3. Geographic segmentation of consumer preferences
We review evidence from multi-country stated-preference surveys of urban rice consumers in SA and SEA to further deepen our understanding on geographic segmentation of preferences for intrinsic grain quality attributes and generate empirical evidence for supporting research priorities for targeted rice breeding.
3.1. Data collection and sampling
Door-to-door, face-to-face interview surveys were conducted in 24 key urban cities in seven countries from May 2013 to January 2014. Multi-stage sampling was used in all countries. Stratified sampling was first employed followed by random selection of primary sampling units (PSU) within each stratum. Households were then selected within PSU using systematic sampling by locating a starting point (e.g. village hall) and following the right hand rule from the starting point and after every successful interview until the target number of interviews from each starting point was achieved.
Target respondents included males and females between 18 and 64 years old from different socio-economic classes (SEC). SEC, as an indicator of household's affluence level, is country-specific. It is a spectrum whereby a household will fall into one of the classes A, B, C, D or E with SEC A as the most affluent and SEC E on the other end of the spectrum. Target respondents should be involved in grocery decision-making and in cooking or preparing meals for the household. With this criterion, majority of respondents (90%) were females. Respondents were interviewed using a structured questionnaire translated in the local languages. Spontaneous and aided types of questions were asked to elicit information. Visual stimuli were used for aided questions. In the preference elicitation question, a photo card was used to verify responses pertaining to grain size and shape to have a common reference point among respondents and increase comparability of results. The response options for grain size were short, medium and long to extra-long grains. The response options for shape were slender, medium and bold grains. Respondents who mentioned aroma were also asked to describe the fragrance that they prefer to further elicit aroma descriptors.