Blog posts on statistics

Posts selected fromManagement Blog - Engineering Blog - Investing Blog and other blogs

• The Art of Discovery with George Box

  Great video of George Box on the process improvement process.

  continue reading: The Art of Discovery with George Box
• There is No Such Thing as “True Unemployment Rate”

  As Dr. Deming said: “there is no true value” of any measured process. The results depend on the process which includes the operation definitions used.

  continue reading: There is No Such Thing as “True Unemployment Rate”
• 200,000 People Die Every Year in Europe from Adverse Drug Effects – How Can We Improve?

  I think interactions is a hugely important area that needs a great deal more research. Doing so is very complex, which means it isn’t surprising so much more work is needed. The work of my father (and George Box and others) on multi-factorial experimentation is a powerful tool to aid this work (and that connection is likely one of the reasons I find the area of interactions so interesting – along with the realization there is so much benefit possible if we focus in that area more). Previous post on this Curious Cat Science and Engineering blog: Introduction to Fractional Factorial Designed Experiments.

  The human and financial costs of adverse effects are very high. That is why the discovery of new medicines is increasingly focused more on predicting possible adverse effects at the initial stages of developing a new drug.

  continue reading: 200,000 People Die Every Year in Europe from Adverse Drug Effects – How Can We Improve?
• Analyzing Data Requires an Understanding of the System Generating the Data

  If you bring in a new effort to focus on customers and solicit more feedback if you don’t get an increase in complaints that is likely not an indication of success but an indication of failure. One of the easiest way to reduce the number of complaints counted is to make complaining, in a way that is counted, difficult.

  Using data to measure the effectiveness of new efforts is important. But you need to understand the risks of being led astray. That risk is much greater if those analyzing the data are not intimately familiar with the processes generating the data and the operational definitions used to collect the data.

  continue reading: Analyzing Data Requires an Understanding of the System Generating the Data
• Myth: If You Can’t Measure It, You Can’t Manage It

  Dr. Deming did very much believe in the value of using data to help improve the management of the organization. But he also knew that it wasn’t close to enough. There are many things that cannot be measured and still must be managed. And there are many things that cannot be measured and managers must still make decisions about.

  I wrote a post on my Curious Cat Management Improvement blog about how to manage what you can’t measure (in 2010).

  continue reading: Myth: If You Can’t Measure It, You Can’t Manage It
• Podcast with Lynda Finn: The Value of the Simple Run Chart

  Looking closely at a specific result is helpful if the result is “special”; has an identifiable special cause for that result. But in most cases the result is do to common causes of the system and the most effective strategy to improve is to look at all the data and seek to find improvement to the overall system that will change the overall performance of the system.

  continue reading: Podcast with Lynda Finn: The Value of the Simple Run Chart
• Data is Important and You Must Confirm What the Data Actually Says

  If you do not have operational definitions for data provided in your organization, as you often will not, you need to figure out if the data is worth trusting (possibly by visiting the gemba) or whether you need to establish operational definitions and make sure new data is collected using the operational definitions.

  Often the variation in the collection of data will hamper your ability to use PDSA. If you practiced PDSA as you should, this wouldn’t be a problem, because you would make sure data had operational definitions.

  Without reliable data, the conclusions that will be drawn from results ads a great deal of potential for error to your effort.

  continue reading: Data is Important and You Must Confirm What the Data Actually Says
• How to Use Data and Avoid Being Mislead by Data

  One of the four areas of Deming’s management system is “understanding variation.” The core principle underlying that concept is using data to improve while understanding what data is and is not telling you.

  The mistakes in interpreting data are very often related to mistaking natural variation in data as meaningful. Combining this with our brains ability to find patterns (even from random data) and confirmation bias this creates problems.

  Data can’t lie, but people can be mislead and they can even mislead themselves by misinterpreting data.

  continue reading: How to Use Data and Avoid Being Mislead by Data
• Statistical Techniques Allow Management to do a Better Job

  For all four groups of people [management, statistical administration, research, front-line workers], the statistical method is more than an array of techniques. It is a mode of thought-sharpened thinking. It helps anyone in the four groups, be he a machine operator or an executive, to make better decisions, and to do his work better, than he could do otherwise.

