Resumen de Consumer preference heterogeneity towards olive oil virgin extra: hypothetical and non-hypothetical choice experiments
Ahmed Yangui
The present dissertation aim a contributing to both agricultural economics and marketing literature by addressing specific issues related to discrete choice models and choice experiments. More precisely, this thesis focuses on two main issues: 1) new tools to tackle with preference heterogeneity; and 2) new response formats to allows researchers to take into account the information provided by no chosen profiles. These two issues have generated three studies, which form the main core of this thesis: two are related to issue 1) (Chapters 2 and 3), while the third one is related to the issue 2) (Chapter 4). In the first one, we evaluate consumers' preferences heterogeneity using a methodological framework with two novelties over past studies: 1) it accounts for both preference heterogeneity around the mean and the variance of random parameters; and 2) it considers both socio-demographic characteristics of consumers as well as their attitudinal factors. Results suggest that the proposed framework significantly increase the good-of-Fit and provides more useful insights for policy analysis. The most important attribute affecting consumers' preferences towards extra virgin olive oil are the price and the product's origin. The consumers perceive the organic olive oil attribute negatively, as they think that it is not worth paying a premium for a product that is healthy in nature. The second paper hypothesizes that differences in consumers' personality traits, such as food-related personality traits, purchasing habits and lifestyles, affect consumers' preferences for extra virgin olive oil. The methodological framework is based on the specification of an extended hybrid choice model (HCM), which was estimated following a two-step procedure. In the first step, a structural equation model was estimated to test hierarchical relationships between latent variables to explain purchasing intentions towards an organic olive oil. In the second step, the resulting latent variables were introduced in a random parameter logit (RPL) model to investigate the main determinants of consumers' choices related to extra virgin olive oil. The results from this study reinforce the need to include the psychological characteristics of consumers to better explain how individuals make food choices and to better understand the decision maker's process. Interestingly, Catalan consumers perceive a disutility from the organic attribute compared to other production system alternatives (conventional and PDO), while subjective norms and a higher perception of behavioural control only partially mitigate this effect. Environmental or health concerns seem to not be relevant to consumers' choices related to organic olive oil as the conventional olive oil is already perceived as a healthy product per se. In the third paper, we compares the ability of hypothetical and non-hypothetical choice experiment respect to incentive compatible ranking conjoint analysis and incentive compatible sequential best worst scaling. The comparison done in terms of estimated partworths, internal and external predictive power, estimated WTP, and participants' response consistency. In general, the results reveal higher preferences regularity between the respondents across the different treatments implying not statistically difference in the marginal participants' WTP. Additionally, the participants behave similarly whether there are asked to choose or to state their most preferred through the two ranking elicitation mechanism. The best worst scaling (BWS) format has been revealed to outperform the other formats in terms of predictive power as its cognitive process seems to better fits the natural tendency of humans at identifying the extreme values.
