The role of dual mechanism control in paranormal beliefs: Evidence from behavioral and electrical stimulation studies

Abstract

Paranormal believers’ thinking is frequently biased by intuitive beliefs. Lack of inhibition of these tempting beliefs is considered a key element in paranormal believers’ thinking. However, the brain process related to cognitive control and dual mechanism control (DMC) in paranormal believers is poorly understood. Given the critical role of cognitive control in paranormal beliefs, the purpose of this study was to investigate the role of DMC in paranormal beliefs in two behavioral and brain stimulation experiments. In the first experiment, participants were screened based on The Revised Paranormal Belief Scale (RPBS) and were divided into two groups with skeptics and paranormal believers and performed the AX continuous Performance task (AX-CPT). In the second experiment, paranormal believers were randomly assigned into one of three groups of rDLPFC, rIFG, and Sham stimulation (2 mA, 20 min) while performing the AX-CPT task. The results of the first experiment showed that there was a significant difference between groups (believers vs. skeptics) in the BX, BY trials, and AX and BY trials in RT and accuracy respectively. Additionally, the results of the second experiment showed that there was a significant difference between the rDLPFC and Sham groups in the BY trial on RT. There was also a significant difference between the rDLPFC and rIFG groups in the BX trial. In addition, the results showed a significant difference between the rDLPFC and rIFG groups in the BX, and BY trials on accuracy. These findings highlight the relationship between DMC, paranormal beliefs, and the underlying neural mechanisms.

Keywords:

• Dual mechanism control
• tDCS
• paranormal beliefs

Reviewing Editor:

• Jim Grange, Keele University, United Kingdom

Subjects:

• Cognitive Psychology
• Parapsychology
• Social Psychology
• Psychological Science

1. Introduction

Paranormal beliefs encompass a wide range of beliefs, entities, practices, and phenomena that defy the established laws of science (Tobacyk & Milford, 1983). Examples of such beliefs include traditional religions, extrasensory perception, witchcraft, superstitions, telekinesis, spiritualism, magical thinking, and precognition (Wilson, 2018). Irwin (Irwin, 2009) has posited that individuals who hold paranormal beliefs may exhibit cognitive deficits. Supporting this hypothesis, several studies have identified cognitive impairments among paranormal believers, including deficits in inhibitory control (Narmashiri et al., 2023a, 2023b; Wain & Spinella, 2007), critical thinking (Hergovich & Arendasy, 2005), perceptual decision-making (Figure 1A; Narmashiri et al., 2023c), probability misjudgment (Musch & Ehrenberg, 2002), working memory and inattention blindness (Richards et al., 2014), reasoning skills (Dagnall et al., 2007), imagination (Sarbin, 2004), cognitive and belief bias (Narmashir, 2017; Narmashiri et al., 2018; White et al., 2021). Human cognition is inherently susceptible to biases, but some individuals, particularly paranormal believers, are more prone to these cognitive biases (Van Elk, 2015; Willard & Norenzayan, 2013). Cognitive and belief biases represent systematic errors in thinking that can lead to unfounded tendencies, attitudes, and beliefs, influencing decision-making, reasoning, perception, and cognition (Buss, 2005). Willard and Norenzayan (Willard & Norenzayan, 2013) argue that these biases, rooted in brain functions, form the biological basis of paranormal experiences, giving rise to paranormal agents as a result of correlated biases. Intuitive beliefs often color the thinking of paranormal believers (Prike et al., 2017). Given that many paranormal beliefs lack empirical support when subjected to analytical scrutiny, it is reasonable to expect a negative correlation with an analytical thinking style (Svedholm & Lindeman, 2013). Indeed, research has shown that lower levels of paranormal belief are associated with higher levels of analytical thinking (Riekki, 2014). However, the dual-process theory of reasoning interaction, particularly in the context of resolving interpretational inconsistencies, remains a topic of debate (De Neys, 2012; Gervais & Norenzayan, 2012; Pennycook et al., 2012). The central question revolves around whether these two cognitive processes (intuitive and analytical thinking styles) compete with each other or if there exists a specific regulatory mechanism governing these reasoning processes. In this context, inhibitory control emerges as a potential suppressor of the intuitive thinking style, effectively reducing the inclination toward paranormal beliefs (Riekki, 2014).

