PA 5.2 Quality Control

Purpose

The purpose of Quality Control is to statistically manage and control the test process. At this level, test process performance is fully predictable and stabilized within acceptable limits. Testing at a project level is performed using statistical methods based on representative samples in order to predict product quality and make testing more efficient.

质量控制的目的是系统的管理和控制测试流程。在这个水平，测试流程性能被充分的预测和稳固在一个可接受的限制内。项目级别的测试被执行使用系统的方法，基于典型的抽样方法来预测产品质量和使测试更加高效。

Introductory Notes

Quality Control consists of the procedures and practices employed to ensure that a work product or deliverable conforms to standards or requirements. In a broad view, the Quality Control procedures and practices can also be applied to the processes creating the product, thereby creating a feedback loop in line with the prevention-oriented and optimizing approach of TMMi level 5. At TMMi level 5, organizations use Quality Control to drive the testing process.

质量控制由过程和时间组成，这些过程和实践保证了工作产品和交付物满足标准和需求。从一个更大的角度来看，质量控制流程和实践能被应用到产品制作的过程中。从而做一个反馈环来保持和面向预防和优化的方法一致。组织使用质量控制来驱动测试过程。

Process Quality Control is supported by statistical techniques and methodologies. The basis for process quality control is viewing the testing process as a series of steps, each of which is a process in itself with a set of inputs and outputs. Ideally the output of each step is determined by rules, procedures and/or standards that prescribe how the step is to be executed. Practically speaking the outcome of a step may be different than expected. The differences are caused by variations. Variations may be due to human error, influences outside of the process, unpredictable events such as hardware/software malfunctions and so forth. If there are many unforeseen variations impacting the process step, then the process will be unstable, unpredictable, and uncontrollable. When a process is unpredictable then it cannot be relied upon to give quality results.

质量控制过程被支持通过统计技术和方法。质量控制流程的基准是检查测试过程作为一系列步骤。理论上讲每一步的输出被明确通过规则，过程，和标准（规定步骤如何被执行）。实际来说，每一步的输出可能与期望的不同。这些不同是变量造成的。变量可能是认为错误，过程外部影响，不可预测的事件比如软硬件不工作了等等。如果有很多不可预测的变量影响流程步骤，这个流程就会不稳定，不可预测，不可控制。如果一个流程不可预测那么它不能被信赖给出高质量结果。

An organization that controls its processes quantitatively will be able to do the following:  Determine the stability of a process  Identify the process performance within the defined natural boundaries  Identify unpredictable processes  Identify the improvement opportunities in existing processes  Identify the best performing processes

组织定量的控制流程将会做以下事情：明确流程的稳定性。明确过程的性能在定义的边界值内，明确不可预测的流程，明确对现有流程的提升机会，明确最好执行流程。

Process quality control involves establishing objectives for the performance of the standard test process, which is defined in the Test Lifecycle and Integration process area. These objectives should be based on the defined test policy. As already stated in the Test Lifecycle and Integration process area, multiple standard test processes may be present to address the needs of different application domains, test levels, lifecycle models, methodologies, and tools in use in the organization. Based on the measurements taken on test process performance from the projects, analysis takes place and adjustments are made to maintain test process performance within acceptable limits. When the test process performance is stabilized within acceptable limits, the defined test process, the associated measurements and the acceptable limits for measurements are established as a baseline and used to control test process performance statistically. The test process capability of the organization’s standard test process, i.e., the test process performance a new project can expect to attain, is now fully understood and known. As a result, the deviations from these expectations can be acted upon in a project early and consistently to ensure that the project performs within the acceptable limits. The test process capability can be used to establish unambiguous quantitative test process performance objectives for the project.

