Citation Context

... [LK94], [LLWW95], [Li98]) 6. Definition of measurement frameworks for one or more structural properties, which provide guidelines how these properties can, in principle, be measured [EKS94], [HM95], =-=[BDW98]-=-, [BDW99]. 7. Criticism/theoretical analysis of existing measures and measurement frameworks; in particular, there is an interest in defining, for measures of various structural properties, necessary ...

Citation Context

...his paper, we distinguish between design properties, design principles, and design methods. Object-oriented design properties characterize the resulting design. Examples are coupling [8] and cohesion =-=[7]-=-. Object-oriented design principles prescribe “good” values of the design properties. Examples are low coupling and high cohesion, as advocated in [14], [22]. Object-oriented design methods prescribe ...

Citation Context

...ent variables used here (Section 5.3) are based on a previous study performed by Li et al. [25]. In their work, the concept “maintenance effort” was reified as the number of lines of code changed. In future work, we are planning to use other more representative measures, such as “time to understand, develop and implement modifications” [1]. The achievement of the construct validity for our independent variables (the metrics in Section 3) would be desirable but is beyond the scope of this paper. However, some of our metrics are extensions of CK metrics that have been theoretically validated in [8, 9, 10]. Internal Validity. Our empirical study cannot be considered a controlled experiment, since all subjects took part in the development of the two systems. However, we tried to minimize the bias, selecting two subjects who defended (before the study) the pattern-oriented approach and two others who defended the aspect-oriented approach. Despite the effect caused by the subjects learning as the study proceeded, the AO system, which was developed first, showed better results. External Validity. The limited size and complexity of the system and the use of student subjects may restrict the extrapol...

Citation Context

...his is based on the widely used and acceptedsassumption a (cohesion) metric that better explains the presence of faults is a bettersquality indicator and therefore a better constructed measure (e.g., =-=Briand et al. 1998-=-,sBriand et al, 2001, Gyimothy et al. 2005, Aggarwal et al, 2007, and Marcus et al.s2008).sThe major contributions of this paper are as follows:s1. Introducing an MMI-based cohesion metric (LSCC) that...