Correlation is a mathematical concept. In a multi-variable environment, correlation is the statistical technique that helps find out if and to what extent pairs of variables depend on one another. A very simple and oft-quoted example is that of height and weight as variables. Taller people have a tendency to be heavier but this relationship between the two is hardly uniform and seldom perfect. In fact, there are many cases that can outright violate the above statement. Correlation would be find how the two variables depend on each other.

However, in larger data sets, there are a number of variables and correlation is more than a little difficult, mostly because many correlations aren’t qualitatively anticipated and can be unprecedented. Good correlation techniques can help make greater sense out of the data available.

As is common among statistical techniques, correlation is limited to certain types of data- those that are quantitative in nature and where it makes sense to correlate how the quantitative values are dependent on one another. It is not applicable in cases of categorical data like colours, gender etc.

The crucial result of any correlation technique is the correlation co-efficient, say r. The value of r ranges from -1 to +1.  A correlation coefficient of 0 indicates complete independence of the variables. More the dependence, closer is the value of r to +1 or -1.  Negativity or positivity of r refers to negative or positive correlation.

Course Duration:-2h 21m