Reforming Sectarian Beliefs in Iraq: Winning the Peace

Sectarian violence continues in Iraq affecting regional and world security. Neuroscience techniques are used to assess the mentalizing process and counter-arguing in response to videos designed to prevent extremist radicalization. Measurement of neural activity in brain Regions of Interest (ROI) assists identification of messages which can promote favorable behavior. Activation of the Medial Prefrontal Cortex (MPFC) is associated with message adoption and behavior change. Public Service Announcements (PSAs) have not been effective in reducing violence in Iraq. This study demonstrates that the four PSAs investigated in this study do not activate the MPFC. The RLPFC is a brain ROI associated with counter-arguing and message resistance. This study demonstrates that reduction in activity in the Right Lateral Prefrontal Cortex (RLPFC) is associated with decreased sectarianism. Engagement was measured and is associated with activity in the frontal pole regions. We introduce Functional Near-infrared Spectroscopy (fNIRS) to measure the neural activity of highly sectarian Iraqis in response to these anti-sectarian messages. Neural activity was measured while viewing three PSAs and a fourth unpublished video. All four videos are intended to reduce sectarianism. A novel sectarianism scale is introduced to measure sectarian beliefs before and after the messages. This sectarian scale has high internal consistency as measured by Cronbach’s alpha. Measured activation of brain ROIs are correlated with changes in the sectarian scale. Twelve Sunni and twelve Shi’a Iraqis participated in the study. Subjects were shown the four videos in randomized order, while equipped with a fNIRS neural imaging device. All four videos produced significant engagement. None of the videos reduced sectarianism nor caused brain activation of www.scholink.org/ojs/index.php/csm Communication, Society and Media Vol. 4, No. 1, 2021 9 Published by SCHOLINK INC. adoption. This is consistent with the widely held Iraqi public perception that the PSAs are ineffective. Only one video, which was un-published, caused reduced sectarian beliefs. This un-published fourth video was associated with decreased counter-arguing. Counter-arguing is associated with message resistance.

study. Of the 24 participants, 12 were Sunni Muslim and 12 were Shi'a Muslim. All subjects were assessed to hold high sectarian bias using the sectarianism scale described below. Participants were recruited by phone to participate in a video marketing study. The authors obtained permission to use the data for secondary analysis for the purpose of this paper. Sample size was determined based on the maximum number of participants that could conduct the study, given limited time, facilities, and resources. A priori power analysis using effect sizes in previous social neuroscience work with type I error of 5 percent and 80 percent power suggests a sample size of 14 participants.

Materials
Neural measurements were conducted with a NIRSport imaging unit, which is a Functional Near-infrared Spectroscopy (fNIRS) device manufactured by NIRx (nirx.net/nirsport). This device was selected because of its small size, portability, and greater ecological validity than other neural imaging alternatives. The NIRSport unit has eight light sources and eight light detectors configured to measure the prefrontal cortex over 23 channels of data. One drawback of the NIRSport is limited spatial coverage and the inability to measure activity in deep brain structures. Measurements include the medial prefrontal cortex, frontal poles, dorsal and lateral prefrontal cortex. These regions in the outer cortex and are several centimeters in size, within the measurement resolution of fNIRS of 1-3 cm. Four headcaps of different sizes where used to affix the light sources and detectors to participants. Spatial positioning was standardized using the 10-10 UI external positioning system. Light intensity was emitted at 760 nm and 850nm wavelengths with a 7.81Hz sampling rate. Four video commercials intended to reduce viewer sectarianism using different persuasion tactics were used a stimulus in this study.
Video One, labelled "Chess", is an animated chess game. Opposing pieces represent Al-Qaeda and Shi'a militia. The animated chess pieces from opposing sides fight each other with graphic violence.
The video ends with the two militia leaders (Kings) walking off the board together. This video suggests participants in sectarian violence are being a used as pawns in a chess game. This is meant to be a shocking and provocative appeal, motivating viewers to challenge their sectarian beliefs and seek more information. Video Two, labelled "Nice", shows two elementary school boys as friends. The boys pray together. Time is forwarded to show the friends grown as adults. It becomes apparent that one friend is Sunni and the other is Shi'a. This demonstrates the similarity between Sunni and Shi'a sects. This is intended as a positively framed message encouraging sectarian unity. Video Three, labelled "Grad", shows the graduation of a group of students. A young female graduate and her proud mother are portrayed. Then there is an explosion. The video switches to the graduate's mother crying over her deceased daughter. The video narrates "ISIL destroys families". This is meant as a shocking and provocative appeal. Video Four, labelled "Basheer", is a monologue by popular Iraqi comedian Ahmad www.scholink.org/ojs/index.php/csm Communication, Society and Media Vol. 4, No. 1, 2021 12 Published by SCHOLINK INC.
Al-Basheer. Basheer argues at length that Iraq should disband Shi'a militias and avoid sectarianism.
This is meant to be a somber and conciliatory appeal, focusing on similarities and Iraqi national identity over sectarian identities.

