Information Technology Reference
In-Depth Information
experience level of the person doing a task. The other component is the environment
factor that influences this productivity level. An environment factor could be usage
of productivity or automation tools and experience of the test team in using these
tools. Thus we can see that productivity varies from person to person, from team to
team, and from organization to organization. The more productive the test team is,
the less the effort will be and vice versa.
4.2.2
TPA Calculation Details
TPA calculation is a complex task at best. But I will try to make it as simple as pos-
sible. I will also try to make it short.
TPA calculation consists of knowing the number of function points and the
type of these function points.
Sometimes function point information may not be available to the test man-
ager. In such instances, he will have to make a rough estimate of the number of
function points in the system so that he can calculate his TPA. One method is to
count the number of logical conditions/interfaces and multiply this figure by 30.
So,
Number of function points
=
number of logical conditions
×
30
4.2.2.1
Dynamic Test Point Characteristics
Dynamic test points are influenced by many factors. These factors can be cat-
egorized as function and quality requirements dependent. Quality requirements
factors are further dependent on usage intensity and quality level. Quality require-
ments are derived from the test strategy factors. As is obvious, usage intensity is
a factor that denotes how much a function will be used by end users. Some func-
tions may be sparingly used while some functions may be extensively used by end
users. For calculations, a general rule is that 25% of functions can be considered
as having high, 50% as medium, and 25% as low usage intensity. Quality level is
the factor that determines to what extent the whole application will be tested. For
general-purpose applications not requiring transactions and where end users do
not expect a high level of reliability, software testing can be up to moderate level
and no extensive testing may be required. But for applications requiring a high
level of reliability (i.e., a mission-critical application with transactions), software
testing should be done extensively to uncover and fix as many critical defects as
possible.
With regard to function-related factors, there are three factors in this category:
interfacing, uniformity, and complexity. Interfacing is a factor that is influenced by
logical data sets, which in fact are the number of points of contact (parameters) with
other functions. The more parameters a function has through which data passes,
