Metrics

1
Metrics

• CSSE 372, Software Project Management
• Mark Ardis, Rose-Hulman Institute
• January 10, 2005

2
Outline

• What to measure?
• How to measure?
• Presenting the data
• Analyzing the results

3
What to measure?

• Product
• Size of artifacts
• Complexity of artifacts
• Process
• Effort
• Time
• Staff

• Quality
• Defects
• Repairs
• Usability

4
Quiz Questions 1-3
List data to be collected
5
Cartoon of the Day
6
How to measure?

• Passive
• measure artifacts
• log activities via tools
• Data collection forms
• part of the process (change requests)
• logs
• special requests

7
Quiz Question 4
Decide how to collect data
8
Presentation of data

• Tables
• Graphs
• Bar charts
• Line charts
• Pie charts
• Equations

9
Tables

• Simplify some types of analysis
• Easily converted to other forms (graphs, input to
  programs)

10
Bar Charts

• Good for comparisons
• Good for trends in small datasets

11
Line Charts

• Good for showing trends, even for large datasets
• Need to sort data first

12
Pie Charts

• Good for comparisons

13
Equations

• Good for deriving consequences
• May be used to validate data

e mc2
14
Quiz Question 5
Decide on presentation style
15
Analysis of results

• Pareto analysis
• 20 of the activities cause 80 of the problem
• Comparison to standards
• industry baseline
• rules of thumb
• Trend analysis
• compare behavior over time