SemOpenAlex |

SemOpenAlex

Matches in SemOpenAlex for { <https://semopenalex.org/work/W4232503257> ?p ?o ?g. }

Showing items 1 to 47 of 47 with 100 items per page.

W4232503257 abstract "Full text Figures and data Side by side Abstract eLife digest Introduction Results Discussion Materials and methods References Decision letter Author response Article and author information Metrics Abstract Socially-conveyed rules and instructions strongly shape expectations and emotions. Yet most neuroscientific studies of learning consider reinforcement history alone, irrespective of knowledge acquired through other means. We examined fear conditioning and reversal in humans to test whether instructed knowledge modulates the neural mechanisms of feedback-driven learning. One group was informed about contingencies and reversals. A second group learned only from reinforcement. We combined quantitative models with functional magnetic resonance imaging and found that instructions induced dissociations in the neural systems of aversive learning. Responses in striatum and orbitofrontal cortex updated with instructions and correlated with prefrontal responses to instructions. Amygdala responses were influenced by reinforcement similarly in both groups and did not update with instructions. Results extend work on instructed reward learning and reveal novel dissociations that have not been observed with punishments or rewards. Findings support theories of specialized threat-detection and may have implications for fear maintenance in anxiety. https://doi.org/10.7554/eLife.15192.001 eLife digest Around the start of the twentieth century, Pavlov discovered that dogs salivate upon hearing a bell that has previously signaled that food is available. This phenomenon, in which a neutral stimulus (the bell) becomes associated with a particular outcome (such as food), is known as classical conditioning. The network of brain regions that supports this process – which includes the striatum, the amygdala and the prefrontal cortex – seems to work in a similar way across most animal species, including humans. However, humans don’t learn only through experience or trial-and-error. We do not need to burn our hands to learn not to touch a hot stove: a verbal warning from others is usually sufficient. Experiments have shown that giving people verbal instructions on how to obtain rewards alters the activity of the striatum and prefrontal cortex. That is, the instructions interact with the circuit that also supports learning through experience. But is this the case for learning how to avoid punishments? That process depends largely on the amygdala, and it is possible that systems designed to detect threats may be less sensitive to verbal warnings. To address this question, Atlas et al. taught people to associate one image with a mild electric shock, and another with the absence of a shock. After a number of trials, the relationships were reversed so that the previously neutral picture now predicted a shock and vice versa. Telling the participants about the reversal in advance triggered changes in the activity of the striatum and part of the prefrontal cortex. By contrast, such warnings had no effect on the amygdala. Instead, the activity of the amygdala changed only after the volunteers had experienced for themselves the new relationship between the pictures and the shocks. A key next step is to find out whether this distinction between the two types of learning signals (those that can be updated by instructions and those that cannot) is specific to humans. While the current study relied upon language, there are other methods that could be used to explore this issue in animals. Furthermore, knowing that the human brain has a specialized threat detection system that is less sensitive to instructions could help us to understand and treat anxiety disorders. Atlas et al. hope to test this possibility directly in the future. https://doi.org/10.7554/eLife.15192.002 Introduction Neuroscientists have built a rich understanding of the brain systems that govern learning from punishments and rewards, which seem to be largely conserved across species (Balleine and O'Doherty, 2010; Haber and Knutson, 2010). Yet humans possess a distinctive capacity to learn from socially-conveyed rules and instructions, which is an important example of a broader capacity, shared by many species, of learning from events other than simple reinforcers (Tolman, 1949). Most of us do not need to get burnt to avoid putting our hands on a hot stove: A verbal warning serves as a