Rolf Degen summarizing... Our memory helps preserve a rosy view, wiping out inconvenient truths about ourselves

Optimistic Amnesia: How Online and Offline Processing Shape Belief Updating and Memory Biases in Immediate and Long-Term Optimism Biases. Ziqing Yao, Xuanyi Lin, Xiaoqing Hu. Social Cognitive and Affective Neuroscience, nsab011, January 27 2021. https://doi.org/10.1093/scan/nsab011

Rolf Degen's take: Our memory helps preserve a rosy view, wiping out inconvenient truths about ourselves. https://t.co/Xn0TYMNLZs https://t.co/OqSl8sKCdP

Abstract: When people are confronted with feedback that counters their prior beliefs, they preferentially rely on desirable rather than undesirable feedback in belief updating, i.e., an optimism bias. In two pre-registered EEG studies employing an adverse life event probability estimation task, we investigated the neurocognitive processes that support the formation and the change of optimism biases in immediate and 24-hour delayed tests. We found that optimistic belief updating biases not only emerged immediately, but also became significantly larger after 24 hours, suggesting an active role of valence-dependent offline consolidation processes in the change of optimism biases. Participants also showed optimistic memory biases: they were less accurate in remembering undesirable than desirable feedback probabilities, with inferior memories of undesirable feedback associated with lower belief updating in the delayed test. Examining event-related brain potentials (ERPs) revealed that desirability of feedback biased initial encoding: desirable feedback elicited larger P300s than undesirable feedback, with larger P300 amplitudes predicting both higher belief updating and memory accuracies. These results suggest that desirability of feedback could bias both online and offline memory-related processes such as encoding and consolidation, with both processes contributing to the formation and change of optimism biases.

Keywords: optimism bias, belief updating, offline processing, P300, motivated cognition

Discussion

Encountering feedback that challenges one’s prior beliefs, people preferentially rely on desirable

than undesirable feedback to guide belief updating, i.e., the optimism bias (Dricu et al., 2020;

Sharot & Garrett, 2016; Sharot et al., 2011). Here, we provide novel evidence that optimism

biases are partially driven by shallower encoding and inferior memories of undesirable versus

desirable feedback, i.e., an optimistic amnesia effect. Moreover, we observed that optimistic

updating biases became larger over time, with preferential retention of updating in the desirable

condition and declined updating in the undesirable condition. Desirability of feedback

consistently modulated parietal P300 brain activities that may indicate encoding depth, with

larger P300s for desirable than undesirable probability feedback.

The present research provides the first evidence that optimism biases become larger over

24 hours. This finding is noteworthy because it suggests that the desirability of feedback not only

influences online attention/encoding-related processes but also biases offline consolidation

processes. A closer inspection of our data suggested that over time, belief updating, and

memories of desirable feedback were largely preserved, whereas updating and memories

significantly declined for undesirable feedback. These findings contribute to a growing literature

suggesting that motivation (e.g., valence, reward) could bias offline consolidation processes and

then influence long-term judgments (Payne & Kensinger, 2018; Rasch & Born, 2013; Stickgold

& Walker, 2013).

Our findings that belief updating and memories changed more significantly in the

undesirable but not in the desirable condition provide additional evidence that optimism bias is

primarily driven by insufficient updating when receiving undesirable feedback (Eil & Rao, 2011;

Sharot et al., 2011). Prior research found that self-related undesirable updating was not only

lower than self-related desirable updating but also lower than other-related updating in general

(Kuzmanovic et al., 2016). Moreover, aging participants showed reduced belief updating

following undesirable feedback compared to young adults, leading to larger optimism biases

(Chowdhury et al., 2014). In contrast, patients with major depressive disorder or individuals with

high functioning autism showed enhanced belief updating toward undesirable feedback relative to

healthy controls, leading to smaller optimism biases (Garrett et al., 2014; Korn et al., 2014;

Kuzmanovic et al., 2019). These findings, together with our novel results on delayed belief

updating and memory biases, consistently suggest that insufficient updating in response to

undesirable feedback is a fundamental mechanism that drives immediate and long-term optimism

biases.

Tracking ERPs allows us to investigate how the desirability of feedback biases

information processing along millisecond temporal scale. We found that the desirability of

feedback significantly modulated P300, and to a less extent, the LPP, but not the FRN. As one of

the most investigated ERP components, P300 has been associated with a range of cognitive

processes, including context updating, motivational salience, evaluation and categorization,

encoding depth, etc. (Azizian & Polich, 2007; Polich, 2007, 2012). In the present study, enhanced

parietal P300s to desirable versus undesirable feedback suggested that participants preferentially

encoded desirable feedback, which then exerted a greater impact on subsequent belief updating

and memory performance. Regarding the LPP effect, prior research suggested that LPP may

reflect in-depth elaboration of motivationally salient stimuli (Hajcak & Foti, 2020). Indeed,

multilevel analyses with trial-level data showed that enhanced P300 and LPPs to feedback

predicted larger belief updating and more accurate memories of feedback probabilities,

substantiating the putative role of the P300/LPP in encoding and elaboration processes (Kamp et

al., 2015; Otten & Donchin, 2000; Otten & Rugg, 2001; Rigney et al., 2020). These ERPs results

also suggest that differential processing of desirable and undesirable feedback can occur rapidly

after the initial valence processing.

We hypothesized that the frontocentral FRN would encode the desirability of feedback,

with larger FRNs elicited by undesirable versus desirable feedback (Heydari & Holroyd, 2016;

Yeung & Sanfey, 2004). However, the desirability of feedback was not observed to modulate

FRN. The insensitivity of FRN to feedback valence in the belief updating task raises the

possibility that estimation errors and reward prediction errors could reflect distinctive

computational processes of error tracking (Sharot & Garrett, 2016). Specifically, FRNs are

typically observed in reward processing tasks during which feedback conveys monetary gains and

losses (Hajcak et al., 2006; Heydari & Holroyd, 2016; Proudfit, 2015), whereas feedback in our

task indicated numerical discrepancies between estimations of probabilities. Specifically,

participants in the belief updating task needed to calculate the discrepancies between feedback

probabilities and their initial estimations to guide belief updating. Such high-level inferential and

calculation processes might make FRNs insensitive to the desirability of feedback in the present

context. Moreover, FRNs have been suggested to be involved in both error tracking and

behavioral adjustment. For example, in reward tasks, FRNs elicited by undesirable feedback (e.g.,

a loss) could guide behavioral adjustment to avoid losses (Cohen et al., 2007; Holroyd & Coles,

2002; Hu et al., 2015; Walsh & Anderson, 2012). However, in our belief updating task,

participants preferentially used desirable rather than undesirable feedback to guide belief

updating. The FRNs may reflect complex motivational-cognitive processes including both error

tracking (in response to both desirable and undesirable feedback), and the signaling of behavioral

adjustment (i.e., in response to desirable feedback). A mixture of these motivational-cognitive

processes may thus lead to comparable FRNs in both desirable and undesirable conditions.

