AN INVESTIGATION OF MOTIVATIONAL BIAS IN SUBJECTIVE PREDICTIVE PROBABILITY DISTRIBUTIONS*

Abstract

ABSTRACT In this paper the problem of conscious motivational bias in the assessment of subjective predictive probability distributions is explored. The primary purpose of this research is to discover how subjects express their bias. Results of an experiment are reported in which subjective distributions were assessed by subjects, first in the absence of a biasing incentive and then in the presence of an incentive to bias upward coupled with an incentive to be “credible.” In the biased task, subjects reported manipulating up to five characteristics of their distribution, such as location, shape, and dispersion. Furthermore, subjects reported using combinations of the characteristics strategically. Dominant strategies included shifting the mode upward and redistributing probability to the left, shifting the mode upward a small amount and redistributing probability to the right, and shifting the mode upward and tightening the distribution. The location of the mode was mentioned first by nearly all subjects, suggesting that they first locate their biased distribution and then manipulate other characteristics to express their bias, need for credibility, and other concerns.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 13 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.