Physiology predicting ideology: The relationship have not fully replicated in more recent, well-powered replications

Physiology predicts ideology. Or does it? The current state of political psychophysiology research. Kevin B Smith, Clarisse Warren. Current Opinion in Behavioral Sciences, Volume 34, August 2020, Pages 88-93. https://doi.org/10.1016/j.cobeha.2020.01.001

Political scientists are increasingly adopting psychophysiological research modalities to investigate the biomarkers of political attitudes and behavior. A good deal of this research focuses on the sympathetic branch of the autonomic nervous system. This makes a good deal of sense as psychophysiologists have long associated the sympathetic nervous system (SNS) with the sorts of implicit emotional-cognitive processing theorized to underpin a range of political attitudes. This review assesses the literature examining the relationship between political attitudes and individual-level variation in SNS activation, especially in response to disgust/threat stimuli where non-physiological research provides the basis for a strong a priori hypothesis for the existence of such a relationship. The empirical record for this relationship proves to be mixed, with a number of studies supporting the base theoretical expectations, but failed replications and questions about what is actually being measured also raising questions about the generalizability of the findings.

Introduction

A considerable research literature suggests that political attitudes and behaviors are genetically, which is to say biologically, influenced. Well-powered analyses using a variety of methodological approaches—including twin studies, adoption studies, and genome wide association studies—converge on the inference that a non-trivial amount of variation in political beliefs and behaviors systematically maps onto genetic variation [1, 2, 3, 4]. While the evidence of biological influences on political phenotypes is persuasive, the specific downstream mechanisms explaining this link have only recently begun to be systematically investigated. Ideology is a complex social phenotype. Its heritable components are almost certainly polygenic in nature, and the biological mechanisms that presumably mediate between genes and political beliefs almost certainly interact with environmental influences in ways that are far from fully understood. How does biology actually influence political traits? Currently, the only honest answer to this question is that we are not completely sure.

While no comprehensive, universally accepted answer exists to this question, a rough theoretical model emerged over the past decade or so to guide investigations of the link between biology and ideology. Succinctly, this model assumes genetic variation leads to individual-level differences in the physiological systems that not only play a key role in extracting and processing information from the external environment, but also in generating automatic emotional and behavioral responses to a given environmental situation or stimuli. The essential idea is that genes build biological information processing systems, there is individual-level variation in those systems due to both genetic and environmental influences, these individual-level variations lead to differences in implicit emotional and cognitive responses to environmental stimuli, and those differences reflect physiologically instantiated predispositions that at least partially drive political preferences [5, 6, 7].

An obvious general hypothesis generated by this framework is that individual-level differences in physiological responses to particular stimuli should predict political beliefs and behavior. An initial tranche of mostly small-N studies focusing on the sympathetic branch of the autonomic nervous system (ANS) reported evidence of exactly such links, though the reported physiology/ideology relationships centered more on certain attitudes associated with social conservatism rather than ideology more broadly conceived. Those relationships, however, have not fully replicated in more recent, well-powered replications. This raises questions not just about the specific relationships tested, but the broader theoretical framework generating the hypotheses.

We review this research and, especially in light of recent replication failures, examine implications for future research. We suggest one way forward is to narrow both the theoretical and empirical approach, focusing on physiological responses to more narrowly targeted stimuli and how those responses do or do not predict more specific political attitudes.

The ANS as a basis to investigate Links between physiology and ideology

The primary purpose of the ANS is to maintain homeostasis between external and internal environments through the regulation and coordination of bodily functions like digestion, respiration, and cardiac activity. These regulatory functions are largely automatic and implicit and occur outside of conscious awareness. The ANS has two primary branches, the sympathetic nervous system (SNS; colloquially known as the ‘fight or flight’ system) and the parasympathetic nervous system (PNS; the ‘rest and digest’ system).

Most research targeting the ANS to investigate the physiological underpinnings of ideology has focused on the SNS. This is in no small part because of the availability of a well-validated measure of SNS arousal, electrodermal activity (EDA), that can be obtained reliably and relatively cheaply using little more than a bioamplifier and sensors capable of capturing skin conductance.

