The effects of placebo brain stimulation on the sense
of agency and neural response to errors
ABSTRACT
Beliefs and expectations regarding cognitive enhancement can exert a powerful influence on behavior and performance. However, little is known about specific effects of belief in cognitive enhancement through brain stimulation on the sense of agency (i.e., the extent to which actions are attributed to oneself) and on neural markers of distress in response to errors. In the present EEG study we used a fake brain stimulation device in a placebo (i.e., cognitive enhancement), nocebo (i.e., cognitive impairment) and a control condition (i.e., no effect on cognitive performance), while participants performed a Flanker task. First, the sense of agency over errors was substantially lower in the experimental conditions than in the control condition, with impairment suggestions yielding more external attributions than enhancement suggestions. Second, the amplitude of the error-related negativity (ERN) was significantly higher in the placebo compared to the control and the nocebo condition, and in the nocebo condition a correlation between sense of agency and the ERN amplitude was found. These findings demonstrate that violated expectations about cognitive enhancement cause personal distress and our study highlights the potential of placebo brain stimulation as a powerful technique in cognitive experimental research.
Keywords: cognitive performance; expectancies; brain stimulation; sense of agency; error-related negativity
INTRODUCTION
Cognitive enhancement enjoys a strong societal and scientific interest. An online survey by Nature among 1,400 people from 60 countries revealed that 20% of the responders had used cognitive enhancers for non-medical purposes, mostly to boost concentration, to increase focus or to enhance memory (Maher, 2008). Next to traditional stimulant drugs such as Ritalin, promising new techniques and devices, such as neurofeedback and brain stimulation devices (e.g., transcranial magnetic stimulation; TMS, transcranial direct current stimulation; tDCS) gain increasing attention (Sahakian & Morein-Zamir, 2011) and people seem to have an unconditional faith in these techniques to boost brain capacity (Rusconi & Mitchener-Nissen, 2014). This is probably partly due to the persistent urban myth that humans only use 10% of their brain capacity (Lilienfeld, Lynn, Ruscio, & Beyerstein, 2011), which leaves 90% to be additionally exploited by ingenious techniques such as tDCS.
The aim of the present study is to investigate whether belief in the potential of cognitive enhancement through neurostimulation devices affects the sense of agency and neural responses to errors. We will address the influence of
expectancies by using a suggestibility manipulation and a placebo vs. nocebo brain stimulation device. Importantly,
the use of neurostimulation devices raises ethical issues with regard to cognitive enhancement (Bostrom & Sandberg, 2009; Heersmink, 2015), specifically related to transferred feelings of agency (i.e., the sense the I am in control of and responsible for my own actions) (Bostrom & Sandberg, 2009, p. 326). We therefore investigate whether the sense of agency over one’s own actions is affected by the use of neurostimulation, as achievements and failures cannot solely be attributed to oneself, but also to the external substance or device. Moreover, prior expectations may also directly modulate the affective responses in relation to one's cognitive performance, as for suggestions of impairment, failures can be attributed to an external source, possibly attenuating the associated distress, whereas for suggested enhancement, one may expect improved performance instead, and hence experience more distress from failures. We thus assessed whether belief in impaired vs. improved cognitive performance also affects neural responses to errors.
The theoretical background of our study is provided by the predictive processing framework, according to which internal predictive models are continuously monitored and updated based on sensory input from the external world, following Bayesian inference principles (Kilner, Friston, & Frith, 2007; Srinivasan, Laughlin, & Dubs, 1982). Based on prior beliefs, predictions are generated to 'explain away' sensory input and prior beliefs are updated based on prediction error signals arising from the comparison between one's prediction and the actual sensory input. Within this framework, perceptual or cognitive processes can be framed in terms of the relation between the strength of one's prior beliefs, predictive signals, the sensory input and the prediction error signal. The placebo effect, which can be defined as the change in internal states, behavior or performance as a result of the manipulation of beliefs and expectations through suggestion (Michael, Garry, & Kirsch, 2012), has indeed been framed in terms of the predictive processing framework, with strong expectations of symptom improvement modulating incoming sensory information (i.e., reduction of experienced pain; Buchel, Geuter, Sprenger, & Eippert, 2014). Interestingly, although research on placebo analgesia has been found to exert significant effects on a behavioral level (e.g., longer endurance of pain),
placebo manipulations to enhance cognitive functions have generally not shown effects on objective performance (Hajcak, Moser, Yeung, & Simons, 2005; Iani, Ricci, Gherri, & Rubichi, 2006; Schwarz & Buchel, 2015).
However, the absence of a behavioral effect does not connote the absence of any placebo effect with regard to cognitive manipulations, as the effects mainly appertain to the subjective domain (Schwarz & Buchel, 2015). For instance, the subjective feeling of control over one’s performance or actions - a concept that has been referred to as the ‘sense of agency’ (Chambon, Filevich, & Haggard, 2014) is likely to be affected by the suggestion of an external agent (i.e., placebo). Sense of agency over an action arises when an individual attributes that specific action to its own act (i.e., internal attribution) rather than to some other agent (i.e., external attribution). Conceptualized in the predictive processing framework, internal efferent signals are used to predict the sensory consequences of one's actions (Haggard & Chambon, 2012; Moore, Wegner, & Haggard, 2009; Voss et al., 2010). When sensory feedback from one's actions confirms the predictions, this increases the sense of agency over these actions whereas a mismatch suggests the sensory information is caused by an external agent (Haggard & Chambon, 2012). Building on this model, several studies have emphasized the contribution of external cues to experienced self-agency. For instance, priming of action-relevant concepts (Wegner & Wheatley, 1999) or performance expectancies (Custers, Aarts, Oikawa, & Elliot, 2009) has been found to enhance the sense of self-agency. Other studies have demonstrated that contextual cues can also attenuate the sense of agency when these external cues provide an alternative attributional source for observed actions (e.g., Moore et al., 2009).
