2017

EXPERT ANSWERS

Who’s accountable?

Q: I need a point clarified about accountability in a learning organization. In a recent QP article ("Know Thyself," April 2009), Robert Warda wrote, "there is no shortage of accountability … [it] just manifests itself in different forms."

Such as what? What does accountability look like? How do you foster an environment that supports an understanding that to take risks means that sometimes we fail, yet still hold individuals accountable for poor performance? Maybe the better question is: How do you differentiate underperforming employees from people who took risks and who might not have succeeded?

Andrea Hufstedler
Quality assurance
Boehringer Ingelheim Roxane Inc.
Columbus, OH

A: You raise a very interesting and difficult question about accountability—one that is still very much on the minds of leaders across all types of organizations.

To me, accountability falls into two broad categories: behavioral and results-oriented. Behavioral accountability includes categories you typically see in situations involving HR (policies or procedures). As you know, there are countless books and papers written about this, and it was not really related to the focus of my article. But, my two cents is that the best organizations use methods that are meant to correct the behaviors and not be punitive in any way.

My article mentions accountability in terms of being results oriented. In that case, the focus is on what was learned and what will be done differently the next time around. Implicit in that statement is that the person who made the "mistake" will continue to be accountable to get the expected result. Their engagement and obligation continues. The role of the leader in that situation is to help ensure employees learn the correct lesson and that the outcome isn’t repeated.

In learning-oriented organizations, the person who erred would also be responsible for sharing his or her failures with the rest of the organization so others can also learn from it. Being "obsessed" with failure means we highlight and celebrate them as opportunities to learn and fix issues, particularly if we can avoid them in the first place.

Remember, it was Thomas Edison who said, "I have not failed. I’ve just found 10,000 ways that won’t work."1 The parts of this process that are critical are the dissemination of lessons learned (good and bad) and the mentoring that takes place to continue to work on the problem and not make the same errors more than once.

Robert P. Warda
Administrator, research operations
Mayo Clinic
Rochester, MN

Reference

  1. QuoteDB, www.quotedb.com/quotes/1351.

 

Reduction explanation

Q: As a quality professional, I see "cost reduction" as a broad term that offers many opportunities for organizational improvement—from design concept to product disposal. But in the real world, cost reduction and workforce reduction have become synonymous. Why? I don’t think management knows the difference.

Inam Ur-Rahman
Quality system specialist
Philadelphia

A: I agree with your broad-scoped definition of cost reduction. In truth, there are multiple components behind an organization’s expenses and its opportunities to reduce those expenses.

To some degree, it appears your question is rooted in, or springs from, a perspective of a manufacturing environment. After spending 25 years in manufacturing, I’m all too aware that cost reduction does, indeed, have a broad definition. This isn’t the case in the service sector, however, where so much of an organization’s budget depends on labor.

One way to get a perspective on this question is to consult the Bureau of Economic Analysis, which is a warehouse for Gross Domestic Product (GDP) data.1 Take a look at the manufacturing sector, and the light starts to shine on the case in point.

Comparing the GDP numbers for 1959 and 2008, you find some very sobering data. In 1959, manufacturing was 26.1% of the total GDP. In 2008, manufacturing was 11.5%. This represents an approximately 56% reduction of the GDP coming from manufacturing. There are many conclusions that can be drawn from this, but one thing that stands out in particular is that more of our economy is rooted in the service sector. 

For another perspective on this question, especially if you are in the service industry, consult your organization’s CFO. Ask the CFO about budgets and what percentage of the operating budget is devoted to labor. You may be surprised by the answer.

In some service areas, labor may occupy in the neighborhood of 80% of the budget. When this is the case, cost reductions need to be associated with labor. If a business is not manufacturing oriented, there are no raw materials—none to procure, none for which you can substitute a less-expensive alternative.

Management is in the difficult position of needing to make the hard decisions to keep organizations afloat in difficult times. As such, it needs to work with what it has. If the bulk of the budget is occupied by labor costs, when it comes time to reduce the budget, there is no alternative.

Keith Wagoner
Certified quality engineer
Charlotte, NC

Reference

  1. Bureau of Economic Analysis, "Gross Domestic Product: First Quarter 2009 (Final)," www.bea.gov/newsreleases/national/gdp/2009/pdf/gdp109f.pdf.

For more information

 

Finding value

Q: In the July 2008 edition of QP, the author of the letter "Predictable confusion" uses the term "confidence interval" and provides the following example: "A 95% confidence interval on a mean signifies there is a 95% chance the population’s average will fall within values predicted by the sample data."

What raises my curiosity is that the use of the words "chance" and "will fall" can be a little confusing. Why? Because if we have a certain population, it will have a certain average. It is fixed, even if we don’t know its exact value. If we choose any interval, we can’t really calculate the chances of an average value falling within its range. Average value either is or isn’t in that interval.

I have read that although a Bayesian approach allows for the expressions used in the letter, the classic approach does not.

Andrzej Proniewicz
Student of quality management
Poznan University of Technology, Poland

A: When speaking of populations, you are correct; there are no chances or percentages. But confidence intervals and prediction intervals are calculated from a sample and not the population. Remember, statistics is about making estimates of the population based on a sampling of data.

In reading your question, I now realize that even though it is acceptable to say there is a "95% chance that the population mean will fall within the confidence interval," the reverse is a more accurate statement. Here’s how it should be phrased: "There is a 95% chance the confidence interval will surround or include the population mean."

You are correct in saying a population will have a certain average. If you take the time to measure the entire population, you can calculate that average. When taking a sample, however, the sample average will not precisely equal the population average. The confidence interval is a function of the sample average, the sample variation, sample size and a degree of confidence (in this case, 95%).

Let me further try to explain this by using an example. Let’s say you are doing a study to determine the average height of a Polish citizen. You measure the height of two Polish citizens and obtain an average. How confident are you that your calculated sample average is representative of the population average?

You’re unsure, so you increase the sample size of your study and measure every Polish citizen except for the ones that are currently out of the country. Now how confident are you that your sample average is representative of the population average?

Calculating a confidence interval on your sample mean will quantify the concept just described. Almost all statistical software will be able to do this calculation for you, and you should be able to find out how to do this in most statistical textbooks. If the university has a copy of Juran’s Quality Handbook, the section titled "Basic Statistical Methods" has a very good explanation of confidence intervals.1

Dave Carroll
Quality assurance engineer
Panduit Corp.
Tinley Park, IL

Reference

  1. Joseph M. Juran and A. Blanton Godfrey, Juran’s Quality Handbook, fifth edition, McGraw-Hill Professional, 1998, section 44, p. 44.42.

Average Rating

Rating

Out of 0 Ratings
Rate this article

Add Comments

View comments
Comments FAQ


Featured advertisers