Some macaques and capuchins are able to discriminate appearance from reality; seems that the evolutionary apparition of this cognitive ability was sooner than commonly believed

Are monkeys able to discriminate appearance from reality? Marie Hirel et al. Cognition, Volume 196, March 2020, 104123. https://doi.org/10.1016/j.cognition.2019.104123

Abstract: The understanding that the perceptual appearance of the environment can differ in several ways from the reality underlies the ability to discriminate appearance from reality. Being able to realize when a misperception can lead us to behave in inappropriate ways confers an evolutionary advantage and may be a prerequisite to develop a Theory of Mind. Understanding that our own perception can differ from reality seems indeed necessary to attribute to others perceptions or beliefs different than ours. This appearance-reality discrimination ability has recently been demonstrated in great apes but no information is currently available regarding this ability in other nonhuman species. In a comparative study, we tested Tonkean macaques (Macaca tonkeana), an Old World primate species, and brown capuchins (Sapajus apella), a New World primate species. We provided monkeys with two experiments using visual illusions of size and quantity to test their ability to discriminate appearance from reality, with an experimental setup similar to the one developed by Krachun et al. (2016) on chimpanzees. A large number of brown capuchins, from different ages and both sexes, as well as two Tonkean macaques succeeded in the two experiments. By ruling out all alternative explanations (i.e. visual tracking or associative learning), our study brings the first evidence that some Old World and New World monkeys are able to discriminate appearance from reality. Our results suggest moving the evolutionary apparition of this cognitive ability earlier in time. Finally, it suggests that humans could share more Theory of Mind components with more nonhuman species than we previously thought.

4. Discussion
This study brings the first evidence that monkeys are able to discriminate
appearance from reality. All our subjects passed the preference
tests of both the Lens and Mirror experiments and thus demonstrated
a strong natural preference for larger or higher number of
food items, as well as a good size and quantity discrimination ability.
Then, in the Lens experiment, all but two subjects of each species
passed successfully the basic test and the tracking control tests, demonstrating
that they were capable of ignoring the misleading appearance
of the magnified grape to choose the truly bigger one, without
being able to visual track the grapes (see Fig. 2). Afterwards, the large
number of successful brown capuchins, from different ages and both
sexes, in the last learning control brings strong evidence that they are
able to discriminate appearance from reality. In Tonkean macaques, our
results are less clear, with only one individual succeeding in the avoidlens
control (see Fig. 2). Thus, we are unable to say with confidence
whether or not this species possesses appearance-reality discrimination
ability. However, our procedure with lenses might have not allow us to
highlight AR discrimination ability in a larger scale in this species and
could provide results underestimating it (see methodological issues
discussed below). This is supported by Tonkean macaques’ good performances
in the Mirror experiment. Indeed, the two Tonkean macaques
and all but one brown capuchin that were tested in this second
experiment passed the visual tracking and the avoid-mirror tests. By
validating all controls, they demonstrated they are able to discriminate
perceived from real quantity.
We found that all except two individuals who were successful in the
Mirror experiment had also succeeded in the Lens experiment.
Successes in both of our experiments provides evidence for a general
understanding in our animals, that perceptual appearance of objects
can be different in several ways from reality. Nevertheless, we should
approach this conclusion with caution for the few individuals that
needed more than two sessions to succeed. Even by limiting the number
of sessions allowed to reach a success, our and other procedures of
progressive sessions generate an increasing probability of false positives.
To account for this multiple testing issue, one option is to look out
group performances. In this respect, brown capuchins succeeded significantly
in all tests of the Lens and Mirror experiments. For Tonkean
macaques, their succeeded significantly in all tests except for the avoidlens
tests n°1 and n°2. These results confirm our inability to conclude of
an AR discrimination ability presence for Tonkean macaques with the
Lens experiment. However, we still able to maintain confidence in our
results demonstrating AR discrimination ability in both experiments for
brown capuchins and in the Mirror experiment for Tonkean macaques.
In summary, even if more subjects of both species are necessary to make
a reliable conclusion at a species scale, results of the Mirror experiment
confirmed those of the Lens experiment and together suggest that these
monkeys could be able to discriminate appearance from reality.
The ability to discriminate appearance from reality was demonstrated
in great apes by previous studies of Karg et al. (2014) and
Krachun et al. (2009, 2016). Now, our study suggests the presence of
this ability in Tonkean macaques, an Old World monkey species, and
brings strong evidence in brown capuchins, a New World monkey
species. Therefore, the hypothesis of the emergence of AR discrimination
ability in a common ancestor dating back to at least the capuchin
monkey genus is conceivable. However, our results certainly need to be
strengthened by further research on more individuals and other species
of macaques to make a more reliable conclusion. Since only two Tonkean
macaques were compared to eight brown capuchins in both experiments,
alternative hypotheses of convergent evolution cannot be
completely ruled out. Thus, instead of looking at phylogenetic relatedness
only, our interest should go towards the socio-ecological aspects
that are characteristic of these species in order to improve our knowledge
about the development of such cognitive abilities during evolution.
Indeed, as example, many similarities were found between genus
Sapajus and Pan: long life spans, explorative and manipulative behaviours,
tool using, omnivorous diet, socially complex behaviours such
as coalitions and cooperation and so forth (Visalberghi & McGrew,
1997).
One of the major aims of this study was to compare performances of
monkey species with those of chimpanzees (Krachun et al., 2016) using
a same experimental paradigm. Like chimpanzees, brown capuchins
performed high scores in the two AR experiments and passed each test
within a small amount of sessions. Not all of them succeeded but the
proportion of successful brown capuchins is still greater than observed
in chimpanzees. Hence, the performance of brown capuchins is likely to
be comparable to those of chimpanzees in these AR discrimination
experiments. It brings more evidence that monkeys can perform like
apes in demanding cognitive tasks, for example like prerequisite ToM
tests, contrary to previous scientists’ common beliefs. These previous
studies that found differences between performances of monkeys and
apes in cognitive tasks, have normally compared apes that are highly
habituated to experiments with naive monkeys (Amici et al., 2010). In
addition to using similar standardized paradigm, our study tested
highly habituated monkeys. Thereby, a comparison between our results
and those on chimpanzees appears more reliable and appropriate.
Several recent studies on cognitive abilities comparing different
monkey and ape species found no strong evidence of a difference between
their performances (Amici et al., 2010; Meunier, 2017; Schmitt
et al., 2012; Tomasello et al., 1998). Instead, Amici et al. (2010) revealed
a link in their results between cognitive capacities and social
organisation: performances of species living in systems with fissionfusion
dynamics (e.g. chimpanzees, bonobos, orangutans, and spider
monkeys) exceeded those of species living in more stable groups (e.g.
long-tailed macaques, gorillas and capuchin monkeys). Thereby, the
type of social organisation seems to impact the physico-cognitive skills
of primates (i.e. spatial memory, quantities, causality, etc.) and may
provide better predictor of performance in some cognitive tasks than
phylogenetic relatedness (Schmitt et al., 2012). This led to the development
of the social intelligence hypothesis, which claims that primate
intelligence evolved in response to challenges of living in large and
complex groups (Byrne & Bates, 2010; MacLean et al., 2012; Schmitt
et al., 2012). However, both for Tonkean macaques and brown capuchins,
this explanation does not apply because they are not living in
fission-fusion dynamics but instead in stable female-bonded groups
with comparable social complexity, and are described as socially tolerant
(Hare et al., 2003). These other social features could then interfere
in their cognitive abilities’ development. Thus, as socio-ecological
conditions might better explain the presence of some cognitive abilities,
instead of comparing monkeys and apes, we should focus on comparison
between particular species. Studying with the same paradigm other
nonhuman primates either closely related or presenting different social
organisations (e.g. lemurs, white-faced capuchins, or long-tailed macaques)
should provide insights about the origin of this ability and
could help to clarify the social intelligence hypothesis (Byrne & Bates,
2010; Humphrey, 1976; MacLean et al., 2012; Schmitt et al., 2012).
The positive results in both experiments could be explained by
several alternative mechanisms of learning. First of all, an association
learning between large or big food item and negative outcome could be
a possibility. However, it seems to be an unlikely explanation because
our successful subjects have chosen the item that appeared larger or
bigger in the avoid tests, and they continued to choose the bigger or
larger amount of items in each warm up preference trials before each
session. Secondly, some might argue that successful individuals could
have done reverse contingency learning, i.e. choose the smaller food
item to obtain the larger one. However, this assumption is highly unlikely
for the following reasons. First, the avoid control test clearly rules
out this possibility. In fact, to validate this control, subjects needed to
choose the apparent bigger or larger food item to obtain the truly bigger
or larger one, and not the apparent smaller one as in previous tests.
Second, several studies have already demonstrated that this type of
contingency learning for both size and quantity of items is really difficult
for these species, even for great apes, and require hundreds of
trials to succeed in it (Anderson, Hattori, & Fujita, 2008; Boysen,
Berntson, & Mukobi, 2001; Krachun et al., 2009; Vlamings, Uher, &
Call, 2006).
In the Lens experiment, a few subjects did not succeed in different
tests. First, a significant left pointing bias can explain the failure of two
brown capuchins, one in the basic and the other one in the seen
tracking test. The fact that this bias was not present in the initial conditions
might indicate they do not understand the situation in the
follow-up conditions and thus could also suggest that a more ecological
procedure may have been necessary for these individuals for AR discrimination
to be revealed. Second, Tonkean macaques and brown capuchins
seem to have failed the Lens experiment in the last learning
control because they have performed learning strategies based on different
stimuli. Two Tonkean macaques and one brown capuchin obtained
really low scores, demonstrating a significant preference for the
truly smaller grape. This could be explained by a learning process to
avoid the magnifying lens or to choose the grape that appeared with the
size of the small one through the lenses. The other five unsuccessful
Tonkean macaques and one brown capuchin obtained medium scores,
which reflect a learning to choose the minimizing lens or the grape that
appeared with the size of the bigger one through the lenses. Because
neither the minimizing lens nor the grape appearing the size of the
bigger one through lenses were no more available in this control, individuals
could not anymore use these stimuli to succeed and so they
chose randomly. Moreover, these individuals who failed the test still
choose preferentially the bigger grape in warmup preference trials
before each session, suggesting their failure was not due to a change of
preference to smaller grapes.
In order to compare performances with those of chimpanzees, we
carried out a second avoid-lens test at the end of the Lens experiment
using exactly the same procedure as Krachun et al. in their study of
2016. Tonkean macaques’ results obtained in the avoid-lens test n°2 are
quite inconsistent with those of avoid-lens test n°1. In fact, one individual
succeeded in both but two individuals that failed avoid-lens
test n°1 then succeeded in the avoid-lens test n°2, within one or two
sessions only. For these two subjects, we carried out one more session
with the unseen test procedure, in order to check whether by changing
again sizes and lenses used, they will succeed (as before in the same
unseen tracking test) and thus demonstrate a truly understanding of the
illusion phenomena. However, both of them drastically failed by never
choosing the truly larger grape in this session. These results highlight a
learning strategy they must have developed during the four sessions of
the avoid-lens test n°1. Because avoid-lens test n°2 was carried out after,
individuals had the possibility to learn the new rule (larger grape
placed behind magnifying lens). In addition, the only subject that
passed the first avoid-lens control test needed four sessions to succeed
in this last control. His result should be interpreted with caution because
he could have also learned, faster than the others, i.e. within only
three sessions. Surprisingly, the other Tonkean macaques and brown
capuchins that have failed did not realise this potential quick learning:
they both failed the avoid-lens n°1 and the avoid-lens n°2. A comparison
between avoid-lens n°1 performances of Tonkean macaques and brown
capuchins with avoid-lens test performances of chimpanzees in the
study of Krachun et al. (2016) seems thus more reliable.
Instead of an absence of AR discrimination, individuals’ failure
could be interpreted by other explanations corresponding to methodological
issues like too many experimenter’s manipulations or isolation
from their social group. Our test procedure could have required too
much concentration, motivation and short-term memory for these animals,
difficulties faced by Krachun et al. (2016) in their study on
chimpanzees. Notably, poor performances of Tonkean macaques (poor
success ratio, higher means of sessions to succeed in each test) compared
to brown capuchins and chimpanzees may support this assumption:
as we carried out half-sessions for brown capuchins, they had
twice as many demo trials as Tonkean macaques and chimpanzees,
hence more opportunities to understand the effects of stimuli illusion.
Moreover, all brown capuchins came to the experimental room at least
once per day, whereas with some Tonkean macaques, several days
sometimes separated two successive sessions because they did not come
back in the experimental room regularly. Because of those delays,
Tonkean macaques might have faced memory challenges about the
impact of illusion. Our results could thus underestimate their AR discrimination
ability.
Given the pointing bias of some brown capuchins and the large
failure of Tonkean macaques in the Lens experiment, ecological relevance
of our paradigm has to be questioned. The better success ratio
in the Mirror experiment could be due to the applied illusion. The
mirror effect indeed creates a false presence of two raisins, which do not
exist in reality. Whereas lens is modifying properties of grapes that
already exist, it does not create new food items. Thereby, it is maybe
easier to understand the effect of a mirror than the subtle modification
of existent items created by lenses. One alternative paradigm to test size
illusion understanding could be the one developed by Karg et al.
(2014). In their comparative study between great apes and children on
AR discrimination, they created a size illusion using completion illusion,
i.e. part of an object is partially occluding by another, in order to
inverse size relation. As Tonkean macaques and brown capuchins are
semi-arboreal primates, they might more regularly face size illusion
challenges by completion illusion because of the dense vegetation in the
canopy rather than by distortion.
Further research using different complementary types of illusion,
e.g. colour or auditory illusions, could now bring more information
about the scope of the AR understanding of these species. Another
notable result is that several capuchins of various ages succeeded. Thus,
AR discrimination ability seems to appear early in the development of
brown capuchins. Further experiences on even younger individuals will
be helpful to determine if the development of AR discrimination occurs
around the same development stage as for human children and with the
same developmental pattern (e.g. Hansen & Markman, 2005; Karg
et al., 2014; Moll & Tomasello, 2012).
To conclude, our results support those of apes, demonstrating that
this AR discrimination ability is not unique to great apes. As we have
now demonstrated that even some monkey species seem to possess this
understanding, it seems these abilities are perhaps more evolutionary
ancestral than previously thought. More research on different monkey
species with similar standardized procedure is now more than ever
necessary to better understand evolution of our remarkable social
cognition.