Specifically, expectancy manipulations have been shown to affect the sense of agency, an extreme illustration of which is provided by instances of hypnosis, as the induction of hypnosis is typically characterized by a diminished sense of volition and control (i.e., a reduced sense of agency) over actions performed under hypnosis (Polito, Barnier, & Woody, 2013; Woody & McConkey, 2003). Similarly, placebos can also cause a misattribution of agency, as central to placebo manipulations is also the suggestion that an external object (e.g., a placebo pill or a brain stimulation device) affects one's behavior. For instance, a placebo cognition-enhancing drug (Clifasefi, Garry, Harper, Sharman, & Sutherland, 2007), as well as placebo thought-transferring device (Swiney & Sousa, 2013) induced the misattribution of the subjectively experience performance improvement and generated thoughts, respectively, to the external agent. Notably, misattributions occurred more often for negative, compared to positive or neutral thoughts (Swiney & Sousa, 2013), consistent with literature on the well-documented self-serving bias (Mezulis, Abramson, Hyde, & Hankin, 2004). With respect to placebos, Gibbons and Gaeddert (1984) already proposed that the extent to which people make external causal attributions is generally larger for inhibiting than performance-enhancing suggestions, reflecting the self-serving bias. Indeed, in a recent study it was found that in a vigilance task, a (subjective) negative effect on cognitive performance tended to be attributed to either a placebo pill or sham magnetic stimulation (Szemerszky, Domotor, Berkes, & Koteles, 2016). Thus, a placebo can provide an alternative causal attribution for an outcome that is perceived to be threatening with respect to one’s self-esteem (Gibbons & Gaeddert, 1984).
Following this line of reasoning, it can be expected that a placebo / nocebo manipulation not only influences the sense of agency over one’s performance but also the affective responses to one's performance, especially with respect to negative outcomes (i.e., errors). The present study therefore also focuses on the effects of expectancies about cognitive enhancement (vs. impairment) on the error-related negativity (ERN), a measure reflecting the personal distress caused by performance errors. The ERN occurs approximately 100ms after making an error and has been localized to the anterior cingulate cortex (ACC; Gehring, Goss, Coles, Meyer, & Donchin, 1993). Although the ERN was originally thought to primarily denote a cognitive response to errors, it has recently been proposed that the ERN also reflects affective or motivational aspects of task performance. For instance, increasing the significance of errors by either including monetary rewards, performance evaluation (Hajcak et al., 2005) or increasing personal relevance (Gentsch, Ullsperger, & Ullsperger, 2009), enhances the ERN amplitude. Additionally, Inzlicht and Al-Khindi (2012) used a placebo misattribution paradigm that allowed participants to attribute negative affect on incorrect trials to a placebo anxiety-inducing beverage, which significantly reduced the ERN. Moreover, individual differences in sensitivity to negative affect have also been related to the magnitude of the ERN, for instance in patients with anxiety disorders, who typically exhibit enhanced ERN amplitudes (Gehring, Himle, & Nisenson, 2000) and psychopaths who show reduced ERN responses (Von Borries et al., 2010). Finally, several studies have shown that the ERN is sensitive to perceived accuracy rather than actual behavioral accuracy (Scheffers & Coles, 2000) and that prior beliefs (expectancies) about intentional control can influence neural responses to errors (Rigoni, Pourtois, & Brass, 2015; Rigoni, Wilquin, Brass, & Burle, 2013). Together these studies show that the ERN can be used as a proxy for personal distress in response to errors. Accordingly, in the present study we investigated whether a placebo vs. nocebo manipulation results in opposite effects on the ERN amplitude in a cognitive control task.
In sum, previous studies have shown that prior beliefs and expectations can influence perceptual and cognitive processes, as exemplified by the placebo effect and conceptualized in a predictive processing framework (Buchel et al., 2014; Corlett, Taylor, Wang, Fletcher, & Krystal, 2010; Schjoedt et al., 2013). More specifically, prior beliefs and expectancies have been shown to influence the neural response to errors (i.e. the ERN; Gentsch et al., 2009; Hajcak et al., 2005; Rigoni et al., 2015; Rigoni et al., 2013), and moderate the sense of agency over intentional actions (Clifasefi et al., 2007; Custers et al., 2009; Moore et al., 2009; Swiney & Sousa, 2013; Szemerszky et al., 2016). In the present study, based on the alleged ‘power’ of cognitive enhancement and neurostimulation devices, and following the predictive processing framework, we tested two main hypotheses regarding the effects of tDCS brain stimulation while participants performed a Flanker task (Eriksen & Eriksen, 1974), used as a measure of cognitive control. Following the observation that the valence of an outcome has a strong effect in placebo-effects and self-attribution (Gibbons & Gaeddert, 1984), we included a nocebo (i.e., impairment) condition in addition to the placebo condition and investigated whether this suggestion resulted in effects in the opposite direction. First, we hypothesized that suggestions of cognitive enhancement would lead participants to feel in control, which should be reflected in a high sense of agency. In the cognitive impairment condition, on the other hand, participants could externally attribute errors to the brain stimulation, which should be reflected in a decreased sense of agency. Second, we expected that the two suggestions (i.e., placebo vs. nocebo manipulation) would result in opposite effects on the ERN in response to errors compared to a control condition in which no suggestions were made. That is, suggestions about enhancement are expected to increase the amplitude of the ERN because errors are more surprising and distressing since one was supposed to commit fewer errors, while suggestions about impairment are expected to decrease the amplitude of the ERN because errors are less surprising and less self-threatening (since they can be externally attributed), relative to the control condition. Placebo suggestions about cognitive enhancement / impairment are thus expected to affect the personal significance of performance, which should be reflected in the sense of agency and the ERN amplitude. Thereby this study should yield exciting new insight into the effects of belief in cognitive enhancement through neurostimulation devices on performance experience.
METHODS Participants
Initially, thirty-one healthy participants participated in the experiment (mean age = 32.3 years, range = 18-64 years, 24 females) for which they received a financial remuneration. Participants were recruited through a local newspaper advertisement as well as the online participant pool of the University of Amsterdam. Exclusion criteria were a history of brain related abnormalities and past knowledge of or experience with tDCS. Due to equipment malfunction and/or excessive signal loss, 8 participants were excluded from analyses, leaving a final sample of twenty-three participants (mean age = 28.9 years, range = 18-61 years, 18 females). The study was approved by the local ethics committee at the Psychology Department of the University of Amsterdam (Project # 2016-SP-6649) and all participants were treated in accordance with the Declaration of Helsinki.
Design
The current study used a within-subjects design with the following conditions: placebo/frontal tDCS, control/no tDCS and nocebo/parietal tDCS. As dependent variables, we focused on sense of agency over committed errors and amplitude of the ERN signal. The order of the suggested stimulation conditions was counterbalanced across participants.
