Area-of-influence (AOI3) experiments measure the effect of a single weed on crop growth at intervals away from the weed plant. Effects of treatment variables, e.g., weed species or control measures, on the AOI of a single weed can be estimated. AOI experiments can be analyzed by regression of crop growth on distance from the weed plant, but this analysis violates an important regression assumption: independece of observations. Statistical dependence can occur among successive observations along the row because uncontrolled sources of variation are likely to act in similar ways on adjacent individuals. Multivariate analysis of variance (MANOVA) is a statistical technique that accounts for dependencies among crop growth measurements along the row. The technique tests three hypotheses: first, that different treatments cause weed AOI to differ in spatial distribution of competitive effects; second, that different treatments cause weed AOI to differ in size; and third, that the weed has an effect, i.e., crop growth near the weed differs from growth away from weed. MANOVA can be applied to most common experimental designs, e.g., randomized blocks or split plots, and can be implemented on various mainframe and microcomputer statistical packages.