The various faces of scientific productivity: a contingency analysis