Procedure
Prior to viewing any videos, all participants completed a survey measuring their level of sectarianism, defined as a bias toward their own sect, Sunni or Shi'a Islam, and a prejudice toward other sects. The survey consisted of eight questions, rated on a scale ranging from 0 (strongly disagree) to 10 (strongly agree Participants were randomly assigned to a permutation, constrained such that for all 6 permutations beginning with a given video, half would be viewed by a Sunni and half would be viewed by a Shi'a.
Immediately following each video, participants were asked to self-report on level of identification, response efficacy, and perceived effectiveness. This was done using a single item, 5-point Likert scale.
Two weeks following initial data collection, the same eight-question sectarianism survey was re-administered to assess change in sectarianism following video intervention.

Neural Data Processing
fNIRS data was pre-processed using NIRSLab. Event markers captured the beginning and end of each video and baseline period. The length of each video is known, so reliable start/stop times for each baseline and video stimuli are recorded. Discontinuities in neural signals were removed. Several of the relevant channels were over-saturated. Four of these were repaired by interpolating over no more than 15 frames (approximately 2 seconds). The others were discarded from analysis. This resulted in the exclusion of 2 participant's data, leaving usable data for 22 participants. Band pass filtering was applied to all signals with a low cutoff frequency of 0.01 Hz and a high cutoff frequency of 0.2 Hz. This effectively removed biometric noise in the neural signal data. A Principal Component Analysis (PCA) algorithm was used to identify and remove motion artifacts from the data. Light intensity was converted into a percentage change in oxygenated hemoglobin concentration relative to baseline using the modified Beer-Lambert law. The first and last 7 seconds of each video stimulus was truncated to account for delayed response due to the participant's hemodynamic response function. Data for each channel were exported to a text file for statistical analysis using R. R is a free, popular, and open source statistical computing and graphics language and environment. Since fNIRS measures a relative frequency, an absolute activity measure was estimated by scaling the neural signal for each fontal pole between zero and one by subtracting the minimum value from the beginning of the baseline period through video completion, then dividing by the range in maximum to minimum over the same time period. For each frame (data point) the left value was divided by the right value and this value was then divided by the sum of the two values. The log of the resulting value was used for calculations.

