# An Exploratory Study of Honey Consumption Preferences: Insights from a Multi-Model Approach in Kosovo

## TL;DR

This study explores how people in Kosovo consume and value honey, finding that local production, knowledge, and trust influence their choices and willingness to pay.

## Contribution

The novelty lies in combining multiple statistical models to analyze honey consumption behavior in Kosovo.

## Key findings

- Honey consumption is high, with a strong preference for locally produced honey.
- Consumer willingness to pay varies based on knowledge, income, and trusted information sources.
- Perceived quality and product attributes like botanical and geographical origin significantly influence purchasing decisions.

## Abstract

This study examines consumer behavior, preferences, and knowledge regarding honey in Kosovo to inform more effective production, marketing, and policy strategies. Data were collected from 503 respondents through an online questionnaire and analyzed using a combination of artificial neural networks (ANN), decision tree modeling (CHAID), and ordinal logistic regression. The results show a high prevalence of honey consumption, strong preference for locally produced honey, and significant variability in consumer willingness to pay (WTP) based on knowledge, income, and trusted information sources. ANN identified recommendations and product familiarity as primary predictors of WTP, while the decision tree highlighted knowledge and income as key variables for segmentation. The ordinal logistic regression confirmed the importance of perceived quality and product attributes, particularly botanical and geographical origin, in shaping purchasing decisions. The use of complementary statistical models enhanced both predictive power and interpretability. The findings highlight the crucial role of consumer education and trust cues in fostering sustainable honey markets in Kosovo.