, we propose a new regret-scaled multinomial logit model that takes the level of dominancy within a choice task into account in order to scale each choice task appropriately. Results are shown based on both simulated and empirical data.

February 24
Title Community preferences for the allocation of donor organs for transplantation
Presenter

Kirsten Howard, PhD

http://www.unisa.edu.au/Research/Institute-for-Choice/Our-people/

Affiliation University of South Australia
Abstract

Background. Demand for organs for transplant exceeds supply. There is an ongoing debate about the relative weighting that should be given to different allocation criteria. Little is known about the relative weight the community places on various allocation criteria. This study aims to determine community preferences for organ allocation. Methods. Community respondents recruited from a web-based panel chose which patient received a transplant in 30 scenarios presenting two hypothetical patients. Patients were described in terms of age, sex, previous transplants, whether they or family were registered donors, had caring responsibilities, adherence, time on waiting list, estimated survival and quality of life (QOL) with and without transplant, comorbidities, and lifestyle factors, such as smoking. Analyses were conducted in NLOGIT 5.0, using a mixed-logit model.

Results. Two thousand fifty-one respondents aged 18 to 83 years completed the survey. All attributes significantly influenced recipient choice except sex and having diabetes. Younger patients were preferred over older patients. Family member donor registration, having caring responsibilities, and longer time on waiting list increased priority. Pretransplant life expectancy was valued more highly than posttransplant life expectancy; 1 year less of pretransplant life expectancy required an increase of 1.49 years in posttransplant life expectancy to compensate. Posttransplant QOL was valued more highly than pretransplant QOL.

Conclusion. Lower pretransplant life expectancy (need) was more important than higher posttransplant life expectancy (utility). Although current allocation algorithms are consistent with community preferences for prioritizing children and time on the waiting list, favouring patients with high predicted posttransplant survival as potential recipients may not be aligned with community preferences.

February 3
Title Bayesian Imputation of Non-Chosen Attribute Values in Revealed Preference Surveys
Presenter

Simon Washington, PhD

http://scholar.google.com.au/citations?user=yka86dQAAAAJ&hl=en

Affiliation Queensland University of Technology
Abstract

Obtaining attribute values of non-chosen alternatives in a revealed preference context is challenging because non-chosen alternative attributes are unobserved by choosers, chooser perceptions of attribute values may not reflect reality, existing methods for imputing these values suffer from shortcomings, and obtaining non-chosen attribute values is resource intensive. This paper presents a unique Bayesian (multiple) Imputation Multinomial Logit model that imputes unobserved travel times and distances of non-chosen travel modes based on random draws from the conditional posterior distribution of missing values. The calibrated Bayesian (multiple) Imputation Multinomial Logit model imputes non-chosen time and distance values that convincingly replicate observed choice behavior. Although network skims were used for calibration, more realistic data such as supplemental geographically referenced surveys or stated preference data may be preferred. The model is ideally suited for imputing variation in intrazonal non-chosen mode attributes and for assessing the marginal impacts of travel policies, programs, or prices within traffic analysis zones.

2014

September 11
Title Language comprehension as a case study of real-time decision making and its implications for behaviour.
Presenter

Ina Bornkessel-Schlesewsky, PhD

http://people.unisa.edu.au/Ina.Bornkessel-Schlesewsky

Affiliation Professor, School of Psychology, Social Work & Social Policy, University of South Australia
Abstract

Comprehending language is essentially a problem of perceptual decision making at multiple levels. This follows from the inherent ambiguity of the linguistic input: as successful comprehenders, we routinely classify sounds uttered by different individuals and under different acoustic circumstances as belonging to the same category; we segment a continuous speech stream into words and infer complex meanings from underspecified sentences (e.g. understanding that "It's cold in here" can be a request to close an open window). Perhaps even more strikingly, we perform this complex task on a millisecond timescale, effortlessly comprehending several words per second. There is evidence to indicate that the mechanisms supporting ambiguity resolution in language share characteristics with mechanisms identified in research on human decision making (e.g. fast and efficient heuristics rather than full algorithmic processing that takes into account all possible sources of information), thus suggesting that language could be used as a model system for more general decision making processes. Moreover, by combining methods that track language comprehension in real time (EEG, eye-tracking) with explicit decision tasks, we have been able to shed light on the time course of (implicit) perceptual decision making during language processing and how it translates into conscious decisions and behaviour. In this talk, I will discuss some of the pertinent results from this literature and will also present some initial findings which suggest that EEG measures during online language processing may serve as reliable predictors of behavioural outcomes.