sufficient threat. How does instructed knowledge influence our subsequent responses to ongoing threats in the environment? We tested whether dynamic feedback-driven aversive learning is modulated when individuals are informed about contingencies in the environment. It is unknown whether instructions integrate with the systems that support error-driven learning, as most reinforcement learning models fail to account for the potential influence of explicit information, despite the fact that verbal information shapes responses across nearly all domains, including aversive learning (Wilson, 1968; Phelps et al., 2001; Costa et al., 2015; Mertens and De Houwer, 2016). We sought to characterize computationally the contribution of instructed knowledge to dynamic aversive learning in humans. Our aim was to determine whether instructed knowledge influences the neural mechanisms of aversive learning, or whether separate neural systems process feedback-driven and instructed knowledge. Recent studies in the appetitive domain suggest that instructions about rewarding outcomes modulate learning-related responses in the striatum (Doll et al., 2009; 2011; Li et al., 2011a) and ventromedial prefrontal cortex (Li et al., 2011a) and that this modulation might depend on the prefrontal cortex (Doll et al., 2009; 2011; Li et al., 2011a). In the aversive domain, such interactions, by which instructed knowledge might help to overcome learned expectations of threat, are of particular importance due to their relevance for anxiety and post-traumatic stress disorder. However, no studies have tested whether instructions have the same effects on dynamic aversive learning, which is known to depend on the amygdala (Maren, 2001) but also involves the striatum (Seymour et al., 2004; 2005; Delgado et al., 2008b) and ventromedial/orbitofrontal cortex (VMPFC/OFC) (Phelps et al., 2004; Kalisch et al., 2006; Schiller et al., 2008). Instructions might modulate learning in the amygdala as well as the striatum and VMPFC/OFC, or amygdala responses might be insensitive to cognitive instruction (Ohman and Mineka, 2001), as suggested by theories of automatic threat detection in the amygdala (Ohman, 2005). All participants performed a Pavlovian aversive learning task in which one image, the Original conditioned stimulus (CS+), was paired with mild electric shock (the unconditioned stimulus [US]) on 30% of trials, and a second image, the Original CS-, was not paired with shock. Contingencies reversed three times. Participants assigned to an Instructed Group were informed about initial contingencies and instructed upon reversal (Figure 1), whereas participants in an Uninstructed Group learned through reinforcement alone. Models were fit to skin conductance responses (SCRs), a traditional measure of the conditioned fear response in humans. We combined quantitative modeling of behavior with functional magnetic resonance imaging (fMRI) to examine how instructions influence brain responses and skin conductance responses (SCRs) during fear conditioning. We evaluated quantitative learning models, fit to SCR, and confirmed model conclusions with task-based fMRI analyses. We focused on responses in the amygdala, striatum, and VMPFC/OFC. We hypothesized that instructions about contingencies would modify learning-related signals and brain responses during fear conditioning, and that, as in the appetitive domain, this modulation would involve the prefrontal cortex. Figure 1 Download asset Open asset Experimental design. (A) Prior to the conditioning phase of the experiment, participants in the Instructed Group saw each image and were informed about initial probabilities. Participants in the Uninstructed Group also saw the images prior to the experiment, but were not told about contingencies. (B) Participants in both groups underwent a Pavlovian fear conditioning task with serial reversals. There were three reversals across the duration of the task, leading to four continuous blocks of twenty trials. In each block, one image (the conditioned stimulus, or CS+) was paired with a shock (the unconditioned stimulus, or US) 30% of the time, leading to 4 reinforced trials and 8 unreinforced trials, whereas a second image (the CS-) was never paired with a shock. Images were presented for 4 s, followed by a 12-second inter-stimulus interval. (C) Upon each reversal, the Instructed Group was informed that contingencies had reversed. Button presses were included to ensure participants were paying attention to the instructions but had no effect on the task itself or task timing. Instructions were always immediately followed by at least two unreinforced presentations of each CS before the new CS+ was paired with a shock. The figure presents one of two pseudorandom trial orders used during the experiment (see Materials and methods). https://doi.org/10.7554/eLife.15192.003 Results Subjective ratings Upon completion of the experiment, participants reported the number of perceived contingency reversals and retrospectively rated shock expectancy and affect in response to each image. Participants recognized that three reversals had occurred (M = 3.22, SD = 0.79), and there was no group difference in the estimated number of reversals (p>0.1). Two-way ANOVAs revealed a main effect of Group on reported affect (F(1,68) = 8.57, p<0.01), such that the Uninstructed Group reported less positive affect for both stimuli. There were no Group differences in shock expectancy ratings, nor were there any main effects of Stimulus (Original CS+ vs Original CS-) or Group x Stimulus interactions on either outcome measure. Instructions influence skin conductance responses We tested whether participants showed differential SCRs to unreinforced CS presentations during fear acquisition (i.e. larger responses to the Original CS+ than Original CS-), and whether responses were modulated after participants were instructed that contingencies had reversed (Instructed Group) or after they received a shock paired with the new CS+/ previous CS- (Uninstructed Group). As reported in Table 1, both groups showed differential responses that reversed in response to contingency changes throughout the task (CS+ > CS-; ß = 0.04, t = 5.82, p<0.0001). Differential responses were larger in the Instructed Group than the Uninstructed Group (ß = 0.03, t = 3.98, p=0.0002), and SCRs habituated over time (ß = -0.07, t = 10.59, p<0.0001). Table 1 Group differences in differential SCRa. https://doi.org/10.7554/eLife.15192.004 AnalysisModelInterceptStimulus (Original CS+ > Original CS-)Reversal effect (Original contingencies vs. Reversed contingencies)Stimulus x Reversal Interaction (Current CS+ > Current CS-)TimeAll participants (n = 68)Within-subjects effects, controlling for group (first level)ß = 0.22n.s. (p = 0.06)n.s. (p = 0.445)ß = 0.04ß = -.07t = 12.37t = 5.82t = 10.59p <0.0001p<0.0001p<0.0001Effect of group (second level)n.s. (p = 0.11)n.s. (p = 0.15)n.s. (p = 1.0)ß = 0.03n.s. (p = 0.09)t = 3.98p = 0.0002Learners only (n = 40)Within-subjects effects, controlling for group (first level)ß = 0.27ß = 0.01n.s. (p = 0.95)ß = 0.06ß = -.08t = 11.62t = 2.79t = 6.5t = -11.18p<0.0001p = 0.0082p<0.0001p<0.0001Effect of group (second level)n.s. (p = 0.13)n.s. (p = 0.20)n.s.(p = 0.35)ß = 0.04ß = -.02t = 3.67t = -2.53p = 0.0007p = 0.0157Instructed Group learners (n = 20)Entire taskß = 0.30n.s. (p = 0.30)n.s. (p = 0.55)ß = 0.10ß = -0.09t = 8.93t = 5.35t = -9.26p<0.0001p<0.0001p<0.0001Entire task, second half of each runß = 0.27ß = 0.02ß = 0.04ß = 0.12n.s. (p = 0.69)t = 7.82t = 2.15t = 4.18t = 6.19p<0.0001p = 0.0314p<0.0001p<0.0001Trials following the first reversalß = 0.28n.s. (p = 0.65)ß = -0.16ß = 0.09ß = -0.07t = 6.89t = -2.36t = 4.19t = -5.05p<0.0001p = 0.0184p<0.0001p<0.0001Trials following the first reversal, second half of each runß = 0.22n.s.(p = 0.81)n.s. (p = 0.30)ß = 0.09n.s. (p = 0.09)t = 5.67t = 4.49p<0.0001p<0.0001Uninstructed Group learners (n=20)Entire taskß = 0.23ß = 0.01n.s. (p = 0.54)ß = 0.03ß = -.06t = 7.30t = 2.09t = 4.58t = -6.10p<0.0001p = 0.0368p<0.0001p<0.0001Entire task, second half of each runß = 0.22ß = 0.02ß = 0.02ß = 0.04n.s.(p = 0.678)t = 6.86t = 3.18t = 2.44t = 5.45p<0.0001p = 0.0015p = 0.0149p<0.0001Trials following the first reversalß = 0.21n.s. (p = 0.58)n.s. (p = 0.49)ß = 0.01ß = -.04t = 5.28t = 1.67t = -2.49p<0.0001p = 0.0967p = 0.0127Trials following the first reversal, second half of each runß = 0.19n.s. (p = 0.65)n.s. (p = 0.21)ß = 0.02ß = 0.03t = 5.83t = 2.04t = 2.88p<0.0001p = 0.0413p = 0.004 aThis table presents results of linear mixed models that included normalized skin conductance response (SCR) as a dependent measure. In the Instructed Group, contingencies and reversals are coded relative to instructed reversal. In the Uninstructed Group, contingencies and reversals are coded relative to reinforcement (i.e. reversals occur when the previous CS- is paired with a shock). Within groups, we analyzed SCRs across the entire task as well as following the first reversal (without the acquisition phase). We also examined responses within the second half of each run, as well as across all trials (including trials that immediately followed reversals). These results reflect group-level effects and group differences across all participants. However, a subset of participants did not show differential SCRs prior to the first reversal. As our primary research question concerns the effects of instructions on the neural systems of aversive reversal learning, the strongest tests are in those individuals who learn contingencies prior to the first reversal. Given this, we restricted our analyses to 'learners.' As described in Materials and methods, we defined learners as those individuals who showed greater SCR to the CS+ relative to the CS- in late acquisition (the second half of the first run; 20/30 Instructed Group participants, 20/38 Uninstructed Group participants). In this subset of learners, the main effects and interactions on SCR reported above remained significant and increased in magnitude (Figure 2A,C; Table 1). Importantly, both the Instructed and Uninstructed Groups showed SCRs that were responsive to changing contingencies when we examined each group separately, and when we restricted analyses to trials that followed the first reversal (Table 1), suggesting that effects were not driven entirely by initial learning. The Uninstructed Group showed significant reversals when we analyzed trials from the second half of each block in this post-acquisition analysis, and marginal effects when we included all post-acquisition trials (see Table 1). This is expected given the low reinforcement rate (33%) that we used in this study, and is consistent with somewhat slow learning. The quantitative models and fMRI analyses reported below focus on the 20 learners in each group. Combined fMRI analyses that include data from the full sample are entirely consistent with these findings and are reported in figure supplements and source data. Figure 2 Download asset Open asset Effects of instructions on skin conductance responses (SCR) and aversive learning. Mean normalized skin conductance responses (SCRs) as a function of group and condition. Both groups showed significant reversals of SCR responses throughout the task (p<0.001; Table 1), and effects were larger in the Instructed Group (see Table 1). Error bars reflect within-subjects error. (A) Mean SCR in the Instructed Group as a function of original contingencies. Runs are defined relative to the delivery of instructions. (B) Dynamics of expected value based on fits of our modified Rescorla-Wagner model, fit to SCR in the Instructed Group. Fitted model parameters were consistent with SCR reversing almost entirely in response to instructions (ρ = 0.943). This timecourse was used in fMRI analyses to isolate regions involved in instruction-based learning. (C) Mean SCR in the Uninstructed Group as a function of original contingencies. A new run is defined when the previous CS- is paired with a shock. (D) Dynamics of expected value based on the model fit to SCR from the Uninstructed Group. This timecourse was used in fMRI analyses to isolate regions involved in feedback-driven learning in both groups. https://doi.org/10.7554/eLife.15192.005 We also tested whether instructions immediately update autonomic responses by examining instructed reversals in Instructed Group learners. Each reversal featured a delay between instruction delivery and reinforcement of instructions (Figure 1C): Following instructions, each CS was presented without reinforcement at least twice before the previous CS- was paired with a US. We compared responses during these post-instruction, pre-reinforcement windows with an equivalent number of trials prior to each instruction (e.g. the last two CS+ and CS- trials in each phase). We found a significant effect of instructions on differential responding (ß = 0.04, t(19) = 3.50, p= 0.0024), such that SCR responses reversed immediately after each instruction, before any actual reinforcement was delivered. Feedback-driven learning is modulated by instructions We were interested in understanding how instructions shape error-driven learning and the development of expectations. To this end, we focused on the computation of expected value (EV) and used a simple Rescorla-Wagner model modified to capture flexible effects of instructions and test for dissociations. Our quantitative models assume that SCR at cue onset reflects EV (see Materials and methods). Shocks were incorporated as reinforcements, and thus positive EV corresponds to an expectation for a shock. Consistent with standard Rescorla-Wagner models, EV updates in response to prediction error (PE), and the speed of updating depends on learning rate (α). We focus on correlations with EV in this manuscript, due to concerns of algebraic collinearity when EV, shock, and PE are included in the same model. Models fit to the Uninstructed Group’s SCRs isolate learning-related processes that respond to reinforcement history alone (i.e. feedback-driven learning), since this group was not informed about cue contingencies or reversals. Models fit to the Instructed Group’s SCRs, however, should capture the immediate effects of instruction reported above. To acknowledge this flexible effect of instructions, we modified the standard Rescorla-Wagner model. We introduced an instructed reversal parameter, ρ, which determines the extent to which EV reverses upon instruction (see Materials and methods). If ρ = 1, the EVs of the two CSs are swapped completely when instructions are delivered, whereas if ρ = 0, each CS maintains its current EV and the model reduces to a standard experiential Rescorla-Wagner model. The best-fitting parameters when fit across Instructed Group subjects revealed that EV reversed almost completely at the time of instructions in the Instructed Group (ρ = 0.943; Table 2), suggesting that instructions immediately influence EV/SCR, as illustrated in Figure 2B. When we fit the same model to SCRs from the Uninstructed Group (for whom the additional effect should not be observed since no instructions were given) estimates were indeed consistent with associations not reversing at the time when the instructions would have been delivered (ρ = 0.0; Table 2). The resulting time course, which captures slower reversals of EV based on purely feedback-driven learning, is depicted in Figure 2D. We also fit the model to individual participants in both groups (see Materials and methods) and found that instructed reversal parameters (i.e. ρ) differed significantly as a function of Group (Instructed Group > Uninstructed Group, t(38) = 6.53, p<0.0001; Table 2). Table 2 Quantitative model of instructed learning: Rescorla-Wagner modulated with instruction parameter (ρ). https://doi.org/10.7554/eLife.15192.006 GroupAnalysis typeαρDevianceInstructed Group learners (n = 20)Across-subjects analysis0.0610.94360.83Within-subjects analysisM = 0.07, SD = 0.09M = 0.69, SD = 0.32M = 2.73, SD = 0.88Uninstructed Group learners (n = 20)Across-subjects analysis0.042040.86Within-subjects analysisM = 0.11, SD = 0.22M = 0.10, SD = 0.25M = 1.98, SD = 0.93 Neural correlates of feedback-driven and instructed aversive learning Our goal was to determine whether and how the neural systems that support feedback-driven aversive learning are modulated by instructions. Thus our computational neuroimaging analyses proceeded in two stages, guided by the quantitative models reported above. First, we examined each group separately using the model fit to behavior in that group. Thus we used the model depicted in Figure 2D to isolate neural correlates of feedback-driven learning in the Uninstructed Group, and the model depicted in Figure 2B to isolate neural correlates of instruction-based learning in the Instructed Group. Next, we directly compared the two sources of learning in the Instructed Group, since these participants were exposed to both forms of feedback (i.e. instructions about contingencies and reversals, as well as experiential learning from reinforcement between each reversal). In each analysis, we focused on results in amygdala, striatum, and VMPFC/OFC (see Figure 3—figure supplement 3) to determine whether these a priori regions of interest (ROIs) were sensitive to feedback-driven learning and/or whether they updated with instructions. Finally, we tested the conclusions from quantitative models with task-based analyses that relied strictly on our experimental design, thus eliminating the influence of assumptions derived from our models. Neural correlates of feedback-driven aversive learning We first focused on the neural correlates of experiential learning by examining responses in the Uninstructed Group. Regressors were based on the best-fitting parameters from the model fit to the Uninstructed Group, thus isolating feedback-driven EV (Figure 2D; see Materials and methods). ROI-based analyses within the Uninstructed Group revealed a main effect of Region (F(2,40) = 5.13, p=0.011). Post-hoc t-tests revealed that this was driven by positive correlations between feedback-driven EV and responses in the amygdala (bilateral: t(1,19) = 4.21, p=0.0005; Left: t(1, 19) = 3.94, p=0.001; Right: t(1,19): = 3.72, p=0.001) and striatum (bilateral: t(1,19) = 2.31, p=0.0017; left: t(1, 19) = 2.58, p=p =0.018; right: p=p =0.063). Voxel-wise FDR-corrected results confirmed ROI-based findings, and isolated additional correlations in the VMPFC/mOFC (see Figure 3A, Figure 3—figure supplement 1, and Figure 3—figure supplement 1—source data 1 and 2). The striatum and amygdala both showed positive correlations with EV, associated with increased activation for the stimulus currently predicting an aversive outcome. The VMPFC/mOFC showed negative correlations with EV, consistent with prior work showing increased VMPFC/OFC responses to conditioned stimuli predicting safe, relative to aversive, outcomes (Schiller et al., 2008). Additional regions that correlated with feedback-driven EV are reported in Figure 3—figure supplement 1 and associated source data. Figure 3 with 3 supplements see all Download asset Open asset Neural correlates of expected value. (A) Neural correlates of feedback-driven expected value (EV) were isolated by examining correlations between the timecourse depicted in Figure 2D and brain activation in response to cue onset in Uninstructed Group learners (n = 20). Top: ROI-based analyses (see Figure 3—figure supplement 3) revealed significant correlations with feedback-driven EV in the amydala and striatum. Error bars reflect standard error of the mean; ***p<0.001; *p<0.05. Bottom: Voxel-wise FDR-corrected analyses confirmed ROI-based results and revealed additional correlations in the VMPFC/OFC, as well as other regions (see Figure 3—figure supplement 1, Figure 3—figure supplement 1—source data 1, 2). (B) Neural correlates of instruction-based EV were isolated by examining correlations between the timecourse depicted in Figure 2B and brain activation in response to cue onset in Instructed Group learners (n = 20). Top: ROI-based analyses revealed a significant negative correlation with instruction-based EV in the VMPFC/OFC. Bottom: Voxel-wise analyses confirmed these results and revealed strong positive correlations in the bilateral striatum, as well as the dACC, insula, and other regions (see Figure 3—figure supplement 2 and Figure 3—figure supplement 2—source data 1 and 2). We did not observe any correlations between amygdala activation and instruction-based EV. https://doi.org/10.7554/eLife.15192.007 Instructions shape responses in VMPFC/OFC and striatum We used parameters from the best-fitting instructed learning model to isolate the neural correlates of aversive learning that updates expected value on the basis of instructions in the Instructed Group. In this model, EV updates immediately with instructions (Figure 2b), consistent with the SCRs measured in the Instructed Group. ROI-based ANOVAs revealed a significant effect of Region (F(2,40) = 6.49, p=0.0038), driven by significant negative correlations with EV in the VMPFC/OFC ROI (t(1,19) = -2.61, p=0.0173; see Figure 3B). Voxelwise FDR-corrected results also revealed robust activation in the bilateral caudate, which showed positive correlations with instructed EV (see Figure 3B, Figure 3—figure supplement 2, Figure 3—figure supplement 2—source data 1, and Figure 3—figure supplement 2—source data 2). Additional regions that tracked instructed EV in whole brain analyses and results across the entire Instructed Group are reported in Figure 3—figure supplements 2, 3 and associated source data. Feedback-driven versus instruction-based learning within the Instructed Group The preceding results, from separate groups using EV signals driven by instructions or feedback, suggest that responses in the amygdala, striatum, and VMPFC/OFC were driven by reinforcement in the Uninstructed Group, while only VMPFC/OFC and striatal responses were sensitive to instructions the Instructed Group. To test for formal dissociations, we directly compared the neural correlates of instructed and feedback-driven aversive learning in Instructed Group participants, who were exposed to both instructions and experiential learning. This within-subjects analysis ensures that potential dissociations indicated above are driven by differences in the computational sources of neural activation, rather than differences in performance between the groups. To isolate brain responses that were sensitive only to reinforcement (despite the presence of instructions about contingencies and reversals), we used the feedback-driven EV regressor generated from the model fit to the Uninstructed Group’s behavior, which takes advantage of the fact that trial sequences were identical for both groups. The instruction-driven EV regressor was generated