Blog Image

Artur Nilsson's research blog

New research on bullshit receptivity

Comments on new research Posted on Mon, July 08, 2019 01:40:23

The notions of ”alternative facts” and fake news have rapidly become viral. Although research on receptivity to falsehoods is useful, there is also a problem here. These notions are often used for ideological rather than scientific purposes—the real facts of the ingroup tribe are pitted against the lies of the other tribes. We need more research that focuses not on what facts people subscribe to but on how they engage with evidence and arguments, and how to promote a more scientific (as opposed to ideological or tribalist) attitude among the public.

One interesting new line of research focuses on the notion of receptivity to bullshit, which the philosopher Harry Frankfurt famously defined (in his book ”On bullshit”) as a statement produced for a purpose other than conveying truth (e.g., persuading or impressing others).

One type of bullshit is that which emerges when someone does not really know the answer to a question but tries to say something that sounds convincing anyway in order to come off as competent. An example is the student who is bullshitting to try to pass an exam by trying to write something that sounds good to fool the teacher. This is a type of bullshit focused on self-promotion. It has been addressed a recent paper by Petrocelli (2018).

Another type of bullshit is the political bullshit. This is the type of bullshit that results when a person says whatever s/he can to place his or her own party or ideology in the best light possible and persuade others or convince them to join him or her. This is the type of bullshit that often makes political debates and opinion journalism so predictable and boring—facts are tortured and twisted to fit into an ideological ”box”, and the whole thing is more a game of trying to ”score” a goal on the opposite team and getting cheered on by your own team than a serious engagement in a rational debate in which you are open to pursuing the truth and learning something new. This type of bullshit is focused on promoting an ingroup cause or ideology rather than the self.

It is, however, the pseudo-profound bullshit that has been the main focus on recent research.


Receptivity to pseudo-profound bullshit

Pseudo-profound bullshit is composed of sentences designed to sound intellectually profound, through the use of buzzwords and jargon, that are actually vacuous. This type of bullshit has a long history in intellectual (or pseudo-intellectual) circles. There has even been a culture of bullshitting in some academic circles, particularly in some quarters of continental and postmodern philosophy. For instance, see this funny Youtube-clip, in which the philosopher John Searle recounts a conversation in which the famous postmodernist Michel Foucault says that in Paris you need to have at least 10% that is incomprehensible in your writings to be considered a serious and profound thinker. The postmodern movement was also the target of the infamous hoax perpetrated by the physicist Alain Sokal, who was able to publish an article crammed with bullshit in a leading postmodern journal. This is how Sokal described the article when he made the hoax public:

I intentionally wrote the article so that any competent physicist or mathematician (or undergraduate physics or math major) would realize that it is a spoof … I assemble a pastiche — Derrida and general relativity, Lacan and topology, Irigaray and quantum gravity — held together by vague rhetoric about “nonlinearity”, “flux” and “interconnectedness.” Finally, I jump (again without argument) to the assertion that “postmodern science” has abolished the concept of objective reality. Nowhere in all of this is there anything resembling a logical sequence of thought; one finds only citations of authority, plays on words, strained analogies, and bald assertions.“

Another prominent source of pseudo-profound bullshit is New Age literature, particularly in the alliance between pseudo-science and spirituality that has come to be symbolized by the well-known New Age guru Deepak Chopra. A Swedish book called ”Life through the eyes of quantum physics” that recently hit the best-seller lists provides an almost parodic illustration of this sort of pseudo-profound bullshit. This book is full of vague Chopraesque claims about quantum consciousness and its ”scientifically proven” power to shape reality, including preventing serious illnesses such as cancer, promoting success in life, altering the magnetic field of the earth, and causing miracles. The authors not only did lacked knowledge of the basics of quantum physics, they had no interest in it either (as interviews have made apparent)—their interest was in selling New Age spirituality with the help of pop-bullshitting about quantum physics and superficial narratives about Eastern spiritual wisdom.

The reason that pseudo-profound bullshit is so pernicious is in part, I suspect, that it plays on the human yearning for a deep sense of mystery and understanding of the cosmos. Our existential predicament is mind-boggling and anxiety-provoking, and it is comforting to believe that there are gurus or other authorities out there with a deeper sense of the truth, and to therefore attribute your own lack of ability to understand what they say to our own ignorance.