The more maladaptive forms of narcissism (hypersensitivity, willfulness) declined across life & individual autonomy increased; later-born birth-cohorts were lower in hypersensitivity & higher in autonomy

Chopik, William J. 2019. “Longitudinal Changes and Historic Differences in Narcissism from Adolescence to Older Adulthood.” PsyArXiv. December 10. doi:10.31234/osf.io/bf7qv

Abstract: In the debate about whether or not narcissism has been increasing in recent history, there is a lack of basic information about how narcissism changes across the adult lifespan. Existing research relies on cross-sectional samples, purposely restricts samples to include only college students, or follows one group of individuals over a short period of time. In the current study, we addressed many of these limitations by examining how narcissism changed longitudinally in a sample of 747 participants (72.3% female) from age 13 to age 77 across six samples of participants born between 1923 and 1969. Narcissism was moderately stable across the lifespan (rs ranged from .37 to .52), to a comparable degree as other psychological characteristics. We found that more maladaptive forms of narcissism (e.g., hypersensitivity, willfulness) declined across life and individual autonomy increased across life. More later-born birth-cohorts were lower in hypersensitivity and higher in autonomy compared to earlier-born birth-cohorts; these differences were most apparent among those born after the 1930s. The results are discussed in the context of the mechanisms that drive both changes in narcissism across the lifespan and substantive differences in narcissism between historical periods.


Discussion

The current study examined changes in narcissism from age 13 to age 77 across six different studies of human development. Narcissism was moderately stable across the lifespan, to a comparable degree as other psychological characteristics (Roberts & DelVecchio, 2000). We found that more maladaptive forms of narcissism (e.g., hypersensitivity, willfulness) declined across life and autonomy increased across life. Later-born birth-cohorts were lower in hypersensitivity and higher in autonomy, which is consistent with one study on historical changes in narcissism (Wetzel et al., 2017) but also in strong contrast to studies suggesting that later-born birth-cohorts are higher in narcissism than earlier-born birth-cohorts (Twenge, 2006; Twenge et al., 2008). The current study is one of the most comprehensive examinations of narcissism changes ever conducted and included data from individuals born throughout a 46-year period (from 1923 to 1969).
The current findings are consistent with tenets of the social investment model of personality development (Roberts et al., 2005). Specifically, although narcissism may serve some protective role for well-being in emerging adulthood (Hill & Roberts, 2012), high levels of maladaptive forms of narcissism are incompatible with the age-graded social roles and expectations that individuals adopt throughout the lifespan. Psychological characteristics are thought to change in response to the complex interplay of individuals functioning within the demands and expectations of increased responsibility and maturity. As a result, maladaptive forms of narcissism that serve as a barrier to success in work, life, and love are abated throughout middle age and older adulthood. Forms of narcissism that enhance successes in these domains are likely to be cultivated. For these reasons, maladaptive forms of narcissism tend to decline across life and adaptive forms of narcissism tend to increase across life.

Birth-cohort Effects on Narcissism

That later-born birth-cohorts were lower in hypersensitivity and higher in autonomy was also perplexing as it goes against a narrative that recent birth-cohorts have experienced a monotonic rise in narcissistic traits across the past century (Twenge et al., 2012a, 2012b; Twenge et al., 2008). It is not necessarily surprising that maladaptive narcissism is declining while autonomy is increasing, as these two are inextricably linked (Deffler, Leary, & Hoyle, 2016). Although some research has found similar (decreasing) levels of maladaptive narcissism, the reasons behind such differences have not been thoroughly examined. There are many substantive and non-substantive differences between birth-cohorts and previous work has tried to identify a number of social indicators for why birth-cohorts might psychologically differ (e.g., technological changes, changes in parenting styles; Konrath et al., 2011). Exactly how these differences translate into mean differences in some forms of narcissism and not others is an exciting direction for future research. Of course, as with gender, the samples were not perfectly comparable with respect to ages distributed across the birth-cohorts. For example, some samples (e.g., the Radcliffe Sample) had only participants who were middle age and older. Other samples (the Block and Block Sample) had only participants who were young adults. Thus, sample/birth year and age are somewhat conflated in the current study. Thus, we interpret caution when drawing conclusions from the birth-cohort analyses.

