Applying the exact randomization test in experimental economics

✍ Scribed by Robert Moir

• Publisher: John Wiley and Sons
• Year: 1998
• Tongue: English
• Weight: 126 KB
• Volume: 9
• Category: Article
• ISSN: 1180-4009
• DOI: 10.1002/(sici)1099-095x(199801/02)9:1<93::aid-env289>3.0.co;2-k

✦ Synopsis

The relatively recent movement to test economic theory and policy in the laboratory has placed new demands upon researchers' data analysis skills. Without any idea of the distribution of errors about the mean, researchers are often forced to resort to a battery of tests in case some of the assumptions for a particular test are not met. Their eorts are further hampered by the small sample size generally associated with laboratory data. The exact randomization technique permits the researcher to make use of standard parametric tests (avoiding the less powerful non-parametric techniques) with few restrictive distributional assumptions. In this paper, the exact randomization t-test is used to analyse data from a common-pool resource experiment. The model used to generate payos in the laboratory shares the characteristics of many environmental situations, particularly those associated with sustainable development. The results indicate that monitoring, on its own, serves to make people more competitive over resource use. Only when monitoring is backed up by some punishment scheme such as sanctioning, do groups achieve some success in governing themselves.