Hardware Reference
In-Depth Information
Science), we exploited the opportunity of promoting a debate with a panel of experts
from industry (9 specialists) and academia (16 researchers) during a workshop hold in
Zaragoza (Spain) in 2007. After an intensive session and refining and consolidating
conclusions (reviewed by participants), 23 factors arose as possible causes of prob-
lems in software testing real practice. Although it was an interesting result, we de-
cided to check if software professionals in Spain really confirm such factors were
applicable to their professional environment. A questionnaire was created (see Table
5) where respondents had to indicate if they consider each item an effective factor of
influence; they also ranked influence in a three-level scale: total/partial/ none.
Table 4.
Factors of influence as items of questionnaire with ID for factorial analysis
Id Factor
Q1.1. When delays or finance problems appear, it is usual to shorten quality and testing effort.
Q1.2. It is not strange a QA position disappears transferring people to software development roles
Q2.1
Testing is not creative: it is something annoying and not attractive which you have to do. Even it is
negative (looks for defects) and destructive (goes against developers' work)
Q2.2
It is an area without good opportunities of career development or promotion.
Q2.3
Career development in testing does not guarantee the same salary or conditions as in other
professional careers in software development (even you may expect worse conditions).
Q2.4
It is not usually recognized this work on testing, it is not usually accounted to be paid by customers,
it is usually an internal service with no direct relationships with customers, etc.
Q2.5
Low level testers do not require a university degree, maybe only a basic professional education so
this tend to project an image of not attractive professional career
Q3.1
Many IT university graduates have not attended specific training on testing
Q3.2
Many IT professionals have not also received specific training on testing
Q3.3
Courses on software testing are not usual in company training programs for software professionals
(more focus on technology, new versions of products or in software development methods)
Q3.4
Testing is not a hot topic in universities: many teachers mention it and encourage students to do but
few of them understand the correct philosophy and techniques of testing.
Q3.5
Specific testing training tend to focus on unit/detailed testing while functional/system testing is
addressed as a marginal topic
Q3.6 Low importance or absence of specific training/qualification in testing (materials, certifications, etc.)
Q4.1 Junior tend to focus on programming and code: reject to work in other activities like testing
Q4.2
Many managers did not attend good training on software testing so they do not appreciate its interest
or potential for efficiency and quality
Q5.1
People tend to execute testing in an uncontrolled manner until the total expenditure of resources in
the belief that if we test a lot, in the end, we will cover or control all the system
Q5.2
It is not usual to plan and design efficient cases with minimum cost or to link tests to priorities or
risks; there is not control on incurred risks depending on tests, no control of evidences, etc.
Q5.3
Test design usually means a rework of what analysts did not completed or documented because
testing is totally dependent on a good requirements specification
Q5.4
Software test phase is located at the end of the project suffering shortened schedule due to delays of
the previous development phases and the impossibility of postpone delivery to customer
Q5.5
Relationship between software models and testing is not exploited, specially for test design: "testing
is something we do at the end once we have code”
Q5.6
It is not usual to design tests once we have a specification (although it is possible to do it in parallel
with analysis): in fact, they document knowledge on functionality and requirements
Q6.1 Market is not mature enough so certain software quality problems are not sufficiently penalized
Q6.2
Best business is possible when customer pays maintenance of defects delivered with the developed
software (getting money for repairing defects one has created)