The SNS represents a good target for testing links between physiology and ideology because psychophysiologists have long associated SNS activity with various aspects of automatic emotional-cognitive processing [8]. Some of those processes are, prima facie, good candidates to map onto political attitudes and behaviors. For example, recent literature reviews suggest the attitudinal and behavioral differences between liberals and conservatives are systematic, are anchored in traits such as negativity bias and in-group versus out-group bias, and almost certainly have a neurobiological basis [9]. That neurobiological basis clearly encompasses the ANS because it is already associated with some of these traits. It is well established, for example, that negative stimuli evoke a greater SNS response than non-negative stimuli, and this includes stimuli such as negative news stories that are relevant to politics [10, 11, 12, 13]. So variation in skin conductance seems to reliably capture (among other things) individual-level variation in negativity bias, and individual-level variation in negativity bias is widely hypothesized to systematically co-vary with political beliefs. This is all consistent with the hypothesis that SNS response to particular types of negatively valenced stimuli will predict political beliefs.

Exactly such arguments have already been made in relation to specific categories of aversive stimuli, especially threat and disgust. Numerous studies using non-physiological (self-report) measures have repeatedly found disgust and threat sensitivity correlate with political attitudes [14, 15, 16, 17, 18, 19], and a meta-analytic review encompassing 134 samples from 16 countries concludes there is consistent evidence that both subjective perceptions and objective experiences of fearful or threating stimuli correlate with conservatism [20]. Theoretically, this relationship is assumed to be anchored in evolved implicit processes [21]. In short, there is consistent and persuasive evidence that variation in threat and disgust responses as captured by self-report batteries is predictive of political attitudes. As there is little doubt such stimuli also evoke SNS arousal [17,22], a clear physiology-based hypothesis is suggested, that is, that individual-level variation in EDA response to such stimuli should predict political beliefs. There now exists a fairly extensive literature focused on testing these sorts of hypotheses that, in effect, combine what is known about how responses to non-political stimuli map onto political beliefs, and how the SNS is known to respond to similar sorts of stimuli.

We analyze whether the positive relation between education and health is causal; no causal effect found; result is extremely robust to changes in the main specification and using other databases

Education and adult health: Is there a causal effect? Pedro Albarrán, Marisa Hidalgo-Hidalgo, Iñigo Iturbe-Ormaetxea. Social Science & Medicine, February 7 2020, 112830. https://doi.org/10.1016/j.socscimed.2020.112830

• We analyze whether the positive relation between education and health is causal.
• We combine multi-country data from two cross-sections of EU-SILC.
• We use exogenous variation in compulsory schooling induced by school laws.
• We find no causal effect of education on any of our several health measures.
• The result is robust to changes in the main specification and using other databases.

Abstract: Many studies find a strong positive correlation between education and adult health. A subtler question is whether this correlation can be interpreted as a causal relationship. We combine multi-country data from two cross-sections of the European Union Statistics on Income and Living Conditions (EU-SILC) survey and use exogenous variation in compulsory years of schooling across countries and cohorts induced by compulsory schooling laws. We find no causal effect of education on any of our several health measures. This finding is extremely robust to different changes in our main specification and holds using other databases. We discuss different explanations for our results.

Keywords: HealthEducationInstrumental variables

The Manosphere, a conglomerate of predominantly Web-based misogynist movements roughly focused on men's issues, is characterized by a volume of hateful speech appreciably higher than other Web communities

From Pick-Up Artists to Incels: A Data-Driven Sketch of the Manosphere. Manoel Horta Ribeiro, Jeremy Blackburn, Barry Bradlyn, Emiliano De Cristofaro, Gianluca Stringhini, Summer Long, Stephanie Greenberg, Savvas Zannettou. arXiv Jan 21 2020. https://arxiv.org/abs/2001.07600

Abstract: Over the past few years, a number of "fringe" online communities have been orchestrating harassment campaigns and spreading extremist ideologies on the Web. In this paper, we present a large-scale characterization of the Manosphere, a conglomerate of predominantly Web-based misogynist movements roughly focused on men's issues. We do so by gathering and analyzing 38M posts obtained from 7 forums and 57 subreddits. We find that milder and older communities, such as Pick Up Artists and Men's Rights Activists, are giving way to more extremist communities like Incels and Men Going Their Own Way, with a substantial migration of active users. We also show that the Manosphere is characterized by a volume of hateful speech appreciably higher than other Web communities, including those, such as Gab, known to be hateful. Overall, we paint a comprehensive picture of the evolution of the Manosphere, highlighting the relevance of studying these communities using computational social science methods.

