Sunday, March 1, 2015

Important Software Test Metrics and Measurements

In Software projects ,it is most important to measure the quality ,cost and effectiveness of the project and the processes. Without measuring these ,project can't be completed successfully.


What are Software Testing Metrics?
A Metric is a quantitative measure of the degree to which a system, system component, or process possesses a given attribute.



Suppose, in general, “Kilogram” is a metric for measuring the attribute “Weight”. Similarly, in software, “How many issues are found in thousand lines of code?”, here No. of issues is one measurement & No. of lines of code is another measurement. Metric is defined from these two measurements.

Test metrics Could be:

  • How many defects are existed within the module?
  • How many test cases are executed per person?
  • What is  the Test coverage %?
Who is accountable for For Metrics: In general , Test Lead /Manager are responsible for generating the test metrics.
Advantages:
  1. Take the decision for next phase of activities such as, estimate the cost & schedule of future projects.
  2. Understand the kind of improvement required to success the project
  3. Take decision on process or technology to be modified etc.
  4. Test Metrics are the most important to measure the quality of the software.
If we don't have any software metrics in our project ,then how can we measure the quality of work done by Software Tester/QA. 
For Example: A Test Analyst has to,
  1. Design the test cases for 5 requirements
  2. Execute the designed test cases
  3. Log the defects & need to fail the related test cases
  4. After the defect is resolved, need to re-test the defect & re-execute the corresponding failed test case.
In above scenario, if metrics are not followed, then the work completed by the test analyst will be subjective i.e. the test report will not have the proper information to know the status of his work/project.

If Metrics are involved in the project, then the exact status of his/her work with proper numbers/data can be published.

I.e. in the Test report, we can publish:
1. How many test cases have been designed per requirement?
2. How many test cases are yet to design?
3. How many test cases are executed?
4. How many test cases are passed/failed/blocked?
5. How many test cases are not yet executed?
6. How many defects are identified & what is the severity of those defects?
7. How many test cases are failed due to one particular defect? etc.
Based on the project needs we can have more metrics than above mentioned list, to know the status of the project in detail.
Based on the above metrics, test lead/manager will get the understanding of the below mentioned key points.
a) %ge of work completed
b) %ge of work yet to be completed
c) Time to complete the remaining work
d) Whether the project is going as per the schedule or lagging? etc.
Based on the metrics, if the project is not going to complete as per the schedule, then the manager will raise the alarm to the client and other stake holders by providing the reasons for lagging to avoid the last minute surprises.

Examples of Software Testing Metrics:
Let’s take an example to calculate various test metrics used in software test reports:
Below is the table format for the data retrieved from the test analyst who is actually involved in testing:
Example of Software Testing Metrics
Definitions and Formulas for Calculating Metrics:
#1) %ge Test cases Executed: This metric is used to obtain the execution status of the test cases in terms of %ge.
%ge Test cases Executed = (No. of Test cases executed / Total no. of Test cases written) * 100.
So, from the above data,
%ge Test cases Executed = (65 / 100) * 100 = 65%
#2) %ge Test cases not executed: This metric is used to obtain the pending execution status of the test cases in terms of %ge.
%ge Test cases not executed = (No. of Test cases not executed / Total no. of Test cases written) * 100.
So, from the above data,
%ge Test cases Blocked = (35 / 100) * 100 = 35%


#3) %ge Test cases Passed: This metric is used to obtain the Pass %ge of the executed test cases.
%ge Test cases Passed = (No. of Test cases Passed / Total no. of Test cases Executed) * 100.
So, from the above data,
%ge Test cases Passed = (30 / 65) * 100 = 46%
#4) %ge Test cases Failed: This metric is used to obtain the Fail %ge of the executed test cases.
%ge Test cases Failed = (No. of Test cases Failed / Total no. of Test cases Executed) * 100.
So, from the above data,
%ge Test cases Passed = (26 / 65) * 100 = 40%
#5) %ge Test cases Blocked: This metric is used to obtain the blocked %ge of the executed test cases. A detailed report can be submitted by specifying the actual reason of blocking the test cases.
%ge Test cases Blocked = (No. of Test cases Blocked / Total no. of Test cases Executed) * 100.
So, from the above data,
%ge Test cases Blocked = (9 / 65) * 100 = 14%
Test metrics graph 3
Test metrics graph 4
#6) Defect Density = No. of Defects identified / size
(Here “Size” is considered as requirement. Hence here the Defect Density is calculated as number of defects identified per requirement. Similarly, Defect Density can be calculated as number of Defects identified per 100 lines of code [OR] No. of defects identified per module etc.)
So, from the above data,
Defect Density = (30 / 5) = 6
#7) Defect Removal Efficiency (DRE) = (No. of Defects found during QA testing / (No. of Defects found during QA testing +No. of Defects found by End user)) * 100
DRE is used to identify the test effectiveness of the system.
Suppose, During Development & QA testing, we have identified 100 defects.
After the QA testing, during Alpha & Beta testing, end user / client identified 40 defects, which could have been identified during QA testing phase.
Now, The DRE will be calculated as,
DRE = [100 / (100 + 40)] * 100 = [100 /140] * 100 = 71%
$8) Defect Leakage: Defect Leakage is the Metric which is used to identify the efficiency of the QA testing i.e., how many defects are missed / slipped during the QA testing.
Defect Leakage = (No. of Defects found in UAT / No. of Defects found in QA testing.) * 100
Suppose, During Development & QA testing, we have identified 100 defects.
After the QA testing, during Alpha & Beta testing, end user / client identified 40 defects, which could have been identified during QA testing phase.
Defect Leakage = (40 /100) * 100 = 40%
#9) Defects by Priority: This metric is used to identify the no. of defects identified based on the Severity / Priority of the defect which is used to decide the quality of the software.
%ge Critical Defects = No. of Critical Defects identified / Total no. of Defects identified * 100
From the data available in the above table,
%ge Critical Defects = 6/ 30 * 100 = 20%
%ge High Defects = No. of High Defects identified / Total no. of Defects identified * 100
From the data available in the above table,
%ge High Defects = 10/ 30 * 100 = 33.33%
%ge Medium Defects = No. of Medium Defects identified / Total no. of Defects identified * 100
From the data available in the above table,
%ge Medium Defects = 6/ 30 * 100 = 20%
%ge Low Defects = No. of Low Defects identified / Total no. of Defects identified * 100
From the data available in the above table,
%ge Low Defects = 8/ 30 * 100 = 27%
Test metrics graph 5
Conclusion:

The metrics provided in this article are majorly used for generating the daily/weekly status report with accurate data during test case development/execution phase & this is also useful for tracking the project status & Quality of the software.