This article is the second in a series to explain how Lean Six Sigma methods can be used to overcome the seven barriers listed in a recent T+D Magazine article.  This article specifically addresses the first barrier,“(it is) Too difficult to isolate training’s impact on results versus other factors influence”.

Suppose the result identified by leadership is, “Reduce lacerations due to sharp edges on unfinished product”.  Safety experts have selected proper personal protective equipment, but it is not being used.  The training department has been asked to produce a program on laceration awareness.

The first step in some organizations would be to begin gathering information.  Some would begin researching existing programs.  Other information would be gathered by interviewing safety experts and first line supervisors.  In a traditional approach, participants in the training would be evaluated on their knowledge on lacerations as they entered the course. Before the participants left the session, they would be evaluated again to determine how much information was learned.  Six months later, the same evaluation tool could be used to determine how much information was retained long term.  An aggressive statistical analysis could be completed to see if there was a correlation between the retention scores and safety incidents.

A training professional quickly sees other factors that influence the number of safety incidents.  Some other factors could include, “personal protective equipment is not available”, “production supervisors do not enforce the rules”, “there are no reminders for the operators”, “we are getting injuries from other causes”, and so on.

With Lean Six Sigma methods, the black belt focuses on the total process.  This total process is defined long before the training program is written.  Some steps in the process may include, complete the task analysis, write the materials, train instructors, certify instructors, schedule classes, reinforce learning on the job, modify job instructions, ensure personal protective equipment is available, audit compliance, etc.

It is obvious there is “noise” in the system.  That means there are many factors involving members of different departments.  A valid criticism is that there is no tool to measure the effect of the training.

Assuming this process, and keeping in mind that the goal is to reduce injuries, there may not be any reason to identify the contributions of training; Leadership is concerned with reducing injuries.  Most organizations operate at an approximate three sigma level.  Moving to a four sigma level will see an approximate 90% reduction in failures. For minor lacerations, this may be a significant improvement which meets the goals of leadership and safety experts. The team can move into the Control Stage to ensure that all changes are permanent.

However, if the consequences were more serious, it may be very necessary to improve the process further. We may need a 99.9% reduction in failures.  This means moving from a three sigma level to a six sigma level. Now it becomes critical to isolate the effects of the training and other steps of the process.

At this point, we determine which of the factors is contributing to the successful program, and which factors need improvement.  The Six Sigma expert now applies Design of Experiment concepts. 

The team identifies the most likely contributing factors.  In Design of Experiments, a set of screening experiments are conducted to determine the factors that are providing the largest contribution.  At this point, we may identify those factors fairly easily.  However, we may have missed a factor, or we failed to determine the interaction of different factors.  A training program alone may result in an improvement of 100%.  Likewise, a reinforcement program alone may result in an improvement of 100%.  However, a training program combined with a reinforcement program could result in improvements of 600% or more when taken in combination.  Advanced statistical tools provide the analysis to determine such contributions. These tools will also identify if we have failed to identify a key factor.

 

By focusing on results, we first evaluate the total process.  We establish our metrics. When simple process improvements meet the goals, we make those changes permanent in the Control Stage.  If however, we need drastic improvements, we move to sophisticated statistical tools.  We design experiments, conduct screening experiments, and if necessary move to higher order experiments.  In this way we can apply the appropriate level determining contributing factors, and if necessary, apply the tools necessary for previously unimagined improvements. The effort is driven by the results stated by leadership.