Check also Inside the World of ‘Femcels.’ Isabelle Kohn. Mel, Feb 10 2010. https://melmagazine.com/en-us/story/femcels-vs-incels-meaning-reddit-discord

And Male Sexlessness is Rising, But Not for the Reasons Incels Claim. Lyman Stone. Institute of Family Studies, May 2018. https://www.bipartisanalliance.com/2018/05/male-sexlessness-is-rising-but-not-for.html

And Does anyone have the right to sex?Amia Srinivasan. London Review of Books, Vol. 40 No. 6 · 22 March 2018, pages 5-10. https://www.bipartisanalliance.com/2018/03/does-anyone-have-right-to-sex.html

Moral incongruence, self-perceived addiction and problematic pornography use

Lewczuk, K., Glica, A., Nowakowska, I., et al. Evaluating Pornography Problems Due to Moral Incongruence Model. J Sex Med 2020;17:300–311. https://www.sciencedirect.com/science/article/pii/S1743609519317837

Abstract
Introduction  To date, multiple models of problematic pornography use have been proposed, but attempts to validate them have been scarce.

Aim  In our study, we aimed to evaluate the Pornography Problems due to Moral Incongruence model proposing that self-appraisals of pornography addiction stem from (i) general dysregulation, (ii) habits of use, and (iii) moral incongruence between internalized norms and behavior. We investigated whether the model can be used to adequately explain the self-perceptions of addiction to pornography (model 1) and a broader phenomenon of problematic pornography use (model 2).

Methods  An online, nationally representative study was conducted on a sample of 1036 Polish adult participants, of whom, 880 declared a lifetime history of viewing pornography.

Main Outcome Measure  The outcomes were self-perceived pornography addiction, problematic pornography use, avoidant coping, frequency of pornography use, religiosity, moral disapproval of pornography, and related variables.

Results  Our results indicated that avoidant coping (an indicator of general dysregulation), frequency of pornography use (indicator of habits of use), and the distress connected with incongruence between own sexual behavior and internalized norms, attitudes and beliefs positively contributed to self-perceived addiction (model 1) as well as problematic pornography use (model 2). This broadly confirms the basic shape of the PPMI model. There were, however, notable differences between the models. Moral incongruence related distress was only weakly related to self-perceived addiction (β = 0.15, P < .001), with a stronger relation for problematic pornography use (β = 0.31, P < .001). When controlling for other factors, religiosity weakly predicted problematic pornography use (β = 0.13, P < .001), but not self-perceived addiction to pornography (β = 0.03, P = .368). Frequency of pornography use was the strongest predictor of both self-perceived addiction (β = 0.52, P < .001) and problematic pornography use (β = 0.43, P < .001).

Clinical Implications  Factors proposed within the PPMI model are distinctly relevant intervention targets, and they should be considered in the process of diagnosis and treatment.

Strengths & Limitations  The presented study is the first to evaluate PPMI model. Its main limitation is that it has a cross-sectional design.

Conclusion  The PPMI model is a promising framework for investigating the factors related to self-perceived addiction and problematic pornography use. Despite the differences between the models and in the strength of specific predictors, (i) dysregulation, (ii) habits of use, and (iii) moral incongruence all uniquely contribute to self-perceived addiction and problematic pornography use.

Discussion

The presented work is one of only a few attempting a nonfragmentary assessment of the validity of any model of pornography addiction, problematic pornography use or problematic sexual behavior, and the first to do so for the PPMI model. On a general level, our results confirmed the appropriateness of the basic shape of the model to depict the structure of predictors of both self-perceived addiction to pornography (model 1, Figure 1) and problematic pornography use (model 2, Figure 2). However in some places, our results diverge from the predictions stemming from the model, and there are at least several specific but important issues that require consideration as well as have potential implications for the shape of the model and future research.

As described previously, the analysis reported in the present study was based on 3 paths proposed within the PPMI model: dysregulation path (as indicated by avoidant coping), habits of use path (indicated by frequency of pornography use) and moral incongruence path (operationalized by religiosity, moral disapproval of pornography use, and moral incongruence–related distress). Overall, the results showed that all 3 paths uniquely and significantly contribute to explaining both self-perceived addiction and a broader set of symptoms that fall under the label of problematic pornography use. Moreover, our results confirmed that problematic pornography use symptoms are distinct from simple declarations of being an addict. The correlation between these 2 constructs was r = 0.55. Based on our results, none of the 3 paths postulated within the model can be reduced to the other or eliminated without deterioration in the quality and predictive value of the model. This confirms the basic prediction stemming from the PPMI model.³ Estimated models explained a significant portion of variance in self-perceived addiction (33.9%, model 1) and problematic pornography use (35.9%, model 2).

Conclusions regarding each of the 3 paths of the model are delineated in the following section.

