In the context of numerical modeling for the assessment of a radioactive waste repository in a tight clay formation, the question arises whether a standard two-phase flow formulation is sufficient to capture the dissipation of gas generated by the degradation of engineered barrier components, or whether more sophisticated constitutive relations, including those accounting for geomechanical effects in deformable materials, need to be incorporated. The appropriateness of using effective geomechanical and two-phase flow parameters in a performance assessment model is examined by inverse modeling, which is used to interpret the experimental data from oedometric gas injection experiments on core samples. Effective parameters are estimated using both a conventional two-phase flow formulation as well as alternative constitutive models that account for the volumetric deformation of the specimen. The conventional model is capable of reproducing the pressure and gas flow data observed in the experiment. However, axial strain data reflecting the expansion and compression of the specimen in response to gas injection are not reproducible by the simple pore deformation model. Joint inversions of all available flow and deformation data performed with different effective stress models lead to similar matches. This indicates that (a) the differences between the conceptual models are either minor, (b) they can be compensated by estimating effective model parameters, or (c) that the calibration data do not contain sufficient information to discriminate amongst the alternative models. Nevertheless, the use of a two-phase flow model with a refined pore deformation relationship appears appropriate for performance assessment calculations.