GUI Design Considerations

Users

• Usually not mathematical statisticians
• Complexities should be kept out of sight
• Friendly interface
• Focus on scientific goals, not formulas

Problems/Scenarios

• ``What if?'' structure
• Adaptability needs
  • Solve for n
  • Get power for given n
  • Find detectable effect size
• Meaningful ``effect size'' inputs
• Fitness for use in complex problems--e.g., multifactor ANOVA

NEXT: Excel Methods
PREVIOUS: Target power
TOP: GUIs for Power and Sample Size