Resumen
Different software product quality models interpret different amounts of information, i.e., they can capture and address different manifestations of software quality. This characteristic can cause misleading statements and misunderstandings while explaining or comparing the results of software product quality assessments. A total of 23 previously identified distinct software product quality models are analysed on how they handle the abstract notion of quality, and a taxonomy on the quality manifestations that the individual software product quality models are able to capture is established. Quality models that are able to solely describe the quality manifestation of the source code are attractive due to their full automation potential through static code analysers, but their assessment results ignore a huge part of software product quality, which is the one that most impresses the end user. The manifestations of software product quality that address the behaviour of the software while it operates, or the perception of the end user with regard to the software in use, require human involvement in the quality assessment. The taxonomy contributes to interpreting the quality assessment results of different quality models by showing the possible quality manifestations that can be captured by the identified models; moreover, the taxonomy also provides assistance while selecting a quality model for a given project. The quality manifestations used for the quality measurement always need to be presented, otherwise the quality assessment results cannot be interpreted in an appropriate manner.