The threat of weighting biases in environmental decision analysis

We investigate the existence of biases, in particular the so called splitting bias, in a real environmental decision analysis case. The splitting bias refers to a situation where presenting an attribute in more detail may increase the weight it receives. We test whether the splitting bias can be eliminated or reduced through instruction and training. The case was the regulation of a lake-river system and the test group consisted of local residents and university students. The test groups carried out attribute weightings with different value tree structures. Our results show that the splitting bias remains a real threat. In the weighting experiments most students avoided the bias. However, nearly all of the local residents showed a systematic splitting bias so that the total weight of the attributes grew in proportion to the number of sub-attributes included. We discuss ways to eliminate the biases by balanced problem structuring.