Logit and Tobit Model Analysis for Determinants of Common Bean Seed Replacement and Its Rate in Southern Ethiopia

Abstract

The objective of this study was to identify determinants of haricot bean seed replacement and rate of its use in Boricha district of Southern Ethiopia. Data were collected from 162 randomly selected common bean farmers through multistage sampling techniques. Inferential statistics namely the Logit model and the Tobit model were used to identify determinants of seed replacement among respondents and the extent of replacement rate respectively. Among 162 farmers, 105 replaced certified seed of haricot bean at least once in the past three years whereas 57 farmers did not use certified seed in the same period but used home saved seeds. On average seed replacers change their seed stock 1.82 times every three years. The results revealed that, household size, livestock ownership, frequency of contact with extension agent, membership in social organizations, distance from formal certified seed and respondents’ opinion on affordability of certified seed price were positively and significantly influenced seed replacement. Negative but significant influence of household head age on seed replacement decision was found. The rate of replacement of certified seed was influenced positively by perception about seed price, haricot bean farming experience, cash income from farm, contact with development agent, livestock ownership, membership in social organizations, and distance from formal certified seed source while age of household head negatively influence replacement rate. Therefore, the findings suggest that seed replacement and its replacement rate should be improved by enhancing asset ownership, delivering input on-time, improving packaging volume based on farmers demand and providing extension service.

Keywords: Common bean farmers, Logit, Multistage, Seed replacement, Tobit