The main research problem is the low productivity of small scale maize farmers largely as a result of low adoption rates of recommended practices that could enhance yield levels and improve their incomes and livelihoods, and the inability of extension workers to effectively influence farmers’ decision making process by their lack of appropriate predictive extension planning tools. The problem faced by extension workers is the lack of understanding of the wants and needs of farmers, their preferences and behavioral inclinations towards agricultural innovations. In order to contribute to the understanding of farmers’ behavior change, the study sought to compare the relative influence of personal and social characteristics of farmers with intervening variables as conceptualized by Düvel and Ajzen’s subjective norms, with the objective of determining their predictive potential of farmers behaviors for extension program planning purposes. A secondary objective was to search for additional variables to explain farmer’s adoption behavior by exploring the predictive value of the subjective norm concept The study was carried out in the Leribe and Maluti-a-Phofung districts of Lesotho and South Africa respectively. A structured questionnaire with a Sotho translation was used to collect data from 107 farmers randomly selected from the districts and administered by trained extension staff. The data collected was analyzed using the social sciences (SPSS). In determining the relationships between the independent and dependent variables, Chi-square test of independence, correlation and regressions analysis were used.
In all 10 independent and nine intervening variables were selected for the study. The independent variables were location of the farm, membership of farmers association, gender, age, level of education, experience in farming, off-farm income, amount of time spent farming, total farm size and area under maize cultivation. The intervening variables were efficiency perception, need compatibility, need tension, awareness and prominence constituted cognitive aspects derived from Düvel’s Model and the social dimensions adapted from Ajzen’s subjective norm concept were, important people, extension agents, close friends and membership of farmers association.
The results suggest that farm size and area under maize cultivation were the only variables that showed any consistent influence with adoption of recommended maize agronomic practices namely: use of improved seeds, lime and fertilizer applications. The association between the remaining variables seems to be more dependent on the type of recommended practice. For example location was found to be significantly associated with the adoption of fertilizer and top-dressing practices but not with lime and the use of improved seed. All the remaining independent variables gender, age, educational level, experience and time spent on the farm appear not to have any significant influence on the adoption of the recommended practices at five percent level of probability. Compared to the independent variables, five out of the nine intervening variables, namely: prominence, awareness, need compatibility, efficiency perception and need tension were consistently found to be highly significantly associated with the adoption of all the four recommended maize agronomic practices at 5% level of probability. On the other hand, the subjective norm variables did not show any consistent association with adoption behaviors of respondents In general the analysis suggest a lower than expected contribution to variation as the results contradicts the hypothesis that farm and farmer characteristics influence adoption behavior. This is supported by the fact that, except for top- dressing where the characteristics of the farmer and farm contribute about 40% to the explanation to total variation, the rest all fall below 20 %. In contrast, the evidence shows that the intervening variables – those with cognitive dimensions, showed a high degree in explaining variation in the adoption behavior in all the production practices studied. The power of explanation ranged from 49% in the case of adoption of improved seed practices to 77.7% for the use of lime. The results provide strong evidence in support of the contention that, the intervening variables of cognitive in nature, are the likely precursors of decision making through which the influence of independent variables become manifested in behavior. The results also show that need, perception and knowledge related variables mediate between intentions, personal variables and the environment and decisions on adoption. This study confirms and opens the way for the search for more intervening variables with the potential to extend the epistemology of extension science.