TMMI_测试过程改进框架_优化级别8_质量控制2

Specific Practices by Goals

SG 1 Establish a Statistically Controlled Test Process

建立一个统计控制测试流程

A statistically controlled test process is established whereby baselines that characterize the expected test process performance of the organization’s standard test processes are established and maintained.

基于基准的一个统计控制流程被建立。这个基准描述了组织标准测试流程的期望的测试过程性能被建立和维护

SP 1.1 Establish test process performance objectives 建立测试过程性能目标

Establish and maintain quantitative objectives for test process performance. Note this specific practice correlates to and builds upon the SP 1.1 Establish test measurement objectives of the Test Measurement process area at TMMi level 4. Example work products 1. List of test processes identified for test process performance analysis 2. Quantitative objectives for the organization’s test performance Sub-practices 1. Study the business needs and objectives regarding product quality and test process performance 2. Study the test policy with respect to the defined test goals and test performance indicators Refer to the Test Policy and Strategy process area for more information on business needs and objectives, test goals and test performance indicators. Explicitly select test processes from the organization’s set of standard test processes that are to be included in the set of statistically controlled test processes. Typically it will not be possible, useful, or economically justifiable to apply statistical management techniques to all test processes of the organization’s standard set of test processes. Selection of the test processes is based on the needs and objectives of both the organization and projects.

为测试流程性能建立和维护定量的目标。备注这个指定的实践关联和建立在SP1.1建立测试测量目标的上。例子工作产品：1. 测试流程列表，基于确认的测试流程性能分析。 2.组织测试性能的定量目标。

子实践：1. 学习需要的业务和基于产品质量和测试过程性能的目标。2. 学习测试方针涉及到的定义的测试目标和测试性能指标。参考测试方针和测试过程域来获得更多商业需求和目标的信息，测试目标和测试性能指标。明确地从组织的标准测试流程选择测试过程，被包含在统计控制测试流程集合中。使用统计管理技术到所有的测试流程是无用或者说经济的。测试流程的选择是基于组织和项目的需求和目标的。

3. Define the organization’s quantitative objectives for test process performance in cooperation with relevant stakeholders Objectives may be established directly for test process measurements (e.g., test effort and defect removal effectiveness) or indirectly for product quality measurements (e.g., reliability) that are the result of the test process. 4. Define the priorities of the organization’s quantitative objectives for test process performance in cooperation with relevant stakeholders, e.g., customers and end users 5. Resolve conflicts among the test process performance objectives (e.g., if one objective cannot be achieved without compromising another objective) 6. Revise the organization’s quantitative objectives for test process performance as necessary Examples of when the organization’s quantitative objectives for test process performance may need to be revised include the following:  Based on findings and recommendations from regular test process assessments. (Refer to the Test Organization process area at TMMi level 3 for more information on test process assessments. Note that at TMMi level 5 test process assessments, both formal and informal, are typically performed more frequently.)  When the organization’s business objectives change  When the organization’s (test) processes change  When actual test process performance differs significantly from the objectives

3.与相关利益人配合为测试过程性能定义组定量目标。目标可能被直接建立作为测试过程测量或者非直接的作为产品质量测量（测试过程的结果）。4，定义测试流程性能的组织定量目标的优先级，与相关利益人一起。比如：客户和用户。 5，解决测试流程性能目标的冲突。6，修改组织的测试流程性能的定量目标，如果必须的话。关于什么时候应该修改测试流程性能的定量目标。包括：基于常规测试过程评估的发现和建议。组织的业务目标改变，测试流程改变，实践测试流程性能明显和目标不同。

SP 1.2 Establish test process performance measures 建立测试流程性能测量

Establish and maintain definitions of measures that are to be included in the organization’s test process performance analyses. Refer to the Test Measurement process area at TMMi level 4 for more information about selecting and establishing measures, especially SP1.2 Specify test measures. Example work products 1. Definitions for selected measures for test process performance Sub-practices 1. Select measures that provide appropriate insight into the organization’s test process performance Interim measures for each lifecycle phase may need to be selected, as appropriate, to monitor progress early toward achieving the objectives. 2. Identify measures that are appropriate for statistical management, e.g., those that are controllable 3. Review the set of test process performance measures for statistical management 4. Incorporate the selected measures into the organization’s set of common test measures 5. Revise the set of test process performance measures as appropriate

建立和维护测量的定义，组织测试流程性能分析包含的测量。参考TMMI4的测试测量流程域。例子工作产品：定义测试流程性能选择的测量。 子实践：1.选择测量，提供合适的观点到组织测试流程性能，每个生命周期阶段的临时测量应该被适当选择。在早期监控进程以达到目的。 2. 明确适合统计管理的测量，比如，可控的测量。 3重审测试流程性能测量。 4 把选择的测量集成到组织常用测试测量中。 5.修改测试流程性能测量。

SP 1.3 Establish test process performance baselines 建立测试流程性能基准。

Establish and maintain the organization’s test process performance baselines. The organization’s test process performance baselines are a measurement of test performance for the organization’s set of standard test processes at various levels of detail, as appropriate. There may be several test process performance baselines to characterize test performance for subgroups of the organization, e.g., application domains, complexity, project size and lifecycle models. Example work products 1. Baseline data on test process performance Sub-practices 1. Collect and analyze measurements from projects Refer to the Test Measurement process area for more information on collecting and analyzing data. 2. Establish and maintain the organization’s test process performance baselines from the collected measurements and analyses Test process performance baselines (typically including minimum and maximum tolerances) are derived by analyzing the collected measures to establish a distribution and range of results that characterize the expected performance for selected test processes when used on an individual project in the organization. 3. Review the validity and get agreement with relevant stakeholders about the test process performance baselines 4. Make the test process performance baselines available across the organization The test process performance baselines are used by the projects to estimate the upper and lower boundaries for test process performance. (Refer to SP 1.4 Apply statistical methods and understand variations for more information on upper and lower boundaries of test process performance.) 5. Revise the set of test process performance baselines as appropriate Examples of when the organization’s test process performance baselines may need to be revised including the following:  When the organization’s business objectives change  When the organization’s (test) processes change  When actual test process performance differs significantly from the baselines

建立和维护组织测试过程性能基准。组织的测试过程性能基准是测试性能的测量，有多个测试过程性能基准来为组织的子部门描述测试性能。比如，应用域，复杂度，项目大小，生命周期模型。

例子工作产品：1. 测试流程性能的基准数据

子实践：1. 从项目搜集和分析测量，参考测试测量过程域获得跟多关于搜集和分析数据的信息。2.建立和维护组织的测试流程性能基准，通过搜集的测量方法和分析测试流程性能基准。3. 重审合理性和与相关利益人获得对于测试流程性能基准的一致观点。 4.是测试流程性能基准在整个组织有效。这个基准在项目使用来评估测试流程性能的范围。5.修改。

SP 1.4 Apply statistical methods to understand variations 应用统计方法来理解变量

Establish and maintain an understanding of the performance variations of the selected test processes using the selected measures. Understanding variation is achieved, in part, by collecting and analyzing measures so that (common) causes of variation can be identified and addressed to achieve predictable performance. Example work products 1. Collected measures 2. Upper and lower boundaries of test process performance for each measured attribute of each selected test process 3. Test process performance compared to the upper and lower boundaries of test process performance for each measured attribute of each selected test process Sub-practices 1. Establish trial upper and lower boundaries (control limits) for test process performance using suitable historical data Boundaries of an attribute are the range within which variation normally occurs. All processes will show some variation in measures each time they are executed. Generally control limits (upper and lower boundaries) are set to 2 or 3 sigma. Control limits may also be set to 1 sigma, which is the most restricted interval. The control limits help the Test Process Group to separate the signals from the noise. The variation of data points inside the control limits is due to noise in the process (common cause or normal variations). 2. Collect data, as defined by the selected measures, on the test processes as they execute Data is collected and validated from projects across the organization according to the selected procedures. The data is stored in the test measurement database.

建立和维护选中的测试流程的性能变化的理解，使用选择的测量方法。

3. Calculate the boundaries of test process performance for each measured attribute Examples of where boundaries are calculated include the following:  Control charts  Histograms  Run charts  Confidence intervals  Prediction intervals 4. Identify causes of variation The criteria for detecting causes of variation are based on statistical theory, e.g., 3-sigma control limits, experience, and economic justification. An important activity during the identification of causes of variation is to determine if a process variation is caused by special circumstances (an assignable or specific cause) which may be linked to a specific time or location, or by the variation inherent in the nature of the process itself (common or natural cause). In general it is recommended to first solve assignable causes, since these may partly decrease common causes. 5. Analyze the cause of test process variation to determine the reason the anomaly occurred Examples of techniques for analyzing the reasons for causes of variation include the following:  Cause and effect (fishbone) diagrams  Designed experiments  Control charts (applied to input or underlying test sub-processes)  Subgrouping Note that some anomalies may simply be extremes of underlying distribution rather than problems. Refer to the Defect Prevention process area for more information about analyzing the cause of an anomaly. 6. Determine what corrective action should be taken when causes of variations are identified Refer to the Test Process Optimization process area for more information about taking corrective action. 7. Recalculate the upper and lower boundaries for each measured attribute of the selected test processes as necessary 8. Record statistical management data in the organization’s measurement repository Refer to the Test Measurement process area for more information about managing and storing data, measurement definitions, and results.

SP 1.5 Monitor performance of the selected test processes 监控选择的测试过程的性能

Monitor the performance of the selected test processes to determine their capability to satisfy their process performance objectives and to identify corrective actions as needed. The intent of this specific practice is to determine statistically the process behavior exhibited by the process and appraise the probability that the process will meet its process performance objectives. Example work products 1. The process performance for each test process 2. Boundaries of test process performance for each selected test process compared to its established objectives3. For each test process, documentation of actions needed to address deficiencies in its process performance Sub-practices 1. Compare the boundaries of the measured attributes to the test process performance objectives This comparison provides an appraisal of the test process capability for each measured attribute of a test process. 2. Periodically review the performance of each selected test process, its capability to be statistically managed and appraise progress towards achieving the test process performance objectives 3. Identify and document test process capability deficiencies 4. Determine and document actions needed to address test process capability deficiencies

监视选中测试流程的性能来决定他们的能力来满足流程性能目标和验证需要的正确行为。