Regarding the delayed optimism biases, although our results suggest that time delay and

offline processes contributed to the enhancement of optimism biases, it remains unknown

whether sleep or wakefulness may differentially influence belief updating and memory biases. On

the one hand, enhancement of optimism biases may be time-dependent rather than sleepdependent. Alternatively, given that sleep plays an important role in memory consolidation

(Rasch & Born, 2013), sleep-based consolidation may be necessary for optimism biases to

change. Future studies shall directly link sleep, offline consolidation processes, and optimism

biases to test this novel hypothesis. 

We could thus find no link between handedness and depression

Packheiser, Julian, Judith Schmitz, Gesa Berretz, Lena S. Pfeifer, Clara C. Stein, Marietta Papadatou-Pastou, Jutta Peterburs, et al. 2021. “Handedness and Depression: A Meta-analysis Across 87 Studies.” PsyArXiv. January 28. doi:10.31234/osf.io/2hqd3

Abstract: Alterations in functional brain lateralization, often indicated by an increased prevalence of left- and/or mixed-handedness, have been demonstrated in several psychiatric and neurodevelopmental disorders like schizophrenia or autism spectrum disorder. For depression, however, this relationship is largely unclear. While a few studies found evidence that handedness and depression are associated, both the effect size and the direction of this association remain elusive. Here, we collected data from 87 studies totaling 35,501 individuals diagnosed with depression disorders to provide a precise estimate of differences in left-, mixed- and non-right-handedness between depressed and healthy samples. We found no differences in left- (OR = 1.04, p = .384), mixed- (OR = 1.64, p = .060) or non-right-handedness (OR = 1.05, p = .309) between the two groups. We could thus find no link between handedness and depression on the meta-analytical level.

To Avoid More Political Violence, Allow Americans To Escape Each Other's Control - Let people join with the like-minded to reject officials and laws that don’t suit them and to construct systems that do

To Avoid More Political Violence, Allow Americans To Escape Each Other's Control. J D Tuccille. Reason, Jan 19 2021. https://reason.com/2021/01/19/to-avoid-more-political-violence-allow-americans-to-escape-each-others-control/

Let people join with the like-minded to reject officials and laws that don’t suit them and to construct systems that do.

Excertps, full text and links in the original URL:

[...]

We've built toward this point for years. While the Trumpists' storming of the Capitol was an unprecedented rejection of the established procedures for transferring power, it built on trends. From the contested, but peaceful, 2000 election, to the boycotting of Trump's 2016 victory by dozens of Democratic members of Congress as other opponents rioted blocks away, Americans have moved toward belief in the legitimacy of elections only if their side wins. At some point, we were going to see an outright refusal to accept a loss, which is what occurred on January 6.

And there's no reason to expect that people will lose their distaste for political defeat in future political contests.

How could Americans be accepting of electoral losses when many view their opponents as immoral and unpatriotic and see them as enemies of the country—to the point that the major factions are defined by their hatreds? "Democrats and Republicans … have grown more contemptuous of opposing partisans for decades, and at similar rates," notes a November 2020 paper on political sectarianism. "Only recently, however, has this aversion exceeded their affection for copartisans."

To a large extent this is because politics has become combat, with election victors using their control of government agencies to torment losers.

"It is more and more dangerous to lose an election," economist John Cochrane, a senior fellow of the Hoover Institution and an adjunct scholar of the Cato Institute, wrote in September. "The vanishing ability to lose an election and not be crushed is the core reason for increased partisan vitriol and astounding violation of basic norms on both sides of our political divide."

No sane people would consent to a political system that works as a weapon against them; they would try to escape its power. One of the virtues of the original decentralized American republic and its federalism was that if you didn't like the laws and rulers where you lived, you could go elsewhere.

"Foot voting is still underrated as a tool for enhancing political freedom: the ability of the people to choose the political regime under which they wish to live," George Mason University law professor Ilya Somin wrote in a 2012 paper since expanded into a book. "When people are able to choose their governments, political leaders have stronger incentives to adopt policies that benefit the people, or at least avoid harming them. And the people themselves are able to select the policies they prefer."

The "people" Somin references aren't the amorphous masses discussed in Social Studies classes as marching to the polls to jam the alleged will of the winners down the throats of the losers. He means individuals turning their backs on governing systems they dislike and picking those that better suit them.

But, as Chapman University law professor Tom Bell—another advocate of political choice—points out in his 2018 book Your Next Government?, "the United States has in recent decades failed to take states' rights seriously, making federal law supreme even in minutely local matters."

Moving does little good when the laws and "vanishing ability to lose an election and not be crushed" (as Cochrane put it) follow you.

Even reviving federalism would accomplish little when many states have larger populations than the whole country did at its founding and the major political divides run not between states or regions, but between urban and rural areas. Within localities are many people who feel trapped by circumstances in "enemy territory," subject to hostile rulers and laws they despise.

How do we make more palatable a political system that functions as a death match between mutually loathing factions who believe themselves—with reason—to be in peril when their enemies win control?

"If every man has freedom to do all that he wills, provided he infringes not the equal freedom of any other man, then he is free to drop connection with the state—to relinquish its protection and to refuse paying toward its support," Herbert Spencer famously argued in 1851. Fundamentally, Spencer wanted the right to exit that Somin favors, but without the physical migration of foot voting, [...].

But Somin not only favors radical decentralization to minimize the costs of migration, he also discusses arrangements whereby "individual citizens can change government service-providers without a physical move." Bell, too, believes that "for the same reason that nation states should and generally do allow the unhappy residents to emigrate, more consent-rich governing services would doubtless guarantee the freedom of citizen-customers to exit to other legal systems" without moving their locations.

In 2001, Swiss economist Bruno Frey proposed what he called functional, overlapping, competing jurisdictions—basically, governments that people choose among as if picking club memberships.

Frey echoed Belgian economist Paul Emile de Puydt who, in the 1860 article "Panarchy," advocated a system of non-territorial federalism under which people could freely register their support for, or withdrawal from, any political associations that gain sufficient support. "I hope we can all go on living together wherever we are, or elsewhere, if one likes, but without discord, like brothers, each freely holding his opinions and submitting only to a power personally chosen and accepted," de Puydt wrote.

