“I got your multiple-linear regression analysis right here!”

Leave a comment

Lured by the siren’s song of a free car, and the glamor that came with it, I enrolled as an Advertising Major at Michigan State University.  I looked forward to taking courses on “How to Shoot a Funny TV Commercial,” “Casting Call 101,” “Bombay or Tanqueray?,” “Creative Expense Account Development,” and “An Executive’s Guide to Tipping.”

Oh the humanity!!  I could find nothing that approximated any of these courses.  My first term I had to take Journalism 101 (I lied about being able to type at 45 wpm), Botany (I needed a “science” and it was rumored that the professor “grew his own”), Econ 101 (taught by a hyper adipose professor who would strip down when it got hot, which would terrorize any coed sitting in the first row of the auditorium), Marketing 101 (taught by a professor whose book we had to buy), Accounting for Non-Accounting Majors (the first day we learned what assets and liabilities were, the next day we were studying collateralized mortgage obligations, defalcation, and high premium convertible debentures), and Advertising 101.

I went to college in the turbulent 60’s. While most college students were turning on and dropping out, practicing free love, and barricading the Student Union, I kept my eye on the prize…a career in advertising and a free car every year.  To do this, I had to take some classes which I knew would never have applicability for me after I graduated.  One of these was Statistics 804.  Yes, that was a graduate level course.  The UAW was striking the auto companies when I got my BA.  Big Three ad budgets were being cut, Detroit agencies weren’t hiring, and since Vietnam didn’t want me because of my blown-out knee, I decided to stay on for graduate school.

Statistics 804 was a required course. It’s like a colonoscopy; something that is awful, but you have to do it.  As with the Accounting course, the first day was a walk in the park. Figuring batting averages? No problem.  The difference between a “mean” and “a median?”  The answer was simple enough to write on my forearm if I had to.  Then came a terrifying sojourn into the dark world of Multiple Linear Regression Analysis.  Simply put, how do X and Y affect Z? Or more simply put:

Image

Well, if this was going to get me a free car, I could at least fake it.  After studying the formula as if it were the Rosetta Stone, we were assigned a term project:  Develop a thesis that can be proven by the use of multiple linear regression analysis.  During my college years I worked as a repo man for Sears.  More about that later.  My thesis would be to correlate the number of 90-day past due Sears credit card accounts (X) and the number of accounts in default (Y), to the dollar amount of uncollectible write-offs.  If I could pull this off I could pass the course, and give my results to Sears in hopes of them taking me off repo duty.

I diligently went through the credit records of the local Sears store.  I compiled mountains of scintillating data.  Now to crunch the numbers.  For those among you who may not know this, there was a time before we had high-speed computers.  To analyze the data, we had to go _FLEXO (1)to the Business School Building where we would sit for hours and punch the data onto a paper tape.  Once our tape was done, we would sign up for an appointment to “go online” with some mysterious GE computer in some unknown location.  I carefully put the roll of punch tape in the slot, pushed the button, and watched in amazement as reams of printed analysis spewed from the back of the magic box.  I was interested in only one number; the K value.  If it was less than 3.0, my thesis was valid. My eyes frantically searched for the K value.  There it was!!!  The K value for my thesis was… 29.7?  The horror. There was only one week left in the term.  And…there was only one thing left to do: Fake the data.  I knew that this skill would serve me well in the world of advertising.  After tweaking my X, and massaging my Y, I was able to get a K value of 2.9! Perfect.  I passed the course.  I didn’t learn much about multiple linear regression analysis, but I did learn how to put perfume on a pig.

I told the folks at Sears that the data findings were proprietary and couldn’t be given out. They got their revenge later.

Next:  “Oh My God, He’s Got A Gun!”

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s