# Relationship Between Variables

Analysis of Relationship Between Variables

Statistically exploring the relationship amongst variables is called conducting an analysis for correlation. When researchers look at correlation they are assessing the variables naturally and without manipulation (or controlling). Correlation allows researchers to determine the associative relationship, and predict the influence of variables on another.

Correlation: Uses two variables, either both continuous, or one continuous and the other variable consisting of two values (dichotomous). This method describes the relationship of the variables.

Example: Is there an associative relationship between empathy and happiness (each measured on separate continuous scales).

Partial Correlation: Uses three continuous variables; two variables that are intended to be analyzed, and one variable that you wish to control for. The partial correlation allows the researcher to assess the relationship between two variables whilst controlling the third.

Multiple Regression (Standard): Uses one continuous dependent variable, and two or more continuous or dichotomous independent variables. Multiple regression allows researchers to assess how much the independent variable(s) influences the dependent variable, and to what degree for each of them – highlighting which independent variable(s) is statistically significant. Multiple regression also measures the influence of all the independent variables combined (model).

Example: Does lack of empathy and impulsivity (measured by two continuous scales) predict antisocial behavior (measured by continuous scale)? Of the two independent scales, is lack of empathy or impulsivity the best predictor? Controlling for a third variable (gender), do these scales still offer predictive variance in antisocial behavior?

Logistic Regression: Can be used in various methods (e.g., Binary, Multinomial). Logistic regression analyzes the levels of the dependent variables (in odds) as the value of the independent variable(s) alter.

Example: What are the odds of imprisonment when parents have been incarcerated?