Blog clippings

Feb 27 2019

Analysis of Oscar TV Viewership | Mark Graban | LinkedIn

“Initial analysis says Oscars TV viewers and ratings are up over 2018. But, for context, the ratings were DOWN in 2018 compared to 2017.

Instead of reacting to each new data point, look at trends over time using “run charts” or “process behavior charts” (as I created). Did having “no host” make any statistically significant difference in ratings? No. The numbers are just fluctuating. Don’t waste time cooking up an explanation for ‘noise’ in a metric.”

Mark Graban’s Oscar Viewership Chart

 

 

Source: LinkedIn

Michel Baudin‘s comments: What happened between 2004 and 2005? Visually, the downward step is the one feature of the chart that stands out. If the technique Mark used is valid, there should be assignable causes from that period to explain the drop. If you forget the green and red lines and just look at the time series, however, a smooth, long-term downward trend sounds like a better fit. Perhaps Mark should try more modern trend analysis tools besides the venerable XmR chart.

Following a day of spirited discussion, Mark sent me the data and asked for my own analysis, which is in the chart below. The gray band is a 99% confidence interval, calculated without any knowledge of the nature of the numbers. The points outside the band can be assumed to have assignable causes for their unusual high or low ratings, and I took a look at the press reports on a few. As you can see, there is nothing special happening in 2004-2005. 

The popularity of a show has nothing to do with a critical dimension on a manufactured part, which is supposed to be constant. The plant organization has the authority to do what it takes to keep it that way. There is no reason why show ratings should be constant.

In addition, while a length measurement qualifies as a raw data point, the published number of viewers of a TV show doesn’t. In the US, it is an estimate based on a poll of 5,000 households. Viewership of the Oscars ceremony in 2019 was 11.5% higher than in 2018. On 5,000 data points, it does not look like a fluctuation. By comparison, political polls use samples of 1,000 to 1,500 voters and claim margins of error of ±3%.

The viewership estimates aren’t just numbers in a table, and you can’t interpret them properly without knowing their backstory.

#SPC, #XmRChart, #ProcessBehaviorChart

Continue reading…

By • Blog clippings • • Tags: , ,

Dec 12 2018

Open workspaces and collaboration | E. S. Bernstein and S. Turban | Royal Society

Data capture device

“[…]In two intervention-based field studies of corporate headquarters transitioning to more open office spaces, we empirically examined—using digital data from advanced wearable devices and from electronic communication servers—the effect of open office architectures on employees’ face-to-face, email and instant messaging (IM) interaction patterns. Contrary to common belief, the volume of face-to-face interaction decreased significantly (approx. 70%) in both cases, with an associated increase in electronic interaction. In short, rather than prompting increasingly vibrant face-to-face collaboration, open architecture appeared to trigger a natural human response to socially withdraw from officemates and interact instead over email and IM.[…]”

Michel Baudin‘s comments: I got curious after reading multiple blanket statements on LinkedIn to the effect that open workspaces decrease office productivity. The authors all refer to the same “Harvard study” without giving any details.  Is the Harvard label sufficient to quell any doubts? As the notorious Reinhart-Rogoff paper on austerity shows, it is nothing of the kind. 

On closer scrutiny, the Bernstein-Turban’s study is serious but limited in scope. The readers of readers of readers of their paper draw increasingly cosmic conclusions that the study does not support. To locate it, you must thread your way through multiple layers of papers. Each one simplifies and amplifies the results of the previous ones. In the process, they forget any of the nuances and restrictions of the original authors. 

Continue reading…

By • Blog clippings • • Tags: , ,

Nov 7 2018

Where Problem-Solving Goes Wrong | Gregg Stocker | Lessons in Lean

“[…] Whenever someone asks for input on a problem-solving A3, I tend to look for the red flags or areas in each section where help is most commonly needed. The key is to help people understand that the process is about investigating, reflecting, and learning, not filling in the form. It is far too common, especially early in a person’s development, to force-fit information into the boxes just to appease someone else and show that the process was followed. […]”

Sourced through Lessons in Lean

Michel Baudin‘s comments:  Companies use forms to make teams answer every question. Filling out forms, however, often degenerates into the formalism Gregg describes. Instead of reviewing content, managers just check that the team has entered something in every box.

