Why Follow these Courses?
The question of “How much testing is enough?” is persistent across Department of Defense Test and Evaluation endeavors.
Design of Experiments (DOE) can help decision makers “right-size” a test, providing a framework for building tests, evaluating test plans, and assessing risk. Leveraging statistical techniques in both test design and test analysis helps decision makers reach correct conclusions about system performance.
The Test Science Courses provide a step-by-step guide to advise testers on how to most efficiently Plan, Design, Analyze, and Evaluate a test.
Planning a Test
Planning is a critical element of this process. Test planning involves identifying the test objectives, determining the appropriate outcome measure(s), and identifying what factors may influence those outcomes. A clear goal when designing a test is essential. The objective of any operational or live fire test will drive the associated test design and analysis. The two most common test objectives used in test and evaluation are screening for influential factors and characterizing a systems performance across a set of conditions. These are best applied sequentially; once a screening experiment has narrowed down the set of factors that are likely to matter, a characterization design describes in more detail the performance of the system at each of these relevant conditions.
Designing a Test
The best approach for collecting data from several factors is to conduct a statistically designed experiment. Design of Experiments (DOE) is an experimental strategy in which factors are varied together strategically, instead of one at a time, or test cases being arbitrarily chosen. A test strategy that employs DOE will always provide the most powerful allocation of test resources for a given number of tests. The specific choice of design should be tied to the test objectives and connected to the anticipated analysis. The test design should have a large enough sample size to be useful, be executable in the real world, and support the desired analysis.
How to Choose a Course
Course videos can either be watched in a series from start to finish, or can be watched as a single video on a topic. Watching each video in order walks the tester through the process section-by-section and can provide insight for how to proceed. For help with a specific topic, select the video that seems most appropriate for your testing needs.
Tools
Many courses link to interactive tools for to help the tester design the best test possible. These tools cover a range of topics from checklists that help with planning to statistical packages to help analyze or visualize results. To view the tools associated with a particular video, scroll down to the bottom of the video and click the “Tools” tab, or to see all tools, click here.
Test Planning:
- Introduction to Test Planning
- Identify Test Objectives
- Select Outcome Measures
- Select Factors Affecting Outcomes
Test Design:
- Introduction to Test Design
- Determine Design Goals
- Build the Test Design
- Evaluate the Design
- Manage and Store the Test Data
Conduct Analysis:
- Introduction to Data Curation and Analysis - Coming Soon!
- Make Your Analysis Reproducible - Coming Soon!
- Exploratory Data Analysis - Coming Soon!
- Modeling Data – Inferential Statistics - Coming Soon!
- Modeling Data – Bayesian Statistics - Coming Soon!
Evaluate Performance:
- Evaluating Your Test Objectives - Coming Soon!
- How to Present Your Results - Coming Soon!