DECISION SCIENCES:
MULTIPLE CRITERIA DECISION MAKING: The development of interactive decision aids, to assist in the analysis of decision problems with multiple and conflicting objectives, with particular reference to natural resource management (TJ Stewart, L Scott, A Joubert and others).
DECISION THEORY: General principles of Bayesian statistical analysis; use of these in sequential stochastic optimisation; quality control and reliability studies (TJ Stewart, RG Guo).
OPERATIONAL RESEARCH in general and in development (TJ Stewart, J Nyirenda)
The Decision Sciences Research Group is committed to encouraging and stimulating the rational use of data in decision-making, planning and research, both in the public and private sectors.

ECONOMETRIC MODELLING: Econometric techniques are being used to test theories related to the South African economy in the fields of finance, monetary economics, interest rate theory and stock market research (GDI Barr, CG Troskie). The research of the Econometrics and Financial Analysis Group focuses on macroeconomic model building, with particular interest in the special problems related to the South African economy.

GIS and SPATIAL MODELLING: Statistics of large geoscience datasets. Geographic information systems (GIS). Geostatistics and spatial modelling (C Thiart).

EDUCATIONAL APPLICATIONS: Statistical examination of data pertaining to schools, disadvantaged students and to science education (TT Dunne, L Zacna, K Stielau, F Gumedze).

BIOSTATISTICS: Medical applications of statistics (TT Dunne, F Little, F Gumedze). The objectives of the Biostatistics Interest Group are to develop statistical methodology motivated by medical problems, particularly in the area of Community Health, and to provide statistical support to medical researchers in the form of short courses and consulting.

MULTIVARIATE ANALYSIS: Detection of outliers and influential observations (CG Troskie, TT Dunne, C Thiart, F Gumedze, A Clark); multivariate distribution theory (CG Troskie); multidimensional scaling, correspondence analysis and cluster analysis (LG Underhill, K Ramaboa); robust regression procedures (CG Troskie, C Thiart).