Role of Sensory Evaluation of Pears in Consumer Choice: Evidence From a Field Choice Experiment in Italy (Poster)

Abstract

When making a purchase decision most consumers evaluate a series of intrinsic (e.g. physical attributes of the product) and extrinsic purchase attributes that characterize food products. Both extrinsic and intrinsic attributes are usually highlighted through labels or brands and consumers are indirectly faced with trade-off between these attributes. Previous research studies have used different methodological approaches such as (i) hypothetical and (ii) non-hypothetical choice methods (Gallardo, 2011; Wayua et al. 2009) as well as (iii) a combination of sensory and economic experiments (Combris et al. 2009; Mueller et al. 2010) to investigate the influence of these food product characteristics on consumers’ preferences and willingness to pay (WTP). However, to our knowledge no previous studies have investigated the role of sensory evaluation on consumers’ WTP for a food product using a combination of sensory tests, such as visual and tasting, and a field choice experiment. To fill this void, we designed a field choice experiment aimed at analyzing how the consistency and the importance that consumers place on sensory characteristics affect consumers’ WTP for two different types of pears: Angelys and Kaiser. Specifically, a full labeled design was generated with 3 price levels assigned to the two pear varieties, thus obtaining a total of 2*3=6 possible choices. Using the choice sets obtained from the full factorial experimental design, four hypothetical surveys are carried out, each with about the same number of subjects (about 80 respondents for each treatment). The first treatment is what we term “hypothetical survey”, where we asked respondents to select among three options: two pear varieties at different prices, and one “no buy/choice” option. This constitutes the control group. The second treatment is the “real visual survey” where respondents are allowed allowed to make a visual inspection of the product in a typical supermarket-like shopping behavior before doing the hypothetical choice experiment. A series of questions on consumer acceptability of sensory attributes linked to the appearance of the pear profiles are also asked. The third treatment is the “real taste” survey aimed at collecting consumer choices after respondents taste the two products (but they are not allowed to visually inspect the product on the shelf) and answer a series of questions on consumer acceptability of sensory attributes linked to taste only. Finally, the fourth treatment is the “real visual and taste survey” where respondents are allowed to to make a visual inspection and taste the pear products before the choice experiment. All respondents in the four treatments are also asked to complete a short questionnaire about socio-demographic characteristics.The sensory and choice experiment data were collected in supermarkets in two locations in the north of Italy: Milan (large city) and Verona (medium-size city). We will be analyzing the data to allow us to evaluate consumer preferences and WTP for the two varieties of pears, and evaluate the role of sensory characteristics on consumer preferences. In particular, a mixed logit error component model (MxLEC) will be estimated for all treatments, resulting in a total of 4 estimated models. The final specification of the utility function will include an alternative-specific constant representing the no buy/choice option, the two types of pears, and the price for all treatments. Also, interaction terms between the two types of pears and the socio-demographic information as well as the variables included in the visual and tasting sensory test questionnaire will also be considered in the models. Using the coefficients that we will obtain from the MxLEC models, a distribution of 1,000 WTP values for each pear type will be generated using a Krinsky & Robb (1986) bootstrapping method. Finally, the computational method suggested by Poe et al. (2001) will be used to measure differences of empirical WTP distribution across the four treatments. Relevant methodological and empirical implications can be drawn from this work. From a methodological point of view, investigating the effect of sensory consumer evaluation on consumer choice is an important issue in order to understand how visual and tasting aspects can affect consumers’ preferences and WTP in the case of choice experiment studies. Empirically, we expect to find significant effects of visual and tasting sensory aspects on consumers’ WTP since these attributes have been found to be important aspects in many consumer studies.