Experimental Strategy: Pre-Experimental Planning
From LabAutopedia
Experimental Strategy: Pre-Experimental Planning
Authored by: James N. Cawse, Ph.D.;Cawse and Effect LLC
The general who wins a battle makes many calculations in his temple before the battle is fought. Sun Tzu, “The Art of War”
To get the most useful information out of experimental program, a few hours of thought and preparation at the beginning are worth days or weeks of agony at the end. Far too rarely do experimenters go through a systematic process of defining their problem before rushing into the lab or operation to "get something done". The following is a guide to slow you down for those crucial first hours. The Japanese have the maxim "aim slow, shoot fast"; it seems to work much better than the old slogan "ready, fire, aim". In order to guide the process of defining the experiment, Montgomery introduced a DOE Master Guide in 1993[1] which is updated here for high throughput experimentation[2]. The DOE Master Guide Sheet is designed to be discussed and filled out by the project team. The team should have representation of all the key disciplines involved: engineers, scientists, operators, managers, and process experts. If all the elements on the sheet are discussed and filled in with care, the team should be well prepared to design and carry out the experiment. Do not expect the completed Master Guide to only fill one page! It is presented in abbreviated form here, with detailed discussion in the links. In addition, the Master Guide is very useful at the end of the experiment as an outline for the final report.
Master Guide Sheet
Project Team:
Project Title:
- Objective of the Experiment
- Relevant Background Knowledge
- Resources Available (Personnel, stakeholders, money, machines, materials, time)
|
Factor (units) |
Type (Quantitative,Qualitative, Formulation) |
Normal Level & Range |
Measurement Precision & setting error |
Proposed Settings |
Ease of Adjustment |
|
Objective |
Expected Level and Range |
Actual Response Measured |
Measurement method: speed, precision, & accuracy |
Relationship of response to Objective |
| |
|
Factor (units) |
Type (nuisance, constant, blocking) |
Measurement Precision |
Strategy for dealing with this factor |
Anticipated Effects |
- Nesting:
- Level of interaction and curvature to be searched:
- Restrictions on the experiment:
- Automation to be used and its issues
- Scale of experiments
- Trial Runs?
References
- ↑ Montgomery, D.C., and Coleman, D.E., Technometrics 1993,35, 1-12.
- ↑ J. N. Cawse, Ed., Experimental Design for Combinatorial and High Throughput Materials Development, Wiley-Interscience, 2002.
| Click [+] for other articles on | The Market Place for Lab Automation & Screening | Automation Software |

