Preview Mode Links will not work in preview mode

Welcome to The Deming Institute Podcast page!

 

Feb 27, 2021

In our 46th Deming Lens episode, host Tripp Babbitt shares his interpretation of wide-ranging aspects and implications of Dr. Deming's theory of management. This month he looks at tampering.

Show Notes

[00:00:15]
Deming Lens Episode 46 - The Art of Tampering

[00:01:38]
Shew hart Invents the Control Chert

[00:02:05]
Control Charts - Two Mistakes

[00:02:28]
The Knee-Jerk Reaction to Managing without Control Charts

[00:03:50]
Tampering in Action

[00:05:19]
Tampering - The Funnel Experiment

[00:05:43]
Using the Funnel and Control Charts to Play Golf

[00:09:06]
Overcorrection in Automation

 

 

Transcript

[00:00:15] In the 46 episode of The Deming Lens, we'll discuss the art of tampering.

 

[00:00:29] Hi, I'm Tripp Babbitt, host of the Deming Institute podcast, and this month I wanted to explore tampering. And it really comes to me in a couple of different ways. One is you can identify tampering if you have control charts and also by virtue of the final experiment, whether your system is on target. And I want to go back to a time when I really started to see control charts. I was selling equipment for Mitutoyo here in the U.S. and I saw the front line using the statistical process control charts. And I thought, wow, this is good. And they told me kind of what they did and how they benefited, making better products and things of that sort. But I never really made the association back with control charts being used really for for any data that you have.

 

[00:01:38] And it was more of a handicap than I than I ever realized, that there were more applications for control charts than just the front line. So when we look at control charts that Dr. Shujaat put together and about, I think was 1925, you really had two different types of mistakes that you could identify.

 

[00:02:05] And on page 317 of out of the crisis, the first mistake as a scribe, a variation or mistake to a special cause when in fact the cause belongs to the system, common cause.And then the second mistake you can make is ascribe a variation or a mistake to the system, a common causes, when in fact the cause was special.

 

[00:02:28] Now as a young manager, I was more in charge of inventory, but we would get for an industrial distributor I worked for, we would get monthly reports with all the numbers for that particular month. And it was like getting a report card. You know, you wanted to see how things came out. How are we doing? Where were sales? Were expenses down? Was the inventory up or down? And that was obviously the one I watched more closely.

 

[00:03:02] And, you know, it was interesting because of the reaction that people would have. It was, you know, psychologically, you felt the need to do something if sales were down or expenses were up, and there would always be these knee jerk reactions. And I was in it to you know, I was watching the inventory of inventory all of a sudden shot up. Then, you know, why was shooting up and try and reduce it immediately. And was it until I started using control charts? Was it long after I worked for the industrial distributor? I just hadn't found the application for it at that time.

 

[00:03:50] And I wish I would have I would have understood a lot more about systems and variation and things of that sort. So control charts is the first the first way that you can identify whether you're tampering. So if you can imagine if you're you're looking at a monthly report and you're reacting to each month over month, you're going to make probably one of the two mistakes Dr. Deming and Dr. Shujaat identified. And it you know, it's just the way that we're built and it's typically we'll treat everything as a special cause for the most part, because we don't really you know, when you're looking at data and if you don't understand systems, that's just the way you're going to do it. And that can be, you know, say you got a 50 50 chance maybe of getting it right. But I found that managers, because they have a tendency to do mistake number one, you know, ascribing a variation or mistake to a special cause when in fact it belongs to the system. Everything was treated as special. There had to be something that we had to do in order to fix a problem that really wasn't a problem. Had we understood anything about variation, it's just the knowledge just wasn't there at that time.

 

[00:05:19] So control charts first way, identify tampering in the system and then the second way through the funnel experiment, you know, are you on target? Are you overinvesting off of the target? And I remember some of the early applications when I really started to grasp the concept of Variation.

 

[00:05:43] And actually, one of the early applications I used was in golf, when you have a new tool that you think can help you, you want to play it in as many places as it might make sense and sometimes places where it doesn't make sense. But one of the first applications was in golf playing golf. And I remember when I played on the golf team in college and at that time the rage was visualizations, you know, closing your eyes, visualizing the shot, how you wanted to go after your club, those types of things. However, I wish I would have had control charts because I think I would have looked at golf a little bit differently. Don't get me wrong, visualization is great to use in golf. If you're not using it, use it. But but it just gave me more information.

 

[00:06:36] So I started putting my golf scores on control charts. And of course, you got variation when I play a different course. I had my home course that I would play and then I would have other courses and obviously courses I was unfamiliar with. My scores would be higher because I didn't know where the ball should land and and so forth.

