Skip to main content
SpringerLink
Log in

Experience as a conditioning effect on choice: Does it matter whether it is exogenous or endogenous?

  • Published:
Transportation Aims and scope Submit manuscript

Abstract

Previous choice studies have proposed a way to condition the utility of each alternative in a choice set on experience with the alternatives accumulated over previous periods, defined either as a mode used or not in a most recent trip, or the mode chosen in their most recent trip and the number of similar one-way trips made during the last week. The paper found that the overall statistical performance of the mixed logit model improved significantly, suggesting that this conditioning idea has merit. Experience was treated as an exogenous influence linked to the scale of the random component, and to that extent it captures some amount of the heterogeneity in unobserved effects, purging them of potential endogeneity. The current paper continues to investigate the matter of endogeneity versus exogeneity. The proposed approach implements the control function method through the experience conditioning feature in a choice model. We develop two choice models, both using stated preference data. The paper extends the received contribution in that we allow for the endogenous variable to have an impact on the attributes through a two stage method, called the Multiple Indicator Solution, originally implemented in a different context and for a single (quality) attribute, in which stage two is the popular control function method. In the first stage, the entire utility expression associated with all observed attributes is conditioned on the prior experience with an alternative. Hence, we are capturing possible correlates associated with each and every attribute and not just one selected attribute. We find evidence of potential endogeneity. The purging exercise however, results in both statistical similarities and differences in time and cost choice elasticities and mean estimates of the value of travel time savings. We are able to identify a very practical method to correct for possible endogeneity under experience conditioning that will encourage researchers and practitioners to use such an approach in more advanced non-linear discrete choice models as a matter of routine.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5

Similar content being viewed by others

Notes

  1. Earlier research by a number of authors, including Morikawa (1994), introduced the idea of inertia dummies into a stated preference model to represent the influence of a current revealed preference mode that was chosen. While this has some similarity to the notion of experience in the current period used in our paper, Morikawa’s inertia dummies are added into the standard utility expression as linear additive in contrast to our approach where we condition the entire SP utility expression on experience, which allows also for frequency of use of an RP alternative.

  2. It is also possible to condition each attribute separately by experience to obtain unique conditioning parameter estimates for each observed attribute, as we have done in other studies such as Balbontin et al. (2020).

  3. In Swait (2006), just below Eq. (10), Swait states that “The trick to deriving the model above [Eq. (2) herein] was to assume that the choice probabilities are determined by differences in the transformed utilities (μqjUqj), rather than the original utilities.”

  4. Alternative functional forms were investigated, with the form presented here found to provide a superior overall model fit and significance levels.

  5. The reader is referred to (Bierlaire 2017) for more information on the sensitivity analysis. 500 draws have been used for the sensitivity analysis simulations.

  6. The models presented could also be estimated in Nlogit6.

  7. Most control function models use two step methods. The one step approach is a lucky accident (2SLS) for a linear model with simple endogeneity as shown by Wooldridge in https://www.nber.org/WNE/lect_6_controlfuncs.pdf and in Wooldridge (2010) where he has a section on 2 step estimation of a CF model, the selection model. A referee suggested we should estimate the model system as FIML. This is extremely complex when extending this to a nonlinear model such as a choice model. In the modern literature, there are applications of “2 step residual inclusion” all over the place. But, no one actually derives the FIML form. The extension is not “simple” and usually impossible.

  8. The “residual” in question is the derivative of the log likelihood for the binary choice with respect to the constant. Wooldridge (2010) discussed the probit model, for which the GR is the inverse Mills ratio. For a binary logit model, it is what we have used here.

  9. Correcting the utility functions in contrast to the experience conditioning expression would cause the utility functions to move up and down, but maintaining their shapes/slopes. Correcting the scale factors would lead to changing the sigmoidal function slopes, but not moving the utilities. That would lead to different elasticities. Correcting the utility function (results not reported here but available on request) using an intercept/ASC modifier accounts for the similarity of results across models, which should change using corrections to scale factors. The use of experience in the scale function is arguing that experience makes an individual more or less “utility dispersed”/“certain”, but does not shift the average utility of the mode. By putting the CF in the utility function, however, the practical effect is that experience is having an impact on both the mean and variance. We estimated all models by including the CF residuals into the utility expressions and then in the experience conditioning expression (reported in the paper), and the findings were very similar.

  10. The control function relates to real experience, whereas the choice model is stated preferences and not revealed preferences. This is an important distinction but one that is inevitable if experience data is being used to condition any choice model regardless of where the nature of the choice data. We should mention that the experiments were designed in such a way that the attribute levels offered for existing modes were pivoted around reported RP levels for each respondent and as such there is a string commonality between the RP and SP data. This might be seen as a very useful way of recognising and accounting form, within limits, hypothetical bias in choice experiments. Also, while socioeconomic variables were included in both the RP experience control function and the SP choice model, we did not find any that were statistically significant in the standard part of the utility expression.

References

  • Antolín, A.F., Stathopoulos, A., Bierlaire, M.: Exploratory analysis of endogeneity in discrete choice models. Paper presented at the 14th Swiss Transport Research Conference, Ascona, 14–16 May 2014.

  • Balbontin, C., Hensher, D.A., Ho, C., Mulley, C.: Do preferences for BRT and LRT change as a voter, citizen, tax payer, or self-interested resident? Transportation. 47, 2981–3030 (2020)

    Article  Google Scholar 

  • Ben-Akiva, M.E., Lerman, S.R.: Discrete Choice Analysis—Theory and Application to Travel Demand. MIT Press, Cambridge, MA (1985)

    Google Scholar 

  • Berry, S., Levinsohn, J., Pakes, A.: Automobile prices in market equilibrium. Econometrica. 63(4), 841–889 (1995)

    Article  Google Scholar 

  • Bierlaire, M.: PythonBiogeme: A short introduction. Report TRANSP-OR 160706, Series on Biogeme. Transport and Mobility Laboratory, School of Architecture, Civil and Environmental Engineering, Ecole Polytechnique Fédérale de Lausanne, Switzerland (2017).

  • Dallaert, B., Brazell, J., Louviere, J.: The effect of attribute variation on consumer choice consistency. Marketing Letters. 10(2), 139–147 (1999)

    Article  Google Scholar 

  • Fernández-Antolín, A., Guevara, C.A., de Lapparent, M., Bierlaire, M.: Correcting for endogeneity due to omitted attitudes: Empirical assessment of a modified MIS method using RP mode choice data. J. Choice Modell. 20, 1–15 (2016)

    Article  Google Scholar 

  • Guevara, C.A.: Critical assessment of five methods to correct for endogeneity in discrete-choice models. Transportation Research Part A: Policy and Practice. 82, 240–254 (2015)

    Google Scholar 

  • Guevara, C.A., Ben Akiva, M.E.: Addressing endogeneity in discrete choice models: Assessing control-function and latent-variable methods. In: Hess, S., Daly, A. (eds.) Choice Modelling: The State of the Art and the State of Practice, Chap. 16, pp. 353–370. Emerald, Bingley, UK (2010)

    Chapter  Google Scholar 

  • Guevara, C.A., Hess, S.: A control-function approach to correct for endogeneity in discrete choice models estimated on SP-off-RP data and contrasts with an earlier FIML approach by Train & Wilson. Paper presented at the 2019 Transportation Research Board Conference, Washington, DC, January (2019).

  • Guevara, C.A., Polanco, D.: Correcting for endogeneity due to omitted attributes in discrete-choice models: The multiple indicator solution (MIS). Transportmetrica A: Transp Sci. 12(5), 458–478 (2016)

    Article  Google Scholar 

  • Guevara, C.A., Tirachinia, A., Hurtubiab, R., Dekker, T.: Correcting for endogeneity due to omitted crowding in public transport choice using the multiple indicator solution (MIS) method. Transportation Research Part A. 137, 472–484 (2019)

    Google Scholar 

  • Heckman, J. J. (1976). Simultaneous equation models with both continuous and discrete endogenous variables with and without structural shift in the equations, in Steven Goldfeld and Richard Quandt (Eds.), Studies in Nonlinear Estimation, Ballinger.

  • Heckman, J., Robb, R.: Alternative methods for evaluating the impact of interventions: An overview. Journal of Econometrics. 30(1–2), 239–267 (1985a)

    Article  Google Scholar 

  • Heckman, J.J., Robb, R.: Alternative methods for estimating the impact of interventions. In: Heckman, J.J., Singer, B. (eds.) Longitudinal Analysis of Labor Market Data. Cambridge University Press, Cambridge (1985b)

    Chapter  Google Scholar 

  • Hensher, D.A., Ho, C.: Experience conditioning in mode choice modelling—does it make a difference? Transportation Research Part E. 95, 164–176 (2016)

    Article  Google Scholar 

  • Hensher, D.A., Rose, J.M.: Development of commuter and non-commuter mode choice models for the assessment of new public transport infrastructure projects: A case study. Transportation Research Part A. 41, 428–443 (2007). https://doi.org/10.1016/j.tra.2006.09.006

    Article  Google Scholar 

  • Hensher, D.A., Rose, J.M.: The influence of alternative acceptability, attribute thresholds and choice response certainty on automobile purchase preferences. Journal of Transport Economics and Policy. 46(3), 451–468 (2012)

    Google Scholar 

  • Hensher, D.A., Rose, J.M., Collins, A.T.: Identifying commuter preferences for existing modes and a proposed Metro in Sydney, Australia with special reference to crowding. Public Transportation. 3, 109–147 (2011). https://doi.org/10.1007/s12469-010-0035-4

    Article  Google Scholar 

  • Hensher, D.A., Li, Z., Ho, C.: The role of source preference and subjective probability in valuing expected travel time savings. Travel Behaviour and Society. 2, 42–54 (2015a)

    Article  Google Scholar 

  • Hensher, D.A., Rose, J.M., Greene, W.H.: Applied Choice Analysis, 2nd edn. Cambridge University Press, Cambridge (2015b)

    Book  Google Scholar 

  • Karaca-Mandic, P., Train, K.: Standard error correction in two-stage estimation with nested samples. Econometrics Journal. 6(2), 401–407 (2003)