Limitations and Future Directions

The current study had many strengths. We combined data from six distinct sources to model changes in narcissism over a 59-year period. Nevertheless, there are some limitations that are worth noting.
Why did people change in narcissism over time? Other than the sample imbalance with respect to age and gender (see above), we did not have a consistent set of predictors to model the mechanisms underlying lifespan changes and birth-cohort differences in narcissism over time. Like other traits that function within the social investment framework, there are likely both selection and socialization factors that drive the development of narcissism across the lifespan (Specht et al., 2014; Specht, Egloff, & Schmukle, 2011). For example, people high in autonomy might select environments that are conducive to cultivating their autonomous lifestyle (and avoid situations that dampen it). Likewise, once individuals have selected an environment, there are additional mechanisms that drive stability and change in narcissistic characteristics as these changes are socialized within individuals (Caspi & Roberts, 1999). What are the environments and life circumstances that initiate changes in narcissism? And which environments and life circumstances lead to sustainable changes in narcissism over long periods of time? Although there is a great deal of research highlighting narcissists’ resilience and resistance to critical feedback about themselves (Bushman & Baumeister, 1998), some studies show some promising signs that narcissists can change in response to life events (Grosz et al., 2017). Nevertheless, the mechanisms through which narcissists are able to reflect on their lives in an impartial way and consciously change their thoughts and actions are currently unknown, even though it appears that narcissists do have some accurate self-insight into their personalities (E. N. Carlson, 2013; E. N. Carlson, Vazire, & Oltmanns, 2011). Future research can more formally model individual differences in changes in narcissism and the conditions under which lifespan changes are largest.

* Why do birth-cohorts differ in narcissism over time? Likewise, it is worth noting that our study did find reliable differences in narcissism between birth-cohorts, but in the reverse direction than is typical seen (Wetzel et al., 2017). Just as it is important to examine predictors of lifespan changes in narcissism, it is also important to examine whether changes in narcissism can be attributable to changes in parenting behaviors, shifting demographics and the relative risk of exposure to life events, some broader cultural changes, or another explanation entirely (Konrath et al., 2011; Twenge et al., 2008). Because the studies did not have consistent or, in some cases, any variables to operationalize changes in these variables, we are left with merely descriptive data on how narcissism might change over historical time. We view this as a major limitation to the current report and hope that future researchers can more seriously conceptualize and test why birth-cohorts might be changing in narcissism. That two of the samples were comprised entirely of women is also a considerable limitation, particularly for the birth-cohort analyses. Worth noting, because of the gender differences found in narcissism facets, it is likely that the birth-cohort differences cannot be entirely attributable to the gender composition of the sample. In other words, samples comprising entirely of women were often still higher than samples containing both men and women. Of course, the differences we observed in the current study are likely the combination of substantive changes in psychological and demographic variables. Future research can have more balanced gender designs with respect to gender and other demographic characteristics.

* How do researchers measure narcissism changes in secondary data sources? In the current study, we relied on just one of many models of narcissism—Wink’s (1992a) conceptualization of hypersensitivity, willfulness, and autonomy. The operationalization and measurement history of narcissism is long and controversial, with several alternative models and measures portending to measure narcissism’s core features (Ackerman et al., 2011; Back et al., 2013; Campbell, Bonacci, Shelton, Exline, & Bushman, 2004; Glover, Miller, Lynam, Crego, & Widiger, 2012; Konrath, Meier, & Bushman, 2014; Pincus et al., 2009; Raskin & Terry, 1988; Wink, 1991). Our choice to use Wink’s conceptualization was primarily (and necessarily) driven by an effort to maximize the available data across the six studies.
However, in directly comparing the CAQ scales to more contemporary measures of narcissism (e.g., NPI, NARQ), we found that one facet in particular—autonomy—had relatively small overlap with other measures of narcissism (rs < |.28|). In previous research, the CAQ autonomy scale was often conceptualized as a form of adaptive narcissism (Cramer, 2011a; Wink, 1992a). However, in our very preliminary examinations, it appears that these items capture autonomy as a broader individual difference characteristic than a form of adaptive narcissism per se. Thus, our and others’ necessary reliance on using CAQ autonomy as a form of adaptive narcissism is likely misplaced. This limitation speaks to a larger concern for researchers who use secondary data to examine changes in characteristics that were not intended to be measured in the original data collection. Running additional studies to verify the convergent and criterion validity of researcher-generated scales in the context of secondary data analysis is not particularly expensive; the benefits of adopting such an approach are numerous though. The danger is in liberally using measures form secondary data sources without formally examining their conceptual meaning and empirical validity. Indeed, such ready adoption of measures is likely one of the contributing factors for why some concepts (e.g., adaptive narcissism) might become so broad as to assume related, but technically different concepts (e.g., autonomy)(Haslam, 2016). For hypersensitivity and willfulness, there was substantial overlap with other measures of narcissism, suggesting that these aspects of narcissism may be less problematic to continue using in future research. Nevertheless, more formal efforts to examine how the developmental trajectories of different narcissism inventories vary would be an exciting direction for future research.
Although our use of a quasi-accelerated longitudinal design had many strengths, including measuring changes in narcissism over a 64-year period and across multiple birth-cohorts, its use also presents several unique limitations. For example, no one person was followed from age 13 to age 77, severely limiting our ability to truly isolate lifespan changes in narcissism. Further, because the studies had large gaps in between assessment points, we were limited in our ability to assess more nuanced developmental changes.

European elections & support for governing parties over time: Anticipating being completely satisfied (as compared to completely dissatisfied) is associated with around a 6.5% higher probability of declaring support for such parties

Happiness and Voting: Evidence from Four Decades of Elections in Europe. George Ward. American Journal of Political Science, December 9 2019. https://doi.org/10.1111/ajps.12492

[An earlier working paper version of this article circulated under the title “Is Happiness a Predictor of Election Results?”]

Abstract: There is a growing interest among policy makers in the use of subjective well‐being (or “happiness”) data to measure societal progress, as well as to inform and evaluate public policy. Yet despite a sharp rise in the supply of well‐being‐based policy making, it remains unclear whether there is any electoral demand for it. In this article, I study a long‐run panel of general elections in Europe and find that well‐being is a strong predictor of election results. National measures of subjective well‐being are able to explain more of the variance in governing party vote share than standard macroeconomic indicators typically used in the economic voting literature. Consistent results are found at the individual level when considering subjective well‐being and voting intentions, both in cross‐sectional and panel analyses.


Discussion

Various countries around the world have recently begun to go “beyond GDP” by measuring subjective well-beingon a large scale and using the data (1) as a general measure of societal success and progress, (2) to guide and inform policy decisions, and (3) to evaluate the outcomes of government programs (Durand 2018; O’Donnell et al.2014). These practices are likely to continue to grow,in part because SWB is able to pick up the benefitsof a great deal of government activity that traditional economic outcomes may struggle to (Krueger and Stone2014).19 

Yet despite the recent sharp rise in the supply of SWB-based public policymaking, an open empiricalquestion is whether there is any electoral demand for it.The findings presented in this article suggest there maybe significant electoral incentives for politicians seeking reelection to consider SWB when deciding upon policypriorities.Global, cognitive evaluations of life are currently themost widely used measure of SWB by researchers inthe economic literature as well as by policy makers, but life satisfaction is only one component of SWB. Large-scale data on the emotional states of citizens are becoming more prevalent and are beginning to provide policymakers with a fuller picture of national SWB (Kahneman et al. 2004; Krueger and Stone 2014). Further research should investigate the extent to which measures of posi-tive and negative affect, as well as eudaemonic measures of purpose, are able to add to our understanding of vot-ing behavior.A further dimension of SWB is temporal: Although the main analysis of this article studies voters’ currentlevels of life satisfaction, it may be that future expectation sof life satisfaction are just as—or even more—important in driving vote choice. In a subset of waves of the German SOEP, respondents were asked about their anticipated life satisfaction in 5 years’ time, using the same 0–10 response scale as with current life satisfaction. In Table A1, I find that people’s future life satisfaction dominates current life satisfaction when it comes to predicting support for governing parties within people over time. Anticipating being completely satisfied (as compared to completely dissatisfied) is associated with around a 6.5% higher probability of declaring support for a governing party.20 

Open to further research is the broader questionof what array of determinants of SWB, and potentiallywhich domains of SWB, drive the link between nationalhappiness and election results—and ultimately what incumbents might do to improve their chances of reelec-tion. Although SWB has been shown to be determinedby a host of policy-relevant yet noneconomic variables,including physical and mental health, environmenta lquality, social cohesion, crime and corruption, quality of government services, and education (see, e.g., Clark et al.2018; Diener et al. 1999; Dolan, Peasgood, and White2008), the analysis of Liberini, Redoano, and Proto (2017)suggests voters may also reward/punish incumbent politicians for boosts and dips in their happiness that arecaused by factors outside of government control. Further research may continue to investigate (1) the extent to which voters are able (or willing) to filter which elementsof their well-being provide useful information about thequality and effort of incumbents, and (2) the theoretical implications of this for our understanding of democratic accountability.21 

Although SWB is a stronger predictor of incumbent vote share in general elections than economic growth, un-employment, or inflation, macroeconomic variables are nevertheless significant predictors of government elec-toral success conditional on national happiness. Equally,at the individual level, SWB and personal finances are in-dependently predictive of voting intentions. This suggests politicians face multiple incentives to improve people’s economic as well as broader non economic well-being.Future theoretical work may look to model these dynam-ics within a multitask political agency framework.22 

The data used here are observational, and it isworth reiterating that it is not possible to interpret thee mpirical associations presented in this article causally. 

