Figure 2.

Correlation matrices of 18 variables of three different school types in Kupang (West-Timor) depicts the correlation matrices of the private, the state, and the underprivileged schoolchildren and highlights the differences between the school types. High correlations existed between the two measures of length (total body height and sitting height), and between the five measures related to the nutritional state. The correlations between height and the measures of the nutritional state however, were low and strongly depended on the school type (Table 3). The correlations between height SDS and variables related to the nutritional state seemed to almost completely disappear in the children of the underprivileged school. Also the interaction of the five measures of the nutritional state significantly declined in the underprivileged school children.