Jurasovic_whole_thesis.pdf (3.37 MB)
You're a wizard, Ari : bias due to rewards in binary decision-making
thesisposted on 2023-05-27, 09:14 authored by Jurasovic, EM
We investigated the effect of the factors Reward Type (points vs. money) and Reward Schedule (constant vs. variable) on bias in a computerised binary decision-making task. Participants (19 female; N = 32) were aged 18 to 41 years were asked to identify which colour was in majority in a two-colour visual discrimination task embedded in a computer game. We measured reaction times, accuracy and proportion of bias congruent responses as manifestations of bias in participants' decision-making. An exact binomial test indicated that participants chose the option providing on average larger rewards at a rate greater than chance (p < .001). However linear mixed modelling and model selection provided minimal evidence to indicate that this bias was due to either Reward Type or Reward Schedule. Bias was greater when participants were told a priori about the likelihood of trial occurrences compared to being rewarded for correct trials (p < .001). Future research may test the effect of different reward sizes and true vs. random scoring on bias in the current context. In keeping with the scientific principles of honesty and collaboration, all analysis scripts, raw data, preliminary study plans and final manuscript are available online at the Open Science Framework website.
Rights statementCopyright 2019 the author