Rather, the analysis is focused on determining what bestpredicts the electoral fate of governing parties. Despite this important caveat, however, a causal interpretation of the findings is suggested by Liberini, Redoano, and Proto(2017), who leverage exogenous variation in SWB to demonstrate a causal mechanism between happiness and self-reported incumbent voting intentions. The analysis presented here suggests that this effect is also evident atthe national level, across 15 countries over four decades, in real-world elections.

When male employees are assigned to male managers, they are promoted faster in the following years than they would have been if they were assigned to female managers, due to socialization between employees & managers

The Old Boys' Club: Schmoozing and the Gender GapZoë B. Cullen, Ricardo Perez-Truglia. NBER Working Paper No. 26530, December 2019https://www.nber.org/papers/w26530

Abstract: The old boys’ club refers to the alleged advantage that male employees have over their female counterparts in interacting with powerful men. For example, male employees may schmooze with their managers in ways that female employees cannot. We study this phenomenon using data from a large financial institution. We use an event study analysis of manager rotation to estimate the causal effect of managers’ gender on their employees’ career progression. We find that when male employees are assigned to male managers, they are promoted faster in the following years than they would have been if they were assigned to female managers. Female employees, on the contrary, have the same career progression regardless of the manager’s gender. These differences in career progression cannot be explained by differences in effort or output. This male-to-male advantage can explain a third of the gender gap in promotions. Moreover, we provide suggestive evidence that these manager effects are due to socialization between male employees and male managers. We show that these manager effects are present only if the employee works in close proximity to the manager. We use survey data to show that, after transitioning to a male manager, male employees spend more time with their managers. Finally, we study a shock to socialization within males, based on the anecdotal evidence that employees who smoke tend to spend more time together. We find that when male employees who smoke switch to male managers who smoke, they spend more of their breaks with their managers and are promoted faster in the following years. Moreover, the effects of these smoking manager switches are similar in timing and magnitude to the effects of the gender manager switches.

Dim Ambient Lighting Increases Game Play Duration and Total Spend

Hidden in the Dark: Dim Ambient Lighting Increases Game Play Duration and Total Spend. Jasmina Ilicic, Stacey M. Baxter. Journal of Gambling Studies, December 10 2019. https://link.springer.com/article/10.1007/s10899-019-09921-5

Abstract: It has been suggested that much like commercial environments (e.g., retailing), the situational characteristics of gambling environments form an important determinant of gambling behavior. However, no research has examined whether ambient lighting in gaming venues can have unintended consequences in terms of gambling behavior. The results of three experimental laboratory studies show that game play duration and total spend increase when ambient lighting is dim (vs. bright). Process evidence suggests that this phenomenon occurs as ambient lighting influences risk-taking, which in turn increases game play duration and total spend. Further, evidence is provided that the effect of dim (vs. bright) ambient lighting reduces risk-taking and subsequent game play duration and total spend when an individual’s self-awareness is facilitated (i.e., screening between gaming machines is removed). This research has implications in terms of public policy regarding the determination of minimum lighting levels in venues as a means to decrease gambling-related harm. Moreover, while gaming venues can use these insights and their ambient lighting switches to nudge individuals toward reducing their game play duration and total spend, gambling-afflicted consumers can opt for gambling venues with bright ambient lighting and those without screened gaming machines.

Keywords: Gambling Ambient lighting Risk-taking Self-awareness Game play duration Total spend


General Discussion

We introduced ambient lighting in gaming venues as an important sensory situational char-acteristic that can be easily adjusted to influence gambling behavior in terms of game play duration and total spend. Results of this research provide evidence that dim ambient light-ing can cause increased gambling behavior in terms of game play duration and total spend, whereas bright ambient lighting can decrease game play duration and total spend. We pro-vide evidence of the process underlying this effect, which we confirm is risk-taking. Dim ambient lighting, as opposed to bright ambient lighting, enhances risk-taking propensity, increasing subsequent game play duration and total spend. Further, we propose and dem-onstrate that the negative effect of dim ambient lighting on risk-taking and on subsequent game play duration and total spend is attenuated when self-awareness is facilitated through the absence of screening between machines. In other words, risk-taking and subsequent game play duration and total spend is decreased when individuals are made more self-aware through the absence of screening between gaming machines in dim lighting con-ditions. Our results provide insights into reducing gambling-related harm in gaming ven-ues and also provide interesting insights for public policy related to luminance levels and screening in gaming venues.In light of our findings, gaming venue operators could encourage more responsible gambling behaviors by adjusting their ambient lighting levels to reduce gambling-related harm to their customers and to decrease their risk-taking propensity. As such, we suggest the following guidelines be adopted by gaming venues to encourage responsible gambling behavior:1. Ensure that ambient lighting levels are bright.2. Use lighting adjusters to increase the lux within gaming venues or alter the bulbs used to brighter levels.3. Avoid the use of screening between gaming machines, which increases risk-taking and gambling behavior in dim ambient lighting conditions.We also suggest that our findings have important implications for public policy and regulations in the gaming industry. We suggest that public policymakers review the dim-luminance lighting level requirements for gaming venues. Brighter ambient lighting is suggested in order to reduce the harm associated with dim ambient lighting on gambling behavior. Furthermore, public policymakers should incorporate screening requirements into regulations. Specifically, we suggest the removal of screening between machines to create an open space that enhances self-awareness and reduces gambling-related harm.Research has found that lighting plays a significant role in influencing gambling behav-ior (e.g., Stark et al. 1982; Griffiths and Parke 2003; Spenwyn et al. 2010). Researchers have typically examined the effect of lighting in terms of the colors that are emitted from gaming machines themselves. For example, research suggests that individuals take greater risks and bet more frequently when gambling on machines that omit red, rather than blue, light (Stark et al. 1982). However, our research extends this body of work by providing evidence that the situational characteristics of ambient lighting within gaming venues can influence gambling behavior in terms of game play duration and total spend. This is espe-cially important for gaming venue managers in adjusting lighting levels to nudge more responsible gambling behavior.

Limitations and Directions for Future Research

This research is not without limitations. This research was conducted within an artifi-cial environment in a controlled laboratory setting. Although a controlled environment is desirable to be able to make casual inferences regarding the effects of ambient lighting on game play duration and total spend, the laboratory environment did not represent real-life gaming environments that include sounds, smells, flashing and colored lights from gam-ing machines, and a variety of gaming machine options available. We suggest that future research conduct a field study in which the ambient lighting levels are adjusted in real gam-ing venues, collecting data regarding game play duration and total spend, as well as other key gambling-related measures, such as numbers of bets and types of bets. Although it is acknowledged that field studies are difficult to implement, venue operators may be more willing to cooperate if only one smaller room were used as a test. This would minimise the impact of the study on operations.Our research is also limited as it only included a student sample and was conducted in Australia. We argue, however, that adolescents or young adults are particularly susceptible to the harm associated with situational characteristics that may encourage gaming machine use, as research shows that electronic gaming machine players are more likely to be in the 18- to 24-year-old age group (Delfabbro 2012). We also suggest that Australia is a par-ticularly vulnerable country to examine, as it holds the 2016 record in terms of losses per inhabitant; losing on average A$990 per adult resident, half of which is due to gaming machines (Pancani et al. 2019). Future research should explore the effects of ambient light-ing on gambling behavior across a wider sample of individuals, as well as across a variety of countries, to examine whether there are any age-related or cultural differences that may be observed.Furthermore, our study examined only two levels of luminance (i.e., 500 lux (bright) and 40 lux (dim)). Future research should examine varying levels of luminance to identify the point of luminance at which ambient lighting can have negative effects on gambling-related behavior. This would also have important implications regarding the specificity of luminance levels for public policy regulations.

