Path dependence and biases in the even swaps decision analysis method

Tuomas Lahtinen, Raimo Hämäläinen*

*Corresponding author for this work

Research output: Contribution to journalArticleScientificpeer-review

21 Citations (Scopus)


There are usually multiple paths that can be followed in a decision analysis process. It is possible that these different paths lead to different outcomes, i.e. there can exist path dependence. To demonstrate the phenomenon we show how path dependence emerges in the Even Swaps method. We also discuss the phenomenon in decision analysis in general. The Even Swaps process helps the decision maker to find the most preferred alternative out of a set of multi-attribute alternatives. In our experiment different paths are found to systematically lead to different choices in the Even Swaps process. This is explained by the accumulated effect of successive biased even swap tasks. The biases in these tasks are shown to be due to scale compatibility and loss aversion phenomena. Estimates of the magnitudes of these biases in the even swap tasks are provided. We suggest procedures to cancel out the effects of biases.

Original languageEnglish
Pages (from-to)890-898
Number of pages9
JournalEuropean Journal of Operational Research
Issue number3
Publication statusPublished - 16 Mar 2016
MoE publication typeA1 Journal article-refereed


  • Behavioral Operational Research
  • Biases
  • Decision analysis
  • Path dependence
  • Scale compatibility
  • Trade-off


Dive into the research topics of 'Path dependence and biases in the even swaps decision analysis method'. Together they form a unique fingerprint.

Cite this