Computer-Aided Statistical Analysis of Agricultural Research Data: Multivariate Data Analysis


Regression analysis is one of the commonest analytical tool for carrying out multivariate data analysis in agricultural economic and social research. It involves estimating parameters, computing F-ratio for the analysis of variance for the model and t-values for individual parameters, calculating efficiency indicators of the parameters like the marginal physical product (MPP) etc. Although there are packaged programs for performing the regression analysis, none is particularly suited for agricultural economics research. The packaged programs are also difficult to use by non-programmers and the cost of acquisition tend to be high for most researchers and students of economics. Therefore their use is limited to programming specialists and a few that can afford the acquisition cost. There is therefore, the need to develop a program specifically for agricultural economics, that is more user-friendly and affordable for the average user. The objective of this paper is to report on a regression software developed in BASIC which meets the above constraints. The program allows for the estimation of a linear simple or multiple regression parameters for ten functional models, calculates the means (arithmetic, geometric and harmonic), variance and coefficient of variation. It also calculates the MPP and performs stepwise linear correlation and principal components coefficients of the variables. The program has been tested on standardised data and the output of the program compares favourably with that of packaged programs. It is presently running on a Microsystems.

22nd Annual Conference of the Nigerian Statistical Association, Niger State House of Assembly, Minna, September 22-25
Job Nmadu
Professor of Agricultural Economics and Dean, School of Agriculture and Agricultural Technology

Research interests are economic efficiencies of small scale farming and welfare effects of agricultural interventions.