First evidence of the acoustic correlates of cooperative behaviour: Men with lower‐pitched voices & greater pitch variations were more cooperative; however, testosterone did not influence cooperative behaviours

Does he sound cooperative? Acoustic correlates of cooperativeness. Arnaud Tognetti, Valerie Durand, Melissa Barkat‐Defradas, Astrid Hopfensitz. British Journal of Psychology, December 9 2019. https://doi.org/10.1111/bjop.12437

Abstract: The sound of the voice has several acoustic features that influence the perception of how cooperative the speaker is. It remains unknown, however, whether these acoustic features are associated with actual cooperative behaviour. This issue is crucial to disentangle whether inferences of traits from voices are based on stereotypes, or facilitate the detection of cooperative partners. The latter is likely due to the pleiotropic effect that testosterone has on both cooperative behaviours and acoustic features. In the present study, we quantified the cooperativeness of native French‐speaking men in a one‐shot public good game. We also measured mean fundamental frequency, pitch variations, roughness, and breathiness from spontaneous speech recordings of the same men and collected saliva samples to measure their testosterone levels. Our results showed that men with lower‐pitched voices and greater pitch variations were more cooperative. However, testosterone did not influence cooperative behaviours or acoustic features. Our finding provides the first evidence of the acoustic correlates of cooperative behaviour. When considered in combination with the literature on the detection of cooperativeness from faces, the results imply that assessment of cooperative behaviour would be improved by simultaneous consideration of visual and auditory cues.



Discussion

Several acoustic features influence the perception of how trustworthy and cooperative the speaker is (Belinet al., 2017; Knowles & Little, 2016; Montano et al., 2017; O’Connor& Barclay, 2017; Oleszkiewicz et al., 2017; Ponsotet al., 2018; Tigue et al., 2012; Tsantani et al., 2016). Their influence could stem from the pleiotropic effect of testosterone on both acoustic features and cooperative behaviours (O’Connor & Barclay, 2017). It is unknown, however, whether acoustic features are associated with actual cooperativebehaviour and whether testosterone mediates this association. In this study, we presentevidence that both vocal pitch and its variations are related to cooperative behaviour in anincentivized social-dilemma game: the public good game. However, no effect oftestosterone level on cooperation, or on any of the other acoustic features studied, wasfound. Overall, ourstudy providesthe first evidence of the existence of acoustic correlates of cooperativeness.

Specifically, our results indicate that men’s contributions to the public good aresignificantly and negatively associated with fundamental frequency and significantly andpositively with its variations. When we compared the acoustic traits between conditional cooperators and free-riders (the two main categories as defined in Fischbackeret al.(2001)), we found that conditional cooperators exhibit significantly higher pitch variations than free-riders. Taken together, our results suggest that highly cooperative men have deeper voices and exhibit greater variations in their intonation compared to less cooperative men.The present results are consistent with the only previous study examining jointly the influence of vocal pitchand its variations on the perception of a speaker’s cooperativeness (Knowles & Little, 2016). Indeed, Knowles and Little (2016) found that male voices wereperceived as the most likely to cooperate when they exhibited high pitch variations in combination with a low pitch (although it was found for women’s but not for men’sratings). Vocal pitch and its variation are, thus, associated in the same way with bothcooperative behaviours and perceived cooperativeness. It, therefore, indicates that inferences of cooperativeness from voices might actually facilitate the detection ofcooperative partners. This sets the ground for future research, namely whether particular ombinations of acoustic traits influence ratings of cooperativeness and to which degreethese acoustic cues of cooperativeness are reliable or could be manipulated throughconscious control of the speaker (e.g., by lowering voice pitch or by increasing speech’sintonation).Behavioural decisions in the public good games (contributions to the public good) andin the trust game (amounts sent to the other player) are highly correlated (Galizzi & Navarro-Mart ınez, 2018; Peysakhovich, Nowak, & Rand, 2014), which suggests a strongassociation between cooperativeness and trustworthiness. Hence, similarly to cooper-ativeness, vocalpitchand itsvariationsarealso likelytobe used ascues oftrustworthiness.

In fact, both acoustic features influence perception of how trustworthy a speaker is. For example, lower-pitched male voices and voices with high pitch variations are perceived asmore trustworthy in general (Belinet al., 2017; Oleszkiewicz et al., 2017; Schirmer et al.,2019; Tsantaniet al., 2016) or when trust is linked to the political context (Klofstadet al.,2015, 2012; Tigueet al., 2012). It remains unknown, however, whether these acousticfeatures correlate with actual trustworthiness, and not only with perceived trustworthiness.The existence of vocal cues of cooperativeness could stem from the pleiotropic effectof testosterone on both cooperative behaviours (Burnham, 2007; Diekhofet al., 2014; Reimers & Diekhof, 2015; Takagishiet al., 2011) and vocal pitch (Dabbs & Mallinger,1999; Evanset al., 2008; Putset al., 2012). As testosterone has immunosuppressive effects (immunocompetence handicap hypothesis: Folstad & Karter, 1992; Rantalaet al., 2012;but see: J. Nowak, Pawłowski, Borkowska, Augustyniak, & Drulis-Kawa, 2018), men with(costly) lower pitch might benefit from a higher biological quality (Arnocky, Hodges-Simeon, Ouellette, & Albert, 2018; Hodges-Simeonet al., 2015). In addition, they may also be more socially dominant (Puts et al., 2012; Puts, Gaulin, & Verdolini, 2006; Putset al.,2007). Accordingly, because of their underlying qualities, including access to resources,men with lower pitch would be more cooperative than men with higher pitch becausethey could better afford the costs associated with cooperative behaviours while receiving reputational benefits (Raihani & Smith, 2015; Sylwester & Roberts, 2010; Tognetti,Berticat, Raymond, & Faurie, 2012; Tognettiet al., 2016). This condition-dependent mechanism could ensure the reliability of the vocal cues of cooperativeness. However, in the present study testosterone levels did not seem to affect cooperation or any of theacoustic features studied. Testosterone is a multiple-effect hormone which is influencedby numerous biological and environmental factors and pathways. As such, it is generally difficult to correlate testosterone levels to other biological or behavioural traits. Inaddition, we could only collect one sample of saliva for hormonal assays, which might notaccurately reflect a participant’s basal testosterone level.

Although the present study retains many strengths, it is also subject to several limitations. In particular, it is the first to investigate the existence of acoustic correlates ofcooperativeness in speech production. However, the investigation is restricted to Frenchmen. To provide broader conclusions, it should not only be extended to women, but toother populations as well. In addition, we did not conduct a perceptual study using ourrecordings to examine whether listeners use acoustic features as a social cue in abehavioural economic task. Indeed, we recorded individuals’ free speech due to itsstronger ecological validity (Puts e tal., 2007; Suireetal., 2018), but this type ofrecordingsis not suitable for perceptual studies, as recordings roughly differ in duration and semanticcontent. Finally, we used state-of-the-art methodology in economics to quantify andcategorize individuals according to type (Fischbacheret al., 2001). However, wecompared two categories with unbalanced sample sizes (NFree-rider=11 vs.Nconditionalcooperator=44) and the sample size of the free-riders was limited (although its proportion(18%) mimics the proportion found in the general French population; Frey, 2017). Hence,we cannot exclude the possibility that the acoustic differences found between free-ridersand conditional cooperators arose from this specific and particular sample of 11 free-riders. It seems, nevertheless, unlikely as the results found using this categorization arequalitatively similar to the ones we found using a continuous measure of cooperativeness(i.e., contributions to the public good).

To conclude, the present study provides evidence that at least two acoustic features(vocal pitch and its local variations) could be used as cues of cooperativeness. Facial cuesenable individuals to discriminate between high and low cooperative individuals with anabove chance accuracy (Bonnefon et al., 2013, 2017; Fetchenhauer et al., 2010; Little et al., 2013; Oda, Naganawa, et al., 2009; Reedet al., 2012; Stirrat & Perrett, 2010, 2012;Tognettiet al., 2013) but the accuracy of face-based cooperation detection is rather low (Bonnefon et al., 2017). Hence, by highlighting the fact that cooperativeness is advertisedby several cues across multiple sensory modalities, our findings pave the way for further investigations examining whether the assessment of cooperative behaviour is improved by simultaneous consideration of both visual and auditory cues.

Economic status cues from clothes affect perceived competence from faces — The Habit Does Make The Monk in the End

Economic status cues from clothes affect perceived competence from faces. DongWon Oh, Eldar Shafir & Alexander Todorov. Nature Human Behaviour, December 9 2019. https://www.nature.com/articles/s41562-019-0782-4

Abstract: Impressions of competence from faces predict important real-world outcomes, including electoral success and chief executive officer selection. Presumed competence is associated with social status. Here we show that subtle economic status cues in clothes affect perceived competence from faces. In nine studies, people rated the competence of faces presented in frontal headshots. Faces were shown with different upper-body clothing rated by independent judges as looking ‘richer’ or ‘poorer’, although not notably perceived as such when explicitly described. The same face when seen with ‘richer’ clothes was judged significantly more competent than with ‘poorer’ clothes. The effect persisted even when perceivers were exposed to the stimuli briefly (129 ms), warned that clothing cues are non-informative and instructed to ignore the clothes (in one study, with considerable incentives). These findings demonstrate the uncontrollable effect of economic status cues on person perception. They add yet another hurdle to the challenges faced by low-status individuals.


