Thread: Testing

What are the methodologies you use to write Test Cases

1 equivalance partitioning and boundary value analysis

There are different methodoogies to write test cases, primary being
1. Boundary Value Analysis -- This technique help the tester to come out with the optimum number of test cases based on the upper and lower boundaries of the available test data.
2. Equivalence Partitioning -- This technique help the testers to come out with optimum number of test areas/domains/partions from where the test cases have to be generated

test case means a check list of all the perceptions of a test engineer for testing an appilication. that is we take valid inputs and invalid inputs for checking the application, there are two types of test design techniues followed by the companies.for each requirement of the the customer we write test cases. 1. boundary value analysis
2. euallance class partition

boundary value analysis: whenever there is a range kind of input usually test eng will use this. here they use 7 values eg: input is in b/w 100-200 numbers

up+1, upper bound, up-1, middile lowerbound-1, lowerbound, lb+1
eg:201 200 199 150 99 100 100+1

here valid are 100,101,150,199,199.
invalid are 99,201
we check for these inputs to test that application

equallance class partition

if there are more number of validations for a feature then test eng will choose this type.

eg: input is in b/w 4charaters to 20 characters, a-z,_ &@

valid
4ch word , 5 ch,12ch,19ch,20ch, a-z,_,@
invalid
3ch, 21ch,A-Z, 0-9,other special ch than _,@

this is the brief can u understand if any more ?s ask
madhav_sus@yahoo.co.in

If we considering only writting the test cases for the black box testing folowing methadolgy is used.

1) Boundary value ananlysis
2) Equivalance Partioning
3) Error gussing technique (Which is used by the experince tester only)

Hii

I found out that there are lots of answers regarding how to write test cases & most of them have mentioned the same thing & offcourse, no doubt all of it is correct, but only to the theorotical extent.
In todays world of myriad GUI & complex server & client architecture over a distributed system or over internet, its never enough if you explain how boundary value analysis or equivalence partitioning is done just by taking a input range from 1 to 100 as in geeks interview we are more concerned with perfect interview replies rather than theory
We should always be carefull while mentioning examples of our points.
lets start with boundary value analysis
I will take the example of a very real business domain, telecom billing(I guess most of us are tensed about our mobile usage ))
Now sometimes a scenario occurs when the telecom operator does not create a bill for a postpaid customer. This happens because there is a mimimum threshold value check set in the database such that if the usage amount of that customer dos not cross the threshold amount then the oprator bypass the expenses of generating the bill for that particular customer. This threshold value is set in database & can be modified as per the operator's bisuness policy. Now suppose a operator has set its threshold value to 100rs. The boundary values for the thrshold logic will come as -1,0,.01,.09,99.99,100,100.01,100.1,101.
The test engineer has to every time change the value of the threshold & try to generate a dummy bill for a dummy customer created in his/her test environment & see the behaviour of the system as per SRS

In my next post I will discuss about equivalence partitioning

For any questions or suggestions do reply freely

Hi cindrela,
If i tell you truly , it depends on tester to tester how he/she write the test cases.It depends on his/her creativity also.Otherwise what so ever other answers,they all are right.

With regards,
Brijesh Jain.