VK4

Alignment of Software Quality and Service Quality

Quality is a concept that has been discussed a lot during the last ten or fifteen years. From the beginning the quality concept was mainly used in connection with industry products. But already at that time, the products were accompanied by some kind of service, e.g. installation aid or maintenance. So the quality was assessed both in the product itself and in the service.

From the product area the quality concept and the task of evaluating quality has spread into pure service areas and into software. The development of evaluation models has taken different tracks in the different areas of application. But at the same time the areas of service and of software have approached each other. This is the background that makes this article interesting.

The author poins out that software systems are also service systems. More and more, service becomes an integral part of the products. However, the language used to describe requirements at business level (quality of service systems) and the lagnuage used to describe requirements at system-component level (quality of software components) differ a lot. This study explores if a common ground can be found between quality models developed and used in these two areas.

After discussing different definitions of the word 'service', the author explores the structures of frequently used quality models. Then one model from the service area is aligned with one from the software area.

For the software area the author has chosen a compilation based on the documents IEEE from1992 and ISO from1988. In the description of the IEEE model the author mentions six attributes, but in the comparison the last attribute, Portability, is missing. This seems to be depending on the ISO document, but no explanation of this is found in the article. The author does not give any reason for choosing this model but states that the differencies between software models are not large and that the models follow a common, established ideology.

Concerning service quality models the picture is more complex. Without much discussion the author chooses the SERVQUAL model, described by Zeithaml, 1990.

The result of the comparison between the two models is visualized by a picture with arrows showing "strong relation" and "weak relation" between the attributes. When the same word occurs as an attribute in both models this is considered a strong relation. This occurs only twice, between the words Reliability and Security. Nine relations are considered as weak, but those are not very clear and explanations are missing.

The most interesting observation from this comparison is that the languages used in the compared models are quite different, probably depending on different historical backgrounds. One example given is that in software quality models there are no analogies to the concept of Courtesy. This might, according to the author, depend on that the social dimension is perceived as much more explicit and important in human-human than in human-computer interaction.

The author points out that distinction ought to be made between functional and non-functional quality and between 'what' and 'how' attributes. These distinctions seem to be more common in the software area. But the distinction between core and cue attributes frequently used in business looks new to software quality.

The author's main conclusions are that the correspondence between widespread models of service quality and software quality is incomplete, and that both service management and software engineering would benefit from widening their perspective.

In my opinion, the main contribution from this article is the discussion around different kinds of attributes and their occurence in service versus software quality evaluation models. This might even have been drawn further and resulted in a suggestion of a common model. But even without this the article is well worth reading.