Discussion 
Across studies, we found that economic status clothing cues influenced competence judgements of faces. The effect persisted when faces were presented very briefly (that is, 129 ms), when informa-tion was provided related to the person’s profession and income, when formal clothing was replaced by more casual clothing, when participants were advised to ignore the clothing, when they were warned that there was no relationship between clothing and com-petence before choosing rather than rating faces, and when partici-pants were offered a monetary reward for accuracy. These findings support the notion of uncontrollable effects of minor contextual cues in face perception, and are consistent with a large body of research that finds people spontaneously encode the context surrounding a face when making social judgements10,11,13,19 (while our focus has been on competence judgements, similar, if attenuated, effects can be observed for other traits, such as trustworthiness; see Supplementary Results and Supplementary Fig. 8). 

The strong and persistent effects we observed are consistent with theoretical work16,18 and empirical findings16,17, showing a robust tendency for people of lower economic status to be perceived as less competent and to be disrespected20, often leading to social exclu-sion with detrimental effects on physical and emotional health21. Poverty is a place where many challenges—physical, social and psychological—converge: being perceived as of lower competence and disrespected adds to those challenges, and can exacerbate cognitive load and hamper performance, thereby potentially prov-ing self-fulfilling22,23. 

To overcome a bias, one needs not only to be aware of it but to have the time, attentional resources and motivation to counteract the bias24. In our studies, we warned participants about the potential bias, presented them with varying lengths of exposure, gave them additional information about the targets and offered financial incen-tives, all intended to alleviate the effect. None of these interventions were effective, however. While it is possible that higher incentives and greater experience could reduce the bias, its persistence in the face of our various manipulations is impressive. 

The present findings demonstrate that economic status cues from clothes naturally intervene in people’s assessments of compe-tence. This is consistent with research showing that people associ-ate status with competence in stereotypes of social groups16,25. This strong status–competence association suggests that any attempt at independen manipulation of the apparent competence and eco-nomic status of a person may need to resort to explicit and salient manipulations, rather than fairly subtle cues. 

The poor clothing cues in our studies were benign compared to real-world poverty signals. Recent work has shown that people can accurately guess others’ social class from brief exposure to photos or speech recordings26. We might thus expect people wearing truly impoverished clothes or exhibiting other peripheral signs of poverty to encounter substantial low-competence stereotyping, both when perceivers think fast as well as when they have more time to deliberate.Stereotypes about rich and poor individuals are common, prom-inent and consequential. Just as the clothing cues in our studies led to differential disambiguation of facial competence, views about a person’s economic background can lead to notably different inter-pretations of what is otherwise ambiguous performance27. Beyond their immediate impact, an important question for future research concerns the extent to which we might be able to transcend first impressions.

Our results indicate that conscientiousness is positively associated with wages, while ***agreeableness***, extraversion, and neuroticism are associated with negative returns

McLean, D., Bouaissa, M., Rainville, B., & Auger, L. (2019). Non-Cognitive Skills: How Much Do They Matter for Earnings in Canada?. American Journal of Management, 19(4). https://doi.org/10.33423/ajm.v19i4.2392

Abstract: Evidence from different countries suggests that non-cognitive skills play an important role in wage determination and overall social outcomes, but studies for Canada are scarce. We contribute to filling this gap by estimating wage regressions with the Big Five traits using the Longitudinal and International Study of Adults. Our results indicate that conscientiousness is positively associated with wages, while agreeableness, extraversion, and neuroticism are associated with negative returns, with higher magnitudes on agreeableness and conscientiousness for females. Cognitive ability has the highest estimated wage return so, while significant, non-cognitive skills do not seem to be the most important wage determinant.

Keywords: Management, Labour Market, Returns to Skills, Non-Cognitive Skill, Cognitive Skill, Wage Regressions, Personality Traits, Five-Factor Model


REVIEW OF RELEVANT LITERATURE

Gary Becker set the stage for the present analysis in the 1960s by being among the first to formally
describe the process of engaging in education as an individual investment decision, with pay-offs in the form of higher future wages. This ultimately motivated Jacob Mincer ten years later to develop the famous Mincer Equation, which attributed differences in cross-sectional wages to differentials in
schooling and experience.
In this framework, cognitive ability is seen as an endowment factor which differentiates marginal
productivity, helping to partially explain wage differentials. Cawley, Heckman and Vytlacil (2001)
estimate the magnitude of this effect using the NLSY survey, showing that a one standard deviation
increase in innate thinking ability is associated with between 10% and 15% higher wages. Both Cawley, Heckman and Vytlacil (2001) and Cunha, Heckman, Lochner and Masterov (2006) also suggest that, rather than being a fixed endowment, cognitive ability can evolve over a lifetime, dynamically affecting wages both through increased schooling and higher productivity. More recently, Hanushek, Schwerdt, Wiederhold, and Woessmann (2015) used data from the 2012 Programme for International Assessment of Adult Competencies (PIAAC) survey to examine the relationship between numeracy and literacy abilities and wages. They find evidence of a positive relationship, however concede that the selection process of high-skilled individuals into higher levels of schooling may confound the true effect. Cawley, Heckman and Vytlacil (2001) find that cognitive ability, while statistically a significant predictor of income, explains only a small fraction of the variation in income, suggesting that other, unobserved factors may also play a key role in determining economic success. This then opens the discussion on the ability of other unobserved traits to affect economic outcomes of an individual. Mischel (1973) argued that personality traits are one fundamental factor in helping determine how an individual responds to stimulus in their environment in a consistent way. Thus we could reasonably expect personality traits to form a part of an individual's stock of human capital and to be associated with labour market outcomes.
Non-cognitive skills are notoriously difficult to measure, given the lack of consensus in the literature
on their definition and which proxies are most appropriate. One strategy that has received widespread
acceptance involves quantifying non-cognitive skills using the Five Factor Model (Goldberg, 1971; Costa and McCrae, 1985). The Five Factor model posits the existence of five distinct but not necessarily orthogonal personality characteristics, the Big Five personality traits. These characteristics, Openness to Experience, Conscientiousness, Extraversion, Agreeableness and Neuroticism, are traits that are theorised to help determine a personality that is consistent across time and situation. The Big Five are derived from well-established definitions in the psychology literature dating back to the seminal lexical work by Allport and Odbert (1936), and have seen applications in numerous economic studies attempting to find links between personality and learning, work, romantic and social outcomes. The traits are characterised as relative rigid over the life cycle, particularly by the age of 25-30 (Cobb-Clark and Schurer, 2011).1
Each of the Big Five dimensions in turn are made up of several smaller personality facets, which are
typically more easily measured using survey questions. John and Srivastava (1999), McCrae and John
(1992) and others outline the history of the Five Factor Model and in particular highlight which facets are associated with which personality dimensions. Openness to Experience, for example, is associated with concepts at the intersection of intelligence and ingenuity such as imagination, being artistic and
thoughtfulness. Neuroticism, on the other hand, is associated with facets such as anxiety, selfconsciousness, and vulnerability. Facets for the other three dimensions can be found in Appendix A.
While the Five Factor Model has its criticisms, its robustness, relative ease of interpretation and
applicability to survey questionnaires have made it an attractive option for economists interested in
studying personality. Numerous papers have attempted to tie these personality traits to economic
outcomes, including two early papers by Barrick and Mount (1991) and Tett, Jackson and Rothstein
(1991) who review work linking these characteristics to the labour market. Barrick et al. in particular find that conscientiousness has a strong positive correlation with job performance for all occupational groups, while the impact of the other four traits are differentiated by occupation type. Judge, Higgins, Thoresen Barrick and Murray (1999) use longitudinal data to find that conscientiousness and extraversion are positively related to various metrics of career success, while neuroticism and agreeableness are negatively correlated. Nyhus and Pons (2005), Mueller and Plug (2006), and Heineck and Anger (2010) unanimously find that conscientiousness is the most positively associated with and predictive of higher earnings, while agreeableness and neuroticism may be associated with negative returns. Nyhus and Pons find extraversion to have negative returns, while Mueller and Plug find this is true only for females.
Mueller and Plug are also the only to find a positive estimated return to openness to experience. Almlund (2011) affirm these main findings, suggesting that conscientiousness is the best overall predictor of job performance, though this correlation decreases with increased job complexity. Openness to experience and extraversion are found to be positively associated with job performance but only indirectly through education. A positive relationship between extraversion and education is not a consistent finding, however, as Goldberg (1998) and van Eijck and de Graaf (2004) find a slight negative relationship between educational attainment and neuroticism, extraversion and agreeableness.
Of course, using the Big Five dimensions is also not the only identification strategy used in the
literature. In a highly influential paper, Heckman and Rubinstein (2001) use enrolment in a General
Education Development (GED) program as a measure of non-cognitive skill and find that differences in social skills between GED and high school graduates explain a large proportion of the wage differential between the two groups. Heckman, Stixrud and Urzua (2006) used the Rotter Locus of Control Scale and the Rosenberg Self-Esteem Scale derived from the 1979 National Longitudinal Survey of Youth (NLSY) as their preferred indicators, again finding evidence that non-cognitive skills strongly influence schooling decisions and wages. In another influential paper, Lindqvist and Vestman (2011) use an administrative dataset of Swedish military enlistees to derive a measure of personality as assessed by interviews with professional psychologists. Their results show that individuals who struggle with low earnings and unemployment are systematically characterised by low non-cognitive skill scores. Borghans, ter Weel and Weinberg (2014) use data from the NLSY to determine if differences in non-cognitive skills are able to predict wage differentials between minority and gender groups. In a related line of thinking, they look at whether returns to what they term ‘people skills’ have increased or decreased with time. They find that, between the 1970s and 1990s, non-cognitive skills have become increasingly important determinants of wages, which they attribute to an increased substitution of technology to perform tasks. This line of reasoning is affirmed by Weinberger (2014), Deming (2017) and Edin, Fredriksson, Nybom and Öckert (2017), who each find evidence of returns to non-cognitive skills increasing over time, and their growing importance to working in high-paying managerial occupations.
While the existing literature provides evidence for a wide array of OECD countries, to our
knowledge there is only one paper that examines the data specifically for Canada. Green and Riddell
(2003) use Canadian male respondents from the 1994 International Adult Literacy Survey (IALS) to
conclude that an individual's cognitive abilities are primarily established during formal schooling. In line with the literature, they also suggest that cognitive and non-cognitive skills affect earnings both
independently and through an interplay between them. Our work in this paper thus builds upon the
foundation established by Heckman, Green and Riddel and others that already exists in the literature. We contribute to the knowledge gap in the literature first by directly measuring both cognitive and noncognitive skills and focussing on how cognitive and non-cognitive skills are independently rewarded in the Canadian labour market. To our knowledge, we are also among the first to use the Longitudinal
International Survey of Adults (LISA) to analyse this question, which will allow us to offer further
empirical evidence to the literature from an under-studied OECD country. The advantages of using the LISA survey are two-fold. First, it offers some of the most recent data available for studying earnings and wage data, and will also allow for the robust examination of heterogeneity in cognitive and non-cognitive skill returns across a wide spectrum of Canadian workers.1