Expectancy manipulation
Participants were told that the study aimed to investigate the effect of a completely safe brain stimulation device
(tDCS) that has the power to activate the brain’s unused potential. The alleged effects of stimulation were strongly
emphasized and the differential effects of the stimulation conditions were explained multiple times. First during a telephone screening and then upon arrival at the lab these effects were verbally disclosed. The researcher verbally repeated the expected effects when switching on the tDCS device. In addition, each condition block of the Flanker task included written information on the induced effects (placebo condition: “In this session frontal tDCS is used,
which can improve performance. The stimulation releases extra neural activity in your frontal cortex, which makes the neural processing more efficient if you are sufficiently sensitive to the stimulation. The frontal stimulation can make you feel more energized and active.” Nocebo condition: “In this session parietal tDCS is used, which can impair performance. The stimulation reduces the neural activity in the frontal cortex, which makes the neural processing less efficient if you are sufficiently sensitive to the stimulation. The parietal stimulation can make you feel more tired and dazed.” Control condition: “In this session, no stimulation is used.”). The credibility of the
manipulation was enhanced in multiple ways. Firstly, a telephone screening with exclusion criteria based on a standard tDCS screening form was applied (e.g., history of epilepsy, severe concussion, psychotropic drugs, pregnancy). Secondly, an actual tDCS device was used and sham stimulation (consisting of a 20 seconds ramping up of the current as is common practice in real tDCS studies) was administered at the beginning of the experiment, so that participants would actually experience a slightly tingling sensation on their head. Lastly, at the end of the experiment, participants completed a questionnaire on possibly experienced side-effects of tDCS.
Flanker task
A Flanker task (Eriksen & Eriksen, 1974) was presented on a 60cm computer screen (1920 x 1080 pixels) placed at approximately 50cm from the participant’s eyes and was programmed using Presentation software (Neurobehavioral Systems, Inc.). The task consisted of 8 blocks of 20 trials, with 50% congruent (<<<<<) and 50% incongruent (<<><<) trials, which were repeated over the different expectancy conditions. Stimuli were presented in black (font size 36) on a white background. In order to increase difficulty, the contrast of the central target arrow was lowered (RGB color 235, 235, 235) compared the surrounding distractors. The participant’s task was to indicate in which direction the middle arrow was pointing. Equal emphasis was placed on accuracy and speed in the task instructions. To account for large individual differences in accuracy on the Flanker task, the test phase of the task was preceded by a practice phase, in which an individually adjusted level of difficulty was set. This encompassed modification of the response interval in which participants had to press a key (right or left). The interval was initially fixed at 1000ms and was shortened with 100ms after 10 trials if accuracy was higher than 80% or was extended with 100ms if performance was lower than chance level (50% correct). Thus, the response interval was kept the same if accuracy was within the 60-80% range. If participants failed to respond within the set interval (miss), they received feedback to respond faster. Both errors and misses counted towards the reduction of accuracy. Adjustment continued until performance in 2 consecutive blocks fell within the 60-80% range1. During the actual test phase, accuracy was evaluated after each
block and adjusted in steps of 50ms, following the same criteria as during the practice phase. This procedure ensured that each participant would make a sufficient amount of errors (M = 13.9%, SD = 4.0%). In the test phase, each error was followed by a feedback screen, in which the participant was required to indicate to what extent the response was influenced by the brain stimulation on a 7-point scale ranging from “not at all” to “completely”2, assessing the sense
of agency over the error. Each participant completed the Flanker task in each stimulation condition (i.e., three times) with only the information on the effects of the alleged stimulation differing between conditions.
Experimental setup
For this experiment a combined EEG cap with tDCS electrodes was used. EEG was recorded over the entire scalp at 2048Hz using BioSemi’s Active-Two System (BioSemi, Amsterdam, Netherlands). Horizontal and vertical EOG was measured with electrodes at the outer canthi and above and below the participant’s dominant eye. The frontal tDCS electrode (consisting of a wet 3x3cm sponge electrode) was positioned on top of the EEG cap at the location corresponding to electrode AFz (which was omitted from the EEG; resulting in 63 EEG electrodes) and the parietal tDCS electrode (3x3cm) was positioned in the neck below the EEG cap. Before the start of the experiment, sham stimulation was administered for the participants to show them the effects and sensation of tDCS. The sham stimulation included the ramping up for 20 seconds with a 1mA current stimulation with the anodal electrode placed over the prefrontal cortex and the cathodal electrode in the neck. This sham stimulation was repeated at the beginning of the two stimulation blocks3, after which the device was switched off for the remainder of the block. To ensure that
the tDCS apparatus would not leak any current or distort the EEG signal, the cable of the tDCS device was disconnected during the Flanker task and EEG recording – out of sight of the participant.
Procedure
Participants who responded to the advertisement were first contacted by telephone, through which they received detailed information about the experiment and were screened for participation. Upon arrival at the lab, participants were again informed on the aim and procedures of the study and signed the informed consent form. The participants were seated in an experimental room, in which they carried out all tasks. The EEG cap and electrodes and tDCS electrodes were applied to the participants’ head. At the beginning of the study participants could monitor the EEG signals online on the computer screen and see the effects of eye blinks and facial movement, as well as experience the sensation of the tDCS stimulation (this also contributed to the credibility of the experimental manipulation). Participants then completed a practice block of the Flanker task to determine their individual response interval threshold (see above). After 8 blocks of the Flanker task in the placebo or nocebo stimulation condition, the participant completed three questions on their experience during the tDCS, which served as manipulation check items. Participants then had a 10-minute break to 'ensure the stimulation had completely worn off'. This was then repeated for the other two conditions. After these experimental blocks, the tDCS electrodes and the EEG cap were removed and the participant completed the exit questionnaires to assess demographics, their experience during the experiment (as an additional manipulation check), possible tDCS side-effects and their general level of absorption (Tellegen & Atkinson, 1974) and locus of control (Rotter, 1966)4, as both these measures have been related to individual
differences in susceptibility to suggestibility in the context of paranormal beliefs or mystical experiences (Andersen, Schjoedt, Nielbo, & Sørensen, 2014; Groth-Marnat & Pegden, 1998; Paddock et al., 1998).
2 This 1-7 scale was recoded in order for high and low values to correspond to high and low sense of agency, respectively. 3 For one participant, the impedance continuously remained too high for sham stimulation to be administered. Therefore, this participant did not receive actual sham stimulation in the experimental blocks. Nevertheless, the expectancy manipulation was still successful, as indicated by the manipulation check items, which were above average. We therefore decided to keep this participant in the analysis of the study.
4 We included an adapted version of the Absorption Scale (Tellegen & Atkinson, 1974) with 34 items (Cronbach’s = .93) and an adapted version of the Locus of Control Scale (Rotter, 1966) with 11 items consisting of a pair of statements reflecting an internal vs. external locus of control. Analysis indicated that the reliability of the latter scale was relatively low (Cronbach’s = .52). Deletion of the three items with the lowest loading on the first factor in a factor analysis (PCA) resulted in a moderately reliable Locus of Control scale ( = .65) with 8 items.
