use. On the other hand, Hooks, Napier and Carter's (1983) research shows no correlation

between use of certain farm technologies and resource access.

 There may be some association between building one's confidence and control

beliefs and having the ability to build skills and knowledge by having access to resources

(Sanderson, 2004). For example, high TAP users, who are younger and less educated, do

not have reliable and consistent access to traditional resources. They may also not have

the knowledge about how to access traditional resources. Thus, even if high TAP users

want to use TAP, they may only be somewhat confident and feel somewhat in control of

their ability to do so because of lack of resource access. They do not differ from low

TAP user groups members because low TAP user groups do not necessary need to access

traditional resources about farming.

 Hypothesis Four

4. The theory of planned behavior will better explain why farmers' use TAP than any
 of the other theoretical models.

 My research results do not corroborate this hypothesis, based on the logistic

regression. Socioeconomic characteristics are a better predictor of Cherokee farmers' use

of TAP than either TPB or diffusion innovation variables. Subjective norms, attitudes

and modem self-identity were different between groups. However, none of these

variables individually were significant predictors of behavior in a regression model. The

model as a whole was not predictive of behavior.

 The socioeconomic characteristics, acreage cultivated, hours per week working on

farm, number of years farming and selling of crops were significant predictors of

behavior as a whole. However, none of the variables were individually significant

predictors of behavior. In both regressions the individual variables probably have high