SPURIOUS CORRELATION AND IT'S DETECTION

PARTIAL CORRELATION (DEFINED)

If we think we have a spurious correlation and we think we know what extraneous variable is causing the problem, we can control the extraneous variable and see what the "true" relationship may be between the two variables.

The statistical procedure used for controlling the extraneous variable is called partial correlation. Partial correlation yields a single coefficient that can be interpreted just like a Pearson's correlation coefficient. The Partial correlation coefficient indicates the strength and direction of the relationship between two variables when an extraneous variable is controlled.

CONTROLLING MORE THAN ONE EXTRANEOUS VARIABLE:

In the social sciences, we often have situations where there are several extraneous independent variables that are having an effect on the relationship between two variables that are the focus of our research. For example, if a researcher was examining the relationship between divorce rates and suicide rates, they must also consider and control other extraneous variables such as church membership rates, population growth, and alcohol use.

The task of controlling three extraneous variables while examining the relationship between two focus variables can be accomplished with a procedure called multiple regression. Multiple regression coefficients indicate the strength and direction of the relationship between the focus independent variable and dependent variable while several other extraneous variables are controlled.

If you have time series data multiple regression is often inadequate as the form of the identified relationship may be flawed as there may be a need to include lags. The estimated parameters are usually biased by omitted noise structure both in terms of ARIMA structure and Intervention Variables evidenced by movements in the mean of the errors.

CLICK HERE:Home Page For AUTOBOX