Let me start by saying I am not a composite engineer, nor mechanical engineer, but a frustrated plant operations manager trying to walk between multiple disciplines and form some degree of opinion based on input from aforementioned engineers. We are testing a composite tube with metal ends attached. This part is placed in a torque machine in an effort to derive  multiple sets of data for future development. In this process we have been calculating Young's modulus. My staff tells me they are only considering the area between the two metal ends for the part length. My question is should we consider the entire length which would be the inclusion of the area that has metal glued to it? Hope that makes sense.

well, that all depends on how long the metal bit is I suppose, how accurate you need the results and where you are measuring the (I assume) rotation from to calculate E, the rotation you measure is a function of the effective youngs modulus (E), length and torsional stiffness  

That explanation makes sense to me. We have 4140 steel ends that are 2 piece where the composite tube slides into a groove. Thus we have steel ID and steel OD with composite glued between them. I would assume then that sense this connection point would be significantly more rigid than the composite alone it should not be considered.Thanks.