Statistics Storyland - Part One

Shall I tell you of interesting affairs that take place in the numbers world?

O! Yes, make no mistake, they have lives too and are so full of character,

Humour them and they flow from your head to your hand to your pen to your paper,

And you get one truly great story to amaze everyone; so, why wait to stake your claim to be a storyteller of numbers?

The storyteller of numbers elucidates his state of mind in what comes to be known as RESEARCH,

A research well done holds the complete story of who’s done what, when, where, and how – the rich drama of interplay between the number actors (variables),

So, the storyteller, our resident statistician, tells each VARIABLE before casting it off into the deep,

Go forth, my friend, boldly into the world of statistics; for you are that anchor around which spins my tale,

Variables are details of things that are measured, controlled or manipulated in a statistical research,

They differ notably in the role that they are given to them in the research and the types of MEASURES that are applied to them.

Oh! I dreamt in agony that I died last night as the moon rose; are you keen to know how my dying and the moon rising are associated with each other?

Oh! Won’t she come and rescue me, I’m burning can’t she see; are you now dying to know how much I need to burn before she comes?

Research, as researchers come and go, are of mainly two types: – CO-RELATIONAL and EXPERIMENTAL,

A co-relational research does not control the behaviour of variables, it only measures and probes for relations that may exist between a set of variables,

An experimental research, on the other hand, manipulates some variables in one set and then measures the effects of that manipulation on the other set of variables,

Ultimately, it does come to a pass that causal relations discovered in co-relational studies are best confirmed when data comes from experimental research.

It would seem that there is no colour if nature is to be void of spring’s display; and so, of the roles that variables play, there indeed are definitive shades of character depth and play,

The first distinction which defines the disposition of variables is whether they are INDEPENDENT or DEPENDENT,

Independent variables are those that are manipulated, whereas dependent variables get only measured or registered,

The second distinction categorically defining the innate nature of variables is dependent on how they get measured; a MEASUREMENT SCALE gives them their own true pedigree,

Given that a variable is the detail of a thing; ITEMS of a DATASET are nothing but just an enumeration of all observed values of given detail of various things of the same form,

On the measurement scale, if items take on particular values and no values in-between, they are called DISCRETE, e.g. no of cars; and CONTINOUS is that other ilk who can take on any value in a given range, e.g. temperature and height.

And so, we come to speak of measurement scales now; it is but an allusion to the advancement of stature in social standing of the variables,

Whilst fictional stories hold paean to under, working, middle and upper class, the commune of variables is a step ladder of inheritance, the next rung has all properties of the previous and then some more of its own,

At the lowest level, the NOMINAL variables hold their forte; they contain only discrete measurements for qualitative classification, e.g. male, female for a gender variable,

At the next level, ORDINAL variables contain discrete measurements and hold ranking order of items amongst themselves, e.g. first, second or third for a race-finish variable,

INTERVAL variables are the next improvement, they are continuous and contain not only the rank order, but also quantify & compare size of differences, e.g. degrees celsius for a temperate variable,

And the richest of them all is RATIO variable, which has, on top of everything else, an absolute zero point as well, so you may say easily one value is a multiple of another, e.g. a height variable.