These proposals expand on Spencer's "right to ignore the state" in empowering people to join with the like-minded not just to reject officials and laws that don't suit them, but to construct systems that do.

Their advocates emphasize existing precedents for choice in government. "People choose between governments every time they choose to live in a new city, state, or country," writes Bell. "Businesses and others are often able to choose for themselves which state's law will govern their dealings with each other, even if they do not actually reside in the state in question," points out Somin.

What if Americans could choose governing systems rather than having them jammed down their throats? They could embrace rules as limited or restrictive as they please, programs and policies that suit their tastes, and officials who resist treating election to office as opportunities to punish enemies. [...]

True, American politics has been moving away from allowing exit in recent years, centralizing power so that people can't escape and even attempting to continue taxing those who flee, as California lawmakers propose. [...]

We can have a future of increasing conflict between Americans who hate each other. Or we can make it easier for people to peacefully escape each other's control.

Receptivity to casual sexual requests: The more attractive the requester, the higher the proportion of agreement; heterosexuals were most impacted by the attractiveness of the target

Receptivity to casual sexual requests. John E. Edlund ,Dailyn Q. Clark,Alissa M. Kalmus &Aquene Sausville. The Journal of Social Psychology, Jan 27 2021. https://doi.org/10.1080/00224545.2021.1881030

Abstract: Research has long noted that there are differences between men’s and women’s responses to casual sexual requests. In this study, we sought to replicate and extend the Clark and Hatfield paradigm while exploring the influence of requestor attractiveness, sexual orientation, and two individual difference measures: sociosexuality (which is how open to sexuality a person is) and personal mate value (which is how high quality of a mate the person is). We found that attractiveness matters in the likelihood of a request being accepted (the more attractive the requester, the higher the proportion of agreement); sexual orientation matters for the overall proportion of responses agreed to (heterosexuals were most impacted by the attractiveness of the target), and that sociosexuality moderates the likelihood of agreeing to the requests (such that participants with higher sociosexuality scores were more likely to agree to requests).

KEYWORDS: Sexual orientationcasual sexsociosexualitymate value

Secular rituals might play a similar role to religious ones in fostering feelings of social connection and boosting positive affect

Charles SJ, van Mulukom V, Brown JE, Watts F, Dunbar RIM, Farias M (2021) United on Sunday: The effects of secular rituals on social bonding and affect. PLoS ONE 16(1): e0242546. https://doi.org/10.1371/journal.pone.0242546

Rolf Degen's take: "A secular collective Sunday-ritual fostered social bonding and positive affect to the same degree as its Christian equivalent. https://t.co/EsWZdqQKrA https://t.co/aIAJlJBFCf"

Abstract: Religious rituals are associated with health benefits, potentially produced via social bonding. It is unknown whether secular rituals similarly increase social bonding. We conducted a field study with individuals who celebrate secular rituals at Sunday Assemblies and compared them with participants attending Christian rituals. We assessed levels of social bonding and affect before and after the rituals. Results showed the increase in social bonding taking place in secular rituals is comparable to religious rituals. We also found that both sets of rituals increased positive affect and decreased negative affect, and that the change in positive affect predicted the change in social bonding observed. Together these results suggest that secular rituals might play a similar role to religious ones in fostering feelings of social connection and boosting positive affect.

Discussion

Religious rituals occur in all human societies [77], and they seem to confer various benefits to those who take part [1]. It has been suggested that rituals are evolutionarily adaptive by helping foster social bonds [16]. This hypothesis has received some support from field research on religious rituals [30] and from a large body of research showing social bonds provide health benefits [39–41]. It has also been proposed that attending secular rituals, such as Sunday Assembly meetings, may lead to improved wellbeing (e.g. [53]). However, whether the social bonding effect reported from religious rituals is also seen in secular rituals that mimic the behaviours of religious rituals had not been tested before. This study of participants from Sunday Assemblies provides the first evidence that the fostering of social bonds occurs in a secular ritual setting. We compared this to a matched group of individuals from four Christian churches. The results showed that social bonding improvements are of a similar level at both Sunday Assembly and religious ritual settings.

Follow-up analyses found that the increase in social bonding from before to after the Sunday Assemblies was positively predicted by the change in positive affect, as has been found for churches across the UK [30], but not negative affect. These findings are in line with the ‘broaden and build’ hypothesis, which suggests that positive emotions increase the scope of one’s attention to others to allow for the formation of social connections, which themselves lead to improved mental wellbeing [19, 34, 35]. This hypothesis has also been used to suggest that link between religion and wellbeing stems from changes in positive affect [34, 35], which in turn leads to protective social benefits, such as social support [36].

Stepwise regression analysis found that neither level of spirituality nor level of religiosity played a significant role in social bonding change, despite both variables having been related to wellbeing in the past [78]. This could be the result of the methodology of previous studies, which have often used attendance of religious services as a measure for religiosity itself [1, 79], which could conflate the effect of ritual attendance with religiosity and/or spirituality. Diener and colleagues [36] have noted that the reported relationship between religiosity and wellbeing is conditional on social support and social structure. This may explain why religiosity did not directly predict social bonding change in either Sunday Assembly or church participants. Price and Launay [53] have specifically suggested that future research should account for the length of time one had been attending Sunday Assembly, to see if this could explain the wellbeing effects they reported. In the stepwise regression model, the length of attendance did not add predictive value for the change in social bonding in the Sunday Assembly participants. If, as Price and Launay [53] suggest, the improved wellbeing stems from social bonding, this may suggest that protective effects of participating in secular ritual could occur quickly. We must note, though, that we likely failed to detect this effect for Sunday Assembly participants because there were a number of people attending the ritual for the first time in the Sunday Assembly population, which was not the case with the Christian church participants, for whom we found that length of attendance predicted strength of social bonding. Future research should attempt to account for the effect of newcomers on social bonding during group rituals.

This work is the first to demonstrate that secular rituals, much like religious rituals, promote feelings of social bonding. However, we acknowledge that there are limitations to this study. Firstly, this study was not pre-registered. Given the changes suggested by those promoting Open Science methodologies since the advent of the replication crisis [80–82], the methods and analysis plans could have been registered in advance of conducting the study. Though pre-registration was not done in this case, the full anonymised dataset and the research materials are provided in supplementary materials in accordance with other Open Science practices, and a power analysis was provided to support the sample size used in this study.

Another limitation is that we only conducted research with one type of secular ritual, the Sunday Assembly meetings. Sunday Assembly meetings are not the only secular ritual that mimic religious ritual, with other examples including the Church of Positivism [50]–still active in Brazil–and the Religious Humanism movement in the United States.

