Control and Treated values for Treatment variable

Categories → Causal Wizard Concept , Causal Inference , Variables , Study Design

Control and Treated values of the Treatment variable assign samples to control and treated cohorts respectively.

Control and Treated groups

In causal inference, the terms control and treated are used to describe two groups of subjects (also known as sample-units or observations) that differ in terms of their exposure to a particular treatment or intervention. For example, the control group could be the group of subjects who do not receive the treatment, then the treated group would be the group of subjects who did receive the treatment.

In other words, the control group serves as a comparison or reference group to the treated group, allowing us to assess the effect of the treatment by comparing the outcomes of the two groups. The control group provides an estimate of what would have happened in the absence of the treatment, while the treated group provides an estimate of what happens when the treatment is applied.

For example, suppose we are studying the effect of a new drug on blood pressure. We might randomly assign some participants to receive the drug (treated group) while others receive a placebo (control group). By comparing the changes in blood pressure between the two groups, we can estimate the causal effect of the drug on blood pressure.

Separating the Control and Treated groups

In Causal Wizard, while editing your Study you must define the values of the Treatment variable which define which group to assign each sample unit / subject / participant to. 

  • If the Treatment variable is Binary (0/1) or Boolean (True/False), you must specify which value corresponds to Control and Treated groups
  • If a greater range of numerical values is present, you must specify a threshold to separate the two groups.
    • Values <= threshold will become Controls
    • Values > threshold will become Treated

When you click the Define Groups button in the Wizard page, you will see one of two different views depending on the data type of your Treatment variable.

  • If your treatment variable is numerical, you'll see the Threshold view to pick a threshold. Enter a value in the box and press Enter. The graph will update to confirm your selection. Press Save to close the dialog and remember your threshold value.
  • If your treatment variable is categorical, you'll see a Categorical value selection view. Assign specific values to Treated and Control groups by typing them in the boxes provided, and clicking Add. Press Save to close the view and save your changes.

Threshold view

Categorical value selection view

Change variable type

Note that you can change the variable type by clicking it in the Causal Diagram editor. You'll see a dialog which has a Variable type dropdown. This will change the way control and treated groups are defined and the view displayed to you.

Summary

In summary, the control and treated values of the treatment variable represent the two groups of subjects or observations that differ in their exposure to the treatment, and are used to estimate the causal effect of the treatment.

Related articles
In categories