June 12, 2005
June 12, 2005
June 15, 2005
10.410.1 - 10.410.6
DESIGN OF EXPERIMENTS IN UNDERGRADUATE LABORATORY EDUCATION
Swami Karunamoorthy Saint Louis University, St. Louis, Missouri
Abstract Design of experiments is a necessary skill for a test engineer in an industry. In any engineering program, it is an important learning outcome. In this paper, an emphasis is given to how this skill can be developed in undergraduate laboratory education. Some examples are presented along with theoretical background that can be easily implemented in laboratory courses. It is a viable approach to give an exposure to design of experiments as well as to enhance the learning experience in laboratory education.
Introduction Laboratory experiments are essential part of engineering curriculum. Traditionally, students in a laboratory course would set up an experiment, take measurements, analyze data, plot graphs, and write a report. This approach provides a learning experience on how to conduct experiments and how to analyze data. However, it does not provide an experience in design of experiments. “Design of experiment” means planning the experiment1 and one of the aspects is statistical design of experiment. Statistical design in general implies the estimation of number of measurements or tests required to determine the true mean of a variable being measured. In a typical laboratory experiment a student group would collect only one set of data for a given specimen. But, similar set of data would be available from several groups performing the same test at different times. While it is difficult to apply the design of experiments on the data from any single group, one could pool the data from all groups and apply the principles of design of experiment. So, it is important to understand Pooled Statistics and hence some background on this theory is also included in this paper.
Statistics for Lab Data scatter introduces an uncertain vagueness into the measurement scheme, which requires statistical methods to quantify. Statistics then becomes a powerful tool to interpret and present data. Statistics for laboratory experiments can be broadly classified into three categories.
1. Infinite Statistics: This is based on the assumption that very large amount of experimental data are available. Then the True Mean can be obtained from Normal Distribution Table. 2. Finite Statistics: This is based on the assumption that the available data are not large. In this case, the True Mean can be estimated from the available sample mean with the help of Student’s “t” Distribution Table.
“Proceedings of the 2005 American Society for Engineering Education Annual Conference & Exposition Copyright © 2005, American Society for Engineering Education”
Karunamoorthy, S. (2005, June), Design Of Experiments In Undergraduate Laboratory Education Paper presented at 2005 Annual Conference, Portland, Oregon. https://peer.asee.org/15508
ASEE holds the copyright on this document. It may be read by the public free of charge. Authors may archive their work on personal websites or in institutional repositories with the following citation: © 2005 American Society for Engineering Education. Other scholars may excerpt or quote from these materials with the same citation. When excerpting or quoting from Conference Proceedings, authors should, in addition to noting the ASEE copyright, list all the original authors and their institutions and name the host city of the conference. - Last updated April 1, 2015