Boundary Value Analysis ( BVA) Testing

Boundary value analysis is the method of making sure that behaviour of system is certain for the input and output boundary conditions. the reason why boundary conditions are necessary for testing is because defects could be inserted at the boundaries very easily.

let us see this following example :
Input should be greater than equal to 5 and less than 10
Probably you will write something like
if (input >=5 AND input <10)>
then
do this
else
do some thing else.

So, according to this input values from 5 to 9 are valid, but if you make mistake in specifying the conditions, following things can happen
input >5 AND input <10-------> Input value 10 in invalid now
input <=5 AND input <10> Input value 4 is valid now
input >=5 AND input <=10 -----> Input value 10 is valid now
input >=5 AND input >10 -----> Input value 9 is invalid now

Because it is very easy to introduce defects at boundaries, boundary values are important. So for the above example, at the minimum we should have following test cases for boundaries
4,5,6 and 8, 9, 10
lower_boundary - 1, lower_boundary, lower_boundary + 1 and upper_boundary - 1, upper_boundary, upper_boundary + 1.

what is off-by-one
off-by-one: An exceedingly common error induced in many ways, such as by starting at zero when you should have started at one or vice-versa.

Related Topics:

Unit Testing

The aim of Unit Testing is to take a small piece of software from the application, isolate it from the reminder of the code and check whether it works exactly as you expect. like wise each unit is testing in this manner seperately before intergrating them into modules to test into interfaces between modules, and most of the defects are identified during this unit test.

Unit testing requires drivers and stubs to be written, here the driver simulates a calling unit and the stub simulates a called unit. though the Driver and Stub costs time and money, unit testing provides some unquestionable advantages, this allows automation of testing process, reduces troubles of exploring errors contained in more complex pieces of the application, test coverage is often raised because attention is given to each unit.

Example: if we have 3 units, lets see how to check it..

• check is the error due to unit1
• check is the error due to unit2
• check is the error due to unit3
• check weather it is because of the problem with 3 units
• check is the error due to interfaces between units
• it might be due to defect in the process of testing

Identifying the error (or errors) in the integrated module is much more complicated than first insulating the units, testing each module and integrating them as an integrated unit and testing thee whole system.

Driver and Stub can be reused to reduce the time and cost instead of writing large amount of code.

Related Topics: