|
|
 |
Importance of Test Plans/Test Protocol (with a Template)
|
|
|
|
|
|
|
|
|
"Is there such a thing as experimental design when we build a logisitc regression model and all predictors are catogorical? ...We need to test various combinations of envelope color, size and font and find out which one gives the highest response rate."
Contribute to this Discussion
|
|
|
|
By Kim Niles Test plans, also called test protocol, are formal documents that typically outline requirements, activities, resources, documentation and schedules to be completed. Some form of test plan should be developed prior to any test. The key reasons for developing test plans are: - Preparation: To assure that all reasonable aspects of running a test have been considered.
- Communication and Training: To train those who need to assist with the test.
- Effectiveness: To provide a mechanism for outlining test needs, limitations (listing assumptions), and justification for purposes of setting expectations, acquiring resources, investigating unexpected results, assuring normalcy and effectiveness.
- Legal and Regulatory Prudence: To enable replication and protection of discoveries made, to mitigate potential litigation costs from use of those discoveries, and to help provide evidence to show regulatory bodies of efficacy.
Test Plan ContentTitles and order of sections within a test plan vary significantly from company to company, given differences in test types, strategy, scope and industry. Examples of tests which should have plans or protocol written for them include medical, non-medical, field, lab or production studies, design of experiments (DOE), problem solving, reliability or software regression tests, and design assurance, evaluation or validation tests. Beginning of the Plan (typically background-related information) – Suggested considerations for information to be included in the beginning of the plan are header information, title, date, author, file code, project number, product/device/model description, sub-components of that product actually being tested, to whom the report is written, test number and revision, and references to other related documents. References might include industry standards, brainstorm documents (fishbone diagrams, FMEA/FTA/FMECA studies, related V&V protocols and reports, field history or rejection reports, company procedures, related website links, etc.). Cover pages and a table of contents should be considered for large plans. Most plans contain an explanation of what the test is about (i.e. summary, introduction, abstract, background, purpose, history, etc.). Some plans also include test type and strategy information, definitions, important terms or key words, approval signatures, return on investment and/or other justification information. Template for Test PlanA free Word template for a test plan with common sections already setup accompanies this article. Using details from previous test plans, statistical software, company or project logos, and other materials, you can cut and paste in the template and shorten your test plan development time. Or save yourself and your company even more time, money, risk and related trouble by using the template to assist you in developing your own test plan format. Of course, you will want to remove sections that do not fit the way you do business. |
|
|
Middle of the Plan (typically scope-related information) – Suggested considerations for information to be included in the middle of the plan are a data collection or sampling plan (including sample size, confidence acceptance levels and sampling techniques), test conditions or setup instructions, a test procedure specifying exact measurements to be taken, test monitoring requirements, and resources (equipment and personnel) to be used including responsibilities. Most plans contain some form of flow chart, process or value stream maps, assumptions that need to be made and why (including statistical distribution related as applicable), a work breakdown structure or schedule, and/or how the test factors (response Ys and control Xs) were selected. Some plans include special customer or site-related requests or considerations, measurement system capability information, and how test-related problems are to be reported and corrected (i.e. failure reporting and corrective action system). Also include unresolved issue reporting. End of the Plan (typically analysis-related information) – Suggested considerations for information to be included in the end of the plan are statistical techniques to be used, the hypothesis to be tested, the power of the test matrix and/or why it was chosen (for DOE) and a definition for test success (pass/fail criteria as possible). Most plans include how the data is to be analyzed, contingencies for how to handle different types of preliminary or insufficient results, how the conclusions are to be reported or summarized (i.e. significance with confidence, pass/fail, etc.). Some plans also include references and/or a bibliography, appendices that might be needed, and copyright or proprietary information messages as applicable. Generic Test Plan Advice- Make Them Context Rich: Context-rich test plans contain seemingly unimportant details that often do help reduce confusion, document test details that might be needed, and/or help explain any unexplained results.
- Use Hypertext: It's often helpful to add links to other parts of the same document (i.e. "jump to conclusion," "see details," "see references").
- Add Key Assumptions: Listing key assumptions in the plan can be important for setting reader expectations, influencing approval, assuring management that the test is well thought out, explaining unexpected results, and providing direction for future tests. Example assumptions are as follows:
- "Results are assumed to contain an even distribution of errors."
- "Temperature change stabilization times selected are assumed to be sufficient."
- Convert Test Plan into a Test Report as Possible: Test plans can often be converted into test reports. Some sections only change in tense. Other sections can be reused exactly as they were written.
- Take Extra Data-Related Precautions: Ensure sampling is representative of the population (random and unbiased), personally observe data collection, and make sure the measurement system is capable.
- Develop Test Libraries: Standard tests and/or test results from prior tests (especially reliability tests) should be easily accessed and re-used as applicable and possible.
Conclusion: Test Plans a Sensible StepIt is easy to see how test plans are an important and sensible part of performing an experiment. They save time and money, assist in getting the best results and can facilitate speedy test report writing. About the AuthorKim Niles has more than 17 years process control and improvement experience working with San Diego manufacturing companies in a wide range of industries and disciplines. Currently an officer in three professional societies, Mr. Niles has a master's degree in quality science with an emphasis in Six Sigma from California State University Dominguez Hills. He has a bachelor's degree from San Diego State University through the industrial technology department. He can be reached at kim.niles@cox.net.
Copyright � 2000-2010 iSixSigma – All Rights Reserved
Reproduction Without Permission Is Strictly Prohibited – Copyright Requests
Publish an Article: Do you have a Six Sigma tip, learning or case study?
Share it with the largest community of Six Sigma professionals, and be recognized by your peers.
It's a great way to promote your expertise and/or build your resume. Read more about submitting an article.
"The Bottom Line" Links
|
|
|
