Beware Adding Unnecessary Complexity
An On-Line Debate with Dr. Mikel Harry
I belong to several groups on the Linked In web site. Not long ago, on one of the group sites a member named Victor posted the following query: "In your opinion, what are the greatest benefits of Force Field Analysis?" Several group members posted responses to Victor's question.
Dr. Mikel Harry described himself as "Co-Creator of Six Sigma, National Best-Selling Author and Consultant to World's Top Executives" and posted a lengthy reply that read, in part:
"BEWARE FORCE FIELD ANALYSIS (FFA): Force Field Analysis is a bit of a misnomer. Actually, a Force (F) can be combined with Span (S) and Time (T) to derive the available power. In this context, the power (P) function can be computed as P = (F x S) / T, where (F x S) is also called 'work.' Hence, Power = Work / Time, or what many refer to as 'torque.' That is, how much work can this force accomplish over a given amount of time, analogous to horsepower..."
I decided to join the discussion and posted the following:
Force Field Analysis is a technique that teams employ at the end of their analysis and testing and before introducing process changes. They should sit in the safety of their meeting rooms - before even going public with any proposed changes - and "lob some shells." Force Field Analysis allows them to brainstorm anticipated reactions for or against their proposal(s). Some of those reactions the team can anticipate may be positive; helping forces. ("Great idea! Let's do it! What kept you?...") Other anticipated reactions may be negative; resistance; obstacle forces.
Teams proceed to assign weights to and thereby model those anticipated forces for and against the change. Then, they can plan specific action steps to either weaken obstacle forces, strengthen helping forces, or both, to move in the direction they want to move; i.e., positive change and process improvement. To more specifically answer Victor's original question, I would say that the greatest benefits of Force Field Analysis are the actions taken in the wake of the analysis.
Mikel Harry's ideas about span, time, power, work, torque and horsepower are interesting; but they add no value. Trying to include these additional factors in the analysis yields no better precision than a simple weighted vote to assign values to the anticipated obstacle and helping forces. On the other hand, if one can measure and/or control span, time, power and other factors that may affect the process change, he should skip Force Field Analysis and proceed to Design of Experiments (DOE).
As it relates to the Force Field Analysis technique itself, however, my clients and I have neither the time nor the patience for adding unnecessary complexity to a simple tool.
Dr. Harry felt compelled to respond to my note and wrote a post directed not to Victor's original thread, but to me:
"I really liked your recent post on forces; however, I humbly offer a piece of additional thinking on the subject that will likely make a difference in your conclusion -- since its [sp] possible you missed the full impact of my earlier point.
"In the physical sense, power (P) can be defined as P = F x V, where F is the applied force and V is the velocity. Of course, this equation is well established and discussed in all entry level text books on Physics.
"Let's now apply physics by considering a couple of cases. as we dig into these case [sp], you will come to better understand why use of 'Power' is far superior to simply considering 'Force.'
"Let's consider a situation where the force (F) is high and the velocity (V) is low. In this case the net power is low since P = F x V. Remember, its [sp] power that 'gets the work done,' not just the independent magnitude of applied force. Consequently, when the net power is low, not much 'work' will get accomplished -- even though the force (F) is high.
"A real world example of this would be when a very large man walks very slowly into a small person that is stationary. The effect is negligible -- the small person's position will likely not be changed. On the other hand, if the same large person is running forward at maximum speed; and then strikes a stationary small man, it's obvious the little guy's position will change by a large amount -- just like when a lineman in football contacts a much smaller player on the field.
"Now, reasoning from the flip side, it is also just as obvious that when a very large slow walking man impacts a stationary small person, the resulting collision will not 'send bodies flying through the air.' There will be little (if any) change in the small man's position. In this instance, the energy would be absorbed and dissipated.
"So, the same amount of force (in both situations) is 'mitigated' by the velocity. Therefore, a 'Power Analysis' is superior to a simple 'Force Analysis.' Simply stated, to be strong and fast is far better (if you are a football player) than to be strong and slow. But then again, a relatively weak person that runs very, very fast can also have a 'high impact' on the playing field.
"On the one hand, your post states: 'Teams proceed to assign weights to and thereby model those anticipated forces for and against the change.' Consistent with this message, you also stated: 'Mikel Harry's ideas... add no value. Trying to include these additional factors yields no better precision to the analysis than a simple weighted vote...'
"Well, as you can see from my scientific argument and real-world example above, the inclusion of velocity (V) is much, much more than a 'simple weighted vote,' as you say. Its [sp] actually the way the world works, at least that's what Newton taught us."
I followed Dr. Harry's lead and posted this response:
Mikel Harry --
I enjoyed your recent post about adding power, velocity, mass and other variables to the Force Field Analysis technique. Some of my friends thought your comments about "entry level books on Physics" and "Its [sp] actually the way the world works" were condescending; but I'm sure you didn't mean it that way. In my more than 30 years of helping people learn and apply Deming's principles and statistical methods, however, I can't think of any clients who would have much interest in your slow-walking large man, stationary small person and/or playing fields. Let's consider something real.