Controlling dominant biases and responses is closely linked to inhibitory control, a fundamental aspect of executive functions (Miyake et al., 2000). According to Hood (Hood, 2009), some individuals may struggle to inhibit intuitive thoughts, leading to a heightened sense of the paranormal. Previous studies have indeed found inhibitory impairments among paranormal believers within a single cognitive control paradigm (Hood et al., 2011; Narmashiri et al., 2023a, 2023b, 2023c; Svedholm & Lindeman, 2013). However, a single cognitive control process cannot comprehensively account for all inhibitory effects on behavior (Aron & Verbruggen, 2008). Therefore, the theory of dual mechanism control (DMC) has been proposed to address this issue, significantly influencing the field of cognitive control (Braver, 2012; Braver et al., 2007). In line with the DMC theory, there are two distinct cognitive control modes: proactive control and reactive control. These modes offer flexibility for swift and targeted actions and are instrumental in managing and suppressing irrational beliefs. Specifically, proactive control is employed to proactively prevent interference, as detailed in references (Braver, 2012; Braver et al., 2007; Miller & Cohen, 2001). Conversely, reactive control plays a role in regulating information interference during action execution and serves as a corrective decision-making mechanism during the action itself (Braver et al., 2009).

A noticeable feature of DMC is that proactive and reactive control may vary depending on brain region activity (Braver, 2012). Studies support this differentiation, revealing variations in cognitive control mode utilization (Braver et al., 2009; Morales et al., 2013; 2015; Paxton et al., 2006, 2008). Impaired proactive control has been observed in conditions like schizophrenia (Lesh et al., 2013), which overlaps with paranormal beliefs (Thalbourne, 1994). As a result, cognitive control deficits manifest across various cognitive domains in paranormal beliefs (Riekki, 2014), schizophrenia (Lesh et al., 2011), and reductions in activity within the Dorsolateral Prefrontal Cortex (DLPFC) and right Inferior Frontal Gyrus (rIFG) regions in schizophrenia (Lesh et al., 2013) and paranormal beliefs (Riekki, 2014). Given the association of the DLPFC and rIFG regions with the neural network of cognitive control (Braver et al., 2009; Chiew & Braver, 2013; Niendam et al., 2012), it’s plausible that this network may be impaired in paranormal believers. However, while previous research hints at potential impairments in the DMC among paranormal believers, no clear and direct behavioral studies have explored the roles of proactive control and reactive control in paranormal beliefs. Therefore, our first study aims to investigate the differences in proactive control and reactive control between paranormal believers and skeptics through behavioral assessments. Building upon these findings, our second study delves into the relationship between the prefrontal cortex (PFC) region and proactive control and reactive control by employing transcranial Direct Current Stimulation (tDCS) as a cortical stimulation method (Antal et al., 2014).

Understanding the cognitive control processes underlying paranormal beliefs is not only academically intriguing but also carries significant implications for our comprehension of human cognition. The existing body of research has established a compelling link between paranormal beliefs and cognitive deficits (Irwin, 2009), particularly in single cognitive control paradigm (Hood et al., 2011; Narmashiri et al., 2023a, 2023b, 2023c; Svedholm & Lindeman, 2013). However, a comprehensive exploration of the dual-process theory in cognitive control, specifically the interplay between proactive and reactive cognitive control modes, within the context of paranormal beliefs remains largely uncharted. Our research endeavors aim to fill this critical gap in the literature by conducting two interconnected studies. The first study investigates behavioral differences in proactive and reactive control between paranormal believers and skeptics, employing the AX-CPT task. Building upon these findings, our second study utilizes tDCS to explore the neural correlates of proactive and reactive control in the PFC regions. By bridging the conceptual gap between DMC and paranormal beliefs, our research contributes not only to the understanding of individual differences in cognitive processing but also provides valuable insights into potential intervention strategies for enhancing cognitive control in populations susceptible to paranormal beliefs. Ultimately, these studies aim to advance our understanding of the cognitive mechanisms underlying paranormal beliefs and offer avenues for further exploration.

