How to activate threat perceptions in behavior research: A simple technique for inducing health and resource scarcity threats

Abstract

Understanding our cognitive and behavioral reactions to large-scale collective problems involving health and resource scarcity threats, such as the COVID-19 pandemic, helps us be better prepared for future collective threats. However, existing studies on these threats tend to be restricted to correlational data, partly due to a lack of reliable experimental techniques for manipulating threat perceptions. In four preregistered experiments ( N = 5152), we developed and validated an experimental technique that can separately activate perceptions of personal health threat or resource scarcity threat, either in the specific context of the COVID-19 pandemic or in general. We compared the threat manipulations to a relaxation manipulation designed to deactivate background threat perceptions as well as to a passive control condition. Confirmatory tests showed substantial activation of personal health and resource scarcity threat perceptions. This brief technique can be easily used in online experiments. Distress due to the threat manipulation was rarely reported and easily managed with a debriefing toolkit.

Full text

Vol.:(0123456789)
Behavior Research Methods
https://doi.org/10.3758/s13428-024-02481-6
ORIGINAL MANUSCRIPT
How toactivate threat perceptions inbehavior research: Asimple
technique forinducing health andresource scarcity threats
OzanIsler1 · OnurcanYilmaz2· A. JohnMaule3· SimonGächter4,5,6
Accepted: 16 July 2024
© The Author(s) 2024
Abstract
Understanding our cognitive and behavioral reactions to large-scale collective problems involving health and resource scar-
city threats, such as the COVID-19 pandemic, helps usbe better prepared for future collective threats. However, existing
studies on these threats tend to be restricted to correlational data, partly due to a lack of reliable experimental techniques for
manipulating threat perceptions. In four preregistered experiments (N = 5152), we developed and validated an experimental
technique that can separately activate perceptions ofpersonal health threat or resource scarcity threat, either in the specific
context of the COVID-19 pandemic or in general. We compared the threat manipulations to a relaxation manipulation
designed to deactivate background threat perceptions as well as to a passive control condition. Confirmatory tests showed
substantial activation of personal health and resourcescarcity threat perceptions. This brief technique can be easily used in
online experiments. Distress due to the threat manipulation was rarely reported and easily managed with a debriefing toolkit.
Keywords Threat perception· Experimentalmanipulation technique· Health threat· Resource scarcity threat
Introduction
The COVID-19 pandemicposed significant threats to both
health and livelihoods. How do humans react to such exis-
tential threats? While health and resource scarcity threats
can clearly affect us, often by heightening our perceptions of
and emotional reactions to these threats, little is known about
the systematic effects of these changes on human psychology
and behavior. Most studies in behavioral and psychological
sciences on the effects of health and scarcity threats rely on
surveys and other correlational field data (e.g., Hensel etal.,
2022; Van Bavel etal., 2022), due to a lack of experimental
techniques for reliably increasing the cognitive saliency of
threat perceptions. The present study addresses this method-
ological gap by providing a brief technique that can system-
atically activate perceptions of personal health or resource
scarcity threats in online and laboratory studies, both in the
context of the COVID-19 pandemic and in amore general
threat context.
Previous research on human threat response tended
to focus on the threat of violence rather than health and
resource scarcity threats. This work has identified a broad
range of responses to violence threats including conservative
reactions following terror attacks (Jost etal., 2017; Sibley
etal., 2012); group divisions along ideological lines and
in general (Greenberg etal., 1990; Greenberg etal., 1992;
Greenberg etal., 1994; Greenberg etal., 2001), strengthen-
ing group loyalties (e.g., Van de Vyver etal., 2016); system
justification motives (e.g., Ullrich & Cohrs, 2007; Van der
Toorn etal., 2015); religiosity and patriotism (e.g., Bonanno
& Jost 2006); authoritarianism (e.g., Echebarria-Echabe &
Fernández-Guede, 2006); and support for military spending,
racism, and conservatism (e.g., Craig & Richeson, 2014;
Landau etal., 2004; Janoff-Bulman & Usoof-Thowfeek,
* Ozan Isler
o.isler@uq.edu.au
* Simon Gächter
simon.gaechter@nottingham.ac.uk
1 School ofEconomics, University ofQueensland,
StLucia4072, Australia
2 Department ofPsychology, Kadir Has University,
Istanbul34083, Turkey
3 Leeds University Business School, University ofLeeds,
LeedsLS29JT, UK
4 School ofEconomics, University ofNottingham,
NottinghamNG72RD, UK
5 CESifo Munich, 81679Munich, Germany
6 IZA Bonn, 53113Bonn, Germany

Behavior Research Methods
2009; Nail etal., 2009). Although these findings may be
relevant for other threat types, such as scarcity and health
threats, they are often based on methods predating the open
science movement, which limits their generalizability due
to non-experimental methods, small sample sizes, and lack
of preregistration.
