Variables affecting the levels of Hispanic participation in Texas extension programs

Predicated on the rapidly changing ethnic demographics in Texas and 69.3% parity levels for Hispanic participants in Extension programs, this causal-comparative study examined relationships and causality between institutional, methodological, competency, and demographic variables and levels of Hispanic participation in Texas Extension programs. Parity was used as measure of participation and calculated as a 3- year mean percent of Hispanic participation from 2001-2003. Statewide, parity levels ranged from 38.9% - 99.7% for state goals and from 6% to 251% for statewide initiatives. Hispanic participation in county programs varied from 0% to 409% parity.

Variables were organized categorically in a web-based instrument measuring perceptions through a Likert-type scale. A pilot test resulted in acceptable internal validity of .77 using Cronbach's alpha. Five research questions were posed and four null hypotheses were tested. Confidence levels were set at .05 a priori and data were analyzed through descriptive statistics, bivariate correlations, and linear regression models. A population of 332 county faculty from TCE (1862) and CEP (1890) were included. Using the Hardin-Brashears Bi-Modal method provided a 64% response rate. Tests for non-response error found no significant differences between early and late responders.

Regression models provided statistical significance for institutional (I), methodological (M), and competency (C) variables. Institutional variables were negatively correlated to levels of Hispanic participation, while methodological and competency variables were positively correlated. Regression coefficients for the I,M,C model (R2=.20) were -29.31 for I and 45.19 for C&M variables. For the demographic regression model (R2=.35), four variables were significant, ANR (-36.36), 4-H (20.02), non-urban (-23.08), and committee parity (.10).

It was concluded that agent perceptions of adequate institutional support for Hispanic programming were associated with lower levels of Hispanic participation. Self reported implementation of selected methodologies and perceptions of certain competencies among agents were associated with higher levels of Hispanic participation. Demographic variables associated with high levels of Hispanic participation were committee parity, employer (1890), ethnic minority agents, FCS agents, and 4-H agents. Variables associated with lower levels of Hispanic participation were gender (male), ethnically white agents, non-urban counties, and ANR agents.