Summary of Elapsed Process Variation Coefficients

This table provides a bird's eye view of variability in elapsed process times for baseline and current build stream performance scenarios. This summary is provided to facilitate the identification of scenarios that should be examined due to high variability.The variability for each scenario is expressed as a coefficient of variation (CV). The CV is calculated by dividing the standard deviation of the elapse process time over builds by the average elapsed process time over builds and multiplying by 100.

High CV values may be indicative of any of the following:

1. an unstable performance test.
• may be evidenced by an erratic elapsed process line graph.
  
  
2. performance regressions or improvements at some time in the course of builds.
• may be evidenced by plateaus in elapsed process line graphs.
  
  
3. unstable testing hardware.
   • consistent higher CV values for one test configuration as compared to others across scenarios may be related to hardward problems.

Scenarios are listed in alphabetical order in the far right column. A scenario's variation coefficients (CVs) are in columns to the left for baseline and current build streams for each test configuration. Scenarios with CVs > 10% are highlighted in yellow (10%<CV<20%) and orange(CV>20%).

Each CV value links to the scenario's detailed results to allow viewers to investigate the variability.