Aphantasics recalled significantly fewer object details than controls, and showed a reliance on verbal strategies; showed equally high spatial accuracy as controls, and draw fewer false objects than controls

Quantifying Aphantasia through drawing: Those without visual imagery show deficits in object but not spatial memory. Wilma A. Bainbridge, Zoë Pounder, Alison F. Eardley, Chris I. Baker. bioRxiv, Dec 2019. https://doi.org/10.1101/865576

Abstract: Congenital aphantasia is a recently identified experience defined by the inability to form voluntary visual imagery, with intact semantic memory and vision. Although understanding aphantasia promises insights into the nature of visual imagery, as a new focus of study, research is limited and has largely focused on small samples and subjective report. The current large-scale online study of aphantasics (N=63) and controls required participants to draw real-world scenes from memory, and copy them during a matched perceptual condition. Drawings were objectively quantified by 2,700 online scorers for object and spatial details. Aphantasics recalled significantly fewer object details than controls, and showed a reliance on verbal strategies. However, aphantasics showed equally high spatial accuracy as controls, and made significantly fewer memory errors, with no differences between groups in the perceptual condition. This object-specific memory impairment in aphantasics provides evidence for separate systems in memory that support object versus spatial details.


 ---
Aphantasics draw fewer false objects than controls

Finally, we quantified the amount of error in participants’ drawings from memory by group. AMT workers (N=5 per drawing) viewed a drawing and its corresponding image and wrote down all objects in the drawin gs that were not pre sent in the or iginal image (essentially quantifying false object memories). Significantly more memory drawings by controls contained false objects than drawings by aphantasics (control: 12 drawings, aphantasic: 3 drawings; ...); examples can be seen in Fig. 5. Similarly, significantly more objects drawn by controls were false alarms than those drawn by aphantasics ... This indicates that control participants were making more memory errors, even after controlling for the fewer number of objects drawn overall by aphantasics. Interestingly, all aphantasic errors (se e Fig. 5) were transpositions from another image and drawn in the cor rect location as the original object (a tree from the bedroom to the living room, a window from the kit chen to the l ivi ng room, and a ceiling fan from the kitchen to the bedroom). In contrast, seve ral fa lse memories from controls were objects that did no t exist across any image but instead appeared to be filled in based on the scene category (e.g., a piano in the living room, a dresser in the bedroom, logs in the living room). No perception drawings by participants from either group contained false objects.

People trust happy-sounding artificial agents more, even in the face of behavioral evidence of untrustworthiness

If your device could smile: People trust happy-sounding artificial agents more. Ilaria Torre, Jeremy Goslin, Laurence White. Computers in Human Behavior, December 9 2019. https://doi.org/10.1016/j.chb.2019.106215

Highlights
• Smiling can be heard in the voice without any visual cue.
• This ‘smiling voice’ elicits higher trusting behaviors than a neutral one.
• The higher trust persists even when the speaker is untrustworthy.
• This has implications for the design of voice-based artificial agents.

Abstract: While it is clear that artificial agents that are able to express emotions increase trust in Human-Machine Interaction, most studies looking at this effect concentrated on the expression of emotions through the visual channel, e.g. facial expressions. However, emotions can be expressed in the vocal channel too, yet the relationship between trust and vocally expressive agents has not yet been investigated. We use a game theory paradigm to examine the influence of smiling in the voice on trusting behavior towards a virtual agent, who responds either trustworthily or untrustworthily in an investment game. We found that a smiling voice increases trust, and that this effect persists over time, despite the accumulation of clear evidence regarding the agent’s level of trustworthiness in a negotiated interaction. Smiling voices maintain this benefit even in the face of behavioral evidence of untrustworthiness.

