Relationships Among Variables

www.seeingstatistics.com

Start > Relationship

What is the type of the dependent or outcome variable or criterion variable?

The dependent or outcome variable is the variable that is observed. It is the variable whose values we are trying to predict.

Score or Ratio or Interval or Quantitative

There is a numerical score available for each observation. Examples include height measured in centimeters, score on a psychological test, time to complete a task, and number of problems solved.

Categorical or nominal or qualitative

Each observation is placed in a category. Examples include voting Democrat, Independent, or Republican; being diagnosed with the disease or not; and completing the task or not. Note: For categorical dependent variables, the same statistical procedures are used to answer differences and relationship questions. Selecting this option will switch you to the Differences path of the flowchart.

Ordered or ordinal or ranks

The relative placing of each observation is known. Examples include finishing place in a competition and ranking of height from tallest to shortest (without information on the specific heights).[Note: When the dependent variable has a very non-normal distribution within levels of a continuous independent variable, then nonparametric tests may be appropriate. Convert your data to ranks and then use the appropriate statistical test for rank data.]

Caution: For some relationship questions it is ambiguous which variable is the independent and which is the dependent. In such cases, simply select one of the variables to be the dependent variable for purposes of answering the above question.