The Human Side of Statistical Process Control: Three Applications of SAS/QC You Might Not Have Thought About


When you think of statistical process control, or SPC for short, what industry first comes to your mind? In the past 10 or 15 years, diverse industries have begun to standardize processes and administrative tasks with statistical process control. While the top two bars of the industrial Pareto chart are probably still manufacturing and machine maintenance, in North America, SPC is being used in far more types of work than many people realize.

One reason that researchers and process managers turn to Walter Shewhart’s trusty techniques to distinguish between common cause and special cause variation—in other words, variation that represents unimportant bumps on the road and variation that means that the wheels have fallen off the bus—is that they work far better than gut instinct in many cases. Furthermore, they are defensible, because they are based on sound, scientific practice. What might surprise you is the extent to which social science, human subject research, and administrative fields are making use of SPC.

1. Health Care. This is the area I hear the most buzz about for SPC. A great deal of interesting work is being done in monitoring process variables such as late payments, the number of available beds, payments to providers, script-writing rates for high-risk drugs, pain levels during recovery from surgery, and so on. I can only scratch the surface here. The writing is on the wall about health care becoming still more expensive, data are becoming more and more plentiful, and it is especially easy for problems to hide in massive data. Unpredictable problems = cost. Process control = savings.

2. Customer Service How long did you have to wait in line the last time you were in a big box store? When you last called the cable company? Some companies recognize that if customer service is slow, you will take your business elsewhere. Some are even willing to take action on that knowledge. There are plenty of measureable service quality characteristics that can be tapped into to identify times of day or product lines that are inconsistent, which translates to a better customer experience, and to customer loyalty.

3. Survey Research This is the one I’m most excited about right now. SPC interest in survey research has been on the rise for the past 5 or so years, and I think it’s an area ripe for this sort of analysis. If the news tells you that 52% of Americans who eat raw oysters only eat them to be polite, someone had to get that information. Enter the heroes of the survey world. Survey researchers, the people behind all the Things We Know about How We Are, are applying statistical process control methods with variables related to survey data collection, such as interview length, nonresponse rate, interviewer driving distances, and so on.
In fact, just last month I was teaching a class in statistical process control to students who work with data from face-to-face interviews and mail-in surveys. Survey methodologists have a wealth of questions: Did the surveys go out to the intended recipients? Were they completed in a timely manner? Were the answers honest? Did the interviewer falsify data? Did the machine that reads the forms make an error?

So how is the human side of SPC different? Some of the questions that these industries pose are the same as any SPC application, such as the error rate of the machine reader. But unlike traditional areas of SPC, such as manufacturing, packaging, and repair, the sources of variance in health care, customer service, and survey research are difficult to capture and even more difficult to control. Why? Because People are Different. As soon as there is a human fingerprint on the process variable, there is extra noise in the data.

To illustrate the problem, here is a simplified comparison. What are the root causes of variance in a process for manufacturing a candy bar?

• Temperature of facility
• Temperature of raw product
• Humidity of facility
• Ingredients in product
• Wear state of equipment
• Operator on duty
• Other factors

What are the root causes of variance in a refusal to respond to an interview survey?
• Weather
• Day of the week
• Mood of recipient
• Presence/absence of small children
• Time of day
• Friendliness of interviewer
• Perceived trustworthiness of interviewer
• Perceived motives of survey
• General distrustfulness of others by recipient
• Other factors

While the sheer number of factors may not be very different, your ability to control each group of factors is a world apart. Merely identifying the major factors may be impossible with human subjects research. Some of these factors cannot be controlled, and should be distributed randomly throughout the population, such as whether a recipient has children. They contribute to the common cause variation that produces the up-and-down wiggles of a controlled process. Other factors are more insidious, such as the friendliness of the interviewer.

This is exactly the reason that SPC techniques fit so well into this kind of work. For example, Shewhart charts draw attention to the special causes, while allowing the common cause variation to be what it is—random. It is possible to distinguish between a predictable and an unpredictable process. A Pareto chart can then be used to identify the cause of the problem, or trigger further investigation where no clear cause is apparent.

SPC in the context of social research naturally brings along baggage in the variance, as well. For example, in health care research, prescriptions are measured at the individual script level. But they are nested within patients, who are nested within doctors. There is random variability at each level.

Another example, from survey research, measurements are taken at the individual interviewer level, but the variance in the response (say, time to complete an interview) is a function of an unknown residual error, as well as random variance due to interviewers. Some interviewers talk faster than others, etc. For the variance on the control chart to be an appropriate estimate of the variance of the response, estimates of each of these variance components is needed. Techniques that combine sophisticated modeling techniques with classic Shewhart control charting techniques are definitely on the rise as more situations like these are becoming apparent to researchers.

Do you work in an area that is moving towards SPC implementation? Do you use SPC in nontraditional ways? I’d like to hear some additional examples of SPC in social or human-subjects research

By the way, if you want to learn more about statistical process control, check out our offerings using SAS/QC and using JMP software. See you in class!


About Author

Catherine (Cat) Truxillo

Director of Analytical Education, SAS

Catherine Truxillo, Ph.D. has written or co-written SAS training courses for advanced statistical methods, including: multivariate statistics, linear and generalized linear mixed models, multilevel models, structural equation models, imputation methods for missing data, statistical process control, design and analysis of experiments, and cluster analysis. She also teaches courses on leadership and communication in data science.


  1. I've been teaching SPC for the Service Quality Division of the American Society for Quality for about 15 years. I have a bunch of non-manufacturing examples using Shewhart's methods from healthcare, insurance, call centers, etc. Don Wheeler (colleague of Dr. Deming) also has some great examples of using SPC to find trends in sales data (shows how the standard "this month vs. last month" and "this month vs. same month last year" is a really bad way to evaluate performance). I would be happy to share my experiences, data and examples.

  2. Courtney Dacey on

    It wasn’t – just an odd bit of luck. Showing the choices to a question in random order is a great way to neutralize order effects, the tendency of respondents to pick the first choices in a long list.

Back to Top