Predictive Modeling

Overview

Predictive modeling is a technique that creates a model that can be used to project the future of hypothetical conditions. Models can be used to compare predicted impacts vs. actual impacts of a behavior or to perform an experiment that would not be feasible in the real world. In other words, a predictive model can create a simulation from data and information to predict what may happen in the future: If A equals ____ and B equals ____ then C will equal ____.

Application

Predictive modeling can be applied in a number of ways that can help estimate long-term impacts of activities and occurrences in an area. Understanding and knowing the long-term impacts may guide management decisions to prevent future problems from occurring.

Typically, this method is used to predict aspects such as labor or income; however, it may also be useful for other factors. For example, predictive modeling can help predict the condition of the marine resource at a given time in the future, both with and without the benefit of protection.

Predictive modeling may also be an educational tool for the public and stakeholders because it can demonstrate the impacts of time and specific actions (or lack of action). 

Strengths

  • Helps predict potential outcomes of management decisions

  • Determines possible long-term impacts of policies

  • Identifies potential issues and generates hypotheses

  • Saves money by permitting managers to review the scenarios on a theoretic level without investing in conducting them

Limitations

  • Accurate data not easily available

  • Not all future variables can be considered

  • Levels of reliability vary

  • Can be very costly

Expertise Needed

This technique requires assistance from skilled experts.

Related Partners

NOAA is a federal agency focused on the condition of the oceans and the atmosphere. It plays several distinct roles within the Department of Commerce.

Resources: Books and Publications

  • Starfield, A. and A. L. Bleloch. 1986. Building Models for Conservation and Wildlife Management. New York: McMillan.

  • Starfield, A. M., K. A. Smith, and others. 1994. How to Model It: Problem Solving for the Computer Age. Edina, MN: Burgess International Group, Inc.