Unformatted text preview: ances. Given the small sample
size, and because we are looking for a definite decrease in waste, we use a one-tailed test to increase
the power of the test statistic (perhaps at the expense of stronger assumptions around the impact
5Data collection on this phase also occurred after the researcher had left the plant at the end of the study period of the feedback). Even so, the p-value of the test (0.84) indicates that we fail to reject the null hypothesis with any generally acceptable level of confidence.
Despite the relatively low statistical confidence using this approach. however, it is also observed
that average waste during the operator feedback period dropped about 14%, corresponding to an
overall energy reduction of 11%, which is fairly consistent with previous results from published
studies in other environments. If prior beliefs about the positive causal effect of giving feedback are
incorporated, then results from this field experiment at least argue at least for further testing with more sample...
View Full Document
This document was uploaded on 03/03/2014 for the course ADMIN 1 at Universidad Nacional de Colombia.
- Spring '14