July 29
Title Cooperation and competition in the emergence of a social order 
Presenter David Goldbaum, PhD
Affiliation Associate Professor, Economics Discipline Group, University of Technology, Sydney
Abstract In an experiment, we explore how individuals organize into groups when no fundamental incentives are present.  We link subjects together in a social network with limited ability to observe others.  Subjects receive a list of options having no inherent values, and must either choose one or follow the choice of a network contact.  A subject is rewarded for choosing an option that turns out to be popular, and is paid an additional premium if people choose identically after he or she does.  The optimal order occurs when everyone follows the decision of a single leader, but this equilibrium requires individuals to balance their desire to cooperate with their competitive drive to capture the highest payoff as leader.  The emergence of a social order in this context thus requires subjects to adapt their individual strategies to the nascent advantages gained by others early in the experiment.  To capture this evolution, we model decisions with an experience-weighted attractor having recency, reinforcement, and lock-on biases. We find that subjects are often able to organize into the optimal order, but that considerable heterogeneity in biases exists across individuals.  This heterogeneity, coupled with choice history, determines much of the speed of emergence.
July 9
Title Stochastic variational inference for large-scale discrete choice models using adaptive batch sizes.
Presenter Linda S. L. Tan
Affiliation Department of Statistics and Applied Probability, National University of Singapore
Abstract The mixed multinomial logit (MMNL) model is a popular discrete choice model that captures heterogeneity in the preferences of decision makers through random coefficients. While Markov chain Monte Carlo methods provide the Bayesian analogue to classical procedures for estimating MMNL models, computations can be prohibitively expensive for large datasets. Approximate inference can be obtained using variational methods at a lower computational cost with competitive accuracy. In this seminar, we present variational methods for estimating MMNL models that allow random coefficients to be correlated in the posterior and can be extended easily to large-scale datasets. We explore three alternatives: (1) Laplace variational inference, (2) nonconjugate variational message passing and (3) stochastic linear regression. Their performances are compared using real and simulated data. To accelerate convergence for large datasets, we develop stochastic variational inference for MMNL models using each of the above alternatives. Stochastic variational inference allows data to be processed in minibatches by optimizing global variational parameters using stochastic gradient approximation. We also present a novel strategy for increasing minibatch sizes adaptively within stochastic variational inference.
June 23
Title What determines student satisfaction with university subjects? A choice-based approach
Presenter A/Prof Twan Huybers
Affiliation Deputy Head of School - Research, UNSW - Canberra
Abstract

In this paper, we report on a study of student satisfaction with university subjects and teaching. Quantative analysis of student perceptions of university subjects traditionally has been based on instruments containing a list of items (statements) to which students respond in an item-by-item manner using Likert-type rating scales. The main purpose of this paper is methodological: we propose and apply a novel application of a discrete choice experiment (DCE) to evaluate and measure the individual contributions of various subject and teaching attributes on student satisfaction with higher education teaching experiences. We are not proposing the use of a DCE approach as an alternative to regular classroom evaluation exercises. Rather, we see the approach as complementary, especially because DCEs avoid some well-known issues with rating scales like response styles. A representative Australia-wide sample of university students completed an evaluation ratings task and/or an evaluation DCE task. Our results indicate that the choice-based approach is better able to identify the relative importance of various evaluation items that drive student satisfaction. We also use a latent class analysis to explore differences in effects for sub-groups of students.

Download slides

Download audio

May 27
Title Importance Sampling Squared for Bayesian Inference in Latent Variable Models
Presenter Robert Kohn, PhD
Affiliation Scientia Professor, University of New South Wales
Abstract We consider Bayesian inference by importance sampling when the likelihood is analytically intractable but can be unbiasedly estimated. We refer to this procedure as importance sampling squared, as we can often estimate the likelihood itself by importance sampling. We provide a formal justification for importance sampling when working with an estimate of the likelihood and study its convergence properties. We analyze the effect of estimating the likelihood on the resulting inference and provide guidelines on how to set up the precision of the likelihood estimate in order to obtain an optimal tradeoff between computational cost and accuracy for posterior inference on the model parameters. We illustrate the procedure in empirical applications for a generalized multinomial logit model and a stochastic volatility model. The results show that the IS squared method can lead to fast and accurate posterior inference under the optimal implementation.
May 26
Title Combining facial EMG and eye tracking to explore integral affect in consumer choice
Presenter Prof. Thorsten Teichert
Affiliation Director of the Chair of Marketing and Innovation, University of Hamburg, Germany
Abstract Although affect is acknowledged as an integral part of decision-making, it is largely neglected in consumer choice models. Premises of rational choice still dominate the field of preference elicitation. A methodological explanation for this deficiency is that immediate affect during decisions cannot be traced easily on the spot. Most affect measures focus on self-reports, making them less reliable and in a sense redundant to the observed choice. Hence, it remains unknown if and how immediate affect actually matters in consumer choice. To close this gap, this work tests the existence of immediate affect integral in semmingly trivial consumer choices (e.g. choosing a yoghurt), and identifies its drivers and the context in which affect occurs. In order to acheive this objective, we implement an innovative method, which combines preference elicitation with in-depth process measures. Combination of eye tracking and facial electromyography (fEMG) allows assigning integral affect during decision-making to the observed choice options. An experiment with N=37 participants was conducted to test hypothesized drivers and the context in which affect occurs during consumer decision-making. Results indicate the existence of affect even during seemingly trivial consumer choices, and further point out significant factors in affective choice processes. By uncovering drivers and the context of integral affect in consumer choice, we provide support for future joint investigations of cognitive and affective processes in consumer choice tasks.
April 16
Title The Impact of Macro Socio-Economic Drivers and Fiscal Policy on Expenditure Allocation and Attribute Preferences
Presenter André Bonfrer
Affiliation Professor of Marketing, Research School of Management, College of Business and Economics, Australian National University
Abstract We investigate the effect of macro socio-economic drivers on Australian households' allocation of expenditure in a category (household appliances) and conditional on the allocated category expenditure, preferences for products (clothes washers) within the category. At the category level, we quantify the effect of changes in social mobility, disposable income, housing prices and the 2009 stimulus payments on purchase propensity and expenditure. At the product-level, we investigate how households trade off between price, energy efficiency and loading capacity consitional on allocated category expenditure, measuring non-homotheticity in preferences. We use the model to study a number of hypothetical scenarios, where we stimulate the effect of changes in macro socio-economic drivers and fiscal policies on market structure and revenue.
April 11
Title Structural Choice Modelling of Embedded Experiments: A Case Study of a Voluntary Blood Donation System
Presenter A/Prof Len Coote
Affiliation Associate Professor in Marketing and Marketing Cluster Leader, UQ Business School, University of Queensland
Abstract

