Introduction to Investigation And Utilizing Lean Test Metrics In Agile Software Testing Methodologies

Padmaraj Nidagundi. Int. Journal of Engineering Research and Applications www.ijera.com
ISSN: 2248-9622, Vol. 6, Issue 4, (Part - 1) April 2016, pp.13-16
www.ijera.com 13|P a g e
Introduction to Investigation And Utilizing Lean Test Metrics In
Agile Software Testing Methodologies
Padmaraj Nidagundi
Faculty of Computer Science and Information Technology, Institute of Applied computer Systems, Riga
Technical University, Riga, Latvia.
ABSTRACT
The growth of the software development industry approaches the new development methodologies to deliver the
error free software to its end-user fulfilling the business values to product. The growth of tools and technology
has brought the automation in the development and software testing process, it has also increased the demand of
the fast testing and delivery of the software to end customers. Traditional software development methodologies
to trending agile software development trend have brought new philosophy, dimensions, and processes having
invested new tools to make process easy. The Agile development (Scrum, XP, FDD, BDD, ATDD, ASD,
DSDM, Kanban, Crystal and Lean) process also faces the software testing model crises because of the fast
development of life cycles, fast delivery to end users without having appropriate test metrics which make the
software testing process slow as well as increase the risk. The analysis of the testing metrics in the software
testing process and setting the right lean test metrics help to improve the software quality effectively in agile
process.
Keywords: Agile, Software Testing, Software Test Metrics, Software Error.
I. INTRODUCTION
The rapid growth of software use in daily
life with electronical devices has brought a new
lifestyle flexibilities to all humans. Building
software in a fast way and delivering in a fast way
got a new dimension when Agile Manifesto
established in 2001 and after six years it was re-
introduced as agile methodology. The growing
market demand of the fast delivery of the software
build by build in short iterations deliver the
potentially shippable software features. All software
team keep focused on the goals for the planned short
iteration, a short and tight learning feedback process
helps discover and optimize the process & solutions.
Agile methodology has ten core principles,
and many specific agile development methods. All
focus on the teamwork, collaboration and process
agility with limitations throughout the life-cycle of
the software development.
The meaning of agile or agility is quick. In
software it means a quick focus on short iterations.
Software development and testing can be challenging
in programming for the [1, 2] complex or porotypes
of software development. Agile development has
different approaches and all approaches have a
unique value proposition. The agile software testing
plays a key role in the verification of code and
validation of the software component.
The agile software test process is complex
due to short time, the software needs to be very well
tested in a very short time covering all test cases,
modules, unit tests, integration tests, performance
and including end user beta tests. Agile brings one
Significant gain in software development, test &
delivery speed, but on the other hand it brings the
time constraints that affect the software test process.
My core contribution is to show that having well
identified agile lean test metrics it is necessary to
identify and to apply to agile different development
and testing methodologies.
Specifically:
 Scrum, where any one of the team member can
act and test the software for the errors in the
dedicated sprint time.
 In extreme programming where a developer
writes a unit test and later does an acceptance
testing.
 Behavior design development, where user
stories tested with business requirements.
 In the test-driven development, at first write a
failing test and then make the test pass and
finally refract the code and follow the first stage.
Iterative and incremental features which
drive the development, in which each feature of the
software is developed and tested.
II. THE PROBLEM WE TACKLE
In the software development process,
software testing brings a key benefit to verify and
validate the full software product before giving it to
end customers. In the agile testing process there is a
chance of (a) repeated testing part of the software
that has already been tested (b) without having the
clear path in terms of utilizing the human resources
and budget, (c) increasing complexity of the
software with each release criteria makes the testing
RESEARCH ARTICLE OPEN ACCESS