2. Methods

2.1. Participants

In the first experiment, we assessed the handedness of 88 participants using the Edinburgh Handedness Inventory. Based on the median score (48.50) on The Revised Paranormal Belief Scale (RPBS), we categorized participants into two groups: skeptics (30 males and 14 females) and believers (22 males and 22 females). Skeptics are individuals characterized by strong skepticism toward paranormal phenomena, typically indicated by lower scores on the RPBS scale. In contrast, believers hold strong beliefs in paranormal phenomena, often reflected in higher RPBS scores. The mean (standard deviation) RPBS scores were 34.25 (3.90) for skeptics and 70.75 (12.47) for believers. Additionally, the mean age (standard deviation) for skeptics was 27.00 (5.74), and for believers, it was 27.00 (5.941). In the second experiment, 72 right-handed participants (determined by the Edinburgh Handedness Inventory) were recruited based on their RPBS scores. They were then randomly assigned to one of three groups: rDLPFC (19 females), rIFG (8 females), and Sham (8 females), with RPBS scores averaging 51.50 (standard deviation not mentioned). The mean age (standard deviation) was 19.58 (1.01) for rDLPFC, 27.58 (5.08) for rIFG, and 26.16 (5.77) for Sham. Before participating in the study, all participants reported having normal vision and no history of mental, neurological, personality disorders, acute or chronic diseases, or drug abuse, as indicated in the demographic questionnaire. The experiments were conducted at the social cognition laboratory at University of Tehran. Prior to participation, all subjects provided informed consent, completed relevant questionnaires, and received detailed information about the experimental tasks. The computer screen was positioned at an approximate distance of 60 centimeters from the participants. This study adhered to the principles of the Declaration of Helsinki and was approved by the Ethics Committee of The Institute for Cognitive Science Studies. All participants provided written informed consent before enrollment.

2.2. Measures

2.2.1. The revised paranormal belief scale (RPBS)

The RPBS is the most widely used measure of belief in the paranormal (Dagnall et al., 2022; Drinkwater et al., 2017). The RPBS consists of 26 items grouped into seven subscales: Traditional Religious Beliefs, Psi, Witchcraft, Superstition, Spiritualism, Extraordinary Life Forms, and Precognition. Respondents rate these items on a seven-point Likert scale, ranging from 1 = Strongly Disagree to 7 = Strongly Agree (Tobacyk, 2004). The RPBS presents statements (e.g., ‘If you break a mirror, you will have bad luck’), and participants indicate their level of endorsement using the seven-point Likert scale. The RPBS yields scores for each subscale and can be summed to provide an overall measure of belief in the paranormal. The RPBS subscales and the overall measure demonstrate adequate validity and reliability (Drinkwater et al., 2017; Tobacyk, 2004). A previous study showed excellent internal consistency with a Cronbach’s α value of 0.92 (Coleman et al., 2022). The RPBS serves as the primary questionnaire for assessing paranormal beliefs in the current study.

2.2.2. AX-continuous performance task (AX-CPT)

This task serves as a behavioral measure of the DMC (Cooper, 2016; Marcora et al., 2009) and comprises four trial types: AX (target trial), AY (conflict trial), BX (conflict trial), and BY (control trial). Participants were instructed to retain the cue (either the letter “A” or any letter except “X,” “K,” and “Y” due to their perceptual similarity with "X") in memory until the appearance of the probe (either the letter “X” or any letter except “A,” “K,” or “Y”). Previous studies have employed high-conflict trials (BX, AY) to investigate differences in proactive control and reactive control structures (Braver et al., 2009; Cooper, 2016; Marcora et al., 2009). Baseline measurements in AX-CPT encompass mean accuracy and reaction time, computed for each trial (Ogundele et al., 2011). Chiew and Braver (Chiew & Braver, 2017) introduced performance in the AY and BX trials as the proactive control Index and reactive control Index, respectively. These indices are based on mean error and reaction time (RT) for correct responses. The proactive control behavioral indices (PCB) gauge the proactive control’s relative dominance, assuming a complementary relationship with reactive control (Braver et al., 2009). Proactive indices is computed separately for reaction time and accuracy, resulting in two indices, with values ranging from −1 to +1. A result closer to +1 signifies a stronger inclination toward proactive control, while a result closer to -1 indicates a greater propensity for reactive control (Braver et al., 2009).

