Roger Kurian
Published: 2009
Total Pages: 37
Get eBook
Organizations have several different kinds of measurement programs around cost, time, risk, quality, management buy-in, user buy-in, etc. In order to conduct a measurement program, information on the measurement objectives are captured in the form of metrics. In the past, there have been several problems reported with software metrics. The most critical of these is the fact that these metrics do not provide an easy way to be comprehended and applied. GQM is one such method that helps ask specific questions towards the management goals and provides for metrics that answer these specific questions. Since this metrics plan is defined purely based on management goals, GQM provides for better use of such software metrics by tying these metrics to the management goals. Thus, the GQM methodology provides for better understanding and base-lining the management organization's practices by taking a structured and collective approach towards deriving metrics. The GQM methodology provides for effective metrics that help answer important questions tied to management goals. However, even such an effective metrics program may sometimes not be successful. One of the most important causes is that such programs may be hard or difficult to implement. Variations to the GQM methodology such as continuous GQM and V-GQM are far easier to use and save the management time and effort. The introduction of the GQM methodology to software measurement was a significant step forward in that, when used, it causes software measurement users to consciously consider the connections between what they want and need to accomplish and the metrics used. The continuous GQM variation was developed in conjunction with the automated use of metrics. The V-GQM variation used the idea of continuous GQM, though without requiring automated metrics, to highlight additional benefits of repeated uses of the GQM methodology. However, even with significant advances and improvements in software measurement programs, these programs will often not be successful without positive buy-in from managers and metrics users. Interestingly, managers and metrics users are not always most interested in the quality of the measurement programs. Often, especially, when busy, managers and metrics users really want a measurement program that in many ways is self sustaining and this self sustainability is a natural benefit of the continuous GQM and the V-GQM methodologies. Of course, the most self sustaining metrics programs are in some ways the automated ones. However, especially, from managers' perspectives, the continuous GQM and especially, the V-GQM are very self sustaining from a monitoring perspective. With the V-GQM methodology, once the metrics program is set up, it can run until the goal or goals are obtained. Such a program is ideal from a manager's perspective. This self sustainability of using the continuous GQM and, especially, the V-GQM is an aspect of these methodologies which is a highlight of my using these methodologies in my data warehouse project. The measurement projects were easy for me to implement, and their use was transparent to my manager. By the use of a real-life project, each of the above methodologies have been put to use and benefits derived by management by the use of such methodologies have been documented in this thesis. Other potential benefits to management by the use of such methodologies such as finding out the rate of increase of data and the amount of data in the data warehouse at any given point in time are also highlighted in this thesis. Management/organizations can thus take advantage of these methodologies and apply them to several measurement programs.
