According to Almeida et al. (2016), Google sheets and excel are some of the current spreadsheet applications
that provide inbuilt facilities such as statistical, arithmetic and financial operations.
The application can be used by
programmers to develop spreadsheet programs without training. These applications fail
to provide the errors in the program that can be misleading in the decision-making process in an organization.
Almeida et al. focused on an approach, to develop high-quality
spreadsheet programs that provide the
programmer with the possible faults. Also,
the authors proposed an SS-BDD framework for
developing and conducting spreadsheets test scenarios. The tests used are
relied on Behavior Driven Development (BDD) to test three different
spreadsheets which are taken from various
sources. The accuracy is also tested by
injecting some faults knowingly in the spreadsheets. The primary elements
followed in the spreadsheet testing process are
scenarios to which the spreadsheets are
exposed and the values that are expected
from these conditions.
which are implemented are named as spreadsheets under test (SUT).
scripts communicate with the SUT input cells to verify that the output values
are corresponding to the expected input values.
The execution is
done, and a report is prepared to focus on the results of executed tests.
In 2016, Almeida et al.
conducted the test using three spreadsheets
with an SS-BDD framework with defined scenarios.
The first scenario is error-guessing test
case design technique, in second scenario test script was implemented manually,
all the implemented test scripts were executed.
During the execution of the test scripts,
38 faults were detected as the output of
According to Almeida et al. based on the challenges faced
during the test, the authors suggest that
instead of implementing these tests, the end user programmers must be allowed
to access the spreadsheets directly from the specified scenarios. Also, the
spreadsheet applications such as Excel and Google sheets should be analyzed by incremental graphical feedback
that is connected with the construction process.
The proposed framework can efficiently identify the faults
related to spreadsheets by helping the end user programmers. The authors plan
to implement another technique in the future which generates test scripts and
test inputs with the existing test tools for spreadsheets.
According to Ding et al. (2016), the tests identified do not evaluate all the statements
and other coverages such as modified condition/decision coverage(MC/DC). Also,
the efficiency of metamorphic testing is verified using partial mutation
testing, by applying to one module of ADDA program that is linked to critical
computing task. For a general idea, the testing
was performed with few mutants that are created by mutuation testing tools to
code manually. But, in this study the mutants were easily destroyed by simple tests
due to sceintific software complexity.
The relations identified indirectly
on image data were relatively weak and because of uncertainity of the test
output, testing was unsuccessful. Ding et al. proposed that by focusing on MRs,
the metamorphic testing can be performed and validated.