This paper examines decision trees as a practical and mathematical tool for facilitating effective decision-making in organizational and individual contexts. It explains the structure and construction of decision trees, describes the evaluation process for comparing alternatives, and demonstrates their application through real-world examples such as Gerber Products' PVC safety analysis. The paper argues that decision trees reduce decision-making errors by providing graphical representation, cost-benefit analysis, and empirical data support, making them particularly valuable for complex decisions with multiple alternatives.
Effective decision-making is vital for organizations and individuals alike (Howard, 2001). Good decision-making is an essential skill, but one that often does not come easily. Decision-making involves identifying situations where decisions are necessary, recognizing values implicit within a situation, identifying alternative courses of action, securing factual information, predicting possible consequences, and making decisions based on those potential consequences (Howard, 2002).
A visual representation of the alternatives available is a tool that has been widely used to help facilitate decision-making (Howard, 2002). This graphic or visual representation is often referred to as a "decision tree" and has been used in organizations and educational facilities for years to help individuals formulate the pros and cons of alternative decisions, create a decision-making grid, and evaluate the best alternatives for a particular situation (Howard, 2002).
Decision trees are primary tools that enable decision-making between multiple courses of action (Mind Tools, 2005). A decision tree provides a graphical reference an individual can use to create a balanced picture of the risks associated with choices and the different rewards associated with making a decision one way or another (Mind Tools, 2005). Decision trees have been applied across numerous fields, from business strategy to educational assessment, making them one of the most versatile decision-making aids available.
To create a decision tree, you must first start with a decision, represented by a small square on a piece of paper. From here you can extend lines toward the right of the box that include solutions for the decision at hand. At the end of each line, results are listed. When the results of a decision are not certain, a small circle is placed, representing another decision that needs to be made (Mind Tools, 2005). This graphical representation can be extended to portray as many lines and potential or plausible solutions and future questions as reasonable.
Evaluating the graphical representation is a key component of analyzing the decision tree. Here, the individual can look at the alternatives available and decide which holds the best value or most potential for a positive outcome (Mind Tools, 2005). You can use decision trees to calculate the value uncertain outcomes have to offer, including the value of, for example, new products or services (Mind Tools, 2005). Placing numbers to ideas and concepts will help create valuation and make projecting the best course of action easier.
As part of this process, the individual using the decision tree should also consider writing the cost of each choice available along decision lines and also the benefits (Mind Tools, 2005). The key points of a decision tree include: (1) clearly defining a problem, (2) exploring all possible outcomes and challenging those outcomes, (3) analyzing the full consequences of any decisions taken, (4) providing a clear framework from which one can quantify the potential outcomes in any given situation, and (5) enabling decision-making that is based on formal analysis rather than guessing (Mind Tools, 2005; Buckley & Dudley, 2004).
Companies large and small have used decision trees to make strategic business decisions. Management must often deal with difficult decisions where the alternatives are quite clear but the outcomes are often uncertain (Buckley & Dudley, 2004). Fortunately, decision tree analysis can help in such situations. A simple graphical diagram consisting of "nodes and branches," a decision tree enables a decision maker to select one alternative from many by providing visual representation of the benefits, costs, and problems associated with varying outcomes (Buckley & Dudley, 2004).
Using a tree, managers should have a goal in mind of obtaining the probabilities of what the chances are that each consequence will occur (Buckley & Dudley, 2004). Often, when using a decision tree, individuals have few experiences of similar outcomes they can compare their hypothesis to. Thus, it is sometimes appropriate to use decision trees as a "consensus-building technique" where panels of experts can gather to create average probabilities of one outcome occurring versus another (Buckley & Dudley, 2004).
Decision trees are mathematical in nature, using a technique referred to as mathematical expectation to produce the "most beneficial" alternative in many instances (Buckley & Dudley, 2004). Mathematical expectation involves calculating weighted averages of potential outcomes based on their probabilities, allowing decision-makers to quantify uncertain scenarios. A prominent example illustrates this approach: Gerber Products Inc. recently took advantage of decision tree analysis when deciding whether to use PVC for its products (Buckley & Dudley, 2004). The organization, working in conjunction with other prominent names including the U.S. Consumer Products Safety Commission and Greenpeace, set out to decide whether PVC use was safe (Buckley & Dudley, 2004).
Because Gerber manufactures a variety of products used orally for children, including nipples and pacifiers, it was vital that they evaluate the pros and cons of PVC use and the potential effects such use would have on consumers (Buckley & Dudley, 2004). In this particular case, Gerber found use of a decision tree helpful for evaluating all of the information related to product use, weighing the consequences of each action, and deciding on a prudent course of action (Buckley & Dudley, 2004).
Eight possible results were weighed using a decision tree. During the process, four additional alternatives were discovered. The company was able to realize the worst-case option and also use data to articulate potential positive outcomes using various alternatives (Buckley & Dudley, 2004). In this case, the decision tree also enabled the company to predict the most lucrative solution, thus showing how graphical representation of decisions can enable even strategic managers to facilitate proper decision-making. Decision trees can be used to solve complex problems where classification is necessary, analysis of fixed attributes is required, or where complex descriptions are required (AAAI, 2005). Decision trees generally help reduce the number of errors associated with a poor decision because they offer an opportunity to weed out problematic choices before they are taken.
"Appropriate and inappropriate use cases"
You’re 83% through this paper. Sign up to read the remaining 1 section.
Sign Up Now — Instant Access Already a member? Log inAlways verify citation format against your institution’s current style guide requirements.