An Industrial Case Study on Requirements Volatility Measures

Author(s): Loconsole, Annabella, and Jurgen Borstler
Venue: 12th Asia-Pacific Software Engineering Conference (APSEC'05)
Date: 2005

Type of Experiement: Case Study
Sample Size: 10
Class/Experience Level: Professional


The paper discusses an “industrial case study that investigated measures of volatility for a medium size software project.” Requirements volatility is defined as “the amount of changes to a use case model over time.” Predicting volatility help project managers to take appropriate actions to minimize project risks and set up stable processes. The goals of the study are “to empirically validate a set of measures associated with the volatility of use case modems (UCM)” and “to investigate the correlations between subjective and objective volatility.” The paper concludes a high correlations between the measures of size of UCM and total number of changes which indicates that the measures of size of UCMs are good indicators of requirements volatility. However, no correlations were found between subjective and objective volatility. “The results suggest that project managers at this company should measure their projects because of the risk to take wrong decisions based on their own and the developer’s perceptions.”

What was the software methodology? Rational Unified Process (RUP)
How long was the study? Two iterations performed for the elaboration phase, four iterations for the construction phase, and one each in the inception and transition phases.

Process Outline

  1. Phase 1: Manual collection of data starting from first available revision of the UCMs (dated July 2001 – three months after the beginning of the project)
  2. Phase 2: Distributed the forms and contacted the 10 stakeholders at the company for subjective data which 8 of them responded reliably


  1. There is a high correlation between the size of UCM and the total number of changes to the UCM
  2. There is no signification correlation between the measures the subjective rating of volatility by the subjects