A quantum transition can be seen as a result of interference between various pathways (e.g.Feynman paths) which can be labelled by a variable f . An attempt to determine the value of f without destroying the coherence between the pathways produces the weak value off . We showf to be an average obtained with amplitude distribution which can, in general, take negative values which, in accordance with the uncertainty principle, need not contain information about the actual range of f which contribute to the transition. It is also demonstrated that the moments of such alternating distributions have a number of unusual properties which may lead to misinterpretation of the weak measurement results. We provide a detailed analysis of weak measurements with and without post-selection. Examples include the double slit diffraction experiment, weak von Neumann and von Neumann-like measurements, traversal time for an elastic collision, the phase time, the local angular momentum (LAM) and the 'three-box case' of Aharonov et al.