Book contents
- Frontmatter
- Contents
- List of figures and tables
- Preface
- Acknowledgments
- EXPERIMENTAL METHODS
- 1 Introduction
- 2 Principles of economics experiments
- 3 Experimental design
- 4 Human Subjects
- 5 Laboratory facilities
- 6 Conducting an experiment
- 7 Data analysis
- 8 Reporting your results
- 9 The emergence of experimental economics
- Appendixes: Supplemental materials
- Glossary
- References
- Index
6 - Conducting an experiment
Published online by Cambridge University Press: 05 June 2012
- Frontmatter
- Contents
- List of figures and tables
- Preface
- Acknowledgments
- EXPERIMENTAL METHODS
- 1 Introduction
- 2 Principles of economics experiments
- 3 Experimental design
- 4 Human Subjects
- 5 Laboratory facilities
- 6 Conducting an experiment
- 7 Data analysis
- 8 Reporting your results
- 9 The emergence of experimental economics
- Appendixes: Supplemental materials
- Glossary
- References
- Index
Summary
Learning to conduct economics experiments is like learning to swim or bike; no amount of reading will substitute for giving it a try. You will find it more efficient to learn lab practice by offering to assist someone who does experiments. It is also easy to pick up good as well as bad habits from others. This chapter is intended as a helpful checklist for a beginner. As you learn to conduct experiments yourself, you will develop your own more detailed checklist of do's and don't's.
Lab log
It is a good lab practice to record all experiments in a bound log book by date, purpose, subjects, software, and parameter values. Also note any unusual events (e.g., “Ran with 8 subjects instead of the planned 10 due to no-shows.”) It is amazing how soon we forget what we did, when, and why. A log book is a permanent record of great value.
Pilot experiments
Conducting pilots is usually the only sensible way of developing the design and procedures of new experiments. In conducting pilots, you will discover many things, including ambiguities in the instructions, missing information, unintended leakages of information, too much time for some activities and too little for others, weaknesses or malfunctions in software or random-number–generating devices, insufficiency of assistants, and lack of coordination and timing of your tasks. Some pilot experiments are planned as such, but quite often this designation is applied ex post to experiments in which procedural glitches arise.
- Type
- Chapter
- Information
- Experimental MethodsA Primer for Economists, pp. 74 - 84Publisher: Cambridge University PressPrint publication year: 1994