A. Data collection The study was carried out at Kapasia Upazila of Gazipur district during January to March, 2005. Two villages under Durgapur union were purposively selected for farm-level data collection. As the sampling frame was not readily available, a total of 60 farmers were selected randomly to collect primary data. A predesign and pre-tested survey schedule was used for this purpose. Direct interview technique was followed.
B. Operationalization of the independent variable Age: The age of the farmers was measured in terms of years. The farmers were categorized as young (S35 years); middle (36-50 years) and old (> 50 years) on the basis of their age.
Education: Education was measured by considering the total number of years of schooling completed by a farmer. The education of the farmers was categorized as illiterate (No schooling), Primary level (Class Ito V), Secondary level (Class VI to X), and above secondary level (Class XI and above). The person who can sign only was categorized under the primary level.
Family size: Family size was measured in terms of a number of family members including farmer himself, wife, children and dependents who live under the same roof and share the same kitchen were categorized as small (< 5), medium (5-7) and large (>7). Farm size: The size of the farm was determined by actual land under possession by the respondents and classified as marginal (up to 0.49 ha), small (0.50 to 0.99 ha), medium (1.00 to 2.99 ha) and large (3.00 ha and above). Annual income: The tentative price of farm output (eg. Crops, dairy, poultry and fruits etc.) as well as income derived from other sources like business, jobs of any kind was taken as personal income and was categorized as low (up to Tk. 35000), medium (Tk. 35000 to 80000) and high (above Tk. 85000).
Cosmopoliteness: Cosmopolitenes of the respondents were computed on the basis of the respondents' visit to five different places outside of his village. The respondents indicated whether they visited those places frequently, occasionally, rarely, or never. Weights assigned to those visits were 3, 2, 1 and 0 respectively. Extension Contact: The extension contact of the respondents was computed on the basis of their contact with sources of information. Respondents were asked to determine their extent of contact with those information sources. A 4-point scale such as never, rarely, occasionally and frequently were used to determine the extent of contact. Weights were assigned as 0 for never, 1 for rarely, 2 for occasionally and 3 for frequently. Organizational participation: Organizational participation was measured by computing a formula - Organizational participation = E (level of participation X duration of participation). Weights were assigned as 0 for participation, 1 for a general member, 2 for executive member and 3 for a key position like chairperson or secretary. The duration of participants was scored as 1 for one year, 2 for two years and 3 for three years and above.
C. Operationalization of the dependent variable Farmers' attitude towards agroforestry was considered as the dependent variable. Respondent's attitude towards agroforestry was measured by asking his opinion regarding the socio-economic aspect of their livelihood due to the contribution of the agroforestry system. It was measured on the basis of opinion obtained from the respondents on 10 statements comprising positive and negative following the `Liken' method with a 5-point scale: 'strongly agree"agree', `undecided' disagree' and 'strongly disagree'. A score of 5, 4, 3, 2 and 1 was assigned for strongly agreed, agreed, undecided, disagreed and strongly disagreed statements, respectively. Based on the computed scores, the respondents were classified into three categories such as slightly favorable attitude, moderately favourable attitude and highly favourable attitude. In order to explore the relationship between the selected characteristics of the framers and their attitude towards homestead agroforestry, Pearson's Product Moment Coefficient of correlation (r) was computed.