Statistical Decision Techniques

“Green River Army Depot's main business is the repair and refurbishment of electronics, mainly satellite and communication systems, in partnership with the Department of Defense (DOD). Recently, DOD has put a great deal of effort into continuous quality improvement, focusing on the length of time it takes to complete a project.

Dave Smith, head of the Productivity and Quality Improvement (PQI) directorate is responsible for (among other things) facilitating lean improvement events throughout the Depot. These events bring together cross-functional teams with the goal of streamlining processes. PQI staff guide these teams in value stream mapping, identifying non—value added activities, and redesigning processes to eliminate waste. Davies concerned that his group may not meet the new standards for quality which have specified that 97% of all projects must be completed within 60 days of their start dates.

In preparation for a meeting with the Depot commander, Dave decides to review data on lean improvement events. He finds that of the past 137 projects, 30% went beyond 60 days. The commander suggests that Dave be "creative" in his presentation of the statistics to make it look like the group is actually in compliance with the DOD.

The completion times are very skewed to the high end, which doesn't surprise Dave in the least. After all, no project can take less than 0 time, but a few always seem to go on for a much longer time than planned. He is pleased to find that the average completion time is only 40 days and that the standard deviation is 10 days. Dave knows that a Normal model is a poor representation of the completion times because of the skewness, but decides to use a N (40,10) model. With this model, only about 2.5% of the projects would be expected to take more than 60 days!

He explains his model to this supervisor who is pleased with the results. Even though they know the data don't fit the model well, they also know that the DOD analysts are very familiar with Normal models and will be pleased to know that they are in compliance using it.”

Respond to each question in detail. The best responses will use the case material itself to show what is going on and what is going wrong.

1. Identify the ethical dilemma in this scenario.

2. What are the undesirable consequences?

3. Propose an ethical solution that considers the welfare of all stakeholders.