Figure 1. (A) Experimental set-up. Participants were placed at approximately 40cm from the screen and equipped with the EEG cap with electrodes, connected to the EEG amplifier, as well as the frontal and parietal tDCS electrodes, placed on top of and below the EEG cap. (B) Study flow. Participants received verbal and written instructions about the task and expectancy manipulation (placebo induction) and signed an informed consent. EEG and tDCS electrodes were set up, followed by a practice session of the Flanker task (without stimulation). In the actual test phase, participants completed the Flanker task three times (in randomized counterbalanced order), each time first being informed on the specific stimulation condition (i.e. frontal: improvement; parietal: impairment; no stimulation: no effect). ‘Active’ stimulation sessions were followed by 3 manipulation check question on experience of the tDCS. Participants t hen completed the post-experimental questionnaires and were debriefed about the placebo manipulation. (C) Flanker task procedure. The Flanker task started with instructions and information on the specific stimulation condition, followed by a 2-minute interval in which the stimulation was ‘ramped up’. A trial consisted of a fixation screen (450ms), a black screen (inter-stimulus interval; 200ms), the target presentation and participants had to respond within the individually set response interval by pressing the right or left control key with their right or left index finger, respectively. This was followed by an inter-trial interval (600ms). If the given response was incorrect, a feedback screen appeared and participants had to rate to what extent they felt the error was caused by the brain stimulation (on a 7-point scale). After this, or when they had given the correct response, the fixation screen appeared again and a new trial started.
A final manipulation check was included in which the participants completed two questions on the influence (“ To
what extent did you experience the influence of tDCS on the neuronal energy in your brain?”) and on the efficacy
(“To what extent do you consider tDCS an effective method to enhance or impair brain functioning?”) of the stimulation. The correlation between the manipulation check items was r(21) = .44, p < .05 and the items were combined as a measure for ‘believe in the efficacy of the brain stimulation’. Additionally, participants had to state the purpose of the study in their own words. Finally, the participants received the financial remuneration and were debriefed on the true purpose of the study. This was done carefully as participants might feel uncomfortable regarding the element of deception. They were explained that “your brain was not actually stimulated and that all experiences
you had were self-generated. Placebo has been proven to be a powerful effect for instance in the medical practice and we are now looking at the effects of placebo brain stimulation on cognitive performance. We also looked at the influence on brain processes to assess the underlying mechanisms of the placebo effect. The information you received at the beginning of the experiment is factually accurate. tDCS is indeed a method to activate and deactivate the brain and influence performance. We did however not do this in the current experiment”.
EEG analysis
The EEG data were down-sampled offline to 256Hz, rereferenced to the average signal across all electrodes and filtered between 0.16Hz and 30Hz. Response-trials (miss-trials were excluded from analysis) were segmented into epochs from 100ms before to 300ms after the button-press, baseline-corrected to the signal from -100ms to 0ms and corrected for ocular movements using the semi-manual IRA algorithm in the Brain Vision Analyzer software (Gratton, Coles, & Donchin, 1983). Segments with artifacts in which the signal exceeded the -100µv and +100µv thresholds were manually removed. For the final sample of 23 participants, this processing resulted in an average of 1.3% (SD = 1.2%) of the response-trials being rejected from analysis due to artifacts. Individual event-related potentials (ERPs) were calculated separately for correct and incorrect trials and for each stimulation condition (placebo, nocebo, control), resulting in 6 ERPs per participant. Grand-averages were calculated and peak information for the ERN was extracted for each participant for electrodes Cz, FCz, FC1 and C1 within a predefined time-window ranging from 0 to 150ms. As visual inspection revealed the strongest ERN at location FCz, which is in line with literature (Falkenstein, Hoormann, Christ, & Hohnsbein, 2000; Gehring et al., 1993), the ERN peak amplitude at this electrode was used in further analyses.
Behavioral data analysis
Behavioral data were analyzed with SPSS 20 software (IBM). Before looking into effects of the experimental manipulation on the dependent variables (i.e., sense of agency and ERN), we first checked whether the experimental expectancy manipulation was successful. This was done by interpreting the average score on the combined ‘belief in the efficacy of the brain stimulation’ item, as well as comparing the mean scores on the items assessing the tDCS influence after the experimental conditions between the two stimulation blocks. With regard to the sense of agency, a repeated measures ANOVA with condition (placebo vs. nocebo vs. control) as within-subjects factor was performed, followed by paired t-tests. For the ERN data, a repeated measures ANOVA with condition and correctness as within-subjects factors was performed for ERN peak amplitudes that were calculated in Brain Vision Analyzer, followed by paired t-tests to directly compare conditions. Additionally, correlations between the ERN data and the sense of agency were calculated for each of the two stimulation conditions, to examine whether behavioral responses (i.e., attribution of the error to the brain stimulation device) and neural responses to errors were related. Finally, the amount of errors on the Flanker task (accuracy) was calculated for each condition in order to assess performance effects. Two repeated measures ANOVAs with congruency (congruent vs. incongruent) and condition (placebo vs. nocebo vs. control) as within-subjects factors were executed with the percentage of errors and reaction times (RTs) as dependent measures.
Manipulation check
First of all, our data indicated that the placebo tDCS manipulation was successful, as the mean score on the combined manipulation check items presented at the very end of the study was M = 3.20 (SD = 0.75), indicating that participants judged the brain stimulation to have exerted ‘moderate’ to ‘substantial’ influence on their performance. Importantly, the extent of the influence did not differ between the placebo (M = 3.00, SD = 0.66) and nocebo (M = 3.32, SD = 0.87) condition, as indicated by comparisons between the manipulations check items after the stimulation blocks, t(22) = -1.43, p = .167 Apparently, participants felt they were equally influenced by the tDCS in both the enhancement and impairment condition.
Sense of agency
With regard to the sense of agency, a significant main effect for condition was observed, F(2,44) = 59.94, p < .001, p2 = .731, with post-hoc tests revealing that all expectancy conditions significantly differed from each other; the
sense of agency was significantly higher in the control condition than in the placebo condition, t(22) = 6.76, p < .001 and than in the nocebo condition, t(22) = 10.84, p < .001. Crucially, the sense of agency in the nocebo condition was significantly lower than in the placebo condition t(22) = -4.084, p < .001.
Neural response to errors (ERN)
A repeated measures ANOVA with condition and correctness as within-subject factors was conducted to examine the effect of condition on the neural response to errors as reflected in the ERN amplitude. A trend towards a main effect for condition was observed, F(2,44) = 2.96, p = .062, but this effect did not reach statistical significance. As expected, a main effect for correctness was found, F(1,22) = 100.77, p < .001, p2 = .821, with error-trials (M = -8.00
V) eliciting a significantly larger ERN signal than correct-trials (M = -1.12 V). Crucially, a significant interaction between condition and correctness appeared, F(2,44) = 6.79, p = .003, p2 = .236, reflecting that the ERN amplitude
(i.e., the difference in signal between error- and correct-trials) was significantly influenced by the expectancy manipulation. In order to decompose this interaction effect, the actual ERN signal was calculated for each participant by subtracting the signal on correct-trials from the signal on error-trials. This difference signal was then directly compared between conditions by means of three paired t-tests, which revealed a significant difference between the placebo (M = -7.90 V) and nocebo (M = -6.03 V) condition in ERN amplitude, t(22) = -3.64, p = .001, as well as between the placebo and the control condition, t(22) = -2.66, p = .014. No difference was found between the nocebo and the control condition, t(22) = 1.17, p = .255. These findings indicate that participants exhibited a stronger neural response to errors when they thought their performance would be enhanced by the brain stimulation, compared to when they believed their performance would be impaired by the stimulation or in absence of any stimulation.
RESULTS
Table 1
Descriptive statistics for stimulation influence, sense of agency and Flanker performance
Placebo Nocebo Control
Stimulation influencea (manipulation check) 3.00(0.66) 3.32(0.87) Sense of agencyb 5.09(1.16) 4.04(1.24) 6.82(0.35) Performance Errors (%) Congruent 1.6(1.74) 1.5(1.81) 1.4(1.97) Incongruent 25.5(9.36) 25.5(7.34) 27.9(8.60)
Reaction times Congruent 337(37.6) 339(35.6) 339(36.0)
Incongruent 396(43.1) 399(43.1) 397(46.4)
Note. Displays mean values, with standard deviations given in parentheses. a Measured on a 5-point scale. b Measured on a 7-point scale.
Figure 2. (A) Sense of agency over committed errors, showing a lowered sense of agency in the experimental conditions, and especially more external attribution
(i.e., lower sense of agency) in the nocebo condition, compared to the placebo condition. (B) Error-related negativity (ERN) amplitudes per condition, reflecting the strongest distress response in the placebo compared to the control and nocebo condition. (C) Proportion incorrect responses as a function of expectancy (placebo, nocebo and control) and Flanker congruency (congruent – incongruent), showing a clear congruency effect, but no expectancy or interaction effect.
(D) Correlation between the ERN and the sense of agency in the nocebo condition. A stronger ERN amplitude correlated with more internal attributions (i.e.,
higher sense of agency) and a smaller ERN amplitude was associated with more external attributions (i.e., lower sense of agency). The ERN is defined here as the ERN difference wave (i.e., incorrect – correct trials). Error bars indicate standard errors.
Sense of agency & ERN
Correlations between the sense of agency and amplitude of the ERN signal were computed, in order to investigate whether the strength of the (affective) response to errors was indeed related to the extent to which external attributions were made. To this end, the data was analyzed for correlations between the sense of agency and the ERN for the two stimulation conditions separately. There was no correlation between sense of agency and ERN amplitude in the placebo condition, r(21) = -.21, p = .172 (one-tailed). In the nocebo condition, however, a significant negative correlation between these two variables was found, r(21) = .60, p = .001 (one-tailed), indicating that an internal sense of agency was related to larger ERNs and an external sense of agency to smaller ERNs (figure 2D).
Performance effects
As expected, the results revealed that the expectancy manipulation had no noticeable effects on accuracy in the Flanker task. Specifically, there was no main effect for condition, F(2,44) = 1.79, p = .178, indicating that participants performed equally well in the placebo, nocebo and control condition. The main effect for flanker-congruency was significant, F(1,22) = 241.12, p < .001, p2 = .916, indicating that people made more errors on
incongruent (26.3%) than on congruent trials (1.5%). Moreover, there was no significant interaction effect between condition and congruency, F(2,44) = 2.55, p = .09. For the reaction times (RT), the repeated measures ANOVA revealed no main effect for condition, F(2,44) = 0.30, p = .739, and a strong effect for congruency, F(1,22) = 360.23,
p < .001, p2 = .942, showing that people were significantly slower on incongruent (397ms) compared to congruent
(339ms) trials. There was again no interaction effect between condition and congruency, F(2,44) = 0.24, p = .786.5
5 For the important measures in the current study (i.e., sense of agency and neural response to errors), no interaction effect was found between examined factors and order in which the expectancy conditions were executed (between subjects). Only with regard to reaction times there appeared to be an effect for block order (F(10,34) = 4.64, p < .001, p2 = .577), which was found to arise from a main effect for block order (irrespective of expectancy condition).
Figure 3. (A) Topoplots of averaged voltage topographies across the scalp for the ERN across the three different conditions, all showing a clear frontocentral
distribution of the error-related negativity. Response-locked waveform amplitude at FCz following correct and incorrect responses on the Flanker task for participants in the (B) placebo, (C) nocebo (D) control conditions, showing a clear ERN response across all conditions. (E) Comparison of ERN waveforms in the placebo, nocebo and control condition, showing a larger ERN amplitude for the placebo compared to the nocebo and control condition.
Individual difference measures
Individual differences in level of absorption and locus of control as measured by the absorption scale and the locus of control scale did not correlate with any of the main dependent measures (subjective performance, objective performance, sense of agency and ERN). Further analyses on individual differences were therefore not conducted.
DISCUSSION
In the present study, we investigated the placebo effect with regard to alleged cognitive enhancement and impairment through brain stimulation and linked this to error processing and the sense of agency. Results of the experiment supported the notion that placebo and nocebo suggestions influence the sense of agency over errors (impairment > enhancement) and the (affective) neural response to errors (enhancement > impairment). Furthermore, a relation between the sense of agency and the ERN amplitude was found in the nocebo condition, showing that a lower sense of agency corresponded to a reduced ERN amplitude.
The present findings extend literature on the placebo effect and cognitive enhancement by combining a purely cognitive enhancement manipulation by means of brain stimulation with a neural marker for affective aspects of error processing as well as with attributional processes as an underlying explanatory mechanism. In addition, by including both enhancement and impairment suggestions, we were able to extract distinct directional effects of expectancies on distress and sense of agency with regard to performance errors. Specifically, we provided evidence that violated expectations of increased cognitive efficacy in particular are experienced as distressing, more so than expectations of decrease efficacy reduce distress associated with performance errors. However, in the context of suggested impairment, the modulation of distress in response to errors appears to be dependent on the extent to which negative outcomes are attributed externally rather than to one’s own abilities. To our knowledge, this study was the first to provide evidence for this relationship, which seems highly relevant as it sheds light on how the concepts of agency and responsibility influence personal performance evaluation and how this relation can be modulated by expectancies regarding an external source enhancing or impairing one’s cognitive abilities. Below, we will discuss the neural and attributional effects observed in the present study in more detail.