Figure 1. fNIRS Channels Measuring A Priori Brain Regions of Interest
We focus measurement in three key brain regions of interest. The level of engagement or disgust toward video content was estimated by asymmetry in the frontal poles. People with greater left than right activity experience engagement with stimuli, while people with greater right than left frontal pole activity experience disgust. We use the formula (L/R)/(L+R), where L and R refers to the brain activity in neural channels 11 and 13 respectively as indicated in Figure 1. Message adoption was estimated by activity in the medial prefrontal cortex observed in channel 12. Counter-arguing toward stimulus was estimated by activity in the right lateral prefrontal cortex observed in channel 18.
Four research questions are investigated in this experiment: R1: Are the videos cognitively engaging? We hypothesize that all four videos will cause cognitive engagement in participants as observed by a significant increase in left greater than right frontal pole activation (Channels 11 and 13) during the video compared to the blank screen viewed immediately prior.
R2: Do the videos increase adoption of the message? We hypothesize that the published videos will not increase adoption. Increased adoption will result in decreased sectarianism. Lack of adoption will not reduce sectarianism. Adoption is shown by activity in the medial prefrontal cortex (channel 12).
R3: Do the shocking, provocative videos affect counter-arguing? We hypothesize that the provocative videos, Video 1, the Chess video, and Video 3, the Grad video, will exhibit higher levels of counter arguing. The un-provocative, pleasant videos, Video 2, the Nice video, and Video 4, the Basheer video, will not exhibit counter-arguing. Counter-arguing is shown by activity in the right lateral prefrontal cortex (Channel 18).
R4: Does decreased counter-arguing (Channel 18) increase message integration and adoption (Chanel 12)? We hypothesize that videos with high cognitive engagement and low counter-arguing will exhibit higher activity in the medial prefrontal cortex.
Behavior is not limited to individual regions of interest, but is more reflected in changes in larger circuits. Overall, a pattern of increased engagement, decreased counter-arguing, and increased adoption will likely correlate with decreased sectarianism. The decrease in sectarianism will be measured between the pre-video survey and the two week post-video survey. All statistical tests will apply the student T-test to compare the block average of neural signals while participants viewed stimulus material against those while they viewed blank screens prior to stimulus. The t-test will be corrected for heteroskedasticity.

Results: Summary of Key Findings
Reduced sectarianism is the intended effect of all four videos in this study. The sectarian survey has high Cronbach's alpha equalling 0.78, indicating high internal consistency. There is no observed difference in sectarian score between Sunni and Shi'a. All four videos were engaging (Table 1). There was no effect of engagement on sectarianism ( Table 2). None of the videos shows increased neural activity associated with adoption of new behavior (Table 3). Video 2 shows decreased neural activity association with adoption ( Table 4). None of the videos show an effect of adoption on sectarianism (Table 5). None of the videos show increased neural activity associated with counter-arguing (Table 6).
Videos 2 and 4 show decreased neural activity associated with counter-arguing (Table 7). Only video 4, published. The Basheer video shows decreased sectarianism as a function of decreased counter-arguing (Table 8). We believe this indicates decreased resistance to the message.

Findings
Asymmetric activation of the frontal poles, left greater than right, measured in Channels 11 and 13, indicates engagement. All four videos show significant increase in engagement according to two-sample t-tests comparing baseline and stimulus block-average:

Figure 2. Engagement Measured in Channels 11 and 13
There is no significant effect of engagement on sectarianism in a regression model:  None of the four videos shows increased adoption:  There is no effect of adoption on sectarianism in a regression model: Activation of the right lateral prefrontal cortex, measured in Channel 18, indicates counter-arguing.
None of the videos showed increased counter-arguing ( Figure 4) in a one-tailed, two sample T-test:

Figure 4. Counter-Arguing Measured in Channel 18
Video 2, the Nice video, and Video 4, the Basheer video, shows decreased counter-arguing in a one-tailed, two-sample T-test comparing baseline and stimulus block-averages:  The Basheer video was not a published public service announcement. The Basheer video is the only video to demonstrate the effect of decreased sectarianism. This video is longer and allows the audience to identify with the speaker, a popular comedian and cultural icon, through evocative rhetoric.
One individual scored a statistically significant increase in sectarianism, greater than two standard deviations above the mean. This was a twenty year old Sunni with a bachelor's degree, with no known unusual characteristics compared to the rest of the population.

