Path analysis allows propositions to be represented in context by means of a formal model, which makes possible testing of its various aspects. The technique was developed for genetics by Wright & Li & was made applicable to the social sciences by Boudon & Blalock. It is assumed that variables are all measured on ratio or interval level, that the connections are linear & monocausal, & that there are no interaction effects. remainder variables of endogen variables are assumed not to be correlated to exogene variables. Using these assumptions, the theory & technique are explained. Complications such as those having to do with 2-sided causality & the problematics of unmeasured variables are discussed, & the Blalock procedure (partial r analysis) is compared to path analysis whose advantages offer the opportunity to measure the strength of diverse influences, as well as indirect effects. P. Tiersma.