One avenue for future research is to conduct studies investigating whether the positive health effects found in those who regularly attend religious rituals can also be seen in those who regularly attend Sunday Assemblies or other similar non-religious rituals that mimic the behaviours of religious rituals, compared to those who do not attend such rituals. Examples of such positive health effects are better immune function and lowering levels of all-cause mortality [3–6], depression [7–9, 83, 84] and suicidality [1]. Here, we have examined the role of ritual on social bonding. However, to better understand the mechanisms underlying the protective factors that have previously been related only to religious participation, future research could compare health outcomes from those who attend secular rituals to those who do not, while taking affect and social bonding into account. We also recommend that, much like in our research, social bonding factors be explicitly measured in future ritual and health research, as this may provide a more comprehensive understanding of the mechanism by which ritual attendance appears to improve wellbeing.

Future research can also look more widely at gatherings of secular groups, which are not intentionally ‘rituals’ but nonetheless may create a sense of connection to something bigger than oneself. A variety of gatherings may function as a form of ‘implicit religion’ [85–87], such as sporting events where one feels connected to a team spirit [88, 89], thus creating social bonds in ways similar to religious rituals. Conducting research in such settings would allow us to better understand the nature and effects of ritual-like social bonding in secular contexts.

Europe & marriage of the unequals: Authors confirm the decline of hypergamy as women gain advantage in education, with only 4 countries out of 27 showing some remnants of a male educational lead

Educational assortative mating and the decline of hypergamy in 27 European countries: An examination of trends through cohorts. Dávid Erát. Demographic Research, Vol 44, pp 157–188. Jan 28 2021, doi 10.4054/DemRes.2021.44.7

Abstract

Background: Theories of partner selection emphasize the principal role of available partners in the relationship market. As education is a common socioeconomic attribute through which individuals choose a mate, macrostructure theory highlights the importance of the asymmetric change in educational attainment seen in Europe. As women increasingly participate in tertiary education, this restructuring might result in the decline of traditional hypergamous unions.

Objective: We aim to verify previous results confirming the decline of educational hypergamy and the rise of hypogamy, which has been found to be related to women’s growing educational advantage. We also wish to provide a current picture of this process in Europe by looking at the youngest cohort available during the analysis.

Methods: We pooled nine waves of the European Social Survey and examined trends in seven cohorts. Apart from simple percentage differences in education, we reconstructed the indices of female educational advantage (F-index) and the prevalence of hypergamy (H-index), with provided correlation statistics and fitted linear trend lines.

Results: Our results corroborated the findings of previous multi-country analyses. In nearly all selected countries, women were more present in higher education than men, resulting in a uniform increase in the female educational advantage. Parallel to this, hypergamy declined through the cohorts, which correlated with women’s emerging educational lead.

Contribution: Our results verify the findings from previous years, using a newer dataset and detailed cohort perspective, and confirm the decline of hypergamy as women gain advantage in education, with only four countries out of 27 showing some remnants of a male educational lead.

KEYWORDS: assortative mating, education, hypogamy, macrostructure, partner selection, relationship formation, relationship market

People more receptive to bullshit grossly overestimate their ability to detect it and believe they are better able to detect it than others (i.e., “bullshit blind spot”)

A bullshit blind spot? Dunning-Kruger effects in bullshit detection. Shane Littrell and Jonathan Fugelsang. Waterloo Univ., Jan 2021. https://dlab.sauder.ubc.ca/sjdm/presentations/2020-Poster-Littrell-Shane-BullshitDetect-DunningKruger-Blindspot.pdf

Rolf Degen's take: "People least able to detect bullshit believe they are better at it than everyone else. https://t.co/DQhx0t4kme https://t.co/fknlV8iPgO"

Background

• People are often confident that they are not easily misled (i.e., they have good “bullshit

detectors”). However, their BS detection confidence may not relate to their actual ability.

• Past work has shown that people who engage in bullshitting more frequently are

metacognitively less able to distinguish bullshit from non-bullshit, suggesting that some

people may be unaware of their susceptibility to misleading information.

• We examined the extent to which a person’s confidence in their BS detection abilities is

related to actual accuracy on a BS detection task as well as how they feel their detection

ability compares to the ability of others.

• We also examined the associations of these variables with bullshitting frequency and

intelligence.

• We followed up in Study 2 by investigating whether one’s bullshit detection ability is

perceived to be an intuitive or reflective process.

Changes in sexual behavior during of the COVID-19 pandemic and physical distancing measures in single and partnered participants in Germany, Switzerland and Austria

The disruptive impact of the COVID-19 pandemic on sexual behavior of a German-speaking population. Z. Hille, U.C. Oezdemir, K.M. Beier, L. Hatzler. Sexologies, January 27 2021. https://www.sciencedirect.com/science/article/abs/pii/S1158136020301250

Summary

Objective: The aim of this study was to investigate changes in sexual behavior during of the COVID-19 pandemic and physical distancing measures in single and partnered participants in Germany, Switzerland and Austria.

Material and methods: Participants were assessed in a cross-sectional online survey. Amongst others, sociodemographic data, sociosexual attitudes as well as engagement in a range of sexual activities and practices prior to and during the pandemic were collected. Additionally, for subjects in a relationship, sexual attraction to the partner (feelings of affection during partnered sexual activities, and physical sexual attraction) and relationship satisfaction were measured.

Results: Data of 1017 single and 1498 partnered participants were analyzed. Partnered participants masturbated significantly less during physical distancing measures compared to the period before, whereas single males masturbated more often. Single females masturbate less frequently but this difference was not statistically significant. For both subgroups, the frequency of most sexual activities significantly declined since the beginning of physical distancing measures with anal intercourse in partnered participants being the only exception that showed no significant decrease. In the group of participants in relationships, sociosexual variables and physical sexual attraction to one's partner showed a significant positive relationship to the number of new sexual practices added during physical distancing measures, while feelings of affection during partnered sexual activities and relationship satisfaction did not.

Conclusion: Our data support previous findings showing potential disruptive effects on sexual routines of single and partnered participants by the COVID-19 pandemic and physical distancing measures. Further studies are needed to reveal causal factors and to study long-term effects on mental health and relationships.