Environmental (e.g., scarcity), existential (e.g., mortal-
ity salience), and relational (e.g., separation) threats can
affect social beliefs and outcomes (e.g., Greenberg etal.,
1994; Navarrete & Fessler, 2006; Mikulincer etal., 2002;
Roux etal., 2015; Tybur etal., 2014), but these findings
suffer from similar limitations such as small sample sizes
and unclear experimental manipulations. For instance,
tasks involving viewing disgusting images or recalling
memoriesto measure disgust sensitivity have been used
to activate threat perceptions (Navarrete & Fessler, 2006;
Schaller etal., 2010; Wu & Chang, 2012). However, to our
knowledge, no study has systematically explored the impact
of the presence (vs. absence) of photographic images, used
incentivized tasks, or compared different types of threats
in preregistered, high-powered experiment. In one notable
exception, van Leeuwen etal. (2023) applied varioustech-
niques designed to manipulate pathogen avoidance to assess
their impact on conformity, but found no significant overall
effect. The null result of this comprehensive study under-
scores the need for effective methods to manipulate these
threat perceptions in novel ways.
Reactions to existential health threats, compared to the
threat of violence, may work through additional evolved psy-
chological and physiological mechanisms. The behavioral
immune system (BIS) is thought to have evolved todetect
and respond to pathogens in the environment. Accordingly,
threat detection activates the BIS, eliciting spontaneous emo-
tional reactions that help avoid disease and prevent trans-
mission (Ackerman etal., 2018). For example, the sight and
smell of spoiled food causes disgust, motivating avoidance
of potential pathogens (Terrizzi etal., 2013). This reaction
enhances survival, particularly in regions with high patho-
gen prevalence. However, the BIS can alsofuel undesirable
social behaviors such as xenophobia (Helzer & Pizarro, 2011;
Inbar etal., 2009; Jones & Fitness, 2008; Murray & Schaller,
2012; Wu & Chang, 2012; Faulkneret al., 2004; Fincher &
Thornhill, 2012; Navarette & Fessler, 2006; Terrizzi etal.,
2013). Geographical differences in pathogen prevalence are
associated with more ethnocentric, collectivist, and conserva-
tive social attitudes (Murray etal., 2013; Terrizzi etal., 2013;
Thornhill etal., 2009; but see Horita & Takezawa, 2018).
From this perspective, the COVID-19 threat can be expected
to have resulted in a global conservative shift, but to our
knowledge this possibility has not been experimentally tested
with manipulations reliably activating COVID-19-related
threats (cf. Karwowski etal., 2020b).
The COVID-19 pandemic not only poseda serious public
health threat but, through its negative impact on production
processes, logistics, and financial markets, also a serious
collective resource scarcity threat. Cognitive saliency of
resource scarcity can influence decisions byalteringper-
ceptions of value and impairing cognitive performance
(Mani etal., 2013; Mullainathan & Shafir, 2013; Shah
etal., 2012; Spiller, 2011). Despite its financial implica-
tions, experiments on resource scarcity perceptions are rare,
even beyond the COVID-19 context. While some studies
havefound effects on behavior in economic games, such as
increased selfishness (Roux etal., 2015), these findings have
been hard to replicate (O’Donnell etal., 2021), likely due to
the reliance on small-sample studies with weak manipulation
techniques (see Isler etal., 2023).
Few studies have attempted to experimentally manipu-
late perceived COVID-19threat, but without distinguish-
ingbetween resource scarcity and health threat perceptions.
Cappelen etal. (2021) found that reminding US residents of
the COVID-19 threat increased the tendency to prioritize
societal problems over personal ones. This large-scale pre-
registered experiment aimed to activate COVID-19 threat
perceptions by asking participants two straightforward
questions about the pandemic: (1) “To what extent has your
local community been affected by the current coronavirus
crisis?” and (2) “How long do you expect the current cor-
onavirus crisis to last?” The experimental condition with
these two questions was compared to a control group that
did not recivethese questions. Two other experiments using
a similar approach (Karwowski etal., 2020a; Karwowski
etal., 2020b)involved participants reading three brief press
reports. Two of the reports were on a neutral topic and were
common across theconditions, whereas the third report
was related to COVID-19 for the experimental group and
to climate change for the control group. These reminders
of COVID-19 increased anxiety without changing ideologi-
cal attitudes (Karwowski etal., 2020b) or cognitive perfor-
mance (Karwowski etal., 2020a). Despite these insights,
these studies lack clarity regarding the construct validity of
their threat manipulation techniques. In particular, it is not
known to what extent health threat versus resource scarcity
threat perceptions were activated and whether these con-
textual changes in threat type involved personal or public
threat perceptions. Such systematic tests of the role of deci-
sion context in human threat response require modification
of various aspects of this context (e.g., the type of threat),
which our technique provides.
Current threat manipulation techniques have several limi-
tations. First, common techniques for experimentally activat-
ing threat perceptions are either weak or ineffective. High-
powered preregistered tests of some of the most commonly
used techniques in psychology such as the scarcity lottery

Behavior Research Methods
task (Krosch & Amodio, 2014), the scarcity scale task
(Nelson & Morrison, 2005), and the scarcity consequences
task (Roux etal., 2015) failed to successfully manipulate
resource scarcity perceptions in an ongoing project (Isler
etal., 2024).
Second,while the COVID-19 pandemic likely influenced
bothhealth andresource scarcity threat perceptions, this dis-
tinction is often ignored. Since these perceptions can affect
behavior differently, a more accurate picture of the effects of
COVID-19 can only be drawn if these two threat types are
separately manipulated. It would be an added benefit if the
same techniqueis used to construct these different manipu-
lations, allowing comparability of the observed effects.
