Accurate estimation of soil mechanical properties represents a crucial step for most engineering applications. Both in situ and laboratory testing fundamentally rest on mechanically deforming (actuating) the material and simultaneously measuring its response in terms of displacements and stresses (reactions). Facing this widely adopted scheme, key questions remain unanswered: 1) what is the optimal type and/or mode of actuation that can most effectively extract soil properties; 2) what types of measurements are most useful for inferring material constants? As a first step in the investigation of these questions, an inverse model for the direct simple shear (DSS) test is constructed, wherein measurable responses are used to back-calculate soil properties. Specimens with two different aspect ratios are considered to study the influence of the deformation mode. The effect of the choice of measurements (i.e., which displacements and/or stresses are observed) is explored by assessing inverse model performance considering the DSS test as a boundary value problem, with variable displacement and stress fields, versus the conventional interpretation as an elemental test. Parameter sensitivities and correlation coefficients are employed as quantifiable metrics to compare material characterization based on different aspect ratios and types of measurements, and to interpret the performance of inverse analysis.