What so far I have learnt about Statistics

Let me clarify that I’m not a math guy. I had a traditional fear of the branch of knowledge until I started learning programming, also a second language to deal with mathematical problems, to deal with my geeky interests. No, I’m not a problem solver at all but love the aesthetics how a language translates an abstract idea into concrete one.
If mathematics is compared to pixels, statistics is a photograph captured by a noob, expert or malevolent person.
It’s an art of relating and arranging scattered information into a logical way to visualise those for ‘making’ or solving problems.  
I’m convinced by the concept after reading Naked Statistics by Charles Wheelan— a book that talks about aesthetics of statistics.
It is like a magic show about what an audience is shown, how it is shown and what purposes it is serving to solve a problem or feed malevolent interests.  
For now take the photograph example.
Mr Homeless

The first picture is the photograph of a letter from an alien alphabet set. The second one (anti-clockwise) is a বাসা ভাড়া or to-let notice while the third one is the ‘whole photograph’ that perhaps tells another story. The image also needs some zoom in and out as it was captured with a low res mobile camera. This very photograph can be exploited, like other statistical findings, to serve different interests.
The term ‘whole photograph’ is also a segment or ‘zoom in’ of a bigger picture in a time frame when it had been captured on the particular street. If it was too panoramic, the meaning would certainly have been changed.
On the other hand, this exact moment and scenario might not be reproduced spontaneously. But the message it conveys is pretty universal and reproduced spontaneously in different extents around the planet earth.
This is how I deal with the most basic statistical term distribution in my mind.
If we take too little or too much data that would leave us lost in the wild complacent or frustrated.
Moreover, a statistical solution to real life problems consumes time, efforts, manpower, accurate strategies and money, often beyond our imagination.
To deal with these, statisticians come with different strategies, tools and hacks like probability, polling, regression analysis, p value, and so on, a bag full of terms, with the sole purpose of analysing a distribution to solve or create (in wrong hands and procedures) a real life problem.
And one last thought. It doesn’t matter how a guitarist looks like in videos and posters or what mind blowing techniques (s)he shows, what matters is how (s)he SOUNDS.
So is true for a geek in any branch of knowledge including statistics. It doesn’t matter how smartly one explains a term or solves difficult equations, in the end the relevance of the solution matters.
The book, Naked Statistics, ignited my interest to Statistics and its child Data Science that, however, is not the topic of today’s piece. I’m yet to walk a long way to rein in those wild horses.

Comments