Select outcomes for analysis.


This activity could be triggered by an event (reactive) or could be planned periodically, such as at the beginning of a new phase or task (proactive).

Example Work Products

  1. Data to be used in the initial analysis
  2. Initial analysis results data
  3. Outcomes selected for further analysis


1. Gather relevant data.


Examples of relevant data include the following:
  • Defects reported by customers or end users
  • Defects found in peer reviews or testing
  • Productivity measures that are higher than expected
  • Project management problem reports requiring corrective action
  • Process capability problems
  • Earned value measurements by process (e.g., cost performance index)
  • Resource throughput, utilization, or response time measurements
  • Service fulfillment or service satisfaction problems

2. Determine which outcomes to analyze further.

When determining which outcomes to analyze further, consider their source, impact, frequency of occurrence, similarity, the cost of analysis, the time and resources needed, safety considerations, etc.


Examples of methods for selecting outcomes include the following:
  • Pareto analysis
  • Histograms
  • Box and whisker plots for attributes
  • Failure mode and effects analysis (FMEA)
  • Process capability analysis

3. Formally define the scope of the analysis, including a clear definition of the improvement needed or expected, stakeholders affected, target affected, etc.

Refer to the Decision Analysis and Resolution (DAR) (CMMI-DEV) process area for more information about analyzing possible decisions using a formal evaluation process that evaluates identified alternatives against established criteria