Rafael Wittek: Resource competition and violent conflict
survival. This will occur at a much later stage than in *restrictive" societies, and may be
unnecessary at all, if the resource base could recover during this time.
From what has been said up to now, one general hypothesis can be derived: Socie-
ties marked through partial scarcity will have more violent conflict than societies with
total scarcity. With regard to the two models of scarcity and the independent variables
used to measure them, this assumption has to be specified as follows:
(1) Societies with food stress and social stratification will fight more often than so-
cieties with food stress and no stratification.
(2) Societies with high population densities and social stratification will fight more
often than societies with high population densities and no social stratification.
If we add these two hypotheses to the postulated relationships between the inde-
Pendent variables as stated above, a four-variable model is the result. It is graphically
depicted in Figure 2.
POPULATION
DENSITY
+
STRATI- VIOLENT
FICATION CONFLICT
+
* y
FOOD
STRESS
Fig. 2. Socio-ecological model of violent conflict
Statistica] tests
Both the type of hypotheses and the type of data at hand determine the choice of the
"PPropriate statistical method. In the present case, the method should take into consi-
*ration the interrelationship of the independent variables and the categorical mea-
"urement level of variables. Though for dichotomous dependent variables these condi-