Testing the technique’s applicability beyond the COVID-
19 contextcould reveal whether responses arespecific to
COVID-19 or generalize to othercontexts.
Third, existing studies often overlook the importance of
background threat levels. High background threat levels dur-
ing the pandemic may have resulted in control participants
experiencing threat levels comparable to those subjected to
threat manipulations, compromising the efficacy of experi-
mental tests.
Fourth, and perhaps most importantly, most of the evi-
dence on human threat response is limited to observational,
correlational data, whichprecludes directcausal inference.
Our new technique addresses these limitations. First, rec-
ognizing that baseline threat perceptions were already high
during the COVID-19 pandemic, we developed both threat
manipulations and a relaxation manipulation to reduce pre-
existing threat levels.
Second, we aimed to design manipulationswith strong
(but momentary and psychologically safe) effects on risk
perceptions, addressing the weak manipulations and null
results found in the literature.
Third, we distinguished between various threat types (e.g.,
health vs. resource scarcity, personal vs. public, COVID-
19-specific vs. general) in both devising and testing our cog-
nitive manipulation technique. This enabled systematic com-
parisons of responses to different threat contexts and tested
the applicability of our technique.
Fourth, we tested whether the effectiveness of our tech-
nique varies with individual differences in pre-existing risk
perceptions. In a field experiment, messages emphasizing
self-benefit tended to increase influenza vaccination among
high-risk patients, but only when theyalso perceived them-
selves at high risk (Isler etal., 2020). Since personal health
threat perceptions can motivate vaccination, those who avoid
the COVID-19 vaccine are likely to be less risk-averse than
those who receive it.
Finally, we incorporated various pecuniary and visual
design features to improve the effectiveness of our tech-
nique. Economic incentives have been shown to enhance
attention to and compliance with experimental tasks
(Camerer & Hogarth, 1999; Hertwig & Ortmann, 2001;
Isler et al., 2023), and better compliance with task instruc-
tions can potentially increase manipulation effectiveness.
We conducted four preregistered experiments (N = 5152)
using consistent participant selection criteria across experi-
ments (see Section “Method”). Participants could take part
in only one experiment. The first two experiments were pre-
liminary, testing the effects of the health threat and relaxa-
tion manipulations on threat perceptions, affect, and cog-
nitive performance. Experiment 1 tested the technique for
the first time and compared various incentive schemes for
increasing task compliance. Experiment 2 tested the impact
of visuals on manipulation effectiveness. The final two
experiments validated the effectiveness of the technique with
larger samples. Experiment 3 compared the health threat
and the relaxation manipulations with a control condition
that measured baseline threat levels, used comprehensive
outcome measures that distinguished between personal and
public threat, and explored vaccination status. Experiment
4 validated the results of Experiment 3, introduced an effec-
tive COVID-19 resource scarcity threat manipulation, and
tested the general applicability of these techniques beyond
the COVID-19 context.
The preregistrations, datasets, analysis codes, and manip-
ulation protocol are available at the Open Science Frame-
work (OSF) project site, and the experimental materials are
available in the Supplementary Information. The techniques
developed allow for causal tests of previous correlational
findings on the psychological and behavioral impact of
health and scarcity threats. University of Nottingham and
Kadir Has University provided ethics approvals. Informed
consent was obtainedfrom all participants.
Experiment 1
Method
Experiment 1 compared the COVID-19 health threat manipu-
lation to the relaxation manipulation across three types of
task incentivization using experimental procedures supported
by Qualtrics (https:// www. qualt rics. com/). Equal numbers
of male and female participants were recruited from Prolific
(https:// proli fic. co/), and participation was restricted to UK
residents who were 18 years or older, had English as their first
language,and had Prolific approval rates of 90% or above.
The experiment concluded with a demographic questionnaire
and a debriefing that offered support in case participants
experienced distress due to the experimental manipulations.
The experiment was preregistered at the OSF (https:// osf. io/
vgz9c).

Behavior Research Methods
Materials andprocedures
Experiment 1 used a 2 (cognitive manipulation: relaxation
vs. COVID-19 health threat) by 3 (incentive type: no-bonus
vs. individual bonus vs. lottery bonus) between-subjects
design. To activate perceptions of health threat associated
with COVID-19, a picture of an emergency hospital bed-
room was displayed (see Fig.1a) together with a sentence
prompting participants to “look at the picture and think
about getting very unwell from COVID-19 and needing
emergency help.” A writing task was employedto activate
thoughts of being severely ill due to the coronavirus. Spe-
cifically, ten seconds after the appearance of the picture and
the prompt, four text boxes appeared below the picture ask-
ing participants to describe what could happen to them and
how they would feel in this situation, by typing four full
sentences (one in each of four separate boxes). Responses
to this manipulation were compared with the relaxation
manipulation, which displayed a picture of a typical single
bed (see Fig.1b). Participants were asked to think and write
about lying on their bed at the end of the day and feeling
very relaxed, by typing a full sentence in each of the four
text boxes indicating what could happen to them and how
they would feel. Participants had to enter text into all four
boxes before they could continue with the study. We used a
relaxation manipulation rather than a passive control condi-
tion to prevent any ceiling effects due to the severity of the
ongoing COVID-19 pandemic that could have heightened
baseline levels of risk perceptions. Median response times
in the writing task were 143.7 seconds (s) for the relaxation
manipulation condition and 161.9 s for the threat manipula-
tion condition.