The Impact of the Economy on Presidential Elections Throughout US History: Voters, we find, appear to judge incumbent presidents on the economy all the way back to George Washington

The Impact of the Economy on Presidential Elections Throughout US History. Eric Guntermann, Gabriel S. Lenz & Jeffrey R. Myers. Political Behavior, Jan 27 2021. https://rd.springer.com/article/10.1007/s11109-021-09677-y

Rolf Degen's take: "It's always been the economy, stupid, all the way back to George Washington. https://t.co/1WzdNC8FWe https://t.co/QotZuU5iXG"

Abstract: As numerous studies in the US and elsewhere document, voters often hold incumbents accountable for recent economic circumstances. However, our knowledge of the conditions that allow voters to do so remains incomplete. In particular, most findings about economic voting come from studies of modern economies (post World War II). Modern economies have a host of characteristics that seem to lend themselves to economic voting. Their governments play a large role in the economy and have the Keynesian toolset necessary to influence the economy. Their voters are educated and have access to detailed economic data from ubiquitous media. Are these and other modern conditions necessary for economic voting? Would voters still hold politicians accountable even under adverse conditions? Using economic measures now available back to the 1790s, we study economic voting from the earliest days of the US Republic when none of these conditions were met. Voters, we find, appear to judge incumbent presidents on the economy all the way back to George Washington. Consistent with this pattern, we also find that the economy appears to shape presidents' decisions to run again throughout US history. These findings support recent comparative evidence that economic voting is pervasive across a variety of contexts.

Sperm donation: Mostly altruistic, donors favored anonymity; education level, conscientiousness, empathic concern were positively correlated to donation; age, & conservative and religious values were negatively associated

Motivations and Attitudes of Men Towards Sperm Donation: Whom to Donate and Why? João Areias, Jorge Gato & Mariana Moura-Ramos. Sexuality Research and Social Policy, Jan 27 2021. https://rd.springer.com/article/10.1007/s13178-020-00531-0

Abstract

Background: The widespread access to medically assisted reproduction (MAR) techniques for all women, regardless of any infertility diagnosis, has led to an increased, but as yet unmet, demand for sperm donors in Portugal. For this study, we deployed an online survey to explore men’s motivations for donating and their attitudes toward anonymity and donating for specific groups.

Method: The study’s sample comprised men who were eligible to donate sperm (N = 282). The relationships between these factors and participants’ psychological and sociodemographic characteristics were also explored.

Results: The results mostly indicated altruistic reasons for donating, positive attitudes toward anonymity, and a greater willingness to donate to infertile women. Overall, sexual orientation was not associated with the participants’ attitudes and motivations. Age, education level, conscientiousness, empathic concern, and conservative and religious values were associated with the participants’ motivations and attitudes toward sperm donation.

Conclusion: Recruitment campaigns should therefore consider the specific motivations, attitudes, and psychosocial characteristics of potential sperm donors. Indeed, parenthood is a universal right, so sperm donation should be encouraged, regardless of recipients’ fertility status. Clear information about the identifiability of sperm donors should also be provided.

Discussion

The main goal of this study was to highlight what drives men to donate sperm, their attitudes toward anonymity, and their attitudes toward donating for specific groups, as well as to investigate how sexual orientation, psychological traits, and sociodemographic characteristics influence these motivations and attitudes. For reasons of clarity, the results will be discussed in terms of (i) motivations to donate, (ii) attitudes toward anonymity, and (iii) attitudes toward donating for specific groups.

Motivations to Donate

Consistent with the literature, altruistic motivation seems to be the most significant motivation for donating sperm, emphasizing a desire to help childless couples have children (Ekerhovd et al., 2008; Hedrih & Hedrih, 2012; Thijssen et al., 2017; Van der Broeck et al., 2013) and diminishing the role of financial compensation in return for donation (Ekerhovd et al., 2008; Hedrih & Hedrih, 2012; Thijssen et al., 2017). Motivations to donate were generally not associated with the participants’ sexual orientation, except for the motivation of “knowing one’s fertility,” with heterosexual men assigning more importance to this than other men. Non-heterosexual men’s lack of interest in their fertility status may be because they generally show a weaker intent to parent children and anticipate stigma upon achieving parenthood (Gato, Leal, Coimbra, & Tasker, 2020). Furthermore, many non-heterosexual men seem to be unfamiliar with alternative paths to parenthood (Patterson & Riskind, 2010), and some still envisage parenting as a feminine role (Gato & Fontaine, 2017; Pelka, 2009). Nevertheless, future research should further examine the genesis of these differences.

Age was negatively associated with the motivation of “knowing one’s fertility,” with younger men assigning more importance to this motivation. According to Thijssen et al. (2017), this association is expected because younger, childless men have not yet confirmed their fertility status. Furthermore, the men with a lower level of education also valued this motivation more, so future research should explore this association.

Considering the psychological characteristics of the participants, men with higher levels of Conscientiousness assigned more importance to learning their fertility status. People with high levels of conscientiousness may be more motivated to donate based on willingness to pass on their “good genes” (Thijssen et al., 2017), because this will confirm their purpose and usefulness to society and themselves. Similarly, men with higher levels of Conservative values also ascribed more importance to learning their fertility status. This is unsurprising if we consider that more conservative individuals tend to value social order and assign more importance to traditional institutions (Schwartz, 1992, 1994), such as heteronormative families where children generally have a genetic link with their parents.

People with high levels of empathic concern donate in various contexts, and they are compassionate toward others and seem to be oriented toward alleviating the suffering of others in need (Verhaert & Van den Poel, 2011). This explains the positive association between empathic concern and the motivation to help others have a child, as was found in this study. Conversely, religious values were negatively correlated with this motivation. Even though modern Portugal is a secular country, Catholic values still exert a certain influence (Dix, 2010). In our study, the negative association between adherence to religious values and helping someone to have a child may derive from the fact that Catholicism tends to reject alternative family configurations, including the pursuit of pregnancy through MAR (Rubio, 2015).

Attitudes Toward Anonymity

Willingness to be contacted by the child was the least-indicated attitude, differing from the assertions “parents should disclose DI conception with the child” and “the institution where DI was realized can provide information about me to the child, since that information does not identify me.” This finding agrees with research that found that men tend to avoid donation if their identities may be disclosed (Bay et al., 2014; Thijssen et al., 2017). Nevertheless, a relatively positive view about non-anonymous donation was noticeable, because all participants endorsed disclosing DI conception to the child and releasing non-identifying information about the donor. No differences in attitudes toward anonymity were found as a function of sexual orientation, thus contradicting the study of Freeman et al. (2016), which verified that gay and bisexual men, when compared to their heterosexual peers, were more open minded when it came to anonymity and more willing to have contact with children conceived using their sperm. This study’s results refuted our hypotheses about differences between heterosexual and non-heterosexual men for attitudes toward sperm donation.

