Saturday 21 March 2020

Truths, damn truths and statistics...

Image result for standard deviation

Several years ago, Scrobs was spending several hours a week studying for a Diploma in Marketing.

It was an attempt to leap-frog some contemporaries in a perfectly happy and well-paid job, but one where mundanity was more than apparent, and one could easily be intimidated by the mediocrity of others, if one wanted - which I didn't...

So the spare bedroom became a shrine of learning, and with a small allowance from the County (remember grants - a licence to buy a second-hand Mini,) and a few bob from my management, the work began.

The subjects were interesting and far-reaching, especially the Marketing papers based on Kotler and Management by Drucker - both gurus back then. The course work got off to a cracking start, despite Law and Economics becoming a bit turgid, but the big elephant in the room was Statistics. While I'd never really been an expert with maths, I got a reasonable GCE, and could understand most figures (not the pure maths, or applied maths, which was for people with brains the size of a Sainsbury's and a line of coloured biros in the top pocket - once called a 'nerd pack' by an unkind American.)

So then came Statistics.

At first, it was relatively easy, and the papers went off each week, with all sorts of calcs and figures, and I began to enjoy looking at rows of numbers, and making hypothetical decisions based on the findings. This was all in the late 1970s, so computers were called main-frames, and appeared in 'The Eagle'. But things didn't improve. The elephant produced a child - 'Standard Deviation!

"A quantity expressing by how much the members of a group differ from the mean value for the group".

The calculation was as here: -

\sigma={\sqrt {\frac {\sum(x_{i}-{\mu})^{2}}{N}}}

I think I'll decline to explain what it was all about, as it really doesn't make this yarn any more (or less) interesting, but suffice to say, I just had a mental block on the whole issue! I really couldn't understand it, and tried and tried every evening to fathom the true meaning and even why one really wanted to know what the result achieved! 

But anyway, I learned the calc by rote as it was a cert for a question in the exam, and even that took a long time with just a pocket calculator and a pad and pencil. I used to make up all sorts of mnemonics to remember the sequence, and drag the numbers from the interminable columns of figures, which became a blur after an hour or two. So I could just about do the sums, and achieve a suitable answer (of course, still not knowing why the bloody thing was needed in the first place)!

The day of the exam arrived, and I entered the exam hall muttering the equation over and over so that I could answer the inevitable question. I even panicked directly I sat down and arranged the desk, and immediately wrote the calc down on a piece of paper with a sigh of arrival.

Stone me, when we were all told to turn over the papers, the scheming, sodding examiners had printed all the calcs on the front sheet 'to aid students in their tasks'! 

Bugger! All that work for naught!

So it was ever thus, I passed the paper reasonably well (still not knowing what SD really meant), and went on from there!

But, nowadays with fabulous algorithms for PCs, spreadsheets the size of Wales, the SD is just a click away from an answer, and seeing all these charts concerning Covid 19, and supposed mortality rates, coupled with the Dow Jones graphs going the opposite way, such that new yard-high PC screens need to be invented to show the huge diversion of lines, I sometimes wonder whether it was all worth it!

And I still don't really know what a Standard Deviation does to help anyone!


9 comments:

The Jannie said...

My experience presactly. I'm never stuck by a calculation which has a real application in my ageing real world but my school maths ground to a halt with quadratic equations. I passed GCSE maths in my 50s because by then it made more sense but I must have managed to sidestep the algebra!

A K Haart said...

We used Standard Deviation as a tool, a way to apply some practical quality rules to analytical data. It was fine as long as we didn't try to read too much into it.

Scrobs. said...

Oh heck, quads...:0(

Never mastered 'em, but just managed simultaneous equations!

Scrobs. said...

Ha ha ha, Mr H!

As usual, I blame Excel! I once had a Psion hand-held computer, and built a spreadsheet so large, it took about five minutes to load and edit, and I still got into the overdraft at the end of the month...

goosegirl said...

When I first looked at your post to see the equation I thought it might relate to something that Einstein scrawled on an old beer card relating to something that had to do with relatives, but now I see it because it's that old bell-shaped curve I remember well from my past days at college. The standard is the norm which is represented by a line going straight up from the base of the horizontal axis then goes up to meet the centre of the curve at its mid-line. It clearly shows that anything or anyone deviating from their normal behaviour needs a jolly good slap for nicking all the loo-rolls. I think a barium meal might be in order, but I hope the anti-constipation tablets would not be available.

Scrobs. said...

GG, your definition makes a lot of sense!

I used to think that it was all about straying from the path of human righteousness, then thought that it was no 5 in the English Hymnal, then something scrawled on the fly-leaf of a library book from Wigan.

Then I just gave up...

goosegirl said...

Sirclobs. I may have just got the answer you've been looking for all your life which is "How do you get a lady to fancy you"?

A bell curve is the most common type of distribution for a variable and is thus considered to be a normal distribution. The term "bell curve" originates from the fact that the graph used to depict a normal distribution consists of a bell-shaped line. The highest point on the curve, or the top of the bell, represents the most probable event in a series of data, while all other possible occurrences are equally distributed around the most probable event, creating a downward-sloping line on each side of the peak.

Personally, the graph looks like you tried to entertain a fair maiden in the hope she would have to have a little lie-down somewhere in private but you forgot one essential thing! You were all fired-up but the excess of alcohol took its toll.

Sen. C.R.O'Blene said...

You may think that, Goosey, but I couldn't possibly comment... (We're watching House of cards again - it's rivetting...)!

E-K said...

Totally lost on this !

(As I suspect are most statisticians at the moment.)

I fully expect to see us all here in 18 months - the odds are favourable and I wouldn't bet otherwise.

Then starts the Bring Back Our Pubs campaign !!!!