Instructions regarding incentives were provided to par-
ticipants at the start of the study. Bonus incentive condi-
tions tested whether the effects of the manipulations could
be increased by motivating compliance with and attention
to task instructions using individual or lottery payments.
All participants received a flat fee of £0.50. In addition, par-
ticipants were instructed that the individual bonus scheme
paid £0.50 to everyone who had written four relevant full
sentences and the lottery bonus scheme paid £5.00 to one in
every ten participants if this randomly selected person had
written four relevant full sentences.
Most participants complied with task instructions, as
almost all text boxes (97.6%) contained three or more words.
This measure did not differ statistically between the incentive
conditions (Pearson’s chi-square test: χ2(2, n = 253) = 0.57,
p = 0.752). However, bonus incentives significantly increased
engagement with the task, as the average number of words
written in each box was higher in the two bonus condi-
tions than in the control (MIndividual-Bonus = 12.64 [SD = 5.64]
vs. MControl = 10.50 [SD = 4.41]: t(156) = 2.62, p = .010,
d = 0.42; MLottery-Bonus = 14.94 [SD = 9.19] vs. MControl = 10.50
[SD = 4.41]: t(166) = 3.80, p < .001, d = 0.59).
Threats can increase negative emotions including anger
(Brooks etal., 2020) and disgust (Curtis etal., 2011; Oaten
etal., 2009; Terrizzi etal., 2013). Hence, we elicited vari-
ous self-reported affect measures after the COVID-19 threat
manipulation (for a similar design see also Varma etal.,
2020). Using a scale ranging from 1 (“very slightly or not at
all”) to 5 (“extremely”), participants completed the 20-item
Positive and Negative Affect Schedule (PANAS) (Watson
etal., 1988). As standard, the total score on items describing
positive (negative) affect constituted the positive (negative)
affect score. On the same screen, ratings on two additional
affect items (“disgusted” and “repulsed”) were elicited. The
combined average of these two items was multiplied by 10
to achieve the disgust sensitivity score, which ranged from
Fig. 1 Pictures used in a the health threat manipulation, b the relaxa-
tion manipulation, and c the scarcity threat manipulation conditions.
The picture in c is used in Experiment 4 (see Section "Sample").
Source: Imgbin.com and Commons.wikimedia.org. Note: Changes to
written instructions were also made to distinguish between threat types

Behavior Research Methods
0 to 50 like the positive and negative affect scores (Fincher
etal., 2008; Schaller, 2011; Shook etal., 2019). Using the
positive affect, the negative affect, or the disgust sensitiv-
ity scores as the dependent variable, we estimated three 2
(COVID-19 health threat) by 3 (incentive type) ANOVA
models as exploratory analyses.
Next, on a scale from 0% to 100%, participants com-
pleted, in two randomly presented screens, two questions
about perceived infection risk (“How likely do you think
it is that, within three months from today, you [the average
person in your country] will get infected by the Coronavirus
(COVID-19)?”) and two questions about severity of illness
(“If infected by the Coronavirus (COVID-19), how severely
do you think you [the average person in your country] would
have the illness?”). As preregistered, the combined aver-
age scores on these four questions constituted the perceived
threat score, our key outcome variable that we use as a
manipulation check.
To explore whether the COVID-19 threat affected cogni-
tive performance, participants completed a multiple-choice
version of the first item in the Cognitive Reflection Test
(CRT, Frederick, 2005; Sirota & Juanchich, 2018), which
was modified to make it less familiar to participants by refer-
ring to “a pencil and an eraser” rather than “a bat and ball.”
It has been argued that people faced with mortality threats
exert cognitive effort to avoid thoughts of death, potentially
leading to cognitive resource scarcity and poorer perfor-
mance on tasks such as the CRT (Trémolière etal., 2012,
2014). The evidence for this argument remains limited to
non-preregistered studies with small sample sizes. Our sam-
ple allows for anexploration of this argument in the context
of COVID-19-based mortality threat.
On the next screen, two exploratory self-reported items
about reliance on intuition (“While looking at the picture and
completing the four sentences...to what extent did you rely
on your gut instinct?”) and about perceptions of material and
financial scarcity (“…to what extent did scarcity of material
or financial resources come to your mind?”) were rated on a
scale from 0 (“not at all”) to 5 (“very much”). The first was
similar to the CRT and explored the possible impact of threat
on cognitive processing, whereas the latter was a preliminary
exploration of scarcity threat to be investigated more fully in
later experiments.
As in all subsequent experiments, participants in Experi-
ment 1 completed a comprehensive debriefing, which began
with an open-ended question asking whether thinking about
contracting COVID-19 was distressing. All participants
were provided with emergency contact lines for profes-
sional counseling as well as a Wellness Sheet that provided
concrete directions for calming oneself in the event of
excessive distress reactions. In Experiment 1, only 6.4%
of participants in the health threat manipulation condition
(7 participants) reported experiencing any distress during
the experiment.