For sociodemographic characteristics, age was negatively associated with the attitude that “parents should disclose DI conception to the child,” with younger participants subscribing more to this attitude. According to Riggs and Russell (2011), men under the age of 26 preferred identity-release legislation, and this may reflect in the opinion that parents should disclose DI conception to the child. No psychological characteristics were associated with attitudes toward anonymity, perhaps due to the choice of psychological correlates in this study. Future research should therefore include other aspects, such as neuroticism and agreeableness.

Attitudes Toward Donating for Specific Groups

Regarding attitudes toward donating for specific groups, a positive tendency toward donating to all groups was discernible, and this is consistent with the findings of other studies (Bay et al., 2014; Ekerhovd et al., 2008; Thijssen et al., 2017). Nevertheless, donating to heterosexual women with fertility problems was the group regarded the most positively. This may come from the altruistic motivations to donate that participants reported. Indeed, altruism is based on a desire to help someone to have a child (Ekerhovd et al., 2008; Hedrih & Hedrih, 2012; Thijssen et al., 2017; Van der Broeck et al., 2013), and infertile women may be perceived, a priori, as being at a greater disadvantage.

Empathic concern was positively associated with donating to most recipient groups. This likely derives from the fact that people with high levels of empathic concern tend to focus more on alleviating the suffering of others and showing compassion (Verhaert & Van den Poel, 2011). In contrast, adherence to religious values was negatively associated with donating to all groups. As mentioned earlier, this may stem from the influence of Catholic views about family (Rubio, 2015).

Limitations, Future Directions, and Implications for Practice

Like any research endeavor, this study is not without limitations. More than half of our sample was highly educated, so it was not representative of the wider Portuguese population. Additionally, the imbalance in the number of heterosexual and non-heterosexual men may have further compromised the results and prevented us from drawing conclusions about the influence of this variable. The small magnitude of effects (except in the RM ANOVAs) also raised concerns and somewhat limited generalization of the findings.

Despite these limitations, however, our findings do have implications for practice. By shedding light on the motivations, attitudes, and characteristics of possible candidate donors, the findings of this study may inform future recruitment campaigns. Indeed, based on these findings, it could be argued that campaigns should target specific motivations, attitudes, and characteristics of potential donors. As Ferguson, Atsma, de Kort, and Veldhuizen (2012) reported in the context of blood donation, emphasizing the positive feelings that are associated with donation can increase donation rates, so a similar concept could be applied in the context of sperm donation. For example, in this study, the most frequently reported motivation to donate was to help someone have a child, which is an altruistic motivation. It may therefore be useful to highlight the altruistic nature of donation in any campaign.

This study also contributes to the view that withdrawing donation anonymity, which took place in Portugal in 2018, may not necessarily affect the number of potential donors, thus reflecting the positive trend toward non-anonymity reported by other studies (Bay et al., 2014; Thijssen et al., 2017). However, given a pattern of disengagement when the disclosed information is too personal or when there is a chance of contact with the child, the issue of anonymity should be further debated and clarified. Future recruitment campaigns should therefore stress that upon release of anonymity, the donor bears no obligation or commitment to the child. Likewise, it is crucial to emphasize that parenthood is a universal right in order to encourage sperm donation for any person or couple, regardless of fertility status or sexual orientation.

We showed that 46% of the variability in sleep duration & 44% of the variability in sleep quality is genetically determined; the remaining variation in the sleep characteristics can mostly be attributed to the unique environment

Heritability of Sleep Duration and Quality: A Systematic Review and Meta-analysis. Desana Kocevska et al. Sleep Medicine Reviews, January 27 2021, 101448. https://doi.org/10.1016/j.smrv.2021.101448

Summary: Epidemiological and interventional research has highlighted sleep as a potentially modifiable risk factor associated with poor physical and mental health. Emerging evidence from (behavioral) genetic research also shows that sleep characteristics are under strong genetic control. With this study we aimed to meta-analyze the literature in this area to quantify the heritability of sleep duration and sleep quality in the general population. We conducted a systematic literature search in 5 online databases on January 24th 2020. Two authors independently screened 5,644 abstracts, and 160 complete articles for the inclusion criteria of twin studies from the general population reporting heritability statistics on sleep duration and/or quality, and written in English. We ultimately included 23 papers (19 independent samples: 45,328 twins between 6 months to 88 years) for sleep duration, and 13 papers (10 independent samples: 39,020 twins between 16 and 95 years) for sleep quality. Collectively, we showed that 46% of the variability in sleep duration and 44% of the variability in sleep quality is genetically determined. The remaining variation in the sleep characteristics can mostly be attributed to the unique environment the twins experience, although the shared environment seemed to play a role for the variability of childhood sleep duration. Meta-analyzed heritability estimates for sleep duration, however, varied substantially with age (17% infancy, 20-52% childhood, 69% adolescence and 42-45% adulthood) and reporter (8% parent-report, 38-52% self-report). Heritability estimates for actigraphic and PSG-estimated sleep were based on few small samples, warranting more research. Our findings highlight the importance of considering genetic influences when aiming to understand the underlying mechanisms contributing to the trajectories of sleep patterns across the lifespan.

Keywords: Behavioral GeneticsSleepGenesHeritabilityInheritance PatternsSleep durationSleep Quality

Check also What Do People Know About the Heritability of Sleep? Juan J. Madrid-Valero, Robert Chapman, Evangelina Bailo, Juan R. Ordoñana, Fatos Selita, Yulia Kovas & Alice M. Gregory. Behavior Genetics, Jan 23 2021. https://www.bipartisanalliance.com/2021/01/most-people-do-not-have-accurate.html

Over just 10 years, both implicit and explicit male-science/female-arts and male-career/female-family stereotypes have shifted toward neutrality, weakening by 13%–19%, in all US regions & several other countries

Patterns of Implicit and Explicit Stereotypes III: Long-Term Change in Gender Stereotypes. Tessa E. S. Charlesworth, Mahzarin R. Banaji. Social Psychological and Personality Science, January 27, 2021. https://doi.org/10.1177/1948550620988425

Rolf Degen's take: "Over just 10 years, both implicit and explicit gender stereotypes have have moved toward neutrality worldwide. https://t.co/BzHPXVKVfg https://t.co/Gmpr9bOxEr"

Abstract: Gender stereotypes are widely shared “collective representations” that link gender groups (e.g., male/female) with roles or attributes (e.g., career/family, science/arts). Such collective stereotypes, especially implicit stereotypes, are assumed to be so deeply embedded in society that they are resistant to change. Yet over the past several decades, shifts in real-world gender roles suggest the possibility that gender stereotypes may also have changed alongside such shifts. The current project tests the patterns of recent gender stereotype change using a decade (2007–2018) of continuously collected data from 1.4 million implicit and explicit tests of gender stereotypes (male-science/female-arts, male-career/female-family). Time series analyses revealed that, over just 10 years, both implicit and explicit male-science/female-arts and male-career/female-family stereotypes have shifted toward neutrality, weakening by 13%–19%. Furthermore, these trends were observed across nearly all demographic groups and in all geographic regions of the United States and several other countries, indicating worldwide shifts in collective implicit and explicit gender stereotypes.

