The purpose of this study is to determine spatial pattern recognition of school performance based on
children’s anthropometric and motor skills component. This study involved 94 primary schools with a
total 2237 male students aged 7.30±0.28 years in Pahang, Malaysia. The parameters of anthropometric
(weight and height) and motor component included lower muscular power (standing broad jump),
flexibility (sit and reach), coordination (hand wall toss) and speed (20 meter run) were selected. Cluster
Analysis (CA) and Discriminant Analysis (DA) under Multivariate Method and technique of Kriging
Interpolation in Geographic Interpolation Software (GIS) were used. CA revealed two clusters of school
performance. There are a total 34 high performance schools (HPS) and 60 low performance schools
(LPS). Then, the assigned groups were treated as independent variable (IV) while anthropometric and
motor parameters were treated as dependent variable (DV) in DA. Standard mode of DA obtained
95.74% correctness of classification matrix with three discriminated variables (height, standing broad
jump and 20 meter run) out of six variables. Meanwhile, forward and backward stepwise mode of DA
discriminated only one (standing broad jump) out of six variables with 96.81% of classification
correctness. The map output of Kriging interpolation has shown graphically the pattern of discriminated
variables that greatly influence school performance. It exposed the ability of children motor skills
development in particular region is higher than another region.