Gregg also says nothing about immediate countermeasures to “stop the bleeding.” Assume, for example, that customers start returning defectives. The first step is to prevent more defectives from escaping. Meanwhile, you investigate root causes, implement permanent solutions, and validate them. The point of the process is to go beyond immediate countermeasures and dismantle them once they are no longer useful.

#problemsolving, #A3

By • Blog clippings • • Tags: ,

Nov 5 2018

Vade retro, Pareto ! | Cécile Roche | LinkedIn

Vilfredo Pareto (1848-1923)

“Vilfredo Pareto was a respected economist and sociologist, which few people know. He invented the theory of Pareto optimality, which describes a maximum efficiency of the competitive economy. Better known to the general public for his empirical law of 80/20 (in general, 20% of the causes generate 80% of the effects), his reputation is in reality often reduced to the use of the diagram of the same name (that he did not even invent!)

And that’s where things go wrong. I’m sure you know this famous diagram, the one where you stack problems in pre-labeled columns. Those who know me know that most of the time, these famous diagrams (actually invented by Joseph Juran) generate in me a reflex of distrust…”

Sourced from LinkedIn Pulse

Michel Baudin‘s comments: Cécile Roche’s article has a clever title that works in English as well as in French. The body of the article, however, is in French, and I recommend Google-translating it into the language of your choice. The result won’t be perfect but you will get the gist.

Her experience with Pareto diagrams confirms the points I have been making about them in this blog. I agree with Cécile’s conclusions that they are good for slides but not as drivers for actual change. On the other hand, I don’t believe it has to be that way and I see many practical uses for the analysis behind the diagrams, if not for the diagrams themselves.

Continue reading…

By • Blog clippings • • Tags: , ,

Sep 3 2018

10 Lean Manufacturing Ideas for Machine Shops | Shahrukh Irani | Modern Machine Shop

Machined Workpiece

“Lean manufacturing as it is traditionally practiced is of benefit to machine shops, but the extent of its benefit is often limited. The Toyota Production System on which lean manufacturing is based was designed for assembly plants that produce automobiles by the thousands. While an assembly plant focuses on low-mix, high-volume production, a typical machine shop focuses on high-mix low-volume production. A machine shop and assembly plant cannot expect to realize the same lean benefits with the same lean tools.”

Source Modern Machine Shop

Michel Baudin‘s comments: Shahrukh and I have long disagreed on these issues and I still take exception to many of his general statements. On the other hand, I agree with most of his recommendations.

Machining is not all about High-Mix/Low Volume

While there may be many small machine shops doing high-mix/low-volume production, the bulk of machining is done in larger Automotive and Aerospace & Defense shops that don’t fit this description. Automotive machine shops remove small amounts from hundreds of thousands of aluminum castings/year with product lives on the order of 4 years; aerospace machine shops, large amounts from forgings in exotic alloys in the hundreds/year with product lives in decades.

Continue reading…

By • Blog clippings • • Tags: ,

Aug 5 2018

Why I was wrong about 2- second Lean | Dan Markovitz

 

“It’s time to admit that one of my consulting approaches has failed. I’m a huge fan of Paul Akers’ 2 Second Lean philosophy. It’s simple, easy to understand, and has an intrinsic appeal: “fix what bugs you.” “Figure out how to do your job two seconds faster each day.” Who wouldn’t sign on to a lean program that promotes that mindset?

So off I went to my clients, with Paul’s videos embedded in my PowerPoint presentation, ready to show them how 2 Second Lean is the answer to their productivity problems, their low employee engagement and morale scores, their mediocre customer service, and their too-high defect rates. And I failed.

When I look back at the clients I introduced to Paul’s method, I have to be honest and admit that a more traditional, kaizen event-driven approach would have served them better. 2 Second Lean didn’t hurt them, but they didn’t reap the rewards that they wanted, and that I promised. 

I missed the most obvious fact: 2 Second Lean is simple. But it’s not easy. The commitment required of the president is enormous.”

Sourced from Marcovitz Consulting

Michel Baudin‘s comments:

Is the approach simple or simplistic? Perhaps, the mistake is assuming that a single method is all you need, whether it is the “2-second” approach or the oxymoronic “Kaizen event.” It’s the panacea fallacy. To improve a factory, you need a range of different approaches, from tiny improvements an individual can make within his or her area of responsibility without asking anyone’s permission to projects like line redesigns that may occupy dozens of people for months.

Continue reading…

By • Blog clippings • • Tags:

< 1 2 3 4 5 >»
%d bloggers like this: