Feasibility of Unconscious Instrumental Conditioning: A Registered Replication

Abstract

The extent to which high-level, complex functions can proceed unconsciously has been a topic of considerable debate. While unconscious processing has been demonstrated for a range of low-level processes, from feature integration to simple forms of conditioning and learning, theoretical contributions suggest that increasing complexity requires conscious access. Here, we focus our attention on instrumental conditioning, which has been previously shown to proceed without stimulus awareness. Yet, instrumental conditioning also involves integrating information over a large temporal scale and distinct modalities in order to deploy selective action, constituting a process of substantial complexity. With this in mind, we revisit the question of feasibility of instrumental conditioning in the unconscious domain. Firstly, we address the theoretical and practical considerations relevant to unconscious learning in general. Secondly, we aim to replicate the first study to show instrumental conditioning in the absence of stimulus awareness (Pessiglione et al., 2008), following the original design and supplementing the original crucial analyses with a Bayesian approach (Experiment 1). Should the replication be successful, we will attempt to replicate the effect yet again with improved methods to address the methodological issues related to sensitivity and immediacy (Experiment 2).

Unconscious Knowledge of Rewards Guides Instrumental Behaviors via Conscious Judgments

“…Also, it was robust to an inflation of unconscious accuracy due to regression to the mean effects, because, as we show in the Results section, the majority of trials were attributed to unconscious knowledge (cf. Skora et al, 2020) and conscious knowledge was not more reliable than unconscious knowledge (cf. Shanks, 2017).…”

Conscious and unconscious mental states.

“…Meyen et al assume equal signal to noise ratios in the discrimination and priming tasks, which although simplistic, is a good way of putting the ball in the park of those arguing for the use of an objective threshold. Dienes recommended that the relation between priming and expected discrimination should be empirically calibrated; see Skora et al (2020) for an application. Finally, one potentially useful method for showing performance at objective threshold is to regress priming against discrimination and examine the slope and intercept; a zero slope or especially a non-zero intercept can be taken to indicate subliminal priming (Greenwald et al, 1995).…”

“…That is, regression to the mean can apply to post hoc selection of subsets of trials or participants, leading to artefactual above chance performance on selected trials or subjects (Shanks, 2017). The problem can be addressed by Bayesian methods in the cases of post hoc subject selection (Leganes-Fonteneau et al, 2021) and both trial and subject selection (Skora et al, 2020).…”

Testing theories with Bayes factors

“…If we are going to reject all point H0s because they are not exactly true, we must on the same grounds reject all models whatsoever. When big data are involved, the standard error may be very small; then it will be important to be clear about how H0 should be modeled, maybe for example with a uniform; and conversely, the plausibility of H1 may be zero below a minimally interesting value (see Palfi & Dienes, 2019, for Bayes factors with interval H0s and an example case; see Skora et al, 2020, for another type of example using interval H0's).…”