This paper presents an analysis of core-collapse supernova distributions in isolated and interacting host galaxies, paying close attention to the selection effects involved in conducting host galaxy supernova studies. When taking into account all of the selection effects within our host galaxy sample, we draw the following conclusions. 1. Within interacting, or 'disturbed', systems there is a real, and statistically significant, increase in the fraction of stripped-envelope supernovae in the central regions. A discussion into what may cause this increased fraction, compared to the more common Type IIP supernovae and Type II supernovae without subclassifications, is presented. Selection effects are shown not to drive this result, and so we propose that this study provides direct evidence for a high-mass weighted initial mass function within the central regions of disturbed galaxies. 2. Within 'undisturbed' spiral galaxies the radial distribution of Type Ib and Type Ic supernovae is statistically very different, with the latter showing a more centrally concentrated distribution. This could be driven by metallicity gradients in these undisturbed galaxies, or radial variations in other properties (binarity or stellar rotation) driving envelope loss in progenitor stars. This result is not found in 'disturbed' systems, where the distributions of Type Ib and Ic supernovae are consistent.