The design and analysis of embedded experiments offers the promise of advancing research on decision making and choice. This paper illustrates the application of new model forms, structural choice models (SCMs), for testing behavioral hypotheses using embedded experiments. Donation of blood to the Australian Red Cross Blood Service provides the context for the study. One thousand non-donors completed a discrete choice experiment (DCE) describing the anticipated experience of donating whole blood. The DCE was embedded within a behavioral experiment following a paired-samples, pretest-posttest design. The behavioral experiment manipulated the target beneficiary (self- vs. other-benefit) and the perceived urgency of the need (urgent vs. non-urgent). Choices from the pretest and posttest DCEs are analyzed simultaneously using SCMs. The most general of these models (1) retrieves decision makers’ aggregate preferences for the attributes of the DCE and (2) captures the hypothesized effects of the manipulated variables on donation choices, (3) incorporates latent sources of preference heterogeneity, (4) retrieves the patterns of heterogeneity in relation to the attributes of the DCE, and (5) provides a test of the validity of the global experiment. The results have immediate and practical implications for policy makers at the Blood Service, including marketing and operational decisions. Moreover, the model forms immediately generalise to the analysis of embedded experiments in other contexts.

Download slides

March 19
Title Nudge2: The Multiplicative Effects of Individuals’ Decisions
Presenter Elizabeth Bruch, PhD
Affiliation Assistant Professor, Population Studies Centre, Institute for Social Research, University of Michigan
Abstract The theme of this work is how small changes in the decision architecture may propagate due to feedback between individual behaviour and social environments. In recent years there is increased interest in policy interventions that manipulate the decision-making environment to “nudge” people towards healthier choices. Thus far these interventions have been posed along the lines of putting fruit before fried chicken in school cafeterias, or making retirement savings something people must opt out of rather than opt in. However, this discussion ignores the potentially multiplicative effects that result from the dynamic interplay between individuals’ behaviour and their social environment. Systems science methods such as agent-based modelling aim to capture dynamic, interdependent behaviour but historically they have not been grounded in plausible models of human behaviour. This talk presents ideas and preliminary results from an early-stage research program which aims to: (1) repurpose “cognitively plausible” choice models from marketing and transportation for use in social research; and (2) incorporate the behavioural models into agent-based simulations to explore how key features of individual decision-making shape and are shaped by features of the environment. 
March 14
Title Destination name versus experience type as determinant attributes of Choice: A stated preference analysis
Presenter Harmen Oppewal, PhD
Affiliation Professor, Department of Marketing, Monash University
Abstract The decision process underlying tourists’ vacation choices is often viewed as the choice between competing vacation destinations in terms of their geographical locations. One traditional conceptualisation is that of a funnelling process in which tourists go through a gradual reduction of the number of destination in their choice set. However, the vacation choice process does not necessarily start with a geographical focus. Instead, prospective tourists may be searching for a certain type of vacation activity or “experience” such as outdoor adventure or entertainment and shopping. It is not clear whether the order in which tourists receive information about the vacation destination or the vacation experience affects their decision making. In this study, we investigate the effect of gradually revealing destination attribute information in a stated preference context. We study whether, and to what extent, attribute information received early in the decision process will carry over and have a larger effect on the final choice than attribute information revealed later in the process. We propose that choices will vary depending on which information is received first: vacation destination names or vacation experience types. In an on-line survey residents from Melbourne completed an experimental three-stage task in which they were presented with a vacation choice scenario. Participants either were first exposed to vacation destination names or to types of vacation experience. Further attribute information concerning transport and accommodation. To analyse the data we estimate MNL models allowing for parameters that test differences between the presentation conditions. Study findings suggest that when destination names appear earlier in the choice process they have a larger effect on the final choice than if experience types are presented first.

Areas of study and research

+ Click to minimise