Recent findings

How do you study bullshit receptivity scientifically? First, you need a sample of bullshit sentences. Fortunately, there is a very simple, algorithmic way of constructing such sentences. You let a computer randomly string together impressive-sounding buzzwords into a syntactically correct sequence. There are a number of such bullshit generators available online, including the Postmodernism generator and the Wisdom of Chopra. These sentences are by definintion bullshit, since they are constructed absent concern for the truth.

In a pioneering paper that won them the Ig-Nobel Prize, Pennycook, Cheyne, Barr, Koehler, and Fugelsang (2015) constructed a set of bullshit sentences (e.g., “Wholeness quiets infinite phenomena”) through this method, with a focus on New Age-jargon, and then let people rate how profound they thought these sentences were. The found that receptivity to the bullshit sentences was associated with an intuitive cognitive style, a lack of reflectiveness, supernatural beliefs, and other related constructs. Pennycook and Rand (2019) have later also found that this sort of receptivity to pseudo-profound bullshit plays a role in receptivity to fake news.

My colleagues and I constructed a Swedish measure based on the Pennycook et al. (2015) paradigm. We have used this measure to address, among other things, the debates in political psychology over whether there are ideological asymmetries in epistemic orientations (Nilsson, Erlandsson, & Västfjäll, 2019). We found in essence that social conservatism (and particularly moral intuitions about ingroup loyalty, respect for authority, and purity) is robustly associated with receptivity to pseudo-profound bullshit, consistent with the classical notion of a “rigidity of the right”. Interestingly, we also found a particularly high bullshit receptivity among persons who vote for the green party in Sweden, and a very low bullshit receptivity among right-of-center social liberals.

What are the mechanisms driving these differences? A part of it appears to be a failure to critically engage with information. Like Pennycook and colleagues, we have found that bullshit receptivity is robustly associated with low cognitive reflection, and we have also found it to be negatively associated with numeracy and positively associated with confirmation bias.

But this cannot be the whole story. For example, the greens were close to the average in terms of cognitive reflectiveness in our study. We speculated that their high bullshit receptivity is instead due to a strong openness to ideas that is not always tempered by critical thinking. Interestingly, two papers suggesting that this is indeed a mechanism underlying bullshit receptivity appeared right after our paper was accepted for publication. Bainbridge, Quilan, Mar, and Smillie (2019) found that receptivity to pseudo-profound bullshit is associated with the personality construct “apophenia”—the tendency to see patterns where none exist—which is a form of trait openness. Walker, Turpin, Stolz, Fugelsang, and Koehler (2019) measured illusory pattern perception through a series cognitive tests rather than personality questions but came to a similar conclusion—bullshit-receptive persons tend to endorse patterns where none exist.

There may of course also be other mechanisms that contribute to receptivity to pseudo-profound bullshit. For example, Pennycook and colleagues have suggested that perceptual fluency contributes to receptivity to fake news. It is possible that persons who are commonly exposed to a specific type of pseudo-profound jargon are more likely to be receptive to this kind of bullshit.

Another great addition to this growing body of research is a paper by Čavojová, Secară, Jurkovič, and Šrol (2019), which presents conceptual replications of many of the key findings on receptivity to pseudo-profound bullshit in Slovakia and Romania. I often lament that psychology fails to take the problem of WEIRD samples and studies seriously, but these studies certiainly do. By demonstrating that the research paradigm I have discussed here is meaningful and useful outside of the U.S. and Western Europe, they put this new, fascinating field on firmer ground.

Key papers

___________________________________________________________________________

Bainbridge, T. F., Quinlan, J. A., Mar, R. A., & Smillie, L. D. (2019). Opennes/Intellect and susceptibility to pseudo-profound bullshit: A replication and extension. European Journal of Personality, 33(1), 72-88. https://doi.org/10.1002/per.2176

Čavojová, V., Secară, E-C., Jurkovič, M., & Šrol, J. (2019). Reception and willigness to share pseudo-profound bullshit and their relation to other epistemically suspect beliefs and cognitive ability in Slovakia and Romania. Applied Cognitive Psychology, 33(2), 299-311. https://doi.org/10.1002/acp.3486

Nilsson, A., Erlandsson, A., & Västfjäll, D. (2019). The complex relation between receptivity to pseudo-profound bullshit and political ideology. Personality and Social Psychology Bulletin. https://doi.org/10.1177/0146167219830415

Pennycook, G., Cheyne, J. A., Barr, N., Koehler, D. J., & Fugelsang, J. A. (2015). On the reception and detection of pseudo-profound bullshit. Judgment and Decision Making, 10(6), 549-563. http://journal.sjdm.org/15/15923a/jdm15923a.pdf