Discussion: The Results Are Significant
These findings confirm the impression that public service announcements are not effective in reducing sectarian violence. Function of the RLPFC is complex and incompletely understood. Hutcherson describes the interplay between the ventromedial prefrontal cortex and dorsolateral prefrontal cortex in cognitive regulation generating decision-making (Hutcherson, 2012). The ventromedial and lateral prefrontal cortices cross-talk, arguing and counter-arguing back and forth, counter-balancing each other, resulting in normative processing below the level of consciousness. Anatomic differences in these regions affect cognitive processing (Schmidt, 2018). Attitudes and actions regarding sectarianism are a behavioral state. Transformative processes and influences affecting behaviors include engagement, identification, emotion, narrative immersion, reviewing previous experience, internal dialogue, innoculation, and resolving cognitive dissonance. Cognitive dissonance results from conflicting information, beliefs, or feelings, causing discomfort. Individuals tend to resolve the conflict by various mechanisms to achieve consistency to resolve discomfort (Festinger, 1957). Inoculation involves reinforcing previously held beliefs (Banas, 2010). Inoculation helps protect an individual against un-wanted change. The majority of Iraqis already reject violence and terrorism, and tend to view sectarianism as a recent, externally incited phenomenon, caused by the American invasion. The Basheer video is different from the published videos. Basheer is a popular comedian who addresses the audiences directly in a personal manner. The Basheer video is longer than the published videos. Iraqis identify with Basheer who is popular. Affirmation of core values increases the effectiveness of a message (Falk, 2015). Kaplan reports challenges to establish political beliefs can increase activity in the default mode network. The default network mode is associated with self-representation and disengagement from external reality (Kaplan, 2016). The audience relates to Basheer whose message reinforces negative perceptions of violence and terror. Innoculation of established beliefs against sectarian violence could explain the decrease in sectarianism seen in the Basheer video. Soutschek evaluated the role of Dorsolateral Prefrontal Cortex (DLPFC) in decision-making regarding cooperation. Previous studies, including those using functional magnetic resonance imaging, reported a correlation between cooperative behavior and DLPFC activity (Emonds, 2012). Soutschek used trans-cranial magnetic stimulation to disrupt the DLPFC while playing the prisoner's dilemma game. In the prisoner's dilemma game, the subject risks short-term loss, even if the other player has previously defected, to accrue long-term benefit. To make that strategic decision, the subject has to suppress the impulse to defect. This is an important cognitive control process. Disrupting the DLPFC magnetically disrupts that cognitive process. Incapacitating the DLPFC decreased cooperation rates. This was most pronounced when the partner had previously defected (Soutschek, 2015). This experiment demonstrates that the DLPFC contributes to strategic decision-making in the prisoner's dilemma game. Soutschek infers the DLPFC plays a role in weighing long-term versus short term benefits. Prefrontal cortex is a relatively late evolutionary development, moderating behavior, allowing mutual altruism, among non-relatives, as a basis for cooperation and survival (Soutschek, 2015). "Few things are as fundamental to human www.scholink.org/ojs/index.php/csm progress as our ability to arrive at a shared understanding of the world" (Kaplan, 2016).