2.3. Procedure

2.3.1. Experiment 1

Participants engaged in the AX-CPT task, utilizing the version employed by Marcora et al. (Marcora et al., 2009) (Figure 1A). During this procedure, participants were presented with letter trials lasting 300 ms, each displayed at the center of a black screen. The letters were presented on a cue-probe basis, with a 4200 ms interval between cue and probe presentation. The intertrial interval was set at 1200 ms. Participants were instructed to retain the cue in memory (either the letter "A" or any letter other than "X," "K," and "Y," due to their perceptual resemblance to "X") until they encountered the probe (either the letter "X" or any letter other than "A," "K," or "Y"). When presented with the cue "A" followed by the probe "X," participants were required to respond by pressing the "yes" button. For all other possible cue-probe combinations, participants were instructed to press the "no" button. The inclusion of two distractor letters between each cue-probe pair introduced demands for goal maintenance. AX trials were presented 70% of the time during experimental phases, with each of the remaining experimental conditions appearing 10% of the time. This task typically took approximately 20 minutes to complete.

Figure 1. Task Design, Stimulation Procedure, and Behavioral Performance. A) View of AX-CPT task. Participants completed an AX-CPT task, viewing letter trials lasting 300 ms on a black screen, with a 4200 ms interval between cue and probe presentation. They were instructed to remember the cue (“A” or any letter except “X,” “K,” and “Y”) until they saw the probe (“X” or any letter except “A,” “K,” or “Y”). If the cue was “A” followed by the probe “X,” they pressed the “yes” button; otherwise, they pressed the “no” button. B) Locations of the center of the electrodes of interest (F4 and FC6 according to the 10-10 international system for EEG electrode placement) for the stimulation conditions. The reference electrode was placed over the left supraorbital region. C) Schematic representation of the tDCS procedure. Stimulation sessions lasted 20 minutes, occurring as a single session during AX-CPT task performance. D) Mean reaction time groups in the AX-CPT task. The figure displays the mean RT for skeptics and believers in AX, AY, BX, and BY trials. The x-axis represents the groups, while the y-axis indicates the mean reaction time in AX, AY, BX, and BY trials. Blue bar charts represent skeptics’ performance, and golden bar charts represent the believers’ group performance. E) The figure displays the mean accuracy for skeptics and believers in AX, AY, BX, and BY trials. The x-axis represents the groups, while the y-axis indicates the mean accuracy in AX, AY, BX, and BY trials. Blue bar charts represent skeptics’ performance, and golden bar charts represent the believers’ group performance. F) The figure shows RT (top) and accuracy (bottom) for the groups (skeptics vs. believers) in the proactive indices. The x-axis shows the groups, and the y-axis indicates the RT (top) and accuracy (bottom) in the proactive indices. Blue bar charts represent skeptics’ performance, and golden bar charts represent the believers’ group performance. Error bars represent standard deviation. Symbols *, **, and *** denote statistical significance levels at P < 0.05, P < 0.01, and P < 0.001, respectively.

2.3.2. Experiment 2

In this study, a within-participants, single-blinded, sham-controlled crossover design was employed. First, participants’ handedness was assessed using the EHI, and they were then provided with instructions regarding the RPBS, as detailed in Experiment 1. While tDCS was administered, participants performed the AX-CPT task (Dagnall et al., 2022) (see Figure 1). This experiment had a duration of roughly 30 minutes. Stimulation was maintained at 20 minutes (single session during AX-CPT task performance) with a constant current of 2 mA (current density = 0.057 mA/cm2). The current was gradually faded in and out with an 8-second ramp. For sham stimulation, the same electrode montages were randomly used, lasting for 30 seconds. Sham stimulation followed a similar procedure, except that the current was ramped down and remained off after the initial 8-second ramp-up at the beginning of the 20-minute period.

