Dependent variables are affected by the independent variable and errors. In research, errors are not considered mistakes, rather part of the process. Observational Error can occur in a research. It is the difference between the measured value and the true value of a quantity. The error has two parts, Systematic Error and Random Error. Systematic error is due to extraneous (confounding) variables. These are variables that are not part of the study, but also affect the dependent variable. Random error is random and can occur due to random changes in the individuals being studied along with the random changes in the conditions that the individual is being studied under.

Controlling for Confounding Variables

As stated previously, confounding variables are extraneous factors that affect the relationship between independent and dependent variables.

The effects of the extraneous variables can be controlled by:

  1. Randomly assigning subject to groups.
  2. Selecting all individuals that are homogenous to the confounding variable. For example, if the researcher was concerned that old age may play a factor in measuring the effects of a treatment on memory, then he can select all young subjects. Although it will control the effect of the confounding variable, it will limit generalizability of the findings on older adults.
  3. Build the extraneous variable into the study design. Make it as another independent variable.
  4. Use Analysis of Covariance (ANCOVA) since it controls for effects due to a covariate (extraneous variable). More on this later.