Role of the intrinsic and extrinsic product attributes in the purchase intent of honey

Abstract

Both ancient and modern civilizations consider honey as a food with significant nutritional properties and as a natural product with important therapeutic application, where several factors might lead to honey consumption. Discrete Choice Experiments (DCE) were conducted on a sample of Lebanese honey consumers in order to evaluate the importance of; price, color, texture, and health messages in purchase decisions of honey. Three hundred subjects were asked to choose among different honey profiles identified by different combinations of attribute levels according to their preference. The subjects were also asked to complete a 17-staement questionnaire on health consciousness and to provide information on some demographic and behavioral parameters. The DCE data was analyzed by conditional logit and mixed logit with uncorrelated and correlated part-worths to estimate the relative importance and part-worth of the product attributes and attributes levels, respectively. Cluster analysis was performed on the health statements and then on the health statements and the part-worths calculated by the mixed logit with correlated part-worths models to group subjects with similar preferences and health attitudes-consciousness into clusters. The contribution of demographic and behavioral variables to the purchase intent of honey was also analyzed for each cluster. Part-worth values for the overall sample based on consumers’ responses showed that the purchase decision is dominated by color, health messages and price rather than texture and health attitudes. Color is the most important attribute in determining consumer preference for honey, followed by the health claim “May reduce the risk of certain types of cancers and heart disease because it contains anti-oxidants”. Price was highest for the highest price level 40$-kg indicating that respondents consider price an indicator of the product’s quality. Cluster analysis revealed three clusters when clustering was performed on health statements. Clustering