Keywords: implicit social cognition, gender stereotypes, stereotype change, time series analyses (ARIMA)

Following Twitter: Anger, but not the other constructs, was distinctly reflected in followed accounts, and there was some indication of bias in predictions for women but not for racial/ethnic minorities

Predicting Mental Health From Followed Accounts on Twitter. Cory Costello; Sanjay Srivastava; Reza Rejaie; Maureen Zalewski. Psychology (2021) 7 (1): 18731. https://doi.org/10.1525/collabra.18731

Rolf Degen's take: "The accounts you follow on Twitter could shed some light on your mental health. https://t.co/iLPFnX3LCZ https://t.co/wVX3H6ngxm"

Abstract: The past decade has seen rapid growth in research linking stable psychological characteristics (i.e., traits) to digital records of online behavior in Online Social Networks (OSNs) like Facebook and Twitter, which has implications for basic and applied behavioral sciences. Findings indicate that a broad range of psychological characteristics can be predicted from various behavioral residue online, including language used in posts on Facebook (Park et al., 2015) and Twitter (Reece et al., 2017), and which pages a person ‘likes’ on Facebook (e.g., Kosinski, Stillwell, & Graepel, 2013). The present study examined the extent to which the accounts a user follows on Twitter can be used to predict individual differences in self-reported anxiety, depression, post-traumatic stress, and anger. Followed accounts on Twitter offer distinct theoretical and practical advantages for researchers; they are potentially less subject to overt impression management and may better capture passive users. Using an approach designed to minimize overfitting and provide unbiased estimates of predictive accuracy, our results indicate that each of the four constructs can be predicted with modest accuracy (out-of-sample R’s of approximately .2). Exploratory analyses revealed that anger, but not the other constructs, was distinctly reflected in followed accounts, and there was some indication of bias in predictions for women (vs. men) but not for racial/ethnic minorities (vs. majorities). We discuss our results in light of theories linking psychological traits to behavior online, applications seeking to infer psychological characteristics from records of online behavior, and ethical issues such as algorithmic bias and users’ privacy.

Keywords:emotions, social network analysis, online social networks, machine learning, data science, mental health

Discussion

Our central aim was to understand how mental health is reflected in network connections in social media. We did so by estimating how well individual differences in mental health can be predicted from the accounts that people follow on Twitter. The results showed that it is possible to do so with moderate accuracy. We selected models in training data using 10-fold cross-validation, and then we estimated the models’ performance in new data that was kept completely separate from training, where model Rs of approximately .2 were observed. Although these models were somewhat accurate, when we examined which features were weighted as important for prediction, we did not find them to be readily interpretable with respect to prior theories or broad themes of the mental health constructs we predicted.

Mental Health and the Curation of Social Media Experiences

This study demonstrated that mental health is reflected in the accounts people follow to at least a small extent. The design and data alone cannot support strong causal inferences. One interpretation that we find plausible is that the results reflect selection processes. The list of accounts that a Twitter user follows is a product of decisions made by the user. Those decisions are the primary way that a user creates their personalized experience on the platform: when a user browses Twitter, a majority of what they see is content from the accounts they previously decided to follow. It is thus possible that different mental health symptoms affect the kind of experience people want to have on Twitter, thus impacting their followed-account list. The straightforward ways this could play-out that we discussed at the outset of this paper – e.g., face-valid information-seeking via mental health support or advocacy groups, homophily (following others who display similar mental health symptoms), or emotion regulation strategies – did not seem to be supported. Instead, the accounts with high importance scores were celebrities, sports figures, media outlets, and other people and entities from popular culture. In some rare instances, these hinted towards homophily or a similar mechanism: for example, one account with a high importance score for depression was emo-rapper Lil Peep, who was open about his struggles with depression before his untimely death. More often, however, the connections were even less obvious, and few patterns emerged across the variety of highly important predictors. Other approaches, such as qualitative interviews or experiments that manipulate different account features, may be more promising in the future for shedding light on this question.

Causality in the other direction is also plausible: perhaps following certain accounts affects users’ mental health. For example, accounts that frequently tweet depressing or angry content might elicit depression or anger in their followers in a way that endures past a single browsing session. The two causal directions are not mutually exclusive and could reflect person-situation transactional processes, whereby individual differences in mental health lead to online experiences that then reinforce the pre-existing individual differences, mirroring longitudinal findings of such reciprocal person-environment transactions in personality development (Le et al., 2014; Roberts et al., 2003). Future longitudinal studies could help elucidate whether similar processes occur with mental health and social media use.

In a set of exploratory analyses, we probed the extent to which the predicted scores were capturing specific versus general features of psychopathology. The followed-account scores that were constructed to predict anger captured variance that was unique to that construct; but for depression, anxiety, and post-traumatic stress, we did not see evidence of specificity. One possible explanation is that followed accounts primarily capture a more general psychopathology factor (Lahey et al., 2012; Tackett et al., 2013) but anger also has distinct features that are also relevant. Another possibility is that followed accounts can distinguish between internalizing and externalizing symptoms, and anger appeared to show specificity since it was the only externalizing symptom we examined. The present work cannot distinguish between these possibilities, but future work including more externalizing symptoms may be helpful in differentiating between these and other possibilities.

Relevance for Applications

What does this degree of accuracy – a correlation between predicted and observed scores of approximately .2 – mean for potential applications? First, it’s worth noting that our conclusions are limited to twitter users that meet our minimal activity thresholds (25 tweets, 25 followers, 25 followed accounts), so they may not be applicable to twitter users as a whole, including truly passive users that might follow accounts but not tweet (at all). Even among the users that do meet these thresholds, we do not believe these models are accurate enough for use in individual-level diagnostic applications, as they would provide a highly uncertain, error-prone estimate of any single individual’s mental health status. At best, a correlation of that size might be useful in applications that rely on aggregates of large amounts of data. For example, this approach could be applied to population mental health research to characterize trends in accounts from the same region or with other features in common.