    Article  Google Scholar 

  • Kuksov, D., Villas-Boas, J.M.: Endogeneity and individual consumer choice. Journal of Marketing Research. 45(6), 702–714 (2008)

    Article  Google Scholar 

  • Louviere, J., Train, K., Hensher, D., Brownstone, D., Waldman, D., Greene, W., Ben Akiva, M., DeShazo, J., Carson, R., Cameron, T., Bhat, C.: Endogeneity in discrete choice models. Marketing Letters. 16(3–4), 255–265 (2005)

    Article  Google Scholar 

  • Mariel, P., Hoyos, D., Artabe, A., Guevara, C.A.: A multiple indicator solution approach to endogeneity in discrete-choice models for environmental valuation. Sci. Total Environ. 633, 967–980 (2018)

    Article  Google Scholar 

  • McFadden, D.: Modeling the choice of residential location. In: Karlquist, A., et al. (eds.) Spatial Interaction Theory and Residential Location, pp. 75–96. North Holland, Amsterdam (1978)

    Google Scholar 

  • Morikawa, T.: Correcting state dependence and serial correlation in the RP/SP combined estimation method. Transportation. 21(2), 153–165 (1994)

    Article  Google Scholar 

  • Palma, D., Hess, S., Guevara, C.A., Rizzi, L.I., de Dios Ortúzar, J., Casaubon, G.: When expensive is good: Addressing price-quality association in choice modelling. Working Paper. University of Leeds 2018).

  • Petrin, A., Train, K.: A control function approach to endogeneity in consumer choice models. Journal of Marketing Research. 47, 370–379 (2010)

    Article  Google Scholar 

  • Soofi, E.S.: Capturing the intangible concept of information. Journal of the American Statistical Association. 89, 1243–1254 (1994)

    Article  Google Scholar 

  • Swait, J.: Chapter 9: Advanced choice models. In: Kanninen, B. (ed.) Valuing Environmental Amenities Using Stated Choice Studies: A Common Sense Approach to Theory and Practice. Springer, Dordrecht, The Netherlands (2006)

    Google Scholar 

  • Swait, J., Adamowicz, W.: The influence of task complexity on consumer choice: A latent class model of decision strategy switching. Journal of Consumer Research. 28, 135–148 (2001a)

    Article  Google Scholar 

  • Swait, J., Adamowicz, W.: Choice environment, market complexity, and consumer behavior: A theoretical and empirical approach for incorporating decision complexity into models of consumer choice. Organizational Behavior and Human Decision Processes. 49, 1–27 (2001b)

    Google Scholar 

  • Train, K.E., Wilson, W.W.: Monte Carlo analysis of SP-off-RP data. Journal of Choice Modelling. 2(1), 101–117 (2009)

    Article  Google Scholar 

  • Wooldridge, J.M.: Econometric Analysis of Cross Section and Panel Data, 2nd edn. MIT press (2010)

  • Wooldridge, J.M.: Control function methods in applied econometrics. The Journal of Human Resources. 50(2), 420–445 (2015)

    Article  Google Scholar 

Download references

Acknowledgments

This paper contributes to the research program of the Volvo Research and Education Foundation Bus Rapid Transit Centre of Excellence. The research contribution is also linked to an Australian Research Council Grant No. DP140100909 (2014–2016) on ‘Integrating Attribute Decision Heuristics into Travel Choice Models that accommodate Risk Attitude and Perceptual Conditioning’ and ARC-DP Grant No. (2017–2019) DP170100420 on Business Location Decisions. An earlier version was presented at the 2019 International Conference on Choice Modelling, Kobe, Japan. We thank Angelo Guevara for his comments on an earlier draft as well as three referees and the Associate editor Toshiyuki Yamamoto.

Author information

Authors and Affiliations

  1. Institute of Transport and Logistics Studies, The University of Sydney Business School, The University of Sydney, Sydney, NSW, Australia

    David A. Hensher, Camila Balbontin & William H. Greene

  2. Erasmus School of Health Policy & Management and Erasmus Choice Modelling Centre, Erasmus University Rotterdam, Rotterdam, The Netherlands

    Joffre Swait

Authors
  1. David A. Hensher
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Camila Balbontin
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. William H. Greene
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Joffre Swait
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

David Hensher designed the approach, wrote all of paper except jointly interpreted the results; Camila Balbontin did model estimation and contributed to analysis of findings; William Greene reviewed the entire paper and edited the section on control functions while correcting some notation in the model specification section; Joffre Swait reviewed the entire paper and provided suggestions on how to explain the relationship between scale, alternatives and individuals.

Corresponding author

Correspondence to David A. Hensher.

Ethics declarations

Conflict of interest

On behalf of all authors, the corresponding author states that there is no conflict of interest.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Hensher, .A., Balbontin, C., Greene, W.H. et al. Experience as a conditioning effect on choice: Does it matter whether it is exogenous or endogenous?. Transportation 48, 2825–2855 (2021). https://doi.org/10.1007/s11116-020-10149-1

Download citation

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s11116-020-10149-1

Keywords

Search

Navigation