Meaningful Data

This article has been dedicated to and is inspired by Robina Towers to whom I owe a debt of gratitude...

So, meaningful data... WOW... this is a huge area to explore in one article but will become a boring read if I write more than one article... here goes...

What is meaningful data?

While the term meaningful has different meanings to different people, this phrase pertains to that data which supports (defends) the conclusion(s) one has drawn.

Sounds relatively simple doesn't it?

And, therein lies its deception.

In 1966, when I attended college, we did not have computers or the internet and relied on the Library (Now referred to as an LRC) in which to conduct our research. When we found a piece of data (in the form of a quote or sentence) in order to prove its credibility, we needed to find it in 3 additional reference books.

What is interestingly silly to me is that once we found this same data in 3 other sources then it became “common knowledge” and there was no need (technically) to reference that material with citations; however, our over-zealous professors wanted to make sure that we could properly use MLA or APA protocols in case we decided to pursue a PhD in our respective fields of study.

Of course, they were making a huge conclusion about us without in supporting data but we were paying them to make us “WISE.”

It is here that I need to point out that Rabina Towers questioned some of the data in my Joys of Teaching article and in so doing forced me to realized that I had not established the credibility of my source or of the information, thereby drawing on the reader's emotions to support my conclusions.

And, of course, she was right... hence, this article.

Let's shift our focus to the manufacturing sector of our economy as we journey through the storage fields of meaning data.

In manufacturing in a very simplistic way, we build something, inspect it, check it again, and ship it out and that process is costly and time consuming because it is labor intensive.

A few years ago, it was discovered that we could systematically but randomly select 5 or more sequential items that were being manufactured and measure certain parameters to ascertain it they adhered to specification and if they did, we could then predict that your production would remain the same for the next, maybe two, and in some cases maybe 4 hours.

This process substantially reduced out costs.

However, the blue collar worker, who was the one who took these measurements saw this as something extra he/she had to do without getting any additional pay; therefore, he paid attention and soon realized or had a general idea of what management was looking for, so instead of measuring anymore, he/she simply recorded numbers that were in the expected range, not really knowing if they were that way or not.

Management would then look at the data that had been collected and statistically presented to them and drew important conclusions that were totally predicated upon that data... which in many cases were simply wrong.

This is an example of data that is not meaningful.

Let's move from this field of meaningful data to an educational field of meaningful data, shall we?

It is common practice in Higher Education for each School/Department Dean to administer a Student Survey to its students at the conclusion of each semester.

Sadly, students have to almost be threatened with not getting their final grades to get them to fill out this survey.

WHY? And, why is an IMPORTANT question here...

Because nothing ever changes and the same bad professors that we complain about are still teaching... so, why take the time to complete the survey, they tell me? When forced, they simply circle/check indicating everything was fine.

Deans who get this information and actually read this information (which is rare) call all the faculty members together for a special meeting telling everyone the results, congratulating us on those results, saying each of us deserve a pat on the back.

In some cases, Deans use this information to grant tenure and to secure raises for some faculty members.

These PhDs never stop to consider that the data might not be meaningful...

When I was Dean of several Proprietary Colleges where surveys were the “ritual” to follow at the end of each semester or quarter, I would also visit classes and watch my instructors teach for 30-60 minutes at a time.

For new instructors, I would visit every month the first quarter, then every quarter for as long as they taught for me.

Why did I do this?

Because I wanted to see what my money bought me... and no, I did not base my decision to keep them on the payroll as a result of student surveys.

And yes, it took me an enormous amount of time each quarter to conduct those in-class observations but I saw that as an integral part of my job even though it was not listed as part of my duties.

You, the reader, might find this interesting... or, maybe not so much, but for the last 3 years of my career I taught at Carson-Newman University which is a respected Christian University but not once during those 3 years, did anyone visit my class for any type of observations to determine what kind of instructor I was.

However, they did base their decision to terminate me based upon the email of ONE student who complained that I was being too negative about my employer. Incidentally, the unemployment agency that processed me to receive/not receive unemployment insurance found NO EVIDENCE to support a claim of misconduct when Carson-Newman University tried to defend their decision.

This is an example of not having meaningful data on which to make decisions.

When I taught classes to business and industry in the area of Statistical Problem Solving using SPC charts or Statistical Process Control charts, I would show them how reviewers of data could determine if the data was accurate or not... meaning the employee had “fudged” the data or had not been properly trained.

In short, SPC looks at the average of the data collected, then creates limits of control based upon +/- 3 standard deviations which captures 99.7% of the possible data variation. If they were to conduct what is called ZONE ANALYSIS which breaks down each of the 3 standard deviations, only 68% of the data should fall within the limits of the first +/- standard deviation.

If there is more, then the data has typically been fudged by the employee for whatever reason.

This analysis of meaningful data can take place in manufacturing as well as: non manufacturing, education, healthcare, service, travel, tourism, government, military, transportation, agriculture, legal, CPAs, etc.

In short, it applies EVERYWHERE.

And, since it applies everywhere, there is going to be an everywhere need to make sure the data that is being collected is meaningful...