WOSB submits a man to varying time, problem and social stresses in an experimental "micro-community" of about 8 members, committed to tasks of different types i.e. ou tdoor physical, indoor planning, group discussion etc. No leader is appointed and the group is allowed to throw up its own natural leaders. From observation of the man's group-effectiveness in several different activities under varying degrees of stress it gains a balanced impression of his general group-effectiveness which it checks and co-ordinates with tests of him as an individual either apart from the group or in charge of it. Coordinating group tests with individual tests is an important technical device in WOSB.