Pennycook, G. & Rand, D. G. (2019). Who falls for fake news? The roles of bullshit receptivity, overclaiming, familiarity, and analytic thinking. Journal of Personality. https://doi.org/10.1111/jopy.12476

Petrocelli, J. V. (2018). Antecedents of bullshitting. Journal of Experimental Social Psychology, 76, 249-258. https://doi.org/10.1016/j.jesp.2018.03.004

Walker, A. C., Turpin. M. H., Stolz, J. A., Fugelsang, J. A., & Koehler, D. J. (2019). Finding meaning in the clouds: Illusory pattern perception predicts receptivity to pseudo-profound bullshit. Judgment and Decision Making, 14(2), 109-119. http://journal.sjdm.org/18/181212a/jdm181212a.html
____________________________________________________________________________



Psychology is still WEIRD

Comments on new research Posted on Wed, November 14, 2018 15:20:26

Psychological science is fraught with problems. One of these problems that has recently attracted widespread attention is the proliferation of false positives, which is rooted in a combination of QRPs (questionable research practices), including “p-hacking” (choosing analytical options on the basis of whether they render significant results) and “HARKing” (hypothesizing after the results are known), and very low statistical power (i.e., too few participants). Overall, psychology has responded vigorously to this problem, although much remains to be done. Numerous reforms have been put in place to encourage open science practices and quality in research.

Another problem that has become widely recognized recently is that psychological research often makes inferences about human beings in general based on studies of a thin slice of humanity. As Henrich, Heine, & Norenzayan (2010) noted in a landmark paper, participants in psychological research are usually drawn from populations that are WEIRD (Western, Educated, Industrialized, Rich, Democratic), which are far from representative of mankind—in fact, they turn out to frequently be rather eccentric, even when it comes to basic cognitive, evolutionary, and social phenomena such as cooperation, reasoning styles, and visual perception (see also this interesting preprint by Schultz, Bahrami-Rad, Beauchamp, & Henrich that very thoroughly discusses the historical origins of WEIRD psychology).

The paper by Henrich and colleagues has racked up almost 5000 Google Scholar citations. Yet a recent paper by Rad, Martingano, and Ginges (2018) suggests that the impact of the Henrich et al. paper on actual research practices in psychology has been minimal, at least as indexed by research published in the high-prestige journal Psychological Science. Rad et al. find that researchers persist in relying on WEIRD samples and show little awareness of the WEIRD problem: “Perhaps the most disturbing aspect of our analysis was the lack of information given about the WEIRDness of samples, and the lack of consideration given to issues of cultural diversity in bounding the conclusions” (p. 11402).

Explaining the persistence of the WEIRD problem

How can it be that psychology has responded so vigorously to the problem with false positives, yet so inadequately to the WEIRD problem? Surely both problems are equally serious, are they not? I can think of at least three possible explanations.

1. First and foremost, the WEIRD problem is a manifestation of a much broader problem. It is a manifestation of the lasting influence of the marriage between logical positivism and behaviorism that shaped psychology for almost half a century. Psychological research was supposed to yield universal facts, just like physics, by employing “neutral”, culture-free materials and methods, a quantitative methodology, and hard-core empiricism. Given the vast historical impact of this ideal, it is no mystery that psychology remains both WEIRD and theoretically unsophisticated. This is simply the implicit paradigm under which psychology has operated for more than a century. While the problem with false positives is a problem signaling a crisis within this paradigm, the WEIRD problem is a meta-problem with the paradigm itself.

2. Second, it is possible that researchers do not realize the severity of the WEIRD problem because they are immersed in a homogeneous community of like-minded individuals with similar concerns, and their exposure to other intellectual cultures is limited. Here it is important to note that the WEIRD problem is not limited to participant selection. It is a problem of testing WEIRD theories on WEIRD samples with WEIRD methods. I personally often find psychological theories and concepts US-centric (e.g., the reification of “liberals” and “conservatives” in political psychology or the pre-occupation with the self and neglect of other aspects of the person’s worldview in personality psychology)—which is not surprising given that most of the leading researchers in psychology are from the United States—and I still live in the broader Western cultural sphere.

3. A third possible explanation for the persistence of the WEIRD problem is that there are many practical difficulties involved in conducting research in non-WEIRD contexts. A lot of things could go wrong. You need high-quality translations of research materials. You also need to obtain a reasonable degree of measurement invariance across languages and populations to be able to make meaningful comparisons between them. Even so, the results may not be at all what you expected. Perhaps the theories and instruments do not perform as they are supposed to do. Of course, on a purely scientific basis such findings would be extremely important. But perhaps researchers still find it is easier to just stick to studying well-known populations under well-known conditions in order to more easily find support for their hypotheses and publish their work.

