The first thing to know is what’s a variable. A variable is a factor that can assume a different value.


Example: Gender is a variable that can assume two values, male and female; The review of Amazon is a variable that can assume five values, like 1 star, two stars, and so on; A person's height is a variable that can take values like tall, medium, short.

As we can see in the example, we can have a different variable, depending on their scale. So it’s essential to identify this scale because different variables have different tools.

We can know the different kind of value looking at the table1:

Can I order in Ascending or Descending way?The distance between the two consequent values is the same?
Nominal or categoricalNoNo
OrdinalYesNo
ParametricYesYes
Table1 – Different scale of variable

So if we get at the gender example, we know that I can’t order the value Male and Female, and we can’t calculate a distance. So it’s a nominal or categorical variable.

The review of Amazon is a parametric variable because I can order the value (5 > 4), and the distance between 5 and 4 is the same as 3 and 2.

The Height of a person is ordinal because you can say that Tall is more than medium, but you can’t know if the difference between the value is the same. For example, you can have a tall person of 1,50cm, then a medium of about 1,70cm, and another of 1,73cm.


The equation Y = F(X) states that the Output variable Y is a function of the X variable that is the cause. So we have:

  • Y that is the effect that we want, for instance, lower the time needed to estimate a project;
  • X that is some other Critical to Quality variable, for example, the level of training of the people;
  • F that is what you do to turn X into Y.

So to meet the needed result for Y (that can be a sum of y1..yn variable), you need to look at all the X input x1..xn and even the method F.

At this point is important to know the difference between correlation and causation:

  • Correlation is when two events likely happen simultaneously in a population. For example, buying pizza and Birra at the supermarket can be highly correlated, but one doesn’t cause other event;
  • Causation is when event1 causes event2. For instance, the traffic in a street causes you are arriving late at work.

It is essential to understand how to influence the Y, but knowing the difference is necessary.

Share on: