The Fear of Reliability Statistics

When reading a report and there is a large complex formula, maybe a derivation, do you just skip over it? Does a phrase, 95% confidence of 98% reliability over 2 years, not help your understanding of the result?

Hypothesis testing, confidence intervals, point estimates, parameters, independent identically distributed, random sample, orthogonal array, …

Did you just shiver a bit? 

Or do you know of folks about you that not only do not understand these terms, they do not want to understand.

Signs of Aversion to Statistics

“That result isn’t right.” Actually had this conversation, where presenting results on an experiment to improve factory yield, my boss said that he didn’t believe the conclusion, as it didn’t fit with what he expected. He basically waved his hand over the analysis of measurement error, the data collection procedure, the analysis and conclusions, and said, “Yeah, that’s not right.”

He couldn’t explain it other than he wanted a different result.

“What does this mean?” is an approach that may have two meanings. You did something novel or advanced and your audience truly wants to know more and understand. Said a different way, it may imply they just want to get the overview and just the results — they are not going to sort though the math and analysis to get an understanding.

“I’m off for a weekend in Vegas.” While not a sure sign they do not understand statistics, and it is possible they do enjoy the many entertainment opportunities found in gambling casinos, it may indicate they haven’t grasped the nature of world around them. The law of large numbers requires a lot of numbers before it applies. Just because I’ve lost money on the last 6 trips, that I’m due for a big win, is actually a flawed bit of logic.

We hear that statistics is tough, they didn’t get it in school, or questions about why one sample isn’t enough. We face people that are uninterested, unwilling, or maybe unable to grasp that variance is a measure of dispersion and that shifting the mean about all day long will not help reduce the variability of the results.

It’s not You

Sure many folks, including you would prefer to use a simple easy to calculate value and move on. MTBF is just such a measure, it’s just an average and most of us get that. Beyond just an average, we separate those that ‘get’ statistics and those that don’t.

You are risking confrontation by suggesting that we need to work on reducing the standard deviation, not because it is the right thing to do (how dare you come up with a good idea), it is because you are bordering on topics that your peers do not fully understand.

Learning and master statistical tools takes time and practice for many. And for many they lack the time and interest.

You most likely have taken the time and have found some success with various statistical tools. So, what can you do to continue to make progress improving your product or system?

First, continue to use statistical tools

Second, document and layout the approach, the analysis and the results clearly

Third, find at least one champion that appreciates what you’re doing

Fourth, let the results and benefits speak for themselves (include benefits, results, value created in your reports and presentation — as lower failure rates may go unnoticed otherwise)

Finally, keep looking for ways to apply statistical tools. Get better results and continue to make a difference. Let me know what you have found as resistance to the use of statistics, and what has worked to use statistics to make improvements. What’s working and what’s still a problem? Let’s talk about it.

Fred Schenkelberg is an experienced reliability engineering and management consultant with his firm FMS Reliability. His passion is working with teams to create cost-effective reliability programs that solve problems, create durable and reliable products, increase customer satisfaction, and reduce warranty costs. If you enjoyed this article consider subscribing to the ongoing series at Accendo Reliability.