Software test result analysis and Quality report

As the code is one of the programmers ' results, test reports and quality reports are one of the main outcomes of testers. For a good test report, is based on the correct, sufficient test results, not only to provide the necessary test results of the actual data, at the same time to analyze the results, to find out the nature of the problem in the product, the quality of the product accurately evaluated.

1. Defect Analysis

Analyze the defect to determine whether the test has reached the end of the standard, that is, determine whether the test has reached the user acceptable state. When assessing defects, follow the analysis criteria set out in the defect analysis strategy, the most commonly used methods of defect analysis are:

Defect Distribution report, which allows the defect count to be displayed as a function of one or more defect parameters, to generate a function of defect quantity and defect attributes, such as the transverse distribution of defects in the program module, and the distribution of the severity defects on different causes.

Defect trend reporting, which displays the defect count as a function of time in various states, such as the time distribution of the defect quantity over the entire test cycle. Trend reports can be cumulative, or non-cumulative, can be seen in the trend of growth and reduction of defects;

The Defect age report, a special type of defect distribution report, shows the time the defect is active, showing how long a defect is in a certain state, and thus the progress of dealing with these defects.

Test Results Progress report showing the results of the test process in several versions of the application being tested, as well as the test cycle, showing the results of the test process execution after several iterations of the application and the test life cycle

At the same time, defects can be analyzed at the end of the project to improve the development and testing process, such as:

Understand the effectiveness of your tests with a trend analysis of defects (daily or weekly new defects), or you can see the quality of your tests based on the number of bugs lost and the total number of bugs found. According to the total number of test cases executed, we can calculate the number of test cases, test time and so on for each bug found, and compare and analyze the different stages and modules.

By the number of defects or the distribution of the module, you can master the quality of the program code, such as the number of bugs per thousand lines of code analysis, to understand the program code quality. Analyze the ability or status of the development team to resolve bugs through defects (daily or weekly corrections/closures) trends

2. Overall product quality analysis

The results of the test are collated, summed up and analyzed, generally with the help of Excel file, database and some histogram, pie chart, trend graph and so on to analyze and express, the main methods are comparative analysis, root cause (root cause) search, problem classification, Trend (time series) analysis.

Comparative analysis, the software to perform the test results compared to the standard output, because there may be some of the output can not be directly compared (for example, the date and time of the operation of the record, the path of the running of the record, as well as the test object version data, etc.), it is necessary to process the program.

Root cause (root cause) lookup, "Analysis" is to find out where does not match and indicate the probable cause of the error.

Problem classification, "classification" includes a variety of statistical sub-items, such as the location of the corresponding source program, the severity of the error (hints, warnings, non-failure errors, failure errors, etc.), newly discovered or have logged errors.

Trend (time series) analysis, based on the discovered software defect historical data analysis, forecast the future situation.

Other statistical analysis, through the classification of defects, and then use some mature statistical methods to analyze the existing data to understand the main problems in software development or the main causes of problems, so it is easier to improve software quality.

Reprint: http://www.uml.org.cn/Test/201105041.asp

Software test result analysis and Quality report