This single-page document emphasizes the need for experimentalists and theorists alike to ALWAYS define their strain measure. For every percent increase in strain, the most popular measures of strain will disagree by as much as 1.5%. This might not sound like much, but try running a simple shear Von Mises strain cycle using log strain and engineering strain. You will find that the engineering strain calculation produces anomalous PRESSURES because volumetric strain does NOT equal the trace of strain EXCEPT for logarithmic strain.
You may download the rest of the document here.