Sample
Using G*Power 3.1.9.4 (Faul etal., 2009), we estimated our
required sample size to detect at least a medium-sized main
effect of the manipulations (f = 0.25) in a two-way ANOVA
model (1− β = 0.95 and α = 0.05) to be 251 participants in
total. Experiment 1 was conducted on November 23, 2020.
A total of 253 participants (age: M = 36.9, SD = 13.8) were
recruited (relaxation: no-bonus = 41, individual bonus = 55,
lottery bonus = 48; COVID-19 health threat: no-bonus = 32,
individual bonus = 30, lottery bonus = 47).
Hypotheses
H1: The threat scores will be higher in the health threat
than in the relaxation manipulation.
H2: The difference in threat scores between the health
threat and the relaxation manipulations will be higher in
the bonus conditions than in the no-bonus condition.
Results
Confirmatory tests
No confirmatory evidence was found for either hypothesis.
The preregistered two-way ANOVA model showed (H1)
no main effect of the cognitive manipulations on the threat
scores (F(1, 247) = 0.23, p = .635, ηp2 < .001) and (H2) no
interaction between these manipulations and the incentive
conditions (F(2, 247) = 1.46, p = .235, ηp2 = .012).
Mean threat scores in the health threat (MHT) and the
relaxation manipulation (MR) conditions did not differ across
the no-bonus (MHT = 40.34, MR = 40.85; t(71) = −0.12,
p = .905, d = 0.03) and the individual bonus conditions
(MHT = 38.08, MR = 40.45; t(83) = −0.61, p = .541, d = 0.14).1
In the lottery bonus condition, although the estimated effect
size seems non-negligible (d = 0.37), this difference in threat
scores between the health threat and the relaxation manipu-
lation conditions failed to reach statistical significance at
the 5% level in a two-tailed t-test (MHT = 36.96, MR = 30.96;
t(93) = 1.80, p = .075).
Exploratory analysis
We explored the effect of the COVID-19 health threat
manipulation on scarcity perceptions, affect, and cognitive
1 We report two-tailed t-tests throughout the article.

Behavior Research Methods
performance here and in the subsequent experiments,
because the perception of threat can influence emotions and
decision-making (e.g., Trémolière etal., 2012, 2014).
Threat perceptions Although perceptions of resource scarcity
were on average higher in the health threat than in the relaxa-
tion manipulation (MHT = 3.13, MR = 2.40), this difference also
failed to reach statistical significance at the 5% level in a two-
tailed t-test (t(251) = 1.91, p = .058, d = 0.24).
Affect Compared to the relaxation manipulation, the health
threat manipulation increased negative affect (MHT = 24.27,
MR = 11.85; t(251) = 15.40, p < .001, d = 1.95), decreased posi-
tive affect (MHT = 19.43, MR = 22.35; t(251) = −3.02, p = .003,
d = 0.38), and increased disgust sensitivity (MHT = 16.06,
MR = 10.90; t(251) = 6.42, p < .001, d = 0.81). These affect
measures did not vary significantly between the incentive
conditions (one-way ANOVAs: ps ≥ .255).
Cognitive performance No significant effect of the threat
manipulations was found on either the single CRT item
(MHT = 0.37, MR = 0.30; t(251) = 1.15, p = .253, d = 0.15)
or the self-reported reliance on intuition (MHT = 7.05,
MR = 6.78; t(251) = 0.83, p = .409, d = 0.10).
Discussion
Experiment 1 was a preliminary study designed to detect
medium or larger effects. Individual and lottery bonus incen-
tives increased task engagement, but they did not substan-
tially improve manipulation effectiveness. While the promise
of the lottery bonus incentive schemes should be investigated
further, in the following experiments we continue to use indi-
vidual bonus incentive schemes for their simplicity. The cog-
nitive manipulations changed affect as expected but, except
for the lottery bonus condition, failed to have a discernible
impact on threat perceptions. We surmised that these fail-
ures could stem from three design features: (1) eliciting the
affect measures first, with 22 items in total, could have diluted
the effect of the cognitive manipulations on threat percep-
tion measures; (2) asking about “the average person in one’s
country” to measure threat perceptions could have resulted in
objective risk estimates rather than measures of spontaneous
threat perceptions; and/or (3) showing a picture of the hospi-
tal bed in the health threat manipulation might have limited
the effectiveness of the manipulation by constraining natural
thought processes, thereby diluting negative thoughts. Hence,
in a second preliminary experiment we revised our manipula-
tion check accordingly, placing it immediately after the threat
manipulation, and additionally tested whether the presence or
absence of pictures of beds (as in Fig.1) makes a difference.
Experiment 2
Method
Experiment 2, preregistered at the OSF (https:// osf. io/ au6vj),
compared the COVID-19 health threat manipulation to the
relaxation manipulation condition with or without the use
of pictures.
Materials andprocedures
Experiment 2 used a 2 (cognitive manipulation: relaxation
vs. COVID-19 health threat) by 2 (task type: no-picture vs.
picture) between-subjects design. In the picture conditions,
the cognitive manipulations were the same as Experiment
1 (Fig.1a and b). The no-picture conditions were identical
except that no pictures were displayed and the instructions
were modified by removing any reference to pictures. Median
response times in the writing task were 143.5 s for the relaxa-
tion manipulation and 164.9 s for the health threat manipula-
tion condition. Most text boxes (98.6%) contained three or
more words, with an average of 13.60 words per text box.
