The signal for good news and value of information in human choice behavior

  • Author / Creator
    Sawalha, Jeffrey
  • Surprisingly, pigeons have been shown to sacrifice food rewards to get information about the upcoming outcome when making choices in a delay decision task. For instance, when choosing between an option which offers a reward 100% of the time after a delay (e.g. 10 s), versus an option that gives rewards only 50% of the time after the same delay but gives information about the reward (pecking it leads to a color that signals reward / no reward), pigeons sometimes prefer the 50% option. However, when the information is not immediate (e.g. it takes a few seconds after the peck for the color change to occur) pigeons almost exclusively choose the 100% option. Here, we tested if this same preference for immediate information would be evident in human choice behavior. In four experiments we adapted the pigeon task to a human reaching choice task to test two questions: 1) Do humans prefer information over no information, especially when it is received immediately? 2) Do humans show an asymmetry for information biased by valence, preferring good news and avoiding bad news?
    In Experiments 1 (n = 30) and 2 (n = 30) participants reached to choose between shapes on a touchscreen that either revealed the outcome (turned green for good or red for bad) or not (stayed the same color) and did so either immediately or after a delay (2 s for E1, 4 s for E2). Unlike pigeons, the results show that humans do not prefer information (immediate or delayed), instead choosing almost exclusively based on the probability of payoff. In Experiments 3 (n = 20) and 4 (n = 25), participants again reached to choose between shapes, a “good news” shape that sometimes changed color when the outcome was good, a “bad news” shape that sometimes changed color when the outcome was bad, and a neutral shape which never changed color. Here we see a significant preference to choose “good news” shapes and avoid “bad news” shapes. These results support the Signal for Good News (SiGN) hypothesis which argues that receiving a signal of good news in a 50/50 gamble task is more rewarding because it causes a large positive reward prediction error. However, our results also show that, unlike pigeons, humans avoid bad news, suggesting loss aversion may spill into the information domain.

  • Subjects / Keywords
  • Graduation date
    Fall 2018
  • Type of Item
  • Degree
    Master of Science
  • DOI
  • License
    Permission is hereby granted to the University of Alberta Libraries to reproduce single copies of this thesis and to lend or sell such copies for private, scholarly or scientific research purposes only. Where the thesis is converted to, or otherwise made available in digital form, the University of Alberta will advise potential users of the thesis of these terms. The author reserves all other publication and other rights in association with the copyright in the thesis and, except as herein before provided, neither the thesis nor any substantial portion thereof may be printed or otherwise reproduced in any material form whatsoever without the author's prior written permission.