Limitations: Several Concerns Limit the Ability to Generalize These Results
"Functional Magnetic Resonance Imaging (fMRI) has become the gold standard for in vivo imaging of the human brain. In practice, fNIRS is a more convenient and less expensive technology than Fmri" (Cui, 2011). fNIRS correlates well with fMRI (Cui, 2011). fNIRS is limited relative to fMRI. fNIRS lacks the spatial discrimination of fMRI. The MPFC has voxels of function identified below the level of fNIRS discrimination. For example, activation of the MPFC predicts behavior change, whereas activation of the Dorsomedial Prefrontal Cortex (DMPFC) is more specifically associated with narrative immersion. This is a significant limitation since narrative immersion plays a key role in decision-making. The Ventromedial Prefrontal Cortex (VMPFC) plays multiple, varied roles including social, cognitive, and affective functions (Hiser, 2018). Confirmatory evaluation by fMRI may be necessary for better spatial resolution. Similar limitations likely affect other ROI including the RLPFC.
fNIRS has a decreased Signal-to-noise Ratio (SNR) than fMRI. fNIRS can not access deep brain structures, whereas fMRI can (Cui, 2011).
Measuring behavioral change is problematic. All subjects were male and education levels varied, although this was not controlled. This study assumes a relationship between extreme beliefs and the capacity, willingness, and propensity to commit violence. Implicit is the assumption that reducing sectarianism will reduce sectarian violence. The sectarian survey has not been validated or used in other settings. The study design may have inappropriately increased engagement scores. Videos were contrasted with a blank screen. Instead of a blank screen between videos, it might be useful instead to use videos with some neutral content. The videos do not generate a large range of signal strength to facilitate discrimination. Future studies should create a greater range of output values across input stimuli. The effects of one video lasts well past the baseline/rest period and into the subsequent stimuli.
This may contaminate findings. Dynamic changes occur within the span of a single video, between videos, and beyond. Subsequent events, including sleep, may affect integration of changes.
It is reasonable to question whether any single, brief intervention may change deeply rooted, inter-generational beliefs and biases. Motivational states play a role in behavior change. Falk's smoking cessation studies involved subjects who expressed a desire to change behavior to quit smoking.
Motivation was not addressed in this study. The subjects in this study are highly sectarian, and in focus groups did not express any desire to change that fact. It is reasonable to infer that the subjects were not likely motivated to decrease sectarianism. One subject significantly increased sectarian beliefs for unknown reasons. Emotion plays a role in prefrontal cortical, cognitive processing (Dor'e, 2019*), (Kaplan, 2016 The authors hope this investigation will stimulate discussion and research regarding sectarian violence in Iraq and elsewhere. Future areas of inquiry include exploring and refining methods to study how social messages lead to behaviors, including organized violence. Studying violent acts of terror is limited by ethical and practical constraints. Surrogate violent behavior, such as sport-related fighting, may be an appropriate venue to study violent beliefs and behavior. Naturalistic and retrospective investigations might be useful to gain insight in designing more rigorous studies. Future studies might include characterizing the mental transition from sectarianism to planning to implementation of violence. Negative behaviors may be countered, redirected, and reformed by positive messages which activate the brain. Persuasive messages can change behavior. We seek to reduce sectarian-inspired violence in Iraq. Messages should be legitimate, transparent, and respectful (Aly, 2014). Iraq has an ancient, venerated history of heterogeneous cultures. Iraq has endured millenia of invasions and decades of brutal dictatorship, war, and domestic violence. It would be useful to learn how social media promotes radicalization. Enduring peace, based on pluralistic tolerance, is worthy of our best efforts.
We seek to honor the commitment and sacrifices of those who served to secure the future and freedom of Iraq. The impact of sectarianism in Iraq, the Middle East, and the world, merit further inquiry.

Conclusion
Public Service Announcements (PSAs) do not reduce sectarian violence in Iraq. This experiment measures neural activity in subjects viewing PSAs against sectarianism. Neural activity is correlated with changes in sectarianism. Infrared spectroscopy measures activity in three key brain regions of interest. Activity in these brain regions involve: (1) getting attention, (2) changing behavior, and (3) resisting behavioral change. Three PSA videos fail to activate behavioral change and have no effect on sectarianism. A fourth, un-published video decreases resistance to behavioral change, decreasing sectarianism. This suggests a potential relation between the brain region which changes behavior and the region which resists behavioral change. Messages which activate new behavior need to identified and studied in relation to their effect on sectarianism. Infrared spectroscopy is a convenient, inexpensive, and effective tool to identify persuasive messages to reduce sectarian violence in Iraq.
Cost effective analysis of brain activity can assist selection of messages which may go viral to reduce sectarian violence. This may be a useful tool to ethically influence behaviors which reduce violence and enhance security.