Keywords: TrustSmiling voiceVirtual agents


5. Discussion

Using an investment game paradigm, we found that positive vocal emotional expression – smiling voice – increases participants’ implicit trust attributions to virtual agents, compared with when agents speak with an emotionally neutral voice. As previously observed, the monetary returns of the agent also affected implicit trust, so that participants invested more money in the agent that was behaving generously. Critically, however, there was no interaction between behavior and vocal emotional expression: smiling voice enhanced trust regardless of the explicit behavioral cues that the virtual agent provided to its trustworthiness. The effect of smiling voice in the game, supported by our questionnaire findings, adds to previous studies on emotional expression, showing that the display of a positive emotion increases trust and likeability, even in the vocal channel (Scharlemann et al., 2001; Krumhuber et al., 2007; PentonVoak et al., 2006). Smiling was a consistent predictor of investments overall. That is to say, while participants’ investments were primarily driven by the virtual player’s generosity or meanness, they also overall invested more money in the smiling agents. This contrasts with the predictions of the EASI model (Van Kleef et al., 2010), according to which the display of a positive emotion in an incongruent context (such as the mean behavior condition) should elicit uncooperative behaviors. While Van Kleef et al. (2010) listed social dilemma tasks based on Prisoner’s Dilemma among possible competitive situations, it is possible that participants in an iterated investment game view it as an essentially cooperative task. Specifically,
while typical Prisoner’s Dilemma tasks involve a dichotomous choice (cooperate/defect), in our experiment, even in the mean condition, the agent was still returning a (small) amount of money, which might have been seen as a partially cooperative signal by participants. If participants are reluctant to give up on cooperation — as shown by the fact that investments increase in the second half of the game in the mean condition (Fig. 3) — they might be even more reluctant to give up on partners who seem to encourage them to cooperate, with their positive emotional expression. In Krumhuber et al. (2007), people explicitly and implicitly trusted smiling faces more than neutral faces, regardless of the sincerity of their smile, and genuine smiles were trusted more than fake smiles (Krumhuber et al., 2007). Similarly, Reed et al. (2012) found that people displaying either Duchenne or non-Duchenne smiles were more likely to cooperate in a one-shot investment game (Reed et al., 2012). Thus, displaying an emotion, even a feigned one, might be preferred to not displaying any emotion at all, hence the increased investments to the mean smiling agents. Additionally, participants might have felt more positive emotions themselves upon hearing a smiling agent. In fact, emotional expressions can evoke affective reactions in observers (Geday et al., 2003), which may subsequently influence their behavior (Hatfield et al., 1994), and this ‘emotional contagion’ might be transmitted through the auditory channel as well. If this is the case, participants might have trusted the smiling agents more because feeling a positive emotion themselves might have prompted them to behave in a cooperative manner (Schug et al., 2010; Mieth et al., 2016). These results show similarities with Tsankova et al. (2015), who found that people rated trustworthy faces and voices as happier (Tsankova et al., 2015). Although they addressed the issue from the opposite direction – "Are trustworthy stimuli perceived as happier?" rather than "Are happy stimuli perceived as trustworthy?" – taken together, the studies suggest a bidirectionality in the perception of trustworthiness and cues to positive emotion, congruent with a ’halo effect’ of positive traits (Lau, 1982). The smiling-voice effect suggests that, in the absence of visual information, the audio equivalent of a Duchenne smile might act as a relative ‘honest signal’ of cooperation. As mentioned before, Duchenne smiles are smiles describing genuine happiness or amusement (Ekman & Friesen, 1982). Traditionally, in the visual domain they can be distinguished from other types of smiles because they involve the contraction of the ‘Orbicularis Oculi’ muscle, which is a movement that is notoriously more difficult to fake (Ekman & Friesen, 1982; Schug et al., 2010). Obviously, in the auditory channel it is not possible to detect a genuine smiling voice from this muscular movement. However, it is possible that a smiling voice which sounds happy might be the auditory equivalent of a Duchenne smile. As participants indicated that the smiling voices used in this study did sound happy, it is possible that the expression of happiness and amusement in the speech signal led listeners to believe that the agent could be trusted. A limitation of this study is that no video recordings were taken during the audio recordings of the speakers used in this experiment. This means that, while every effort was was made to ensure consistency in the smile production, it is possible that our speakers might have produced different kinds of smiles. As is well known in emotion theory, smiles can convey many different meanings, and several different facial expressions of smiles are known (e.g. Rychlowska et al., 2017; Keltner, 1995). However, much of the research on the effect of different types of smiles on person perception and decision making has concentrated on the difference between polite (non-Duchenne) and genuine (Duchenne) smiles (e.g. Chu et al., 2019; Krumhuber et al., 2007; Reed et al., 2012). Traditionally, these two are characterised by different muscle activation, with non-Duchenne smiles only activating the Zygomaticus Major muscle, and Duchenne smiles also activating the Orbicularis Oculi muscle (Frank et al., 1993). However, recent studies have suggested that Orbicularis Oculi activation in Duchenne smiles might actually be a by-product of the Zygomaticus Major activation (Girard et al., 2019; Krumhuber & Manstead, 2009). Also, the acoustics of smiling are only affected by activation of the Zygomaticus Major muscle, which contributes to vocal tract shape, but not of Orbicularis Oculi. Following past research that Orbicularis Oculi activation is the only thing that distinguishes Duchenne from non-Duchenne smiles, we would still expect both smiles to sound the same, as the Zygomaticus Major activation would be the same. Still, research on the acoustic characteristics of different types of smiles is lacking. Drahota et al. (2008-04) obtained three different smiling expressions – Duchenne smiles, nonDuchenne smiles, and suppressed smiles – as well as a neutral baseline, from English speakers, and asked participants to correctly identify these four expressions. Participants were only able to reliably distinguish Duchenne smiles from non-smiles, but the majority of the other smile types were classified as non-smiles. Furthermore, they only performed pairwise comparisons between a smile type and a non-smile, but they did not compare differences in identification between two different smile types. Even though they only had 11 participants, which warrants for a much-needed replication of this study, this finding suggests that people might only be able to acoustically discriminate between two categories, smile and non-smile. Similar results were obtained in studies using different types of visual smiles in decision-making tasks. Previous work using cooperative games with Duchenne and non-Duchenne (facial) smiles have shown that people made the same decisions regardless of the type of smile (Reed et al., 2012; Krumhuber et al., 2007). This suggests that people might react according to a broad, dichotomous smile category (smile vs. non-smile), even though the smiles in the experiment stimuli were of different qualities. This corroborates previous findings in nonconscious mimicry, whereby facial EMG recordings were different when viewing a face with a Duchenne smile and a neutral expression, but not when viewing a face with a non-Duchenne smile and a neutral expression (Surakka & Hietanen, 1998). This contrasts with Chu et al. (2019), who
found that participants cooperated more with a confederate expressing a non-Duchenne smile, than with a confederate expressing a Duchenne smile, following a breach of trust. However, in this study the confederate only showed the smiling expression after the cooperate/defect decision was made, whereas in Reed et al. (2012); Krumhuber et al. (2007), as well as in the current study, the smiling expression was displayed before the decision was made. As Chu et al. (2019) point out, this factor might have influenced the decisions and could explain the different behaviors. For example, participants might interpret an emotional expression – such as a smile – after a decision as being an appraisal of that decision. People might put more cognitive effort into understanding this appraisal, as this is essential for shaping future interactions, hence the more accurate discrimination of different smile types. As de Melo et al. (2015, 2013) suggest, a happy expression following the decision to cooperate conveys a different meaning than a happy expression following the decision to defect. This is also consistent with the EASI model (Van Kleef et al., 2010). On the other hand, a happy expression shown before the decision to cooperate / defect might rather convey some information about the emotional state of the person in question, and might be kept independent from that person’s actual behavior in the game. Also, counterparts’ smiles may lead people to anticipate positive social outcomes (Kringelbach & Rolls, 2003). Thus, it seems that the timing of emotional expression in relation to the behavior of interest drastically changes the interpretation of that, and future, behaviors. It would be very interesting to replicate the current experiment with different smiling voices, shown before and after the action is taken in the game. Also, if a similar study were to be replicated, the actual facial expression of the speakers could be recorded in order to determine whether different facial expressions correspond to different auditory smiles, both in terms of objective measures (acoustics) and in terms of perception and behavior correlates in the game. So far, we have compared our results with previous studies that used facial smiles. These comparisons are necessary, as at the time of writing there are virtually no studies that have employed trust games with expressive voices. However, emotional expressions are naturally multimodal, and it is possible that a certain emotion expressed only in the voice might elicit different behaviors than if it were expressed only in the face, or in a voice + face combination. In fact, previous research suggested that an ’Emotional McGurk Effect’ might be at play (Fagel, 2006; Mower et al., 2009; Pourtois et al., 2005). Thus, our current results can only inform the design of voice-based artificial agents, but should not be extended to the design of embodied agents. The results from questionnaires validate the behavioral measures obtained from the investment game. We found that people consistently gave higher ratings of trustworthiness and liking to the smiling agents, and to the agents that behaved generously in the game. Again, the lack of interactions between smiling and behavior suggests that the smiling voice mitigates negative reactions following an untrustworthy behavior. We also found some evidence that individual differences among participants might play a role in trusting behavior, as shown by the 3-way interaction between behavior, game turn, and gender (Section 4.1). The effect of gender on trusting and trustworthiness has been widely studied using game theoretic paradigms, but so far there has been no definite conclusion on whether women trust more / are more trustworthy than men, or vice versa (e.g. Chaudhuri et al., 2013; Bonein & Serra, 2009; Slonim & Guillen, 2010). Our results support previous findings showing that we tend to trust people of the opposite gender more (Slonim & Guillen, 2010), as men in our experiment invested more money than women to the virtual agents, which had a female voice. They also support findings that men trust more than women in general (Chaudhuri & Gangadharan, 2007). However, these conclusions only hold insofar as the generous behavior condition is concerned, as in the mean condition men actually trusted the virtual agent less than women did. A similar behavior was previously observed in Haselhuhn et al. (2015), who found that men showed less trust following a trust breach on the trustee’s part (Haselhuhn et al., 2015). Also, Torre et al. (2018) showed that people who formed a first impression of trustworthiness of a virtual agent punished it when the agent behaved in an untrustworthy manner, by investing less money than to an agent whose first impression was lower. Thus, a ’congruency effect’ might be at play here: our male participants might have formed a first impression of trustworthiness of the female agents (Slonim & Guillen, 2010); when this first impression was congruent with the observed behavior (in the generous condition), the agent received more monetary investments from the male participants. On the other hand, when the first impression was incongruent with the observed behavior (mean condition), it received less (cf. Torre et al., 2018). Participants’ age did not have an effect on the behavioral results from the investment game, but it did influence participants’ explicit ratings of the artificial agents’ trustworthiness, with older people indicating lower trust. This is consistent with the idea that younger people trust technology more, perhaps due to a higher degree of familiarity (e.g. Scopelliti et al., 2005; Giuliani et al., 2005; Czaja & Sharit, 1998). However, we did not match participants’ age – or gender– systematically, so more research is needed on the role of individual differences on trust towards voice-based artificial agents. Finally, speaker identity was varied randomly rather than wholly systematically in our experimental design, and so we included speaker identity as a random rather than fixed effect in our analyses. It is possible, indeed likely, that participants’ trust attributions were influenced by the virtual agents’ unique vocal profiles as well as their behavior and smiling status. In fact, Fig. 7 shows that people invested more money with speaker B2, followed by speakers R1, R2, and B1 (mean overall investments = £5.46, £4.76, £4.11, £3.56, respectively). This is not unexpected: voices carry a wide variety of information about the speaker, such as gender, accent, age, emotional state, socioeconomic background, etc., and all this information is implicitly used by listeners to
form an initial impression of the speaker; a short exposure to someone’s voice is enough to determine if that someone can be trusted (McAleer et al., 2014). For example, in the free-text comments explaining the liking rating to each voice, one participant remarked that smiling speaker B2 “varied in tone and was much more interesting to listen to” and neutral speaker B2 was “calm and convincing”; on the other hand, smiling speaker R2 was “mellow and monotone"and neutral speaker R2 “sounded bored and insincere”. Smiling speaker B1 was “quite annoying” and the neutral version “didn’t seem trustworthy or reassuring”, “sounded too neutral” and even “too fake”. Thus, when designing a voice for an artificial agent, it is important to also keep in mind what effect its specific vocal imprint will have on the user (see als McGinn & Torre, 2019). Nevertheless, any potential between-speaker differences in the current experiment were nested within the effect of smiling voice, as all speakers were recorded in both smiling and neutral conditions