First, the different performance expectancies were found to influence the sense of agency experienced over errors. Suggestions of impairment resulted in a lowered sense of agency over errors, as participants indicated their responses to be substantially influenced by the brain stimulation, compared to the enhancement condition, where participants generally indicated to have caused the erroneous responses themselves. The difference in sense of agency appeared to be specific for errors, as post-experimental questions suggested no difference in the extent to which participants felt globally influenced by the two types of stimulation. These results extend previous studies with reference to the fallibility of the sense of agency (Aarts, Custers, & Wegner, 2005; Moore et al., 2009), in this case specifically showing that people may experience a reduced sense of agency over completely self-generated actions when cues for external influences are provided.
Second, the neural response to errors on the task was affected by the expectancies as well, with larger ERN amplitudes in the placebo condition, compared to the control and the nocebo condition. That is, people’s reaction to errors was stronger when they believed their performance was enhanced by the brain stimulation. As there was no significant difference between the control and the nocebo condition, the results seem to indicate that suggestions of enhancement increase the affective neural response to errors, rather than suggestions of impairment decreasing this response. Following the interpretation of the ERN as a neural signal of personal distress conveying that an outcome is worse than expected (Bartholow, Henry, Lust, Saults, & Wood, 2012; Inzlicht & Al-Khindi, 2012), the present findings demonstrate that violated expectancies about enhanced mental performance cause distress as reflected in an enhanced ERN signal. This finding adds to the existing literature by providing a new instance of increased error significance in the form of positive performance expectations augmenting the ERN response, which has previously been documented for monetary rewards, social evaluation of performance (Hajcak et al., 2005), personal relevance (Gentsch et al., 2009), motivation and personality traits (Pailing & Segalowitz, 2004).
Although we found no significant difference between the nocebo and control condition on the ERN, the direction of the effects, as well as the analysis for linear contrasts (F(1,22) = 7.07, p = .014, p2 = .243), suggest that
the ERN might possibly also be attenuated in response to suggestions of cognitive impairment. This could however not be concluded based on the present results, nor can be concluded that the reduced sense of agency associated with suggestions of externally induced impairment directly causes a decrease of distress from performance failures, as this decrease was not observed. What could be concluded, though, is that in the impairment condition, the sense of agency and the magnitude of the ERN were at least somehow related, implying that the less agency people experienced over errors, the smaller their ERN amplitude. That is, people who experienced a reduced sense of agency over errors and used the brain stimulation as an external ‘excuse’ to explain their errors, also appeared to have a less pronounced (affective) reaction to these errors. The absence of a generally attenuated ERN response in the nocebo condition might arise from the fact that, although the overall sense of agency was reduced, not all participants made external attributions in this condition. Similarly, the reason why the agency-ERN relation was not found in the placebo condition might be due to the fact that hardly any external attributions were made in this condition, i.e., erroneous responses were almost always ascribed to oneself, rather than the brain stimulation. An experiment more strongly
and directly manipulating external vs. internal attributions, rather than providing the possibility for external attribution, might elucidate these proposals.
However, bearing with the fact that a clear attenuation of the ERN in the nocebo condition, as well as an agency-ERN correlation in the placebo condition were not found in the present study, it could also be suggested that there are two different mechanisms underlying the placebo and nocebo ERN effects. In the enhancement condition, the enhanced ERN effect appears to mainly arise from the unsatisfied positive expectations and hence surprising, negative outcomes resulting in increased distress. In the impairment condition, on the other hand, the magnitude of the ERN is mainly dependent on the extent to which errors are attributed to the external source supposedly causing the impairment (i.e., the brain stimulation). That is, suggestions of performance impairment can make errors less distressing, but only when errors are attributed externally. These mechanisms are not mutually exclusive and their consequences – as found in the present experiment – are compatible as well as complementary. Furthermore, framed in terms of behavioral adaptability, the ERN can be interpreted as signaling the need for behavioral adjustment (Hoffmann & Falkenstein, 2012), which is higher when outcomes are unexpected, especially when internally attributed, because then one has feeling that one can indeed adjust, and higher when outcomes are more significant – due to motivation, rewards, or a threat to one’s self-esteem, which is manipulated in our study. Based on this notion, a possible overarching explanation for the present findings is formulated by Maier and Steinhauser (2016), holding that the ERN evaluates the sources underlying errors, taking into account both expectancy and significance of errors, directed at need for behavioral adjustment.
The absence of behavioral performance effects can readily be explained by the nature of the experimental design that was applied. That is, as we used an adaptive response interval threshold that kept accuracy within a 60-80% range in order to obtain a comparable number of errors across participants and conditions (and thus comparable ERN averages), the absence of a behavioral effect is not surprising. Nevertheless, although some instances of placebo effects in cognitive control tasks have been documented (e.g., Clifasefi et al., 2007; Magalhaes De Saldanha da Gama, Slama, Caspar, Gevers, & Cleeremans, 2013), behavioral effects are generally not found (e.g., Iani et al., 2006; Inzlicht & Al-Khindi, 2012; Looby & Earleywine, 2011; Raz, Kirsch, Pollard, & Nitkin-Kaner, 2006; Schwarz & Buchel, 2015).
Relating the present findings to the predictive processing account proposed as an interpretative framework, the following can be stated based on the current results: prior beliefs regarding an external device enhancing or impairing cognitive abilities can substantially influence sensory input concerning performance errors in terms of sense of agency and responsibility, and associated personal distress. By placing the present findings in a predictive processing framework, the study aligns conclusively with the long list of phenomena that have been interpreted in terms of the Bayesian predictive processing perspective, including visual perception (Rao & Ballard, 1999), mystical experiences (Andersen et al., 2014; Schjoedt et al., 2013), delusions (Corlett et al., 2010), placebo analgesia (Buchel et al., 2014) and social cognition (Kilner et al., 2007).
Moreover, returning to the discussion on cognitive enhancement in general, and by means of neurostimulation devices in particular, the present study highlights the relevance of considering consequences in terms of affected feelings of responsibility and negative affect. Importantly, applications of substances and devices should take into account that altered expectations are additionally associated with changes in sense of agency and emotions; (suggestions of) external enhancers and impairers might not only affect performance per se, but also one’s (unconscious) feelings of responsibility over the performance and distress in response to failures, the latter of which is increased for cognitive enhancers. This forms an important but rarely acknowledged drawback of cognitive enhancement substances and devices; people might believe to (or even actually) perform better, but when they are confronted with failures, they experience more distress.
