Objectives
Customer satisfaction for most software systems largely depends on the reliability of the system in question. However, software systems are subject to malfunctions and defections. As a result, a test for any system is necessary to ensure the reliability and performance of the system are as expected by the system developer and designer. This report analyses the competence of one tester to finding defects in a system within a period of 20 weeks. The time of work for both the testing software and the workers is fixed at 25 hours a week. The defects to be fixed were classified to be either hard or easy and minor or major. The hard defects required 5 hours to fix while the easy ones required 2 hours to fix. Excel was used to plot a scatter graph for the data. A defect fitting line was fit to the data to determine the trend of the defects over the 20-week period. The following three metrics were used to measure and analyze the process. An exponential formula was used to predict a number of defects at a future date. Another metric was used to determine the ratio of the total found defects and the fixed defects. This metric is important to determine the sufficiency of the tester within a specified period of time.
Objectives
The purpose of the report was to achieve the following objective:
- Determine the relevance of the defects funds on the actual performance of the system.
- Average time it takes to fix the defects found
- The impact of damage the defects may have on the system
- The determinant of the system reliability
Assumptions
The following assumptions were made when compiling the results of this report:
- The tester was working perfectly and all the defects were the true reflections of defects from the system.
- The time was distributed equally throughout the 20-week period
- The errors were not as a result of errors made by the engineer during the testing period.
Discussion
A tester was used to collect defects data on the defects. The collected information was classified into four groups of data: major, minor, easy and hard. Data were classified to be either major or minor on the basis of its impact on the functioning of the system. Hard and easy defects were classified based on the time it took to correct the type of defects. A total of 300 defects were found, among them, 100 were considered as major and hard while 200 as minor and easy. The figure below is the table with the classified data.
In addition, there is addition column showing remaining major defects and another row for total major defects.
Results
An analysis of the data showed the number of the defects were decreasing across the period of the 20 weeks with week one having the highest number of total defects at 47 and week 20 having the least at 1. However, easy and minor defects were not as consistent as the other the major and hard, as well as minor and hard. The average defects per week were also on the decline with week one having 11.75 and week 20 having 0.25. 31.67% of the defects were categorized as major (both hard and easy to fix). A defect fitting line was drawn into the data using the exponential function y = 60.516e-0.178x. The exponential function shows a decline of the number of defects found with time converging to zero at past week 20. The figure below shows the defects fitting line showing the trend of the defects.
The possible convergence to zero is also supposed by the average number of defects observed per week.
Staff Allocation
Since the major number of defections are considerably large at 31.67%, their impact on the system performance might be considered high. In addition, much of major defects, both hard and easy are clumped within the first three weeks. Consequently, the best strategy would be to solve the major ones first for the first seven weeks first. The following five weeks would then be allocated to the easy ones then resume on the hard ones. Alternating between the major ones and the hard ones would ensure an equal distribution by the last week of working.
Conclusion
System reliability is directly related to a number of defects that may be present in that system. A large number of defects is likely to lead to a conclusion that the system cannot e reliable. A management of defects ensures that a system will work as expected and therefore, translate to a higher customer satisfaction. It is, therefore, necessary to conduct a regular system test to ensure that a system functions the way it’s supposed to. However, challenges such as minimal resources force a strict allocation of resources and require tough decision making to decide what is more important than the other during system testing.
In this report, the system engineer uses the data to determine which category is more likely to affect the performance of the system, this ensures that the most important aspects of the test are given the first priority. For instance, the major category is more likely to affect the system’s performance than the minor category. However, a good number of the major category is in the hard category and may as well take a considerably long period of time fix. This is likely to leave behind a number significantly large number of the minor defects that may have an effect on the performance. As a result, the engineer needs to balance between defects that may have a minor impact and those that have a major impact. To achieve an optimum result for detection of the defects, it would be important to model a poisson distribution for the waiting time across the for possible parameters. The one with the lowest waiting time would be the most suitable to ensure maximum outcome for the defects detection. However, it will be a hard choice to choose one parameter over the other since the underlying factor as well depends on the effect these parameters have on the performance of the system.
Order an Essay Now & Get These Features For Free:
Turnitin Report
Formatting
Title Page
Citation
Outline