 

[00:06:57] But but the thing I really got was kind of I was using the concept of the funnel experiment, which started to make sense to me is when I use my tee shots, you know, adjustments were constant. You know, if you were always to the right, you know, you don't remember necessarily when you're making these adjustments and you can make over investments. But I started keeping data on how far my drives were were fading. I typically hit a fade back in those days. And, you know, who would go anywhere from five to, you know, thirty five yards. And so I started to do things to try and figure out how did I how could I shrink the amount of variation in my tee shots and be able to hit them at least consistently. Because then what you get at least consistent and you know, even if it's a 15 yard fade or slice, whatever you want to call it, then you could predict kind of where you needed to aim. But, you know, but I also had days and if any of you play golf where you're hitting it left and then you're hitting it right and it was like an army marching on the golf course. But but these cuts have really started to to grab at me. And I had opportunities more and more to use control charts in management settings and data and things of that sort. And you start to make better decisions when you're using data now. A lot of the time. And I know what some people are going to say. Well, those are all the results types of data. Yes, they were. But to get people to grasp concepts of variation, just seeing the data on a chart is useful. It is helpful to management to understand systems and variation and things of that sort when they're using it on their own data.

 

[00:09:06] You know, and and as I started to use control charts, I would look back on some of the things that I'd seen and I remember a customer in Kentucky that had bought a machine.

 

[00:09:19] And what it would do is it would drill a hole. Let's say there was probably 26 different stations or 30 to forget how many the exact number of it was. But you would drill a hole in station one, station two would measure it, and then it would make adjustments to station one.

 

[00:09:38] And now I would I know they were having a lot of problems with it. Was it really my machine or anything? Because it was going to replace all these workers? It never did, because they had so many problems with the machine because of overcorrection was was kind of what I remember being the problem with the machine and, you know, off to the Milky Way, so to speak. It was the final experiment at work where where they were tampering, in essence with measurements, with immediate feedback, thinking that that was a good thing. And so you wound up with this overcorrection and and the parts wouldn't come out. Right. And it was I just remember it being a big boat anchor that, you know, the management there really was upset over. But one of the things that I do find is, you know, and putting together my effective executive education program is, as I talked to executives, most of them know about control charts. They do. Many of them, especially manufacturing, see it as something that it is for the front line. But I'll tell you what, to many management executives in service don't know about control charts. They've heard of them, but they don't really understand how they might be able to help them.

 

[00:11:08] And it's one of these concepts that I think the frustration of Dr. Deming after World War II I'm very familiar with because they either believe it's for something for the front line and to get it. And but and they're looking at these data and they're like, I was back, you know, and in the 80s, looking at data and, you know, the highs and lows of the emotion kind of take the day. And that's one of the things that control charts can do, is get you focused on reducing the variation, getting rid of the special causes, you know, and things of that sort and not reacting to mistake. One mistake to that Dr. Shewhart talked about so.

 

[00:12:02] The other thing that I that I've reflected on quite a bit, too, and especially recently, but I remember I took a course back in college and even in my MBA program that had to do with the banking system in the Federal Reserve. And I haven't quite been able to apply all elements of it, although it's something I'm constantly working on is some of these larger systems like this. The Federal Reserve was put in at 1913, something like that.

 

[00:12:33] You know, we're supposed to be an independent group that could kind of guide the economy so we wouldn't have these large gyrations of inflation and things of that sort. But, you know, we were back on the gold standard back in back in 1913. So it wasn't as difficult. And then again, in 1971, when we got off the gold standard and we became fiat currency. This is where the Fed has allowed things to to get out of control over a period of time. And when I look back at some of the data, it almost looks like a random walk since we went off the gold standard, like we lost our target, you know, something to associate fiat money with gold.

 

[00:13:23] And it makes me wonder, are we experiencing right now with our money, you know, M1 and M2, our money supply, a random walk?

 

[00:13:38] Because right now there's a lot of pressure on interest rates to go up and the Fed is doing everything it can to prevent that because we have such a large federal deficit of, you know, 28 trillion dollars that they're artificially manipulating the system or in the words of Dr. Deming, tampering. But those are those are some of the things I'm looking I'd be curious if if people have looked at that or thought about that in application and some of these larger systems like this, because it seems to me.

 

[00:14:17] That we are experiencing a random walk by not keeping to a gold standard and also keeping it on target, knowing that there's pressure to keep interest rates low because of such a high federal deficit. But these are some of the things that you can look at, some of these broader systems. We get all these policy changes and things in government. How can they look at these systems more completely use data to help use this feedback and also some knowledge of tampering and being able to recognize whether you're tampering with the system or not?

 

[00:15:04] Anyway, that's what I want to talk about. This February was was tampering. And if you have any thoughts on tampering or even stories to tell, send them to me at tripp@deming.org.

 

[00:15:21] Hi, this is Tripp Babbitt. One way that you can help the Deming Institute and this podcast is by providing a rating on Apple podcast.