How to Interpret Coefficient Logs

When using statistical software to run statistical models, you are often left with a coefficient log, listing all of the coefficients that correspond to the variables in your study. Interpreting a coefficient log is not easy, as it requires a researcher to understand the mathematical basis of the model used. However, there are some simple techniques to interpreting coefficient logs. By analyzing the coefficients one at a time, you will be able to produce a rigorous analysis of your model’s output.

Instructions

    • 1

      Look at the variable corresponding to the coefficient. There are three possibilities for such a variable. First, it could display no variable, often displayed as “none.” This implies the coefficient you are looking at is the y-intercept of the model, or the beginning point of the model. For example, in the model “Y = axy + bx + c,” “c” is the coefficient corresponding to “none.” What this means in that if all other variables are zero, the output of the model will be the coefficient corresponding to “none.” Second, it could display a single variable. If so, this is the coefficient multiplying a single variable in the model. For the model “Z = axy + bx + c,” this would correspond to the coefficient “b.” Third, the variable could display multiple variables, often in the form “x:y.” This implies the coefficient is multiplying the product of two variables. In the model “Z = axy + bx + c,” this would be a coefficient such as “a.”

    • 2

      Analyze the magnitude of the coefficients. This will let you find the change associated with the variables listed in the coefficient log. Each coefficient will have a different numerical value. These values represent the amount of change in the variable associated with the coefficient for one unit of change. For example, if your model is measuring body language sensitivity and its coefficient log displays a coefficient for 5 for the variable “age,” this means that for every one-unit increase in age (usually in years), there is a 5-unit increase in body language sensitivity.

    • 3

      Create the model from the coefficient log. The coefficients implicitly state the resulting model of your study. The model is essentially “DV = coefficients*variables.” In this equation, “DV” is the dependent variable of the study, “coefficients” are the numerical coefficients listed in the coefficient log and “variables” are the independent variables in the study.

Learnify Hub © www.0685.com All Rights Reserved