2.4. tDCS

Direct current was administered using a battery-driven stimulator (ActivaDose II, ActivaTek Inc., Salt Lake City, UT, USA). Current was delivered through a pair of saline-soaked sponges on rubber electrodes (5 × 7 cm) following the 10-10 international system for EEG electrode placement (Jurcak et al., 2007). Depending on the group, the active electrode was positioned over the right DLPFC (F4) or the right IFJ (FC6h: midway between FC4 and FC6; see Figure 2), while the reference electrode was situated over the left supraorbital region, consistent with prior studies (Feeser et al., 2014; Gómez-Ariza et al., 2017; Noetscher et al., 2014). Stimulation sessions lasted 20 minutes, occurring as a single session during AX-CPT task performance, at a constant current of 2 mA (current density = 0.057 mA/cm2), with an 8-second ramping up and down at the beginning and end of the 20-minute period (Figure 1). Sham stimulation followed the same procedure, with the exception that the current ramped down and remained off after the initial 8-second ramp-up at the start of the 20-minute period.

Figure 2. tDCS Effects on AX-CPT Task in Groups. A) The figure displays the mean reaction time for all groups (rDLPFC, rIFG, Sham) in AX, AY, BX, and BY trials. The x-axis represents the groups, and the y-axis indicates the mean reaction time in AX, AY, BX, and BY trials. B) The figure illustrates the mean accuracy for all groups (rDLPFC, rIFG, Sham) in AX, AY, BX, and BY trials. The x-axis represents groups, while the y-axis indicates the mean accuracy in AX, AY, BX, and BY trials. C) RT and accuracy groups (rDLPFC, rIFG, Sham) in the proactive indices. The figure presents RT (top) and accuracy (bottom) for the groups (rDLPFC, rIFG, Sham) in the proactive indices. The x-axis shows the groups, and the y-axis indicates the RT (top) and accuracy (bottom) in the proactive indices. Error bars represent standard deviation. Symbols *, **, and *** denote statistical significance levels at P < 0.05, P < 0.01, and P < 0.001, respectively.

2.5. Statistical analysis

Before the formal analysis, we conducted exploratory data analysis to identify and remove outliers, mitigating the risk of spurious results. Both experiments involved comparisons of demographic and behavioral variables between groups using independent t-tests and χ2-tests. Statistical analyses were conducted separately for mean reaction times (RT) and accuracy. Initially, we compared the mean RT and accuracy for trials (AX, AY, BX, BY) in the AX-CPT task using repeated-measure analysis of variance (ANOVA). In both experiments, comparisons of proactive control and reactive control indices were conducted using repeated-measure ANOVA, with groups as the between factor and trials as the within factor for both RT and accuracy. If this analysis was significant, we performed a one-way ANOVA for each group (correcting the P-value using Bonferroni correction). When appropriate, post-hoc comparisons were conducted using Bonferroni correction for multiple comparisons. This involved comparing the two groups (skeptics vs. believers) with each other across different sequences (AX, AY, BX, BY) in the first experiment and comparing the three groups (rDLPFC, rIFG, Sham) with each other across different sequences (AX, AY, BX, BY) in the second experiment. Furthermore, proactive indices were compared using ANOVA with groups as the between factor and proactive indices indices as the within factor for both RT and accuracy. We performed statistical analyses using IBM SPSS Statistics version 24 (IBM Inc., New York, USA), with statistical significance set at p < 0.05.

3. Results

3.1. Experiment 1: Behavioral results

RT. The results of the main effect of MANOVA show that there was a significant difference in the RT in different sequences (AX, AY, BX, BY) between the groups (skeptics vs. believers) (Wilks’ Lambda F_4,83=16.56, P = 0.001, η = 0.139). To clarify the differences between the groups, the bonferroni post hoc test was used for paired comparisons. The results showed that there was a significant difference between the groups in the BX, BY trials (P = 0.05), and the results showed a trend toward significance difference between groups in the AY trial (P = 0.06), reflecting a lower tendency to use reactive control in the believers (Figure 1).

Accuracy. The results of the main effect of MANOVA show that there was a significant difference in the RT in different sequences (AX, AY, BX, BY) between the groups (skeptics vs. believers) (Wilks’ Lambda F_4,83=30.183, P = 0.001, η = 0.259). To clarify the differences between the groups, the Bonferroni post hoc test was used for paired comparisons. The results showed a significant difference between the groups in the AX and BY trials (P = 0.05), and there was no significant difference between the groups in the AY and BX trials (Figure 1).

3.1.1. Proactive indices (PCB)

RT. The ANOVA showed a significant difference in the proactive indices between the groups (F_1,86=11.401, P = 0.001, η = 0.117), indicating a lower tendency to use reactive control in the believers (Figure 1).