As the present study focused on error processing, it would be interesting to investigate whether the observed effects can also be found in response to successes, in a reversed fashion. That is, do people also experience less pride and less positive emotions when their success can be attributed externally (i.e., the reverse effect on sense of agency and affective processing), or are the results of the present study specific for negative outcomes, thereby representing a case of the self-serving bias and the general negativity bias in agency (Morewedge, 2009)? This seems rather relevant with reference to the application of cognitive enhancement in practice, as such a reduction of the responsibility over and the appreciation of success would be a considerable negative side-effect of cognitive enhancers, in addition to the increased distress associated with encountered failures.
Finally, the study emphasizes the applicability of brain stimulation paradigms in placebo research, as this technique appears highly suitable and effective in inducing expectancies, especially concerning cognitive enhancement and impairment. Due to the solid reputation and credibility of brain stimulation methods among the general public, as well as the present findings strikingly demonstrating placebo and nocebo effects by means of a
tDCS manipulation, we argue that this method might be a promising and effective technique for placebo research in the cognitive experimental field.
REFERENCES
Aarts, H., Custers, R., & Wegner, D. M. (2005). On the inference of personal authorship: enhancing experienced agency by priming effect information. Conscious Cogn, 14(3), 439-458. doi:10.1016/j.concog.2004.11.001 Andersen, M., Schjoedt, U., Nielbo, K. L., & Sørensen, J. (2014). Mystical experience in the lab. Method & Theory
in the Study of Religion, 26(3), 217-245.
Bartholow, B. D., Henry, E. A., Lust, S. A., Saults, J. S., & Wood, P. K. (2012). Alcohol effects on performance monitoring and adjustment: affect modulation and impairment of evaluative cognitive control. Journal of
Abnormal Psychology, 121(1), 173-186. doi:10.1037/a0023664
Bostrom, N., & Sandberg, A. (2009). Cognitive enhancement: methods, ethics, regulatory challenges. Science and
Engeneering Ethics, 15(3), 311-341. doi:10.1007/s11948-009-9142-5
Buchel, C., Geuter, S., Sprenger, C., & Eippert, F. (2014). Placebo analgesia: a predictive coding perspective.
Neuron, 81(6), 1223-1239. doi:10.1016/j.neuron.2014.02.042
Chambon, V., Filevich, E., & Haggard, P. (2014). What is the Human Sense of Agency, and is it Metacognitive? In S. M. Fleming & C. D. Frith (Eds.), The Cognitive Neuroscience of Metacognition (pp. 321-342). Berlin, Heidelberg: Springer.
Clifasefi, S. L., Garry, M., Harper, D. N., Sharman, S. J., & Sutherland, R. (2007). Psychotropic placebos create resistance to the misinformation effect. Psychonomic Bulletin & Review, 14(1), 112-117.
Corlett, P. R., Taylor, J. R., Wang, X. J., Fletcher, P. C., & Krystal, J. H. (2010). Toward a neurobiology of delusions. Prog Neurobiol, 92(3), 345-369. doi:10.1016/j.pneurobio.2010.06.007
Custers, R., Aarts, H., Oikawa, M., & Elliot, A. (2009). The nonconscious road to perceptions of performance: Achievement priming augments outcome expectancies and experienced self-agency. Journal of
Experimental Social Psychology, 45(6), 1200-1208. doi:10.1016/j.jesp.2009.07.013
Eriksen, B. A., & Eriksen, C. W. (1974). Effects of noise letters upon the identification of target letters in visual search. Perception and Psychophysics, 16, 143-149.
Falkenstein, M., Hoormann, J., Christ, S., & Hohnsbein, J. (2000). ERP components on reaction errors and their functional significance: a tutorial. Biological Psychology, 51(2), 87-107.
Gehring, W. J., Goss, B., Coles, M. G., Meyer, D. E., & Donchin, E. (1993). A neural system for error detection and compensation. Psychol Sci, 4(6), 385-390.
Gehring, W. J., Himle, J., & Nisenson, L. G. (2000). Action-monitoring dysfunction in obsessive-compulsive disorder. Psychol Sci, 11(1), 1-6.
Gentsch, A., Ullsperger, P., & Ullsperger, M. (2009). Dissociable medial frontal negativities from a common monitoring system for self- and externally caused failure of goal achievement. Neuroimage, 47(4), 2023-2030. doi:10.1016/j.neuroimage.2009.05.064
Gibbons, F. X., & Gaeddert, W. P. (1984). Focus of attention and placebo utility. Journal of Experimental Social
Psychology, 20(2), 159-176.
Gratton, G., Coles, M. G., & Donchin, E. (1983). A new method for off-line removal of ocular artifact.
Electroencephalography and Clinical Neurophysiology, 55(4), 468-484.
Groth-Marnat, G., & Pegden, J. A. (1998). Personality correlates of paranormal belief: Locus of control and sensation seeking. Social behavior and personality: An international Journal,, 26(3), 291-296. Haggard, P., & Chambon, V. (2012). Sense of agency. Current Biology, 22(10), R390-R392.
Hajcak, G., Moser, J. S., Yeung, N., & Simons, R. F. (2005). On the ERN and the significance of errors.
Psychophysiology, 42(2), 151-160. doi:10.1111/j.1469-8986.2005.00270.x
Heersmink, R. (2015). Extended mind and cognitive enhancement: moral aspects of cognitive artifacts.
Phenomenology and the Cognitive Sciences, 1-16. doi:10.1007/s11097-015-9448-5
Hoffmann, S., & Falkenstein, M. (2012). Predictive information processing in the brain: errors and response monitoring. International Journal of Psychophysiology, 83(2), 208-212.
doi:10.1016/j.ijpsycho.2011.11.015
Iani, C., Ricci, F., Gherri, E., & Rubichi, S. (2006). Hypnotic Suggestion Modulates Cognitive Conflict The Case of the Flanker Compatibility Effect. Psychol Sci, 17(8), 721-727.
Inzlicht, M., & Al-Khindi, T. (2012). ERN and the placebo: a misattribution approach to studying the arousal properties of the error-related negativity. Journal of Experimental Psychology: General, 141(4), 799-807. doi:10.1037/a0027586
Kilner, J. M., Friston, K. J., & Frith, C. D. (2007). Predictive coding: an account of the mirror neuron system. Cogn
Lilienfeld, S. O., Lynn, S. J., Ruscio, J., & Beyerstein, B. L. (2011). 50 great myths of popular psychology:
Shattering widespread misconceptions about human behavior. Hoboken, NJ: John Wiley & Sons.