A caveat is that the goal of the present study was to focus on followed accounts – not to maximize predictive power by using all available information. It may be possible to achieve greater predictive accuracy by integrating analyses of followed accounts with complementary approaches that use tweet language and other data. In addition, more advanced approaches that would be tractable in larger datasets, such as training vector embeddings for followed accounts (analogous to word2vec embeddings; Mikolov et al., 2006), could help increase accuracy and should be investigated in the future. Likewise, it may be possible to leverage findings from recent work identifying clusters or communities of high in-degree accounts (Motamedi et al., 2020, 2018) to identify important accounts or calculate aggregate community scores, as opposed to the bottom-up approaches to filtering and aggregating accounts used in this study. Future work can examine the extent to which these different modifications to our procedure maximize predictive accuracy.

Another important caveat to consider with respect to possible applications of this work is that this approach is more suited to studying more stable individual differences in mental health rather than dynamic, within-person fluctuations or responses to specific events. This was an aim that was reflected in the design of this study – for example, the wording of the mental health measures covered a broader time span than just the moment of data collection. Followed accounts are likely to be a less dynamic cue than other cues available on social media (e.g., language used in posts). This is not to say that network ties are unrelated to dynamic states entirely, and that possibility could be explored with different methods. For example, rather than focusing on whether accounts are followed or not, researchers could use engagements with accounts (such as liking or retweeting) to predict momentary reports of mental health symptomatology, or they could track users over time to measure new follows added after an event. The present work can only speak indirectly to these possibilities, but exploring approaches that dynamically link network ties to psychological states is a promising future direction for this work.

The present results, and the possibility of even higher predictive accuracy or greater temporal resolution with more sophisticated methods, raise important questions about privacy. The input to the prediction algorithm developed in this paper – a list of followed accounts – is publicly available for every Twitter user by default, and it is only hidden if a user sets their entire account to “private.” It is unlikely that users have considered how this information could be used to infer their mental health status or other sensitive topics. Indeed, even people who deliberately refrain from self-disclosing about their mental health online may be inadvertently providing information that could be the basis of algorithmic estimates, a possibility highlighted by the often less-than-straightforward accounts that the algorithms appeared to use in their predictions. With time, technological advancement, and research, these predictions might become even more accurate using similarly non-obvious cues in their predictions, though we cannot say how much more. In this way, the present findings are relevant for individuals to make informed decisions about whether and how to use social media. Likewise, they speak to broader issues of ethics, policy, and technology regulation at a systemic level (e.g., Tufekci, 2020). The possibility of a business, government, or other organization putting their considerable resources into using public social media data to construct profiles of users’ mental health may have useful applications in public health research, but it simultaneously raises concerns about how that may be misused. Our results suggest that accuracy is too low for such utopic or dystopic ends presently, but they highlight the possibilities, and the need for in-depth discussions about data, computation, privacy, and ethics.

Predictive Bias

Predictive algorithms can be biased with respect to gender, race, ethnicity, and other demographics, which can create and reinforce social inequality when those algorithms are used to conduct basic research or in applications (Mullainathan, 2019). When we probed for evidence of predictive bias for gender, we found somewhat inconclusive results. There was more of a pattern of bias in the smaller holdout dataset than in the combined data. In the holdout data, women showed higher observed levels of internalizing symptoms (depression, anxiety, and post-traumatic stress) than men with the same model-predicted scores. In the larger combined dataset, only post-traumatic stress showed this effect, and to a much smaller magnitude. Confidence bands in both datasets often ranged from no effect to moderately large effects in one or both directions. All together, we took this as suggestive but inconclusive evidence that the models may have been biased. If the pattern is not spurious, one possible reason may stem from the fact that the sample had more men than women. If men’s and women’s mental health status is associated with which accounts they follow, but the specific accounts vary systematically by gender, then overrepresentation of men in the training data could have resulted in overrepresentation of their followed accounts in the algorithm.

We found little to no evidence of bias with respect to race or ethnicity. The relative lack of bias is initially reassuring, but it should be considered alongside two caveats. First, it is possible that there is some amount of bias that we were unable to detect with the numbers of racial and ethnic minority participants in this dataset. This possibility is highlighted by the confidence bands, which (like gender) tended to range from no effect to moderately large effects. Second, it is possible that collapsing into White vs. non-White is obscuring algorithmic bias that is specific to various racial and ethnic identities. Our decision to combine minority racial and ethnic groups was based on the limitations of the available data, and it necessarily collapses across many substantively important differences.

In any future work to extend or apply the followed-accounts prediction method we present in this study, we strongly encourage researchers to attend carefully to the potential for algorithmic bias. We also hope that this work helps demonstrate how well-established psychometric methods for studying predictive bias can be integrated with modern machine learning methods.

Considering Generalizability At Two Levels of Abstraction

To what extent would the conclusions of this study apply in other settings? There are at least two ways to consider generalizability in this context. The first form of generalizability is a more abstract one, associated with the approach. Would it be possible to obtain similar predictive accuracy by applying this modeling approach to new data drawn from a different population, context, or time, developing a culturally-tuned algorithm for that new setting? We believe the results are likely to be generalizable in this sense. We used cross-validation and out-of-sample testing to safeguard against capitalizing on chance in estimates of accuracy. If the general principle holds that Twitter following decisions are associated with mental health, we expect that it would be possible to create predictive algorithms in a similar way in other settings.

A second, more specific way to think about generalizability is whether the particular prediction algorithms we trained in this study would generalize to entirely new samples from different settings. This is a much higher bar, and we are more skeptical that the models trained in this study would meet it. The fact that the models were not interpretable suggests that they may not have been picking up on theoretically central, universally relevant features of psychopathology. Instead, they might be picking up on real, but potentially fleeting, links between psychopathology and Twitter behavior. By analogy, consider differences between a self-report item like, “I frequently feel sad,” and an item like, “I frequently listen to Joy Division.” The first item would probably be endorsed by depressed people in a wide variety of contexts, populations, and historical eras. The second item, however, is deeply culturally embedded – it if is reflective of depression at all, that association would be highly specific to a particular group of people at a particular cultural moment. Even setting aside that Twitter itself is a product of a specific cultural and historical context, our inspection of the followed accounts suggests that they are not reflecting enduring features of psychopathology in a direct, face-valid sense. The associations with particular accounts were real in this data, but as cultural trends change, they may fade while new ones emerge.

Our results cannot speak to this form of generalizability directly, and it would require a new sample and different design to effectively speak to this. One possibility would be to collect several very different samples (e.g., sampled in different years), train models with each, and then evaluate cross-sample predictive accuracy. This would be a much stricter test of accuracy, but it would provide better justification for using model-derived scores in research or application. Such an approach might also be useful for distinguishing which accounts or features of accounts are predictive because of fleeting cultural factors, and which ones reflect stable and cross-contextually consistent associations with psychopathology.