*uyir / sans*and 'life' (in small) to

*vAzhkkai / jeevan*.

The task in hand is to estimate the “value” of LIFE. A simple but profound and most commonly used statistical tool to estimation is Simple Linear Regression Analysis. So to every regression analysis there exist a dependent variable and set of independent variables. It is governed by the principle equation

Y = [beta]n * [X]n + e

In the above equation the dependent variable is Y and the independent variables are represented in matrix notation [X], which means a set of variables X1, X2 …. Xn. So the idea is to estimate the value of the variable Y using already known values of [X]. [beta] is the slope of fit, which implies the dependency of Y over [X]. Finally the factor called ‘e’ which is the error value.

Here, the “unknown” variable LIFE adorns the hat of the dependent variable. To estimate this quantity we perform a regression analysis called life, with lot of hypothesis testing considering all the variables that helps us determine the value of LIFE.

To begin with let us first try and list all the independent variables;Parents, Spouse, Friends, Teacher, Colleagues, Children, Siblings, Nationality,Relatives, education, name, fame, wealth, occupation, Gender, language etc.

Mathematcally;

LIFE = B1*Parents+ B2*Spouse+ B3*Friends+ B4*Teacher+ B5*Colleagues+ B6* Children+ B7*Siblings+ B8*Nationality+ B9*Relatives+ B10*Education+ B11* Name+ B12* B13*Fame+ B14*Wealth+ B15*Occupation+ B16*Gender + B17* Language +e

(hint1) A very important scope of error to any statistical analysis technique is the Assumption of NORMALITY of variables. So beware and do test for normality before performing the Regression Analysis. But most often we assume them to be NORMAL considering the insufficiency of DATA to test.

The next step is to perform hypothesis testing with all the variables and figure out which ones are really SIGNIFICANT variables in estimating the value of LIFE. As a conservative statistician let us consider a 95% significance level. At this level with the available limited DATASET (my experience) the test resulted in the following.

LIFE = B1*Parents+ B2*Spouse+ B3*Friends+ B4*Teacher+ e.

The [beta] matrix or the slope we derive out of the test is what is called as knowledge. That is the virtue of the dependency we have on these significant variables. This measure is used to estimate the future values of the dependent variable.

In any statistical estimation methodology we have an important term called ‘e’ to make the equation sensible and realistic. This measure accounts for all that is unaccounted for n the estimation. This entity can neither be measured nor defined. This, I term it as DIVINITY or GOD.

Am I missing something…….

Oh my god yet another common blunder. Test for independence!!!!!

Oh Jeez! I now realize that I should have performed a multivariate analysis considering the dependency of the variables.