This thesis concerns the development of a hadron calorimeter for future e+ e− collider experiments such as ILC. The detectors at the ILC are optimized for the particle-flow approach, which aims to reach a jet energy resolution of 3-4% by measuring each particle in a jet using the best sub-detector measurement. This can only be achieved by using highly granular calorimeters. The CAlorimeter for LInear Collider Experiment (CALICE) collaboration develops different imaging calorimeters with high granularity. The work in this dissertation presents the analysis of shower shapes with data recorded using a CALICE Analog Hadron CALorimeter (AHCAL) technological prototype. It is a sampling calorimeter made of layers of steel absorbers interleaved with 30 × 30 × 3 mm3 scintillating tiles with Silicon PhotoMultipliers (SiPM)-on-tile readout as active material. The prototype has been operated successfully at CERN in 2018 and acquired sizeable datasets using beams of muons, electrons and pions at different beam energies. Firstly, a successful calibration with more than 99.9% channels of the AHCAL prototype and stable operation over several weeks of testbeam periods is presented. Secondly, GEANT4 simulations are compared to testbeam data as a cross-check to detector calibration. In addition, the conventional calorimeter properties such as the detector response and single-particle energy resolution is evaluated. Using the obtained AHCAL data, this dissertation presents the first three-dimensional hadronic shower model describing the evolution of the average hadronic shower of a pion in the energy range between 10 and 200 GeV. Furthermore, this model holds potential implications for fast simulations and the Particle Flow Approach (PFA). Finally, a comparison is done to validate and understand the picture of an average hadron shower shape by exploiting the Monte Carlo truth information.