The investigations carried out in the course of this thesis had the aim to examine the concept of a physical model of a quality controller for injection molding, which refers to the molded part’s mass, and to expand the model with the quality parameter molded part’s length. The evaluation of the validity of the model was executed with a representative part on the basis of defined experimental planning, in which the two quality parameters’ dependences on the influence parameters holding pressure, nozzle temperature, holding pressure time and mold temperature were identified. The material used in the investigations was an ABS grade. The results of the physical model were compared to relationships determined by a statistical model within a 2^4-factorial design with center point. In doing so, good matches could be found. The other basic issue of this thesis was to repeat the experiments for the physical and the statistical model with the simulation program Autodesk Moldflow Insight (AMI) using the experimental process parameters and to examine how the experimentally observed relations can be predicted numerically. The results generally showed good correlation between experiment and simulation. The calculated curves however differed in the overall level of their absolute values. Subsequently it could be successfully verified, that the calculated data can be calibrated to the level of the real curves, matching these in very good approximation. Therefore, in addition to the execution of the factorial design in the simulation, it is only necessary to determine the real values of the quality parameters of the molded part with the settings of the center point in the course of a single test series on the injection molding machine. The predicted curves can then be shifted to the level of the real relationships using this experimentally determined center point.