Julie Fortune

A comparison of cohesion metrics for object-oriented systems

Abstract Cohesion is the degree to which the elements of a class or object belong together. Many different object-oriented cohesion metrics have been developed; many of them are based on the notion of degree of similarity of methods. No consensus has yet arisen as to which of these metrics best measures cohesion; this is a problem for software developers since there are so many suggested metrics, it is difficult to make an informed choice. This research compares various cohesion metrics with ratings of two separate teams of experts over two software packages, to determine which of these metrics best match human-oriented views of cohesion. Additionally, the metrics are compared statistically, to determine which tend to measure the same kinds of cohesion. Differences in results for different object-oriented metrics tools are discussed.

Semantic software metrics computed from natural language design specifications

An approach using semantic metrics to provide insight into software quality early in the design phase of software development by automatically analysing natural language (NL) design specifications for object-oriented systems is presented. Semantic metrics are based on the meaning of software within the problem domain. In this paper, we extend semantic metrics to analyse design specifications. Since semantic metrics can now be calculated from early in design through software maintenance, they provide a consistent and seamless type of metric that can be collected through the entire lifecycle. We discuss our semMet system, an NL-based program comprehension tool we have expanded to calculate semantic metrics from design specifications. To validate semantic metrics from design specifications and to illustrate their seamless nature across the software lifecycle, we compare semantic metrics from different phases of the lifecycle, and we also compare them to syntactically oriented metrics calculated from the source code. Results indicate semantic metrics calculated from design specifications can give insight into the quality of the source code based on that design. Also, these results illustrate that semantic metrics provide a consistent and seamless type of metric that can be collected through the entire lifecycle.

Team Progress Checksheet: A Study Continuation

Abstract: With the high technology world of today and in an effort to compete on a global level, companies have used teams to tackle complex designs and to continually improve in areas such as quality and service. Large amounts of money are spent on training teams but few organizations evaluate these training efforts once the teams return to the job. In addition, engineering managers need a way to assess teams throughout the project life to ensure team success. One quick and potentially cost effective way to assess teams in these situations is to use a survey. Work has continued on one such survey, called the Team Success Questionnaire (TSQ), to study its psychometric properties. The instrument contains 11 questions pertaining to teaming attributes. The TSQ is very consistent in its measurement as shown by the reliability, presented as a Cronbachs alpha value, of 0.94. It also has construct, concurrent, and discriminate validity. The TSQ correlates with both team performance and group development. The results of this research show that the TSQ would be an appropriate method to assess teams after training and to use by the engineering manager at points throughout the life of the team.

Studying team shared mental models

As technology is used to support team-based activities, one important factor affecting the performance of teams is the kind of mental model shared between team members. This paper describes a novel conceptual graph based methodology to study these mental models to better understand how shared mental models affect performance and other factors of a teams behavior.

SEMANTIC METRICS: METRICS BASED ON SEMANTIC ASPECTS OF SOFTWARE

Software practitioners need ways to assess their software, and metrics can provide an automated way to do that, providing valuable feedback with little effort earlier than the testing phase. Semantic metrics were proposed to quantify aspects of software quality based on the meaning of softwares task in the domain. Unlike traditional software metrics, semantic metrics do not rely on code syntax. Instead, semantic metrics are calculated from domain information, using the knowledge base of a program understanding system. Because semantic metrics do not rely on code syntax, they can be calculated before code is fully implemented. This article evaluates the semantic metrics theoretically and empirically. We find that the semantic metrics compare well to existing metrics and show promise as early indicators of software quality.

Investigation of domain effects on software

It has been stated that there is very less variability in cohesion, coupling and complexity of software packages within specific domains such as Graphical User Interface (GUI). This implies that software metrics show low variability within single domain and high variability between domains. This paper investigates the domain issue by creating hierarchical model of four different domains and two software packages within each domain. Metrics are collected on each package and compared against the domains and packages. Results confirm metrics are not domain centric.