Analysis of extraction cost of quality and testing phase by combining Salleh and Primandaria’s model
Software Testing activities are important to find defects, gain confidence about the level of quality, provide information for decision-making and prevent defects. Nowadays, testing from independent organisations is the best practice to measure the confidence level of stakeholders before deploying a...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | English |
| Published: |
2018
|
| Subjects: | |
| Online Access: | http://psasir.upm.edu.my/id/eprint/69043/1/FSKTM%202018%2058%20-%20IR.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Software Testing activities are important to find defects, gain confidence about the level of quality, provide information for decision-making and prevent defects. Nowadays, testing from independent organisations is the best practice to measure the confidence level of stakeholders before deploying and using a system, especially high impact ones. However, currently, the government agencies of Malaysia do not have any cost estimation models to implement quality and testing phase by independent organisations.
Therefore, the purpose of this study is to analyse and extract the cost of quality and testing phase from the total cost of a software project. The objectives of this study are to design an extraction model to extract the cost of quality and testing phase from the total cost of a software project, to build a prototype for the model and to evaluate it.
This study provides options to extract the cost of quality and testing phase from the total cost of a software project. The options will be based on estimation and/or prediction. Constructive Cost Model (COCOMO) and estimation model by Saleh 2011 were chosen as estimation techniques, while linear regression model was chosen to predict the cost of quality and testing phase. This study used Function Point Analysis (FPA) to measure the size of system.
The best-fitting line that was developed based on four existing projects that implemented the outsourced test team was ŷ = 92,774.32 + 216.04 x̂, where the slope of the line (β) was 216.04, and the intercept (α) was 92,774.32. This study had validated the predicted linear regression by Mean Magnitude of Relative Error (MMRE) and PRED (0.25). The result for MMRE was 0.15 and PRED (0.25) was 1. A small value of MMRE means the estimation is acceptable and PRED (0.25) of 1 means the Prediction Quality is acceptable. |
|---|