format. These interviews asked the organizational leaders to explain what they expect

from county board members and what they want these board members and county boards

to accomplish.

 The second subset of the population was composed of members of local county

Farm Bureau boards. A sample of this sub-population completed the survey derived from

the state leader responses to the interview questionnaires. This portion of the study

determined existing board member leadership behavior. Salant and Dillman (1994)

describe how a sample size is chosen. There are approximately 666 county Farm Bureau

board members in Florida (FFBF, 2004). Using a table provided by Salant and Dillman

(1994), the researcher chose a 50/50 split with a 5% sampling error, a sample of 279

individuals was chosen (p. 55). The researcher was given the names and addresses of all

county board members and randomly chose 279 participants to receive the survey. They

were chosen through a process of systematic sampling, with the first element in the

sample chosen from a random numbers table. After that number was chosen, every other

individual on the list was included in the survey.

 The third subset of the population was a sample of active Florida Farm Bureau

 members. The researcher was given a computer generated random list of active members

 and mailed surveys to 419 of these members. In this study, one of the subsets of the

 population was the active members of Florida Farm Bureau. The size of this subset is

 36,100 (P. Cockrell, personal communication, September 10, 2002). Using a 50/50 split

 with a 5% sampling error, 419 members of this subpopulation were selected to

 participate in this part of the study.