Accuracy. The results showed no significant difference between the groups in the proactive indices (F_1,86=2.193, P = 0.142, η = 0.025).

3.2. Experiment 2: Stimulation results

RT. The results of the main effect of MANOVA show that there was a significant difference in the RT in different sequences (AX, AY, BX, BY) between the groups (rDLPFC, rIFG, Sham) (Wilks’ Lambda F_8,132=3.01, P = 0.004, η = 0.15). To clarify the differences between the groups, the Bonferroni post hoc test was used for paired comparisons. The results showed a trend toward significance difference between the rDLPFC and rIFG groups (P = 0.08), and rIFG and Sham groups in the AY trial (P = 0.08). Additionally, the results showed that there was a significant difference between the rDLPFC and Sham groups in the BY trial (P = 0.05), and there was not a significant difference between the groups in the AX trial (Figure 2).

Accuracy. The results of the main effect of MANOVA show that there was a significant difference in the accuracy in different sequences (AX, AY, BX, BY) between the groups (rDLPFC, rIFG, Sham) (Wilks’ Lambda F_8,132=3.502, P = 0.001, η = 0.17). To clarify the differences between the groups, the bonferroni post hoc test was used for paired comparisons. The results showed that there was a significant difference between the rDLPFC and rIFG groups in the BX trial (P = 0.03). There was a significant difference between the rDLPFC and rIFG groups in the BY trial (P = 0.04), and there was no significant difference between the groups in the AY and AX trials (Figure 2).

3.2.1. Proactive indices (PCB)

RT. The ANOVA showed no significant difference in the proactive indices of the RT between the groups (F_1,86=1.125, P = 0.330, η = 0.032). As a result, RT in proactive indices was not different between groups (Figure 2).

Accuracy. The ANOVA showed no significant difference in the proactive indices of the accuracy between the groups (F_1,69=1.736, P = 0.184, η = 0.04, Figure 2C).

4. Discussion

The first experiment aimed to assess the role of DMC in paranormal beliefs. Results revealed that paranormal believers exhibited weaker performance in the AX-CPT task, particularly in the proactive control strategy, suggesting cognitive control deficits compared to skeptics. While DMC hasn’t been extensively studied in paranormal believers, related research in other disorders has demonstrated differences (Braver et al., 2009; Morales et al., 2013, 2015, Paxton et al., 2006, 2008). The second experiment aimed to investigate the impact of tDCS on rDLPFC and rIFG on paranormal believers’ performance in the AX-CPT task. Following tDCS in rDLPFC and rIFG, the results demonstrated an increase in both proactive control and reactive control strategies in the rDLPFC group compared to the rIFG and sham groups. This suggests that tDCS applied to rDLPFC can enhance DMC in paranormal beliefs, consistent with previous studies (Boudewyn et al., 2019; 2020; Lee et al., 2020; Pulopulos et al., 2020; Schulze et al., 2019).

4.1. Paranormal beliefs and DMC: the role of proactive control and reactive control strategies

Our behavioral results are consistent with prior research on individual differences in related disorders. Pennycook et al. (Pennycook et al., 2014) demonstrated that both paranormal beliefs and traditional religious beliefs negatively impact goal maintenance and conflict resolution during decision-making reasoning. They also found a positive association between analytic thinking and the ability to maintain goals and resolve conflicts. Given these findings, believers may face challenges in maintaining goals and resolving conflicts, potentially favoring an intuitive reasoning style (Riekki, 2014). This suggests a tendency towards employing intuitive thinking for goal maintenance and conflict resolution, indicating potential weaknesses in proactive control strategies among paranormal believers. However, skeptics demonstrate better performance in maintaining goal-related information and retrieving specific information. Drawing from DMC theory (Braver et al., 2007), it can be inferred that skeptics convert their goals into prepared actions. The performance of this group aligns with the notion that these individuals engage in cognitive control based on the anticipated cue information.