质量控制过程包含为标准测试流程的性能建立目标。它被定义在测试生命周期和集成过程域。这些目标应该基于定义的测试方针。就如在测试生命周期和过程域中指出的，多个标准测试流程可能存在来处理各种需求，这些需求来自不同应用域，测试水平，生命周期模型，方法，工具。基于项目测试流程性能的测量，分析存在，修正被做来维护测试流程性能在一个可接受的限制内。当测试流程性能被固定在一个可接受的范围呢，定义的测试流程，相关的测量和接受范围被建立，作为一个基准和用来控制测试流程性能。组织的标准流程的测试过程能力。比如，新项目能够达到的测试流程性能。现在被充分的理解和了解。因此，测试流程的能力被使用来建立清晰的定量的测试流程性能目标。

Product quality control builds on operational profiles [Musa] and usage models of the product in its intended environment to make statistically valid inferences resulting in a representative sample of test cases. This approach, especially useful at the system test level, uses statistical testing methods to predict product quality based on this representative sample. In other words, when testing a subset of all possible usages as represented by the usage or operational profile, the test results can serve as the basis for conclusions about the product’s overall performance. At TMMi level 5, an organization is able to quantify confidence levels and trustworthiness because the infrastructure has been provided to reflect the most frequently requested operations or paths through an operational profile using historical data. Using test data from statistical testing, models such as reliability growth models are built to predict the confidence level and trustworthiness of the system. Confidence level, usually expressed as a percentage, provides information as to the likelihood that the product is defect free. Trustworthiness is defined as the probability that there are no defects in the product that will cause the system to fail. Both the level of confidence and trustworthiness are typically used as exit criteria when applying statistical testing. At TMMi level 5 these factors are used in combination and are usually the main drivers to determine when to stop testing.

产品质量控制基于运营模型和产品的使用模型，在预期的环境来做一个统计地合理的争对测试用例的代表性抽样。这个方法，在系统测试级别特别有用。使用统计测试方法来预测产品质量，基于代表性抽样。换句话说，测试结果能作为产品整体性能的基本定论。在tmmi5，组织有能力来确定自信水平和信任度，因为框架被提供来反映频繁请求的操作或者操作的路径。使用统计测试的测试数据，模型比如可靠性增长模型被建立来预测系统的信心水平和信任度。信心水平。通常用百分比来表达。提供信息来表达产品零缺陷的可能性。可信赖度定义为产品零风险的可能性。信心水平和信赖度通常被使用来作为统计测试的出口标准。在TMMI5.这些因素被用来作为合适停止测试的参考。

Note that addressing product quality control and statistical testing requires a great deal of expertise with statistical techniques including modeling, usage modeling, statistics, testing, and measurements. Specialists must be selected and trained to become leaders in this area of testing.

备注，处理产品质量控制和统计测试需要大量的统计技术经验，包括模型，使用模型，统计学，测试和测量。 员工被选择和培训来成为这个测试领域的领导。

Scope

The process area Quality Control addresses the practices for establishing a statistically controlled test process (process quality control), and testing based on statistical methods and techniques (product quality control). Process quality control strongly builds on the deployed measurement practices from the Test Measurement process area from TMMi level 4. Product quality control builds on the deployed practices from the Product Quality Evaluation process area from TMMi level 4. Both types of quality control make use of available measurement data regarding the test process and product quality from the TMMi level 4 process areas.

质量控制过程域跟踪建立一个统计控制的测试过程的实践，以及基于统计方法和技术的测试。质量控制过程很强的建立在部署测量实践（TMMI4测试测量）。 质量控制基于部署实践。两种类型的质量控制使用可用的测试数据和产品质量。

Specific Goal and Practice Summary

SG 1 Establish a Statistically Controlled Test Process SP 1.1 Establish test process performance objectives SP 1.2 Establish test process performance measures SP 1.3 Establish test process performance baselines SP 1.4 Apply statistical methods to understand variations SP 1.5 Monitor performance of the selected test processes

建立统计控制测试流程。

SG 2 Perform Testing using Statistical Methods SP 2.1 Develop operational profiles SP 2.2 Generate and execute statistically selected test cases SP 2.3 Apply statistical test data to make stop-test decisions

执行测试使用统计方法。