Looby, A., & Earleywine, M. (2011). Expectation to receive methylphenidate enhances subjective arousal but not cognitive performance. Experimental and Clinical Psychopharmacology, 19(6), 433-444.
Magalhaes De Saldanha da Gama, P. A., Slama, H., Caspar, E. A., Gevers, W., & Cleeremans, A. (2013). Placebo-suggestion modulates conflict resolution in the Stroop Task. PLoS One, 8(10), e75701.
doi:10.1371/journal.pone.0075701
Maher, B. (2008). Poll results: look who's doping. Nature, 452(10), 674-675.
Maier, M. E., & Steinhauser, M. (2016). Error significance but not error expectancy predicts error-related negativities for different error types. Behav Brain Res, 297, 259-267. doi:10.1016/j.bbr.2015.10.031 Mezulis, A. H., Abramson, L. Y., Hyde, J. S., & Hankin, B. L. (2004). Is there a universal positivity bias in
attributions? A meta-analytic review of individual, developmental, and cultural differences in the self-serving attributional bias. Psychological Bulletin, 130(5), 711-747.
Michael, R. B., Garry, M., & Kirsch, I. (2012). Suggestion, Cognition, and Behavior. Current Directions in
Psychological Science, 21(3), 151-156. doi:10.1177/0963721412446369
Moore, J. W., Wegner, D. M., & Haggard, P. (2009). Modulating the sense of agency with external cues.
Conscious Cogn, 18(4), 1056-1064. doi:10.1016/j.concog.2009.05.004
Morewedge, C. K. (2009). Negativity bias in attribution of external agency. J Exp Psychol Gen, 138(4), 535-545. doi:10.1037/a0016796
Paddock, J. R., Joseph, A. L., Chan, F. M., Terranova, S., Manning, C., & Loftus, E. F. (1998). When guided visualization procedures may backfire: Imagination inflation and predicting individual differences in suggestibility. Applied Cognitive Psychology, 12(7), S63-S75.
Pailing, P. E., & Segalowitz, S. J. (2004). The error-related negativity as a state and trait measure: motivation, personality, and ERPs in response to errors. Psychophysiology, 41(1), 84-95. doi:10.1111/1469-8986.00124
Polito, V., Barnier, A. J., & Woody, E. Z. (2013). Developing the Sense of Agency Rating Scale (SOARS): an empirical measure of agency disruption in hypnosis. Conscious Cogn, 22(3), 684-696.
Rao, R. P., & Ballard, D. H. (1999). Predictive coding in the visual cortex: a functional interpretation of some extra-classical receptive-field effects. Nature Neuroscience, 2(1), 79-87.
Raz, A., Kirsch, I., Pollard, J., & Nitkin-Kaner, Y. (2006). Suggestion reduces the stroop effect. Psychol Sci, 17(2), 91-95. doi:10.1111/j.1467-9280.2006.01669.x
Rigoni, D., Pourtois, G., & Brass, M. (2015). 'Why should I care?' Challenging free will attenuates neural reaction to errors. Soc Cogn Affect Neurosci, 10(2), 262-268. doi:10.1093/scan/nsu068
Rigoni, D., Wilquin, H., Brass, M., & Burle, B. (2013). When errors do not matter: weakening belief in intentional control impairs cognitive reaction to errors. Cognition, 127(2), 264-269.
doi:10.1016/j.cognition.2013.01.009
Rotter, J. B. (1966). Generalized expectancies for internal versus external control of reinforcement. Psychological
monographs: General and applied, 80(1), 1-28.
Rusconi, E., & Mitchener-Nissen, T. (2014). The role of expectations, hype and ethics in neuroimaging and neuromodulation futures. Frontiers in System Neuroscience, 8, 214. doi:10.3389/fnsys.2014.00214 Sahakian, B. J., & Morein-Zamir, S. (2011). Neuroethical issues in cognitive enhancement. Journal of
Psychopharmacology, 25(2), 197-204.
Scheffers, M. K., & Coles, M. G. H. (2000). Performance monitoring in a confusing world: Error-related brain activity, judgments of response accuracy, and types of errors. Journal of Experimental Psychology: Human
Perception and Performance, 26(1), 141-151. doi:10.1037/0096-1523.26.1.141
Schjoedt, U., Sørensen, J., Nielbo, K. L., Xygalatas, D., Mitkidis, P., & Bulbulia, J. (2013). Cognitive resource depletion in religious interactions. Religion, Brain & Behavior, 3(1), 39-55.
doi:10.1080/2153599x.2012.736714
Schwarz, K. A., & Buchel, C. (2015). Cognition and the Placebo Effect--Dissociating Subjective Perception and Actual Performance. PLoS One, 10(7), e0130492. doi:10.1371/journal.pone.0130492
Srinivasan, M. V., Laughlin, S. B., & Dubs, A. (1982). Predictive coding: a fresh view of inhibition in the retina.
Proceedings of the Royal Society of London B: Biological Sciences, 216(1205), 427-459.
Swiney, L., & Sousa, P. (2013). When our thoughts are not our own: Investigating agency misattributions using the Mind-to-Mind paradigm. Conscious Cogn, 22(2), 589-602.
Szemerszky, R., Domotor, Z., Berkes, T., & Koteles, F. (2016). Attribution-Based Nocebo Effects. Perceived Effects of a Placebo Pill and a Sham Magnetic Field on Cognitive Performance and Somatic Symptoms. Int
J Behav Med, 23(2), 204-213. doi:10.1007/s12529-015-9511-1
Tellegen, A., & Atkinson, G. (1974). Openness to absorbing and self-altering experiences ("absorption"), a trait related to hypnotic susceptibility. Journal of Abnormal Psychology, 83(3), 268-277.
Von Borries, A. K. L., Brazil, I. A., ulten, B. H., Buitelaar, J. K., Verkes, R. J., & De Bruijn, E. R. A. (2010). Neural correlates of error-related learning deficits in individuals with psychopathy. Psychological
Medicine, 40(9), 1559-1568.
Voss, M., Moore, J., Hauser, M., Gallinat, J., Heinz, A., & Haggard, P. (2010). Altered awareness of action in schizophrenia: a specific deficit in predicting action consequences. Brain, 133(10), 3104-3112. doi:10.1093/brain/awq152
Wegner, D. M., & Wheatley, T. (1999). Apparent mental causation: Sources of the experience of will. American
Psychologist, 54(7), 480-492.
Woody, E. Z., & McConkey, K. M. (2003). What we don’t know about the brain and hypnosis, but need to: A view from the Buckhorn Inn. International Journal of Clinical and Experimental Hypnosis, 51(3), 309-338.
