### 3.1.3 How is the Black Box method applied?

The Internet search engine example dealt with a situation, of which we have knowledge about (from friends, books, internet texts, etc.). This knowledge has built a model for us, which can be successfully implemented to use search engines.

What if there is no such a pre built model? It is necessary have to build one. And this requires the following:

• to describe how inputs are transformed into outputs or
• to create a more detailed model, keeping in mind how the object is built and what stays inside. In building the model it is necessary to find the responses of our object to specified inputs. That means, we have to provide some input and read the output, provide another input, read the output, provide, read, provide, read, ...

As one can conclude, there exists an infinite number of different inputs. So, when the number of inputs grows, there is a problem in combining them and then reading the result. How can this be accomplished? What we need is a testing strategy.

What is the goal in testing the Black Box?
The goal is to describe "a system" in such a way that we can use it, without knowing how it works.

Let’s see an example taken from school:
A teacher examining pupils in maths is performing a test, trying to assess the pupils' knowledge by providing them with a quiz. The teacher is aware, that:

• the kids know the whole numbers (integers) up to 100,
• they know the positive numbers only and
• they have learned just how to add and subtract numbers in that area.

What will happen if the teacher sets the following tasks to the pupils?

1.    32+66=?
2.    65-22=?
3.    43-67=?
4.    88+90=?

Obviously, that will be a real problem for the kids. Most of them will probably find the solution of the first two equations. There are exceptions of course. 1

What about the third and fourth question?
These equations do not fit in the pupils' knowledge. They produce results, normally outside their experience.

What kind or results we will get for Questions 3 and 4?
The answer is simple: random. The pupils follow their own methods of analysis to produce an output. As long as the problem is unknown or incorrectly presented to them the teacher should not expect any meaningful answer from them.
This example leads us to the conclusion that we have to know the ranges of the inputs, where our Black Box will behave correctly and ranges where it will not.

What about the environment?
The results of the test will depend on the time it was made, the classroom, the personal mood, etc. (this is the environment), but as long as all the pupils are together and the conditions are the same for all the pupils, we can expect that these factors will influence everyone in the same way. And we can ASSUME (only assume), that the differences between the results of the test depend on the pupils "Black Boxes" only.
Let's turn back to the strategies of applying the Black Box method in detail.

1 http://de.wikipedia.org/wiki/Carl_Friedrich_Gau%C3%9F  - read the story of the young Carl at school during maths hours, viewed: 20-th July, 2008