Monday, October 25, 2004

Cost-Benefit Analysis & Decision Tree

Cost-Benefit Analysis is a tool used to determine what costs are associated with decision making and problem solving. A cost-benefit analysis can be simple basic math or as complex as one presented by Center for Information Technology - National Institutes of Health.

A simple cost-benefit analysis would be how much time would it take to resolve the problem and what is the rate of the person solving the problem.

• If a problem takes 30 minutes to resolve at Technician A’s rate of \$150.00 per hour, then the cost to resolve the problem would be \$75.00.
• If a problem takes 1 hour to resolve at Technician B’s rate of \$100.00 per hour, then the cost to resolve the problem would be \$100.00.
• If a problem takes 4 hours to resolve at Technician C’s rate of \$50.00 per hour, then the cost to resolve the problem would be \$200.00.

So the rates are relevant to the skills of the tech. In this case cheaper rates don’t mean it is less expensive to solve a problem. This is a simple example of cost-benefit analysis.

A more complex and involved cost-benefit analysis would be like what the Center for Information Technology - National Institutes of Health uses. They implemented in their management guide a section that identifies the cost benefits analysis as a part of a requirement that was previously established in the Paperwork Reduction Act (PRA) of 1995 and implemented by the Office of Management and Budget (OMB) in revisions to OMB A-130.

They define Cost-Benefit Analysis as: Preparing an analysis for IT initiatives to demonstrate how the IT resource will meet ICD mission requirements, support ongoing management oversight processes, maximize return on investment and minimize financial and operational risk;

• In section 5.1 of the NIH IT Management Guide the first step is to define the problem.
• In Section 5.5 they use a number of tools and techniques to determine IT performance measures. These tool and techniques are very similar to the one we are studying in this class.
• In Section 5.6 Prepare Cost Benefit Analysis they use this tool as their primary justification for the development or major modification of an IT system. So the Cost-Benefit analysis result plays a major role in determining if the project or problem to be solved will be doable. The key to the cost benefit analysis for the NIH IT is to make it commensurate with the size and complexity of the system.
• In Section 5.6 they refer to a number of governmental and regulatory documents to help guide them in there cost-benefit analysis. This means that the cost-benefit analysis is not as straight forward as simple math. A regulatory requirement could very easily cause the costs of a solution to sky rocket. This is one example of why many government projects have budget over runs and are a sign that a proper or complete cost benefit analysis was not conducted.

A Decision Tree is typically used with deductive logical thinking. It is used with most all the elements are known and the problem has clearly been identified. It is used to help guide the way through the problem to the end result solution. A Decision Tree where the problem is known, a tree is already pre-rendered and can be easily followed.

A Decision Tree can also be used to help state goals and define specific tasks necessary to reach that goal for a particular project and in planning. In this case no decision tree may exist. Therefore it must be constructed.

A Decision Tree can be constructed by using backward planning to build the task lists and their groupings. Backward planning starts with asking the question: What is the finial task to be completed? What are the supporting tasks for that task? You repeat the process for all supporting tasks and sub tasks. The Decision Tree’s backward planning method provides a road map to follow to the finial task.

There are many software programs that can help in the creation of a Decision Tree. The Decision Tree Example below was created in Microsoft Visio. Microsoft Paint Brush and Microsoft Project are just a few applications to help in creating a decision tree. Simple pencil and paper will work too. In Microsoft Word the Outline feature can be used as a quick way of getting your thoughts down and sub tasks to list them. I prefer that method over the Visio method.