3  Selecting Appropriate Analyses

Section in progress

This section covers how to choose appropriate analyses for any number and type of measurements. Think like, flow chart.

Considerations:

• Purpose of analysis.
  
• Types of variables in data set.
  
• Data used in analysis.
  
• Assumptions needed; satisfied?
  
• Choice of analyses is often arbitrary: consider several

Example:

5 independent variables: 3 interval, 1 ordinal, 1 nominal

1 dependent variable: interval

Analysis options

• Multiple regression: pretend independent ordinal variable is an interval variable use dummy (0 /1) variables for nominal variables
• Analysis of variance: categorize all independent variables
• Analysis of covariance: leave variables as is, check assumptions
• Logistic regression: Categorize dependent variable: high, low
• Survival analysis: IF dependent variable is time to an event

Unsure? Do several and compare results.