Cosmological constant I have always wondered about how cosmological constant is characterized. So since it is still a hypothesis you often read the “cosmological constant measured to be ….”. Shouldn't the statement read “cosmological constant calculated to be ….” . Or Is it that such semantics does not matter.
 A: The cosmological constant can be measured just like any quantity.  In fact, the 2011 nobel prize in physics was given for just that.
One of the earlier papers seriously analyzing the issue is linked here.  A more general overview can be found here, on scholarpedia; and a even more general one, targeted at a wider audience from Sean Carroll.  You can also look at this article on measurements with type Ia supernovae, specifically.

Edit:  On the nature of scientific measurement.
In astrophysics, measurements are very rarely done with rulers.  Instead, in the majority of cases, they are done by making comparisons between observations (done with rulers---in this case CCDs, etc) and the predictions of models.  This is the fundamental nature of scientific inference and deduction.  
For example, consider the 'measurement' of mass (any mass):
There is no way to directly measure mass, it can only be inferred using one of many physical laws.  For example, by measuring the force of gravity, and using $F = G \frac{m_1 m_2}{r}$.  Alternatively, one could use a spring of a known strength, measure the frequency of oscillation and use $\omega^2 = \frac{k}{m}$.  Both methods use inference.
As the desired parameter becomes more complex---for example, a dynamic property of the universe (you know, no big deal)---the inferences become more complex.  The state of the art technique for inferring model-dependent parameters based on observational measurements is called Bayesian Inference - which can be used to take into account uncertainty on the models themselves, and make comparisons between models.  In general, scientists use Bayesian Inference implicitly when comparing between different competing models.  One must take into account confidence in a model, in addition to the observational data, to figure out not only what the 'measured values' of parameters are --- but also which models are the most compatible with the data.  This is how competing models are compared.
In the end, there can never be 100% certainty in either a 'measurement' (of a parameter) or a model in-and-of-itself.  Instead a scientist can only become more and more confident in a range of parameters, or a sufficient accuracy of a model.