process more intricate (e) and the testing project
may be collapsed (f) the lack of the road map
increases the risk in the testing process.
Traditional development is the waterfall
software development methodology in which the
software testing process comes to the focus on the
very late stage and we have time a for test plan and
build the strategy. In agile where requirements are
changing more rapidly according to stakeholder
wishes, in a short time testing needs to be carried out
in the right way to deliver a good quality software.
In software testing life cycle, software needs to be
tested in a different way such as in the case of unit
testing, functional testing, performance testing,
UI/UX testing..etc. The goal of the testing process is
to improve the software quality in each iteration. In
this process it’s very important to build the list of
key metrics list which helps to drive the test process
in the right way with a high throughput.
My approach: In agile software testing
process it is very important to test metrics needed to
be identified as early, clearly and to be divided by
test process, in agile where short iteration plays a
key role where this approach can bring the
significant value to improve the test process.
Fig 1: possible way of identifying test metrics in the
test process.
III. IDENTIFYING THE RIGHT TEST
METRICS
Key metrics work as indicators to measure the
performance of the software testing process and
assist in the software development process to
improve the software quality continuously. In the
testing process the right metrics answer all questions
to test software for the next stage. In agile it bridges
the gap between project goal and actual result.
Step 1: Identifying the result/goal of the project. We
must set the goal of products, for example either we
test the software product in manual testing or use
automation tools.
Step 2: Identifying the impact of the next level. If
we use manual testing, what tools, skills, technology
and how much human resource we need.
Step 3: From step 2 we have already got basic
metrics. They grow and at the same time it is at this
stage that we can identify how to divide the project
components (break a large system into a small one)
in terms of architectural level.
Step 4: Once you identified the software
components, now classify the type of testing to
create a baseline you want to perform on (functional,
regression, smoke, API, security, other)
Step 5: In functional, regression, smoke, API,
security, other set the how many test cases you
develop, execute and results out of them in agile
environment.
Step 6: Establish the review process for these
metrics to improve them time to time.
The identification of the right test metrics create the
loop feedback system that give the continuous input
to the test team members to how they are
progressing in each stage.
Let’s focus on the identifying the most common
defect-related terms in agile.
In agile we need to consider all the time how
many errors and defects are found in the short cycle.
As I have highlighted and showed in figure 1, the test
metrics can be further divided into a more sub
category on a deep level according to what kind of a
test needed to be performed on the software
application.
The deep level test metrics may be manual,
automation, performance, unit, integration, and
usability…etc kind of a test or more than this. As we
know the agile development has different
methodologies in itself and the goal is the same to
first identify the right set of software metrics and the
to keep measuring the results.
Table I: identifying the common defect-related
terms.
Terminology Definition Explanation
Error Find an error
and run time or
after completing
the product
Software error
may be run time
or when product
built
Defects Variation at
intervals
expected and
actual result
It’s a mismatch
of the user
requirement
Software Test
Metrics
Measure the
quality of the
project
It is a scale for
the measurement.
A. Test metrics formula generation
In software development and testing
industry there are different test metrics formula.
They are used to evaluate the software quality
process, but the goal is very identical - to keep the
measurement of the software quality in order to get a
deep insight of meaning full data, so use such data to
improve the software testing process in each stage
and give the right impact.

B. Importance of classifying the right test
metrics in agile
The benefit of classifying the right test
metrics should be more than what I have identified
below here.
 To cover the agile manifesto 10 principles.
 To bring out the deep meaningful data from each
testing activity in agile.
 To do the exact type of testing in short iteration.
 To identify' types of defects efficiently in each
short iteration
 To have confidence in testing efforts
IV. INVESTIGATION AND UTILIZING
LEAN TEST METRICS
The term “lean” got more in the focus when
in 1990 Toyota production system used it for a
manufacturing plant, also known as a “lean
manufacturing management philosophy”. The term
“lean” means elimination of waste.
Nowadays “lean” is a trending term in
software development industry, lean is not applied
only in management but also in software
development. Most of the time the term “lean” was
linked with six sigma principles, which is a
combination of reduction of variation willing to
solve the process and business problems. Testing is
a process, and identifying the right test metrics is a
key goal, my focus is more towards on how we can
utilize the lean in principle and practice it to get the
right test metrics in short iterations of the agile.
The main focused goal of lean is to
improve the workflow and eliminate unusual
activities. Lean manufacturing aims to eliminate
seven different types of waste.
Now we can consider these seven
terminologies, compare them and find out how we
can adopt them in agile process.
1) Overproduction – in manufacturing and
producing the item before it is needed to the
customer.
Overproduction in agile testing - in short
iteration is not focusing on what we need to test
and spending more time on unnecessary things
and over producing such as deep documentation.
In reality we need light weight short documents
for short iterations. Load testing took place for
1000 users, but according to the requirement in
that iteration it is needed only for 10 users..etc.
Possible test metrics – test strategy for iteration.
2) Waiting – in the production line goods are not
moving from one place to another for waiting.
Waiting in agile testing - agile focused on better
collaboration, but still we can notice a tester is
waiting for a proper test environment. A
developer is waiting for a designer UI for the
software application.
Possible test metrics – response metrics, volume
measurements metrics, wait time metrics.
3) Transport – It’s a product cost incursion, in
reality which add no value to the product.
Transport in agile testing: spending more time
on test meeting, without any agenda or early
testing without of scope product features or start
testing application when application features are
half built.
Possible test metrics – clearly define a test
component for iteration.
4) Over processing – using the complex or not a
right process.
Over processing in agile testing: one must
choose the right model for testing: manual
testing or automation or mix of both. Which
module of application we automate and which
not is also important to specify.
Possible test metrics – test coverage for the
iteration.
5) Inventory – work in progress or finished
products not well processed directly lead
towards the unnecessary inventory.
Unnecessary inventory in testing –at the
development stage testers are doing nothing, just
waiting for the product ready to test.
Possible test metrics – test resource planning for
the iteration when a project is big and complex.
6) Overproduction – excess motion or
overproduction.
Overproduction in testing – not utilizing the
proper resource or not planned well, so a test
team is over testing on the early stage.
Possible test metrics – proper test planning for
the resources, boundary values for every
iteration.
7) Defects – errors in a product.
Defects in testing – we can identify them as
defects found in the production by the end
customer.
Possible test metrics – software defects found in
each layer of the testing.
While we are identifying the lean seven muda
noticed that, if we use these seven muda in a
correct way in iteration, we lead towards the
new test metrics in agile manifesto ten
principles. The investigation of the metrics
helps to improve the software quality as well as
to increase the software defect traceability in in
agile each iterations.
V. SOFTWARE TEST METRICS
CLASSIFICATION
In agile we clearly classify the lean test
metrics in three category products, processes, and
resources. The aim of the classification of the muda
(waste) is to identify different possible test metrics
in each process, how much accurate we will identify

