The elastic modulus of thin films can be directly determined by instrumented indentation when the indenter penetration does not exceed a fraction of the film thickness, depending on the mechanical properties of both film and substrate. When it is not possible, application of models for separating the contribution of the substrate is necessary. In this work, the robustness of several models is analyzed in the case of the elastic modulus determination of a porous aluminium oxide film produced by anodization of an aluminium alloy. Instrumented indentation tests employing a Berkovich indenter were performed at a nanometric scale, which allowed a direct determination of the film elastic modulus, whose value was found to be approximately 11 GPa. However, at a micrometric scale the elastic modulus tends toward the value corresponding to the substrate, of approximately 73 GPa. The objective of the present work is to apply different models for testing their consistency over the complete set of indentation data obtained from both classical tests in microindentation and the continuous stiffness measurement mode in nanoindentation. This approach shows the continuity between the two scales of measurement thus allowing a better representation of the elastic modulus variation between two limits corresponding to the substrate and film elastic moduli. Gao's function proved to be the best to represent the elastic modulus variation.