Masters Theses

Date of Award


Degree Type


Degree Name

Master of Science


Agricultural Economics

Major Professor

Billy J. Trevena

Committee Members

David Brown, Brady Deaton


The study was designed to develop and test a realistic model of farm management decision making. A general model of decision making comprised decision model that related net farm income, the indicator used for managerial success, as a function of economic resource endowment, the decision criterion and information; a perception of risk model that stated the perceived probability of success to be a function of economic factors affecting that perception; a dynamic learning model that related the results of a previous choice to the perception of alternatives in the next decision. The theory of achievement motivation was incorporated in the decision model and the learning model to explain choice under conditions of uncertainty and persistence of behavior at a particular choice, respectively. The data were obtained from a survey questionnaire and 1974 farm records kept by 108 farmers participating in the Resource Management Program. Measurement of the need for achievement, the fear of failure and information were obtained from the questionnaire. Values for economic resource endowment and net income were found in the farm reco-rds. Regression analysis was used to estimate the effect of the resource endowment, the decision criterion and information on net income and to determine the effect of economic factors affecting the perception of risk. The resource endowment explained the largest amount of variation in net income of any group of variables. Livestock, labor, buildings and credit capital proved significant while land and machinery did not. All but land, machinery and credit capital had the hypothesized positive signs. The decision criterion, while insignifi-cant, was found to have the largest effect, 2638.53, of any variable on net income. The coefficients of the information variables of educa-tion, training and information seeking were not significant. Altogether, the independent variables explained 38.4 percent of the variation in net farm income. The estimation of the perception of risk model was disappointing. Two measures of perceived probability of success were used as the dependent variable in separate estimations. The estimated equations accounted for only 2.5 percent and 3.8 percent of the variation in perceived probability. None of the independent variables were signi-ficant. A Chi-square analysis was used to evaluate the learning model. A contingency table of prior and posterior probabilities of selection of farm plans was devised for those 53 farmers who had altered their original plan. A Bayesian theorem was used to determine the posterior probabilities. The Chi-square analysis showed a significant difference between the probability of selecting a farm plan prior to its imple-mentation and after its alteration. An increase in farm prices took place over the same years as the study. This increase could be the cause of the change in perception of farm plan success. Both learning and increased prices are consistent with the statistical results.

Files over 3MB may be slow to open. For best results, right-click and select "save as..."