Moving forward

The WEIRD-problem needs to attain the same status as the false positives-problem in psychology. As Rad, Martingano, and Ginges (2018) suggest, authors need to do a much better job reporting sample characteristics, explicitly tying findings to populations, justifying the sampled population, discussing the generalizability of the findings, and investigating existing diversity in their samples. Journals and funders need to start encouraging these practices. Given all the work involved in conducting non-WEIRD research and the fierce competition over research funding and space in high-impact journals, we are unlikely to see any real change unless the inclusion of non-WEIRD research will give extra points.

When it comes to the problem with WEIRD perspectives, psychology might need to become more open to scholarship born out of non-WEIRD (particularly non-US) contexts. An increased openness to philosophical, meta-theoretical, historical, and anthropological scholarship in general, which is for the most part completely ignored in psychological science today, would be particularly helpful. That would help us both to address the WEIRD-problem and to make psychology a more theoretically sophisticated science.



The evolutionary foundations of worldviews

Comments on new research Posted on Wed, November 07, 2018 15:27:16

When taking a graduate course on evolutionary psychology a few years ago, I thought a bit about the potential evolutionary bases of worldviews. I was specifically interested in the opposition between humanistic and normativistic perspectives posited by Silvan Tomkins Polarity Theory (more information here) that is encapsulated in the following quotation: “Is man the measure, an end in himself, an active, creative, thinking, desiring, loving force in nature? Or must man realize himself, attain his full stature only through struggle toward, participation in, conformity to a norm, a measure, an ideal essence basically prior to and independent of man?” (Tomkins, 1963).

Evolutionary bases of of normativism and humanism

Drawing on Tomkins’ (1987) notion that “the major dynamic of ideological differentiation and stratification arises from perceived scarcity and the reliance upon violence to reduce such scarcity”, I suggested (in my term paper) that conditions of resource scarcity should have fostered a tough-minded climate where the strong and hostile could prove their worth by contributing to resource provision, and those weak or vulnerable were met with anger, contempt, and disgust. I suggested that humanism is to a greater extent rooted in the problem of forming stable alliances with other persons and groups, which requires interpersonal trust and empathy.


Because psychological traits co-evolve as entire “packages” in response to particular adaptive contexts, it is reasonable to predict that humanism and normativism co-vary with other psychological and physiological traits that also help to solve the respective adaptive problems. Normativism may have co-evolved with other traits that helped to solve the problem of resource acquisition, such as aggressiveness, physical strength and formidability, risk-taking, conscientiousness, persistence, and diligence—this should be true at least among men, who are thought of as the primary resource providers in an evolutionary context. Humanism may instead have co-evolved with traits such as empathy, altruism, agreeableness, and concern for the welfare of individuals, which are crucial for social bonding.

Egalitarianism and upper-body strength

Interestingly, a portion of the aforementioned hypotheses have  subsequently been tested. The results of twelve studies conducted in various countries are reported in a recent paper by Michael Bang Petersen and Lasse Lauritsen titled Upper-body strength and political egalitarianism: Twelve conceptual replications. Drawing on models of animal conflict behavior, Petersen and Lauritsen suggest that attitudes related to resource conflict (i.e., egalitarianism) should be related to upper-body strength among males, which was crucial for the resolution of resource conflicts in our evolutionary past. They argue that “formidable individuals and their allies would be more likely to prevail in resource conflicts and needed to rely less on norms that enforced sharing and equality within or between groups in order to prosper”.

The measures of upper-body strength employed include both self-report measures and objective measures of formidability. The one major limitation of these studies—and this is a major limitation—is that there was, as far as I understand it, no control for significant environmental factors such as time spent in the gym, physical exercise background, occupation, or use of performance enhancing drugs (although other more indirectly relevant variables such as socioeconomic status and unemployment experiences were taken into consideration). Nevertheless, it is interesting to note that the authors find a clear relationship among men (but not women) between physical formidability and social dominance orientation (which encompasses egalitarianism) but not between formidability and right-wing authoritarianism.

Toward an evolutionary understanding of worldviews?