Moral Incongruence Path

People experiencing moral incongruence–related distress reported higher levels of self-perceived addiction and problematic pornography use. This confirms the prediction of the authors of the PPMI model^3,31 regarding the role that moral incongruence plays in shaping the self-appraisals of self-perceived addiction²⁴ and extends it to more general problematic pornography use symptoms. However, the prediction of the model is that moral incongruence should be the stronger predictor of self-perceived addiction to pornography than frequency of use and dysregulation,^3,31 which is not confirmed by our findings. Our results are more in line with recent work showing that frequency of pornography use is a stronger predictor of self-perceived addiction to pornography than moral incongruence²⁶ (refer also to the study by Lewczuk et al²⁷ for an analysis conducted on the same sample as the present study). It is also possible that the lower impact of moral incongruence–related distress on self-perceived addiction is at least partially caused by a slightly lower level of moral disapproval of pornography use in the current Polish sample, compared with, for example, a representative sample of U.S. adults.²⁵ In our study, 20.5% of participants who used pornography in their lifetime agreed that pornography use is morally wrong (answer options ranged from “somewhat agree” to “strongly agree”), while the same answer was given by 24% of Americans. Moreover, based on the same measure, U.S. participants declared to be slightly more religious on average (M = 4.10, SD = 1.95²⁵) than Polish participants in the current sample (M = 3.81, SD = 1.84), which may also explain the weaker impact of the moral incongruence path on self-perceived pornography addiction than the PPMI model based mostly on research performed on the U.S. predicts.

In addition, moral incongruence–related distress was more strongly connected to problematic pornography use than to self-perception of addiction. A possible explanation for this pattern is that, compared with self-perceived addiction, problematic pornography use encompasses a broader group of cognitive and affective consequences and determinants of pornography use. One of them is increased levels of guilt regarding pornography use, which can be a consequence of moral incongruence.²⁰ One of the 5 statements in the BPS,³⁴ which was an indicator of problematic pornography use in our study, reads “You continue to use pornography even though you feel guilty about it.” The relation between self-labeling as an addict and moral incongruence–related distress is theoretically not as close as in other studies, which is reflected by our findings.

Next, our results generally confirmed the specifics of the chain of influence between morality-related variables, although not without a caveat. More religious people were more inclined to see pornography use as morally reprehensible and were more prone to experiencing feelings of incongruence between own sexual behavior and adopted beliefs, attitudes, and norms. The impact of religion was not strong in these cases, as our method of measuring it does not directly invoke a religious context (see the Introduction section for more information on this issue). As expected, distress connected to behavior-attitudes misalignment was determined by 2 additional factors: frequency of the behavior (frequency of pornography use) and restrictiveness of the attitudes (moral disapproval of pornography; refer to the study by Grubbs et al³). However, although religiosity and moral disapproval significantly predicted moral incongruence–related distress, their contribution was somewhat limited. Other possible predictors should be investigated, both connected to other sources of norms that can determine disapproval of pornography, for example, sociopolitical views, religious fundamentalism^53,54 or certain branches of feminism,⁵⁵ as well as variables related to the awareness and sensitivity to own behaviors being incongruent with own beliefs, attitudes, and internalized norms (eg, self, awareness, concern over mistakes, perfectionism, the centrality of the norms that motivate attitudes toward pornography and sexuality). Here, we echo the suggestions that were voiced by other authors in their commentaries for the model.^19,22

In addition, our results showed that, controlling for other variables, more religious people declared higher levels of problematic pornography use. The influence of religiosity on problematic pornography use was weak, but present—which is in agreement with at least a significant portion of previous studies showing a weak, positive relationship between religiosity and problematic pornography use symptoms^25,26 (refer also to the study by Lewczuk et al²⁷). A corresponding relation was not found for self-perceptions of addiction.

Habits of Use Path

Frequency of pornography use was the strongest predictor of self-perceived addiction in model 1 and of problematic pornography use in model 2. This indicates that the self-appraisal of pornography-related problems does not merely rely on perceiving this behavior as transgressing one's personal norms, that is, it is not a function of mere convictions (refer to the discussion in the study by Humphreys⁵⁶). A significant portion of the variance is better explained by the frequency of use, which validates the disorder model of problematic pornography use and is similar to the symptomology of at least some cases of substance use disorders and other behavioral addictions, for which excessive use during at least some part of the course of the disorder is a definitional criterion (refer to the study by Kraus et al¹ and Potenza et al⁵⁷). Frequency of pornography use was also a significant predictor of problematic pornography use, although its influence was slightly weaker than for self-perception of addiction (β = 0.43 vs β = 0.52). This is understandable, given that problematic use has a broader scope than self-perception of addiction, encompassing not only excessive pornography use but also loss of control, using pornography as a coping mechanism and guilt connected to pornography use.³⁴