them that much clear path we get in the agile to
utilize them.
Table II: muda (waist) classification and
characteristic.
Muda classification Characteristic
Products Agile development
activities,
each iteration deliverables
and documentations.
Processes Agile product delivery to
each stage such as test, pre-
production and production.
Resources Agile testing resource
planning for hardware,
software and people.
VI. CONCLUSION
The study explains that it is a new approach
towards adoption of the lean principles in the agile
software testing and identifying the more appropriate
test metrics to improve the software quality. It’s a
never ending search for the lean agile different test
metrics those drive more value for the agile process
itself, further next level research needs to be done.
REFERENCES
[1] Hanssen, G.K.; Yamashita, A.F.; Conradi,
R.; Moonen, L. "Maintenance and agile
development: Challenges, opportunities and
future directions", Software Maintenance,
2009. ICSM 2009. IEEE International
Conference on, On page(s): 487 – 490.
[2] Choliz, J.; Vilas, J.; Moreira, J. "Independent
Security Testing on Agile Software
Development: A Case Study in a Software
Company" Pages: 522 - 531, DOI:
10.1109/ARES.2015.79.
[3] Hayata, T.; Jianchao Han; Beheshti, M.
Facilitating Agile Software Development
with Lean Architecture in the DCI Paradigm,
Information Technology: New Generations
(ITNG), 2012 Ninth International Conference
on Year: 2012 Pages: 343 - 348, DOI:
10.1109/ITNG.2012.1571989.
[4] Karvonen, T.; Rodriguez, P.; Kuvaja, P.;
Mikkonen, K.; Oivo, M. Adapting the Lean
Enterprise Self-Assessment Tool for the
Software Development Domain, Year: 2012.
IEEE Conference Publications, Pages: 266 –
273
[5] Korosec, R. ; Instrum. & Test Syst., AVL
List GmbH, Graz, Austria ; Pfarrhofer, R.
Supporting the Transition to an Agile Test
Matrix. Software Testing, Verification and
Validation (ICST), 2015 IEEE 8th
International Conference, 13-17 April 2015.
[6] E Kupiainen, MV Mäntylä, J Itkonen.
Performance Alignment Work: How software
developers experience the continuous
adaptation of team performance in Lean and
Agile environments. Information and
Software Technology, 2015 - Elsevier.
Volume 64, August 2015, Pages 132–147
[7] H Edison, X Wang, P Abrahamsson. Lean
startup: why large software companies should
care. XP '15 workshops Scientific Workshop
Proceedings of the XP2015. Article No. 2.
[8] P Chandu, T Sasikala. Implementation of
regression testing of test case prioritization.
Volume 8, Supplementary 8, April 2015.

Introduction to Investigation And Utilizing Lean Test Metrics In Agile Software Testing Methodologies

Recommended

More Related Content

What's hot (20)

Viewers also liked (20)

Similar to Introduction to Investigation And Utilizing Lean Test Metrics In Agile Software Testing Methodologies (20)

Recently uploaded (20)

Introduction to Investigation And Utilizing Lean Test Metrics In Agile Software Testing Methodologies