In order to establish that there is genetic covariation (not just covariation in general) between formidability and worldviews, future research needs to do a better job controlling for crucial environmental influences (recent studies have apparently started to do this). Behavioral genetics methods can also be employed to more directly assess genetic covariation. In addition to this, a broader range of worldview dimensions (e.g., normativism and humanism, which are correlated with authoritarianism and social dominance) and physiological predispositions could easily be taken into consideration. Let us hope that this will indeed what will happen over the next years.



The “happiness pie”, genetic and environmental determinism, and free will

Comments on new research Posted on Fri, September 21, 2018 01:25:46

Nick Brown and Julia Rohrer recently posted a new preprint titled Re-slicing the ”Happiness Pie: A Re-examination of the Determinants of Well-being that comments on an influential paper by Lyubomirsky, Sheldon, and Schkade (2005) on the determinants of well-being. Nick Brown is the amateur who debunked the mathematics of happiness (together with the legendary Alain Sokal of the “Sokal hoax”). He has made a name for himself exposing shoddy work in positive psychology. This is another addition to this genre. What is particularly mind-blowing with this one is not just the sheer lack of intellectual sophistication of the criticized paper, but the fact that it has produced a whopping 3000 Google Scholar citations.

The central claim of the Lyubomirsky et al. paper is that roughly 50% of the variance in well-being can be explained in terms of genetic predispositions and that roughly 10% of it can be explained in terms of life circumstances, leaving up to 40% to be explained in terms of intentional activity. This decomposition of the determinants of well-being, which has come to be known as the happiness pie, has become a cornerstone of the self-help and coaching movements, as it appears to suggest that all persons have considerable control over their own well-being.

Problems pointed out by Brown and Rohrer

Brown and Rohrer meticulously pick the happiness pie apart. Here are some of the errors:

1. An additive model that divides the determinants of well-being into three disparate portions (or pieces of a pie) is only meaningful if all the portions are independent of each other. But there is plenty of evidence of interactions between genes, environment, and volitional activity.

2. No evidence is presented that the “leftover” variance after taking genetics and the environment into consideration can be attributed to volitional activity.

3. Measurement error, which attenuates estimates, has not been taken into consideration. If we adjust for this, there will be less “leftover” variance.

4. When the sources for the numbers 50% and 10% are re-examined, these numbers appear to be arbitrary, and how they were derived is not transparent.

5. Particularly the 10% estimate appears based on sloppy reasoning. Countless environmental factors are not measured in the survey studies that this estimate is based on (in utero influence is a dramatic example). Environmental factors are frequently operationalized in terms of demographic variables (which have both genetic and environmental determinants and do certainly not exhaust the full range of relevant life
circumstances).

6. Even if the 50% and 10% figures and the subtraction logic by which 40% of the variance is leftover for intentional activity were correct, this would still just be a population average. It would not imply that each individual has substantial control over his or her well-being.

Moving beyond the Brown and Rohrer paper

The great irony of all this is that there is nothing here—neither in the Lyubomirsky et al. paper nor in the Brown and Rohrer paper—showing that people cannot change their well-being through intentional activity either. The deeper problem is that claims about the effects of will-power (and to an even greater degree claims about free will) do not have anything at all to do with heritability and environmental influences per se. The proper way to scientifically investigate the extent to which people can intentionally change their well-being is to:
(a) recruit a large group of persons who are highly motivated to do what it takes to increase their well-being,
(b) make sure that they have all relevant resources (or at least measure whether they do), including the best forms of therapy or training programs, time, money, etc., and make sure that they actually do what they are supposed to do,
(c) measure changes in their levels of well-being compared to a control group of persons who do not engage in deliberate efforts to change their well-being but who are otherwise comparable to the group of persons who do.

Obviously, this is not easy to do (e.g., how do we make sure that experiment and control group are comparable?), and even if done well, all that we could say is that this is what we can achieve with our current state of knowledge. It is possible that there are effective methods for increasing well-being that have not yet been discovered or are not widely known.

Heritability coefficients and estimates of environmental correlates in a general population are in themselves irrelevant to this question because we do not know whether the persons who participated in these studies did engage in persistent intentional efforts to change their well-being or whether they had access to the best strategies for doing this. These pieces of information could possibly have some relevance if we would know which individuals had the motivation and strategies necessary for deliberate improvement of well-being and which did not. Even so, intentional orientations do not emerge randomly out of thin air—they may be related to traits such as consciousness or openness to change, which have a sizable genetic component—so it might be tricky to find twins who differ enough in this regard.

At any rate, a case can be made that people do at least have the capacity for intentionally changing their well-being if they are motivated to make enough changes to their lives, without resorting to any of the dubious arguments presented by Lyubomirsky.