Dysregulation Path

Avoidant coping style was an indicator of dysregulation in our model. People using an avoidant coping style more frequently were also more inclined to see themselves as pornography addicts and had a higher severity of symptoms of problematic pornography use. This is in line with previous research, which showed the specific importance of an avoidant coping style for problematic sexual behavior.39, 40, 41 This result is also in agreement with studies showing that engagement in sexual behaviors itself can constitute an avoidance strategy (eg, avoiding negative emotions associated with other areas of one's life). However, the impact on avoidant coping for both dependent variables was weak (β = 0.15, P < .001) and was not stronger for problematic pornography use than for self-appraisals of addiction. This can be considered surprising, as problematic pornography use has a pornography-as-coping component (“You find yourself using pornography to cope with strong emotions, eg, sadness, anger, loneliness, etc.” is one of BPS items that operationalized problematic pornography use in our study).

Implications for the Shape of the Model and Future Research

Our findings indicate that the PPMI model can serve as a general model of factors contributing to self-perception of pornography addiction and problematic pornography use. However, the dysregulation path is underdeveloped in the current version of the model. This has also been pointed out by other researchers.¹⁶ This path should be delineated with more detail and extended. In their initial proposition of the model, Grubbs et al³ focused on moral incongruence–related factors describing the dysregulation path with less detail. This approach is understandable as moral incongruence is a central focus of the model. However, as a consequence, the current conceptualization of the PPMI model places all dysregulation-related factors (such as emotion dysregulation, impulsivity, coping, compulsivity) into one general and unspecified category and abstains from depicting mechanisms of influence between these variables, ascribing them differential degrees of importance or depicting relations between dysregulation-related variables and moral incongruence–related variables. Such relations have been proposed by others^16,22 and are also visible in our analysis, as avoidant coping was connected to moral incongruence–related distress (r = 0.21, P < .001) possibly indicating that avoidant coping strategies can serve as a way of dealing with moral incongruence.

As the PPMI model was initially validated in the present study, we postulate that it should be extended and possibly reshaped into an even more ambitious, general model in which dysregulation-related variables will be treated with the same degree of carefulness as morality-related ones. To achieve this, specific models—such as the current version of the PPMI model—should be merged with broader models (eg, I-PACE model^12,58) that go into more detail regarding behavior dysregulation–related factors, but, as of now, neglect the role of morality-related variables. It seems that only this approach would allow for the full picture of factors influencing both lay self-perceptions of addiction and problematic pornography use to be accounted for. These 2 branches of research should not and cannot develop separately because of their possible mutual influence.^16,22 Because of this interdependence, the shape of the moral incongruence path cannot be definitively established when the dysregulation-related side of the model is underdeveloped.

In future studies, other indicators of general dysregulation (eg, impulsivity, maladaptive emotion regulation, perfectionism) should be tested within the PPMI model to extend and provide further support for the discussed framework. Such an extension seems to have been predicted and welcomed by the authors of the model,³¹ which we fully agree with.

Another issue worth pointing out is that our analysis is based on a populational sample. One of the important future directions for further research is to also verify the model based on clinical samples, experiencing a clinical level of symptoms of problematic pornography. This is crucially important because the significance of factors predicting problematic pornography use can change the clinical level, compared with populational investigations. Future studies should also apply the PPMI model to CSBD recognized in the ICD-11^13,14 when screening measures for this disorder become available for use. We agree with other researchers who suggested studying behavior-norms misalignment for sexual behaviors other than problematic pornography use,²⁰ which may lead to an extension of the model to explain general problematic sexual behavior symptoms.

Additional concerns about the issue of operationalizing moral incongruence vs moral disapproval of pornography use (see Material and Methods section) and self-perceived addiction vs disordered pornography use based on formal clinical definitions (such as problematic pornography use, see the Introduction section) were noted in the earlier parts of the manuscript.

The current research extends research on the PPMI model to another cultural context, namely, Polish participants. However, Poland shares cultural similarities with the United States as it is a predominantly Christian country (77.3% of participants in the current analysis declared being Catholic). Future research should further validate the model, based on different religious and cultural circles.

Limitations

Some of the limitations of the present study were already noted (single dysregulation-related factor). We also note that the present work is based on cross-sectional research design, which precludes analyses of directionality or causality. That is, although the present work is consistent with the PPMI, without longitudinal observations that examine trajectories of these variables over time, it is impossible to conclusively evaluate any model of problematic pornography use. Finally, we did not include a definition of pornography for the participants in the online survey.