Next, participants completed a modified manipulation
check involving a question on health threat perceptions and
a question on scarcity threat perceptions on a scale ranging
from 0 to 100: “While making an assessment and trying to
construct sentences…” (1) “...to what extent did risks to
your personal health come to your mind?” and (2) “...to what
extent did scarcity of material resources (such as lack of
goods and services) or scarcity of financial resources (such
as inadequate income or savings) come to your mind?” The
average scores across these two questions about personal
health and resource scarcity threat perceptions constituted
the perceived threat score that we use in confirmatory tests.
PANAS and disgust sensitivity items used in Experiment 1
were elicited next, followed in the second part of the study
by the same CRT item used in Experiment 1.
Participants were paid a flat fee of £0.50. We used random
lottery incentives, a standard protocol for determining indi-
vidual bonus payments in experimental economic research
(Starmer & Sugden, 1991). Accordingly, the participants
were told that there were two parts to the study, one of which
would be selected to determine their additional earnings. If
the first part was chosen (i.e., the cognitive manipulation and
the following manipulation check), then participants earned
an additional £0.50 for writing four full relevant sentences
(i.e., the same as the individual bonus condition of Experi-
ment 1). If the second part was chosen, then participants
earned an additional £0.50 for correctly answering the CRT
item. In the debriefing, 12.8% of participants in the threat
manipulation condition (28 participants) reported experienc-
ing distress during the experiment.

Behavior Research Methods
Sample
Based on exploratory evidence in Experiment 1, we esti-
mated our required sample size using G*Power 3.1.9.4 (Faul
etal., 2009) to detect at least a small-to-medium main effect
of the cognitive manipulations (f = 0.175) in a two-way
ANOVA model (1− β = 0.95 and α = 0.05) to be 427 partici-
pants in total. Experiment 2 was conducted on November 27,
2020. A total of 433 participants (age: M = 37.1, SD = 14.2)
were recruited (relaxation: picture = 105, no-picture = 110;
COVID-19 health threat: picture = 110, no-picture = 108).
Hypotheses
H3: The threat scores will be higher in the health threat
than in the relaxation manipulation.
H4: The difference in threat scores between the health
threat and the relaxation manipulations will depend on
the task type (i.e., picture vs. no-picture).
Results
Confirmatory tests
The cognitive manipulations substantially affected threat
perceptions (H3) but there was no evidence that the pic-
tures enhanced this effect (H4). The preregistered two-way
ANOVA model indicated a main effect of the cognitive
manipulations on the threat score (H3: F(1, 429) = 454.37,
p < .001, ηp2 = .514) with the threat scores in the health threat
manipulation (M = 54.26) being higher than in the relaxa-
tion manipulation (M = 16.73) (t(431) = 21.32, p < .001,
d = 2.05). There was no interaction between the cognitive
manipulations and the task type (H4: F(1, 429) = 1.49,
p = .223, ηp2 = .003).
Exploratory analysis
Threat perceptions The effect of the cognitive manipula-
tions was significant for both the personal health threat
(MHT = 81.89, MR = 16.05; t(431) = 32.20, p < .001) and
the scarcity threat (MHT = 26.63, MR = 17.41; t(431) = 3.70,
p < .001) items comprising the threat score, but the effect size
for the personal health threat item (d = 3.09) was substantially
larger than that for the scarcity threat item (d = 0.36).
Affect Positive affect was lower (MHT = 22.19, MR = 25.55;
t(431) = −4.50, p < .001, d = 0.43) whereas negative affect
(MHT = 23.95, MR = 12.48; t(431) = 16.91, p < .001, d = 1.63)
and disgust sensitivity (MHT = 14.36, MR = 10.98; t(431) = 5.71,
p < .001, d = 0.55) were higher in the health threat manipu-
lation than in the relaxation manipulation. The presence of
the picture (P) had no influence on negative affect, disgust
sensitivity, or cognitive reflection (ps ≥ .609) but increased
positive affect compared to the no-picture (NP) condition
(MP = 24.70, MNP = 23.03; t(431) = −2.21, p = .028, d = 0.21).
Cognitive reflection Neither the cognitive manipulations
(MHT = 0.46, MR = 0.42; t(431) = 0.74, p = .459, d = 0.07)
nor the presence or absence of the picture (MP = 0.44,
MNP = 0.44; t(431) = 0.16, p = .871, d = 0.02) was found to
significantlyaffect performance on the single CRT item.
Discussion
Experiment 2 was the second preliminary experiment. The
revised manipulation check revealed large effects of the cogni-
tive manipulations on both the personal health and the resource
scarcity components of the threat score, though the effect on
personal health threat perceptions was larger. The presence
of the pictures was not found to strengthen (or weaken) the
manipulations. The influence of the cognitive manipulations
on affect measures was consistent with Experiment 1. Like-
wise, no effect on cognitive performance was found.