from the model fit to the Instructed Group’s behavior and reported above. We included both EV regressors in a within-subjects voxel-wise analysis in the Instructed Group. Between reversals, experiential learning would be somewhat correlated across models. Thus, to remove shared variance, we did not orthogonalize regressors in this analysis (see Materials and Methods). A contrast across the two EV regressors formally tests whether each voxel is more related to feedback-driven or instruction-driven EV. Results from this contrast are presented in Figure 4. Voxelwise FDR-corrected analyses revealed that the bilateral amygdala was preferentially correlated with feedback-driven EV, while the right caudate and left putamen showed preferential correlations with instruction-based EV (see Figure 4B, Figure 4—figure supplement 1, and Figure 4—figure supplement 1—source data 1). ROI-based analyses confirmed voxelwise results, with a main effect of Region (F(2,40) = 3.76, p=0.0324; Figure 4A), driven by amygdala correlations with feedback-driven EV but not instructed EV (t(1,19) = 2.57, p=0.0189), although striatal differences between models were not significant when averaged across the entire ROI. Although these effects were somewhat weak when limited to learners, effects increased in magnitude when we examined the entire Instructed Group, including individuals who did not exhibit measurable SCRs (see Figure 4—figure supplement 1 and Figure 4—figure supplement 1—source data 2). When we included all Instructed Group participants, we also observed negative correlations with instruction-based EV in the VMPFC/OFC. ROI-wise analyses across the entire Instructed Group revealed that the main effect of Region (F(2, 60) = 7.12, p =0.0017) was driven by both bilateral amygdala specificity for feedback-driven EV (t(1,29) = 2.93, p =0.0066) as well as VMPFC/OFC specificity for instructed EV (t(1,29) = 2.29, p =0.0293). Figure 4 with 1 supplement see all Download asset Open asset Dissociable effects of instructed and feedback-driven learning in the Instructed Group. (A) ROI-based effects of feedback-driven and instruction based EV signaling within Instructed G" @default.
W4232503257 created "2022-05-12" @default.
W4232503257 date "2016-04-07" @default.
W4232503257 modified "2023-10-18" @default.
W4232503257 title "Decision letter: Instructed knowledge shapes feedback-driven aversive learning in striatum and orbitofrontal cortex, but not the amygdala" @default.
W4232503257 doi "https://doi.org/10.7554/elife.15192.032" @default.
W4232503257 hasPublicationYear "2016" @default.
W4232503257 type Work @default.
W4232503257 citedByCount "0" @default.
W4232503257 crossrefType "peer-review" @default.
W4232503257 hasBestOaLocation W42325032571 @default.
W4232503257 hasConcept C15744967 @default.
W4232503257 hasConcept C169760540 @default.
W4232503257 hasConcept C169900460 @default.
W4232503257 hasConcept C180747234 @default.
W4232503257 hasConcept C2776559556 @default.
W4232503257 hasConcept C2776743251 @default.
W4232503257 hasConcept C2779144063 @default.
W4232503257 hasConcept C2780062018 @default.
W4232503257 hasConcept C2781195155 @default.
W4232503257 hasConcept C513476851 @default.
W4232503257 hasConceptScore W4232503257C15744967 @default.
W4232503257 hasConceptScore W4232503257C169760540 @default.
W4232503257 hasConceptScore W4232503257C169900460 @default.
W4232503257 hasConceptScore W4232503257C180747234 @default.
W4232503257 hasConceptScore W4232503257C2776559556 @default.
W4232503257 hasConceptScore W4232503257C2776743251 @default.
W4232503257 hasConceptScore W4232503257C2779144063 @default.
W4232503257 hasConceptScore W4232503257C2780062018 @default.
W4232503257 hasConceptScore W4232503257C2781195155 @default.
W4232503257 hasConceptScore W4232503257C513476851 @default.
W4232503257 hasLocation W42325032571 @default.
W4232503257 hasOpenAccess W4232503257 @default.
W4232503257 hasPrimaryLocation W42325032571 @default.
W4232503257 hasRelatedWork W1972742627 @default.
W4232503257 hasRelatedWork W1975068431 @default.
W4232503257 hasRelatedWork W1999463885 @default.
W4232503257 hasRelatedWork W2025959734 @default.
W4232503257 hasRelatedWork W2092533470 @default.
W4232503257 hasRelatedWork W2795316889 @default.
W4232503257 hasRelatedWork W2984170056 @default.
W4232503257 hasRelatedWork W3132748226 @default.
W4232503257 hasRelatedWork W4200262660 @default.
W4232503257 hasRelatedWork W4210301867 @default.
W4232503257 isParatext "false" @default.
W4232503257 isRetracted "false" @default.
W4232503257 workType "peer-review" @default.