BP: Test Case weight
4 answers
I'm interested too...
As we have more than one person using RQM, each of us has our own standard of how much to weigh a test case. It's hard to set a rule, because we are not sure which is the right or better way. Hope some one can share some thoughts on this. Thanks.
As we have more than one person using RQM, each of us has our own standard of how much to weigh a test case. It's hard to set a rule, because we are not sure which is the right or better way. Hope some one can share some thoughts on this. Thanks.
Hello all,
Best practices...
Can you all give me some examples on how you assign test case weight? In terms of
Overall
Manual vs Automated
Long Tests vs Short
New Feature vs Regression
etc...
thanks
On 06/18/2010 03:08 PM, itengtools wrote:
this can be quite complicated and subjective. We use weight to convey
complexity and length of time. You would not give a test that ran in 30
seconds the same weight as one that ran for 72 hours. Also a functional
test would be weighted less than a complex system test scenario. It if
very hard to come up with standards as to what is assigned. I have
heard of groups that assign 10 points for each hour a test is expected
to run. We have done something similar but have added another category
to the test case called complexity which we assign a value of 1, 5, 7 as
low medium and high. This times the number of "8 hour shifts" time 100
is our weight as we are a system test shop and most of hour tests run
multiple days. So a hi complexity test case that ran for one day would
get 7 x 3 shifts x 100 or 2100 points. You can subdivide a shift for
shorter runs. This is one way but there are many other ways - I think
the best practice is not to get hung up on the absolute number but try
to get the relative differences to be reasonable.
Hello all,
Best practices...
Can you all give me some examples on how you assign test case weight?
In terms of
Overall
Manual vs Automated
Long Tests vs Short
New Feature vs Regression
etc...
thanks
this can be quite complicated and subjective. We use weight to convey
complexity and length of time. You would not give a test that ran in 30
seconds the same weight as one that ran for 72 hours. Also a functional
test would be weighted less than a complex system test scenario. It if
very hard to come up with standards as to what is assigned. I have
heard of groups that assign 10 points for each hour a test is expected
to run. We have done something similar but have added another category
to the test case called complexity which we assign a value of 1, 5, 7 as
low medium and high. This times the number of "8 hour shifts" time 100
is our weight as we are a system test shop and most of hour tests run
multiple days. So a hi complexity test case that ran for one day would
get 7 x 3 shifts x 100 or 2100 points. You can subdivide a shift for
shorter runs. This is one way but there are many other ways - I think
the best practice is not to get hung up on the absolute number but try
to get the relative differences to be reasonable.
Thanks for sharing your experience, David.
this can be quite complicated and subjective. We use weight to convey
complexity and length of time. You would not give a test that ran in 30
seconds the same weight as one that ran for 72 hours. Also a functional
test would be weighted less than a complex system test scenario. It if
very hard to come up with standards as to what is assigned. I have
heard of groups that assign 10 points for each hour a test is expected
to run. We have done something similar but have added another category
to the test case called complexity which we assign a value of 1, 5, 7 as
low medium and high. This times the number of "8 hour shifts" time 100
is our weight as we are a system test shop and most of hour tests run
multiple days. So a hi complexity test case that ran for one day would
get 7 x 3 shifts x 100 or 2100 points. You can subdivide a shift for
shorter runs. This is one way but there are many other ways - I think
the best practice is not to get hung up on the absolute number but try
to get the relative differences to be reasonable.
On 06/18/2010 03:08 PM, itengtools wrote:
Hello all,
Best practices...
Can you all give me some examples on how you assign test case weight?
In terms of
Overall
Manual vs Automated
Long Tests vs Short
New Feature vs Regression
etc...
thanks
this can be quite complicated and subjective. We use weight to convey
complexity and length of time. You would not give a test that ran in 30
seconds the same weight as one that ran for 72 hours. Also a functional
test would be weighted less than a complex system test scenario. It if
very hard to come up with standards as to what is assigned. I have
heard of groups that assign 10 points for each hour a test is expected
to run. We have done something similar but have added another category
to the test case called complexity which we assign a value of 1, 5, 7 as
low medium and high. This times the number of "8 hour shifts" time 100
is our weight as we are a system test shop and most of hour tests run
multiple days. So a hi complexity test case that ran for one day would
get 7 x 3 shifts x 100 or 2100 points. You can subdivide a shift for
shorter runs. This is one way but there are many other ways - I think
the best practice is not to get hung up on the absolute number but try
to get the relative differences to be reasonable.