As preliminary studies, Experiments 1 and 2 had two
important limitations: (1) the threat score consisted of very
few items, restricting the coverage and the potential reliabil-
ity of the manipulation checks; and (2) the lack of a neutral
control condition made it impossible to assess whether the
effect of the cognitive manipulations was due to the health
threat or the relaxation manipulation or both. To address
these limitations, we conducted a large-scale validation
experiment that included (1) both the main manipula-
tion check used in Experiment 2 and an additional, more
comprehensive one, and (2) both the threat and relaxation
manipulations used in the first two experiments and a pas-
sive control condition to measure baseline threat percep-
tions. In addition, we considered individual differences in
the use of protective/precautionary behaviors (e.g., whether
or not being vaccinated against COVID-19) as reflections of
different attitudes to risk that might influence the effective-
ness of the threat manipulations. Accordingly, in Experiment
3 we tested whether participants who had received a COVID
vaccination revealed higher personal health risk scores than
those who refused vaccination (at the time of the study a
vaccination was available to all UK citizens over 18).
Experiment 3
Method
Experiment 3, preregistered at the OSF (https:// osf. io/
qyc3e), compared the COVID-19 health threat and the
relaxation manipulations to a control condition.

Behavior Research Methods
Materials andprocedures
Experiment 3 compared three conditions in a between-sub-
jects design: the COVID-19 health threat manipulation, the
relaxation manipulation, and the control condition. The first
two manipulations were the same as in Experiment 1. The
control condition was designed to measure the baseline level
of threat perceptions in the sample without any manipulation
(i.e., no pictures or writing task). Median response times in
the writing task were 103.1 s for the relaxation manipula-
tion and 109.2 s for the health threat manipulation. A total
of 94.9% of the text boxes contained three or more words,
with an average of 8.73 words each.
To ensure that the upcoming threat perception measures
were meaningful for participants in the control condition,
we first prompted participants in all conditions to complete
these measures by considering their “current circumstances
and state of mind.” Next, on two subsequent screens, par-
ticipants completed the threat perception measures that were
used as manipulation checks. The main threat measure was
elicited first and included revised versions of the two ques-
tions used in Experiment 2presented in random order on a
scale ranging from 0 to 100: (1) “To what extent do risks to
your personal health come to your mind?” and (2) “To what
extent does scarcity of material resources (such as lack of
goods and services) or scarcity of financial resources (such
as inadequate income or savings) come to your mind?” Aver-
aging the scores on these two items on personal health and
resource scarcity threat provided the main threat score.
The comprehensive threat measure, which was elicited
next, included 16 items on eight threat areas distinguish-
ing not only between health and resource scarcity threat but
also between personal and public threat by repeating each
of the following statements twice, ending it with either “...
for myself” or “...for others in society”: “Because of the
COVID-19 pandemic, there is high risk of...” (1) “not find-
ing enough affordable food or hygiene products....,” (2) “not
getting enough or timely medical help when needed...,” (3)
“unemployment...,” (4) “higher debt...,” (5) “being infected
with COVID-19...,” (6) “becoming severely ill with COVID-
19...,” (7) “being hospitalized due to COVID-19...,” (8)
“dying from COVID-19...” Participants rated how much
they agreed with the statements on a scale from 1 (“strongly
disagree”) to 7 (“strongly agree”). The average scores on
these 16 items constituted the comprehensive threat score
(Cronbach’s α = .899). The same PANAS and disgust sen-
sitivity items used in the first two experiments were elicited
afterward.
A binary (“yes” or “no”) question on vaccination status
was added to the survey for additional exploratory analy-
sis: “Have you been vaccinated against the coronavirus
(COVID-19) with at least one dose?” The CRT item,
explored in previous experiments and in Experiment 4,
was omitted from Experiment 3 by mistake. Participants
were paid a flat fee of £1 for completing the study. In the
debriefing, 3.4% of participants in the threat manipulation
condition (20 participants) reported experiencing distress
during the experiment.
Sample
Because we planned to test our main hypothesis (H5) twice,
with two different manipulation checks, we used a Bonferroni
correction and set α = 0.025. To detect at least a small effect
size (f = 0.10) in a one-way ANOVA model (1− β = 0.95), we
estimated the target sample size in G*Power 3.1.9.4 (Faul
etal., 2009) to be 1779 participants in total. Experiment 3
was conducted on December 2, 2021. A total of 1777 par-
ticipants (age: M = 38.9, SD = 13.5) were recruited (con-
trol = 617; relaxation = 578; COVID-19 health threat = 582).
Hypothesis
H5: The main and the comprehensive threat scores will be
higher in the health threat manipulation than in both the
relaxation manipulation and the control condition.
Results
Confirmatory tests
The manipulations affected threat perceptions as predicted
(H5) (see Fig.2). The preregistered one-way ANOVA models
on both the main (F(2, 1774) = 73.90, p < .001, ηp2 = .077)
and the comprehensive (F(2, 1774) = 7.91, p < .001,
ηp2 = .009) threat scores showed significant differences across
the experimental conditions. The health threat manipulation
increased the main threat scores above both the relaxation
manipulation (t(1158) = 11.18, p < .001, d = 0.66) and the
control (t(1197) = 2.71, p = .007, d = 0.16). The comprehen-
sive threat scores in the threat manipulation condition were
significantly higher than the relaxation manipulation condi-
tion (t(1158) = 3.92, p < .001, d = 0.23) but not the passive
control condition (t(1197) = 1.72, p = .086, d = 0.10). The
relaxation manipulation decreased threat scores compared
to the control for both the main (t(1193) = −8.90, p < .001,
d = 0.52) and the comprehensive measures (t(1193) = −2.34,
p = .020, d = 0.14).