  --- W. Edwards Deming

  Those who continue to thinkW. Edwards Deming was focused on the factory floor alone have missed most of what he proposed. Improving the decision making at the executive level was always Deming’s focus. Continual improvement should be a part of everyone’s job but as executives have more authority the impact of improving their performance multiplies, or stifles, the impact of improvement anywhere else in the organization.

  continue reading: Statistical Techniques Allow Management to do a Better Job
• Understanding Data is Often Challenging

  Using data to understand the system and validate our theories and successful improvements is an important part managing well. In some cases it is fairly easy to understand and collect data that provides a clear and accurate measure of what we care about. But getting data that helps can also be very challenging.

  Creating a management system that aims to use data while focusing on continually improving is a great start.

  continue reading: Understanding Data is Often Challenging
• Stratify Data to Hone in on Special Causes of Problems

  One strategy to help identify special causes so they can be studied and addressed is to stratify your data.

  By stratifying the data you refine your view to make it easier to identify what is causing the problem. Instead of looking at all vehicles and seeking to find the cause they had stratified the data and learned they could exclude looking at most of the processes (those that don’t impact large vehicles). And they then sought to further refine the scope by stratifying the data to further isolate the scope of the investigation. As you refine the scope you can discover what is common just to the population you have isolated by stratifying the data.

  continue reading: Stratify Data to Hone in on Special Causes of Problems
• Dangers of Forgetting the Proxy Nature of Data

  We use data to act as a proxy for some results of the system. Often people forget that the desired end result is not for the number to be improved but for the situation to be improved. We hope, if the measure improves the situation will have improved. But there are many reasons this may not be the case (one number improving at the expense of other parts of the system, the failure of the number to accurately serve as a proxy, distorting numbers, etc.).

  I find something I learned from Brian Joiner an excellent summary – which I remember as:

  Data (measuring a system) can be improved by

  1. distorting the system
  2. distorting the data or
  3. improving the system (which tends to be more difficult though likely what is desired)
     Brian Joiner’s book, 4th Generation Management is a great book for managers.

  continue reading: Dangers of Forgetting the Proxy Nature of Data
• Management Improvement History

  I do believe we need to improve our practice of Quality (and to do that we need to understand what happened in the past and why it was not more successful). The idea that Design of Experiments (DoE) was at the core of some Quality Movement to me is not at all accurate.

  
  In my experience only a few Quality professionals today understand what it means and how it should be applied. The idea that it was common place in the 40’s I seriously doubt (though I don’t have first hand knowledge of this). I find it difficult to believe we would have decided to stop using DoE if it was commonly done previously. The understanding I have from those that should know (like George Box and previously my father – Bill Hunter) is that it was not at all common practice and still is not outside of a few industries and even there it is isolated in the domain of a few experts.I do have first hand knowledge of the 80’s and the idea that we did “employee training in problem solving, team activities and just-in-time inventory” well is not even close to accurate. We sent people to training on these things but other than JIT inventory the effectiveness of these efforts were poor (with a few exceptions that really did well).

  “Quality” is not being practiced anywhere close to the level with which I am satisfied with in more than a few organizations. We have huge improvements to make in the practice of DoE, SPC, process improvement, having decisions made by the appropriate level (as close to the issue as possible), leadership, teamwork, data based decision making, the use of basically all the Quality tools, systems thinking, transformation…

  continue reading: Management Improvement History
• Control Charts in Health Care

  The point of using a control chart, and many of the management improvement tools, are to improve the efficiency and effectiveness of resources spent improving. The trick is not really to improve (that is pretty easy) the trick is to improve quickly and effectively (and in a competitive marketplace to improve more quickly than competitors). Where improvement resources are targeted is critical. In deciding which improvement options to explore it is important to understand the impact on the outcome (in this case the health of the patient).

  continue reading: Control Charts in Health Care
• Statistics for Experimenters – Second Edition

  Complete with applications covering the physical, engineering, biological, and social sciences, Statistics for Experimenters is designed for individuals who must use statistical approaches to conduct an experiment, but do not necessarily have formal training in statistics. Experimenters need only a basic understanding of mathematics to master all the statistical methods presented. This text is an essential reference for all researchers and is a highly recommended course book for undergraduate and graduate students.

  This updates the classic text by George Box, William Hunter (my father) and Stu Hunter.

  continue reading: Statistics for Experimenters – Second Edition