Decomposing bias in different types of simple decisions.

Abstract

The ability to adjust bias, or preference for an option, allows for great behavioral flexibility. Decision bias is also important for understanding cognition as it can provide useful information about underlying cognitive processes. Previous work suggests that bias can be adjusted in 2 primary ways: by adjusting how the stimulus under consideration is processed, or by adjusting how the response is prepared. The present study explored the experimental, behavioral, and theoretical distinctions between these biases. Different bias manipulations were employed in parallel across perceptual and memory-based decisions to assess the generality of the 2 biases. This is the 1st study to directly test whether conceptually similar bias instructions can induce dissociable bias effects across different decision tasks. The results show that stimulus and response biases can be separately induced in both tasks, suggesting that the biases generalize across different types of decisions. When analyzing behavioral data, the 2 biases can be differentiated by focusing on the time course of bias effects and/or by fitting choice reaction time models to the data. These findings have strong theoretical implications about how observed bias relates to underlying cognitive processes and how it should be used when testing cognitive theories. Guidelines are presented to help researchers identify how to induce the biases experimentally, how to dissociate them in the behavioral data, and how to quantify them using drift diffusion models. Because decision bias is pervasive across many domains of cognitive science, these guidelines can be useful for future work exploring decision bias and choice preferences.

Biased sequential sampling underlies the effects of time pressure and delay in social decision making

“…A choice is made once R  reaches one of the two thresholds, normalizing the pro-social threshold to zero and the selfish threshold to a constant, a . An additional feature of the DDM is that there can be an initial bias in the starting point ( R 0 ), often referred to as a response bias 46 , towards selfishness or pro-sociality.…”

Ignoring memory hints: The stubborn influence of environmental cues on recognition memory.

“…An analogous way to think about this is that the average drift rate, collapsed across both item classes, shifts upwards for a liberal bias whereas the average drift rate for both item classes drifts downwards for a conservative bias. For example, White and Poldrack (2014) manipulated both the relative proportion of items classes in mini blocks of trials (e.g., most items in the mini block would be old or studied) and also manipulated the strength or veracity that subjects should demand for one versus the other conclusion (e.g., only respond ‘old’ only if a strong memory signal is perceived). They referred to these as response and stimulus biases respectively, and accordingly the former had a stronger influence on the starting point estimate and the latter on the measured drift criterion.…”

The reliability of criterion shifting in recognition memory is task dependent

“…Drift rates for lures differed when they were in strong-target versus weak-target test blocks, suggesting that people changed their thresholds based on expected strength. Similarly, White and Poldrack (2014) showed the same influence on drift criterion (and drift rates) when asking people before each trial to be either conservative (i.e., expect a strong memory signal before responding) or liberal (i.e., expect a weak memory signal before responding). The dissociation between base rates influencing the starting point, but evidence or strength expectancy influencing the drift criterion, could be one reason effects of these different manipulations might not correlate at the coarse level we are examining.…”

“…A starting point parameter reflects how conservative or liberal one's criterion might be at the start of a trial. Base-rate manipulations have been shown to impact this particular parameter by influencing an initial expectation about what type of response is most likely to be correct, described by White and Poldrack (2014) as response expectancy. In contrast, the drift-criterion parameter in the diffusion model influences the amount of evidence needed to make a particular choice decision, which depends on the accumulation of evidence over repeated sampling after the test item has been perceived.…”

“…We are happy to share these data with anyone who is interested. decision process that unfolds over time (e.g., Criss, 2009Criss, , 2010White & Poldrack, 2014). Drift diffusion modeling, for example, posits two important parameters that might influence a basic SDT measure of criterion setting (Criss, 2010;Starns, Ratcliff, & White, 2012).…”