The manipulated variable in a statistical model is the variable that on being changed causes the value of the response variable to change. The manipulated variable is the independent/explanatory variable in our model. It is under the control of the researcher. The researcher changes the value of the manipulated variable to study the corresponding changes in the response/dependent variable.

**Examples of manipulated variables**:

1. Suppose we develop a linear regression model to study the effect of diet on the weight of individuals. Here the diet is under the control of the researcher. The researcher can choose different diets and observe the effects due to them on the weight of the person. Since the diet is being controlled(“manipulated”) by the researcher, here the diet is the manipulated variable.

2. When we are modelling the effect of hours spent studying by students on the marks obtained in a test, the number of hours spent can be varied, and hence it is the manipulated variable. The marks obtained in the test is the dependent/response variable in this case.

3. When modelling the effect of rainfall levels on the girth of trees, the rainfall levels is the independent (manipulated) variable and the girth of trees is the dependent variable.

**Manipulated Variable vs Control Variable**:

In a statistical model, there are generally 3 kinds of variables – manipulated variable, control variable, and response variable. The manipulated variable must not be confused with the control variable. The value of the manipulated variable is always changed by the researcher but the value of the control variable is always held constant by the researcher. For example, if we are studying the effect of diet on weight loss we must make sure that all the individuals have the same exercise regimen to understand the true effect due to the diet. In this case, the exercise regimen is the control variable.