There seems to be two different approaches that QA engineers and test management tools take to managing tests. The first is the concept that you have independent testcases and each has it's own result record. The second is that you have one testcase and that can be related to multiple results. A good way to refer to these two different approaches is by using the terms independent and linked. Lets take the scenario where you have a regression set and run those regressions for each release of the software.
Test management by the independent method operates by creation of new instances of the testcase for each version of the software released by making a copy of the original regression instance. The final result is having many independent instances from the same original testcase. Each instance has a totally different result from the next hence the description of the relationship as a one to one relation.
On the other hand, in the linked approach, there are several related outcomes, all resulting only from the original testcase. In this approach, only the outcome is linked to the release of the software that undergoes the testing. This is why this association is called a one to many relationship.
Both approaches come with many advantages as well as disadvantages of using them. The QA engineer is forced by the testing tools to use one of these methods, and has to enjoy the advantages, as well as looks for ways to overcome the disadvantages. For instance, the linked approach has a major disadvantage that is invalidation of results. This is due to the fact that any modifications made later in the test may be recorded, invalidating results that were recorded earlier. If the engineer adds a number of stages, the alterations may mean that previous results are now invalid if they were tested with the new procedures.
The independent method also has several disadvantages but invalidity of results is not one of them. This is due to the result only being linked to just one test case meaning changes made only affect the result for that particular instance. Any changes made cannot affect other results of other instances of the testcases. The major disadvantage with this method is that if any changes are required, the engineer has to apply the change to each and every one of the instances manually.
to choose an ideal tool to manage your tests, you should choose one that uses the relates s single result to a particular instance, but also allows the engineer to automatically make modifications to all the testcases he wants to modify without having to do it manually. This way, the engineer can choose which modifications to apply depending on several issues such as whether he has applied that particular change before in the testing process. Hence the changes can only be applied to testcases that have not yet been run, and not to those that have passed or failed. Therefore the validity of the results of the already run testcases is preserved.
With the linked method (one to many) you can get round this issue by branching off the testcase each time you need to make modifications. So each time you make a significant modification to the case you create a new instance and that instance is then related to many results. Again this can become a little difficult to manage as the number of your cases grows. In fact the only way to really handle this effectively is to work with a test management tool that tracks these modification.
Both the linked and independent approaches have their pros and cons; therefore the engineer has to look at several important factors before making a choice. Some of them are the alterations needed and the volume to be run. The user should consider this issue before making any decisions.
About the Author:
There are 2 ways of test management that we would like to discuss with you. Let us give you some information on this topic today. Http://www.testmanagement.com

Google
Facebook
Twitter
Myspace
Yahoo
Digg
del.icio.us
Windows Live
Reddit
Blogger
Rain Concert