On the other hand, it can be noted that proactive control operates early and continuously, providing information at cue presentation and maintaining it until goal-related stimuli emerge. Proactive control predominantly functions as a form of top-down processing. Given its reliance on top-down processing, paranormal believers tend to exhibit unsatisfactory performance in this cognitive approach. They heavily rely on their experiences and knowledge in decision-making and reasoning, emphasizing a more bottom-up processing approach, which is rooted in perceived physical stimuli moving from lower to higher cognitive levels, such as principles and organizing concepts. Top-down processing relies on high-level cognitive functions, existing knowledge, and previous expectations (Demanez & Demanez, 2003). This processing approach is goal-oriented. Supporting evidence can be found in Ochsner, et al. (Ochsner et al., 2009), which indicated the involvement of occipital, temporal, and parietal lobes in bottom-up processing. Conversely, prefrontal cortex play pivotal roles in top-down processing, which is closely associated with higher-level cognitive processes. Cognitive control represents a critical top-down approach in human cognition. Individuals with robust cognitive control tend to base their decision-making and reasoning on an analytical cognitive style, thus reducing the reliance on intuitive thinking. Some studies have indicated underactivation of the frontal lobe in paranormal believers compared to skeptics (Lindeman et al., 2013; Narmashiri et al., 2019, 2022, 2023a; Wain & Spinella, 2007). Consequently, it can be inferred that the reduced utilization of top-down processing and the weakening of the proactive control strategy in paranormal believers may be linked to lower activity in these brain regions crucial for critical cognitive functions.

Emphasizing bottom-up processing and the reactive control strategy may diminish the influence of top-down processes and the proactive control strategy, potentially creating an imbalance between these two cognitive approaches. The top-down proactive control strategy inherently requires more intentionality and cognitive resources for efficient functioning (Braver et al., 2007). Consequently, over time and under certain conditions like brain disorders (e.g., aging, Alzheimer’s, and schizophrenia), the proactive control strategy may become impaired or compromised. This weakening of the proactive control strategy can have social implications for paranormal believers, affecting their goal maintenance and response preparation. Conversely, a weak reactive control strategy also has limitations in social contexts, potentially leading to perceptual-cognitive biases and hindering the ability to learn new rules. Therefore, the complementarity of these two strategies is crucial, as both are necessary for the social well-being of paranormal believers.

Furthermore, these results suggest a potential link between cognitive control deficits and the frontal lobe’s role in paranormal beliefs. Cognitive control deficits have been observed in disorders associated with impaired DLPFC function (Lopez-Garcia et al., 2016). Research indicates that inadequate DLPFC regulation of input from the midbrain dopaminergic system can lead to deficits (Edwards et al., 2010), with dopamine deficiency playing a crucial role in the pathophysiology and symptoms of schizophrenia (Chuhma et al., 2017). Additionally, individuals with ADHD often face challenges in cognitive control (Wang et al., 2013), as evidenced by deficiencies in the AX-CPT task (Wang et al., 2011). Medications used to treat ADHD are thought to modulate activity in the Substantia nigra (SN) and PFC related to cognitive control (Agster et al., 2011), potentially enhancing dopaminergic system function. This leads to the hypothesis that dopamine and PFC modulation could improve cognitive control in individuals with paranormal beliefs, with tDCS potentially aiding in maintaining regular DLPFC activity (Boudewyn et al., 2019).

4.2. Effect of tDCS on improving DMC in paranormal believers

This study is the first to explore the impact of tDCS on cognitive control in paranormal believers. The results indicate that neuromodulation of the rDLPFC influences DMC in paranormal believers. Interestingly, these findings contradict the study by Braver et al. (Braver et al., 2009), which suggested that the left DLPFC is associated with proactive control. Regarding the effects of tDCS on the Stroop task, Vanderhasselt, et al. (Vanderhasselt et al., 2009) reported that the left DLPFC is actively involved in maintaining attention and the proactive control strategy. Braver, et al. (Braver et al., 2009) investigated brain activity changes in the elderly and found that they tend to utilize reactive control strategies in the AX-CPT task. However, after training to employ the proactive control strategy, an increase in left DLPFC activity was observed in the early stages. Additionally, Boudewyn, et al. (Boudewyn et al., 2019) demonstrated that anodal tDCS on the left DLPFC enhances the proactive control strategy. These results appear inconsistent with the DMC theory, which associates the proactive control strategy with DLPFC activation.