Years ago, one of my clients assigned a project to reduce costs and part shortages in maintenance and spare parts inventories. The project team discovered a lot of wasteful, redundant inventory levels and occasional part shortages because their buildings were maintaining separate stocks. Among other improvements, the team decided to propose setting up a central inventory of maintenance supplies and spare parts from which all buildings could draw.
Using the Force Field Analysis technique, the team brainstormed possible reactions to their central inventory proposal. Among the obstacle forces anticipated were concerns about bureaucracy; local managers losing control of their own stuff; and that people would not get needed supplies quickly enough. Anticipated helping forces included better utilization of inventory dollars; having large quantities available if or when needed; and the potential for better teamwork between and among the local managers.
Via voting and ranking, the team assigned weights to all of the anticipated helping and obstacle forces. Both sets of forces totaled a weight of 95; a state of equilibrium. Newton's first law of motion teaches us that a body at rest tends to stay at rest. So, the team took action to strengthen some of the anticipated helping forces and weaken some of the obstacle forces. The project resulted in substantial reductions in costs, part shortages and cycle time.
You seem to promote only Newton's second law; i.e., F = ma (you present it as P = F x V); but how does adding power, velocity, mass, acceleration and other such factors add any value to anticipated reactions like "local managers losing control" or "having large quantities available if or when needed?" I agree with your point that "the inclusion of velocity (V) is much, much more than a 'simple weighted vote';" but why include something more if it lends no better precision to the simple weighted vote?
I would advise against adopting your approach to Force Field Analysis for the same reason I always urge teams to generate overview process maps -- as opposed to detail maps -- in the "Define the project" stage of Dr. Juran's sequence for breakthrough. I want them to save their energy for the work they'll have to do to improve their process, as opposed to falling prey to the detail monster so early in the project.
In closing, I might as well make reference to Newton's third law of motion; i.e., for every action there is an equal and opposite reaction. This applies to the deployment of any statistical method -- basic, intermediate or advanced. In my experience, adding unnecessary detail always results in wasted time and unnecessary delays. Like you said, it's "the way the world works."
I think I hit a nerve. Dr. Harry seemed a little defensive in his next post.
"Certainly, I can see your point of view on the issues mentioned in your last post. It would seem you don't really care for the 'Power Metric' and much prefer to use the 'Force' component. That's fully acceptable, as FFA has been around for a lot of years and used with success (as you graciously point out). At the end of the day, such choices really boils [sp] down to personal preference.
"However, in our experience, after teaching many classes of The Great Discovery, people much prefer the Power Metric -- the opposite of your hypothesis. Our classes normally have a wide range of academic backgrounds and credentials. In particular, they really liked the idea that a strong force with low velocity lacks 'influence power.' This seems to have a strong intuitive appeal among our students.
"It makes sense that you might not like my 'Big Man, Slow Walk' example. However, for the sake of brevity and the small space for postings, I did not give the following details. In our case, we select a 'Big Person' and a 'Little Person' from the audience. We then have them actually execute the four factorial combinations.
"After each trial, the 'Little Person' rates (on a 1-10 scale) the 'feel' upon contact with the 'Big Person.' Upon completion of the 4 combinations, we create the 2 main effect plots and the interaction plot. From this quasi-experiment and resulting graphs, the groups discusses [sp] the outcomes and draws [sp] their conclusions -- which are virtually always the same from class-to-class. In this way, they 'get it first hand thorough direct experience and feelings' versus intellectualized understandings through definitions and the simple power equation.
"Of keen interest, even the more highly educated and technically oriented people really like the exercise because they can 'feel the principle in operation and watch it play-out.' Of course, under these circumstances no one feels 'demeaned' or 'intellectually insulted.' In fact, quite the opposite is true because they inevitably say it gave them a deeper understanding of how the principles can be applied in their home life, work life and professional life.
"I trust you will continue having great success in teaching the FFA tool using your instructional approach and application framework. At the same time, we'll continue following our model that has been refined and perfected using student feedback forms. So, in a nutshell, 'what ever [sp] works for the student (customer) is best'."
I closed out our exchange with the following post:
Mikel Harry --
Thank you for your note and description of the "Big Man, Slow Walk" example that you use in your classes. Your exercise, however, is not an application of the Force Field Analysis technique. Rather, it's a neat 2 x 2 factorial experiment to illustrate the effects of size and velocity on your 1-10 "feel" response.
As I noted in my original post on Victor's thread, "... if one can measure and/or control span, time, power and other factors affecting the process change, I recommend that he skip the Force Field Analysis and proceed to Design of Experiments (DOE)." Your DOE yields not only main effect and interaction plots, but also the opportunity to run replicates and test the significance of those effects. You get none of this from Force Field Analysis.
We apply Force Field Analysis to predict reactions to proposals and plan ways to strengthen or weaken those anticipated reactions for or against the proposal. We experiment to learn.
Copyright 2014 James F. Leonard. All rights reserved.