Partially Balanced Incomplete Randomized Group Design (RAKTLSP) is a design with v treatments arranged into b groups with each group consisting of k treatments (k<v) and there are treatment pairs that appear together in the same group λ_m times. The purpose of this study is to be able to construct an experimental design plan and to be able to carry out ANOVA preparation, follow-up tests, testing model assumptions from the Right Angular Association scheme Partial Incomplete Randomized Block Design. This study uses the Right Angular association scheme which is a 4-class association scheme used to determine the first, second, third and fourth associations of each treatment. Based on the association formed, the number of treatment pairs that appear in each group will be obtained (λ_m, m=1,…,4). The test that was carried out was an influence test on the treatment because only the treatment was important where the treatment was adjusted (adjusted). The assumptions needed are the assumption of residual normality, equality of variance and independence with the Tukey test as a follow-up test. The data used as an example of application in this study is data in agriculture by observing the effect of the type of hebricide on weeds, which consists of three types of herbicides with 4 dose levels, namely the single herbicide Bentazon 400 g.L⁻¹ and MCPA 60 g.L⁻¹ (2; 1; 0.5; 0.25; L.ha⁻¹), mixed herbicide Bentazon + MCPA 460 g.L⁻¹ (2.5; 1.25; 0.625; 0.3215; L.ha⁻¹). The results of the Right Angular Association scheme, Partially Balanced Incomplete Randomized Block Design show that the type of herbicide has an effect on the level of damage to weeds. By using this method it can be seen that the recommended hebricide is type J, namely the single active ingredient MCPA 60 g.L⁻¹ (1 ; L.ha⁻¹)

RAKTLSP, Right Angular Association, Tukey's Test, Normality, Equality of Variances, Independence