It’s worth noting that while tDCS does modulate activity in the stimulated region, it can also have widespread effects on the brain network (Corlier et al., 2019; Tik et al., 2017). Neuroscience studies consistently highlight the crucial role of cognitive control in regulating behavior and cognition. Moreover, recent research suggests that both proactive control and reactive control strategies are pivotal mechanisms underlying tendencies toward paranormal beliefs and related phenomena. Furthermore, the DLPFC is considered vital for inhibition (Vanderhasselt et al., 2009), although its exact role, especially the rDLPFC, remains a topic of debate. In this context, comprehending the rDLPFC’s role in proactive control and reactive control becomes paramount for a deeper understanding of their implications in paranormal beliefs. Several studies support the involvement of the DLPFC in reactive control and proactive control strategies using the AX-CPT task (Gomez-Ariza et al., 2017; Gómez-Ariza et al., 2017; Pulopulos et al., 2020).

In accordance with the DMC framework, the proactive control strategy is associated with stable and early DLPFC activation, while the reactive control strategy manifests as transient DLPFC activation during conflict detection (Braver, 2012). Several studies have indicated rDLPFC involvement in both proactive control and reactive control strategies during the AX-CPT task (Holmes et al., 2005; MacDonald & Carter, 2003; Paxton et al., 2008). Since tDCS enhances rDLPFC excitability in both proactive control and reactive control, its stimulation may impact both strategies. Our findings highlight the essential role of rDLPFC in cognitive control and paranormal believers’ strategies. Additionally, previous studies using a similar task version have shown that participants exhibit slower responses in AY trials compared to other trials, a typical result in healthy adults (Gonthier et al., 2016). Given these findings, it’s plausible that tDCS could affect the proactive control strategy due to the adoption of the reactive control strategy by paranormal believers during the AX-CPT task, resulting in impaired proactive control. This prompts the question of whether applying tDCS to the right rDLPFC in a modified AX-CPT task aimed at encouraging reactive control adoption could significantly influence the proactive control strategy. Further research is warranted to explore tDCS effects on both left and right DLPFC in relation to proactive control and reactive control strategies across AX-CPT task iterations among paranormal believers.

Lastly, the study has some limitations. Firstly, the study design lacked blinding, potentially introducing bias. Secondly, conducting post-electrical stimulation brain marker investigation via neuroimaging techniques was not feasible. Regarding women’s participation in this study, it’s worth noting that prior research has suggested hormonal influences, like estrogen and progesterone (Amin et al., 2006; Solis-Ortiz et al., 2004), on cognitive task performance associated with the prefrontal cortex. Consequently, future studies may benefit from considering menstrual cycle control for female participants.

In summary, it reveals that paranormal believers lack proactive control and demonstrates the effectiveness of tDCS in enhancing rDLPFC activity. These findings underscore the pivotal role of rDLPFC in proactive control and reactive control strategies, paving the way for future investigations into the cognitive mechanisms of paranormal beliefs and tDCS’s impact on cognitive control. The results provide strong evidence for rDLPFC’s involvement in cognitive control among paranormal believers and suggest that rDLPFC stimulation boosts cognitive resource allocation. This study offers valuable insights for further research on the neural mechanisms underlying tDCS’s beneficial effects on cognitive functions related to paranormal beliefs.

Acknowledgments

We thank the participants for their cooperation in this research.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Additional information

Funding

This study was supported by the Cognitive Sciences and Technologies Council (CSTC) (https://cogc.ir/?lang=2 Grant No. of 8349).

Notes on contributors

Abdolvahed Narmashiri

Abdolvahed Narmashiri, delves into diverse fields of neuroscience. His research explores the neural basis of value-based search efficiency, the psychological dimensions of paranormal beliefs, and their influence on cognitive functions.

Javad Hatami

Javad Hatami, a Professor of Psychology at the University of Tehran, is an accomplished academic with a focus on cognitive science, behavioral neuroscience, and developmental psychology.

Reza Khosrowabadi

Reza Khosrowabadi, an Assistant Professor at the Institute for Cognitive and Brain Sciences at Shahid Beheshti University, has research interests that span biosignal processing, pattern recognition, multimodal imaging, and cognitive modeling. Here are some highlights from his academic journey.