Statistical Modeling

Written by Scott Martin
Bookmark and Share

Statistical modeling is a numerical representation of reality that estimates an unknown characteristic. Essentially, it is a way to quantitatively describe and/or predict a physical or behavior phenomenon. Statistical models range from very simplistic to extraordinarily complex, depending on the topic being studied. Some models try to measure the causal relationship between different variables.

Types of Statistical Modeling

One type of statistical modeling that is frequently used in many different disciplines is logistic regression. This type of model looks at the relationship between one or more independent variables and the log odds (a dichotomous outcome). Additionally, this type of statistical model can use categorical and continuous data as predictors.

Another use of statistical model is with structural modeling. In this type of model you describe the ways that you think the variables are related. Then using statistical software, you calculate what this means in terms of the variances and covariances of the models and test to see if they fit the model.

By using statistical modeling, you are able to describe and predict many ways the world works. However, creating a model is not enough. The model must be grounded in theory and must be able to be tested. A statistician is always a good resource when trying to develop or fine tune a model.

Bookmark and Share