Exploratory analysis
Threat perceptions The cognitive manipulations had the
intended effect on personal health threat but not resource
scarcity threat perceptions (see Fig.3). Compared to the
control, the health threat manipulation increased and the
relaxation manipulation decreased personal health threat

Behavior Research Methods
perceptions (health threat: t(1197) = 7.96, p < .001, d = 0.46;
relaxation: t(1193) = −8.90, p < .001, d = 0.51). While there
was a significant difference in scarcity threat perceptions
between the health threat manipulation and the relaxation
manipulation (t(1158) = 3.00, p = .003, d = 0.18), the scarcity
threat perceptions were lower than the control in both manip-
ulation conditions (relaxation: t(1193) = −6.33, p < .001,
d = 0.37; threat: t(1197) = −3.11, p = .002, d = 0.18). The
components of the comprehensive threat score provided
consistent results, indicating that the largest and most con-
sistent effects of the threat manipulation were on personal
health threat perceptions (see Supplementary Information).
Affect The impact of the cognitive manipulations on affect
was weaker but consistent with the first two experiments. In
pairwise comparisons, the difference between the relaxa-
tion and health threat manipulations was significant for dis-
gust sensitivity (MHT = 13.45, MR = 12.38; t(1158) = 2.84,
p = .005, d = 0.17) and negative affect (MHT = 17.64,
MR = 16.17; t(1158) = 3.42, p < .001, d = 0.20) but not for
positive affect (MHT = 27.36, MR = 28.04; t(1158) = −1.48,
p = .139, d = 0.09).
Vaccination Threat perceptions can depend on vaccination
status (Isler etal., 2020). Since Experiment 3 was conducted
when multiple COVID-19 vaccines had been widely available
in the UK for almost a year, compared to participants who
reported having received at least one dose of the COVID-19
vaccine (90.9%), those who remained unvaccinated (9.1%)
might have felt less threatened by the COVID-19 pandemic.
Consistent with this argument, the baseline levels of perceived
health threat, as measured by scores on the personal health
item of the main threat measure in the control condition, were
significantly lower for the unvaccinated (M = 46.35) than t he
vaccinated (M = 55.35), t(615) = −2.56, p = .011, d = 0.37.
Relatedly, the effect of the cognitive manipulations on per-
sonal health threat perceptions was stronger for the vaccinated
(MHT = 67.63, MR = 41.61; t(1049) = 16.80, p < .001, d = 1.04)
than the unvaccinated participants (MHT = 48.17, MR = 39.37;
t(107) = 1.54, p = .125, d = 0.30). Analysis of the comprehen-
sive threat measure supports these results (see Supplementary
Information).
Considering the passive control condition and vaccina-
tion status in a 3 (experimental condition) by 2 (vaccination
status) two-way ANOVA models, there was no significant
difference in affect among the three experimental conditions
Fig. 2 a Main and b comprehensive threat scores for the control, the
relaxation manipulation, and the health threat manipulation condi-
tions in Experiment 3. Error bars indicate 95% confidence intervals
Fig. 3 Components of the main threat perceptions score (i.e., per-
sonal health and resource scarcity threat) in the control, the relaxation
manipulation, and the health threat manipulation conditions. Error
bars indicate 95% confidence intervals

Behavior Research Methods
overall (ps ≥ .530) and no interaction with vaccination sta-
tus (ps ≥ .112). There was no main effect of vaccination
status on either positive affect (F(1, 1771) = 0.05, p = .818,
ηp2 < .001) or negative affect (F(1, 1771) = 3.79, p = .052,
ηp2 = .002), but disgust sensitivity was higher among the
non-vaccinated participants (F(1, 1771) = 11.56, p < .001,
ηp2 = .006).
Discussion
Confirmatory tests showed that the manipulations affected
perceptions of COVID-19 threat as intended: compared to
the control, the health threat manipulation increased and the
relaxation manipulation decreased threat perceptions. The
manipulations had a stronger and more consistent effect on
personal health threat perceptions than on scarcity threat
perceptions. Exploratory analysis revealed small (d < 0.20)
differences in negative affect and disgust sensitivity but no
differences in positive affect between the health threat and
the relaxation manipulations.
Despite the insights of Experiment 3, it remains unknown
whether the effect on personal health threat perceptions is
replicable and effective beyond the COVID-19 context. Also,
an effective resource scarcity threat perceptions manipula-
tion remains lacking. Therefore, we decided to run a fourth
experiment to replicate the effect of the health threat manip-
ulation on personal health threat perceptions and to test a
novel manipulation for activating personal resource scarcity
threat perceptions both in the context of the COVID-19 pan-
demic and in general.
Experiment 4
Method
Experiment 4 compared both general and COVID-19-spe-
cific versions of the health threat and the resource scarcity
threat manipulations to the relaxation manipulation and a
passive control condition. The experiment was preregistered
at the OSF (https:// osf. io/ h24vf).
Materials andprocedures
Experiment 4 included six conditions in a between-subjects
design: (1) the general and (2) the COVID-19-specific per-
sonal health threat manipulations, (3) the general and (4)
the COVID-19-specific personal resource scarcity threat
manipulation, (5) the relaxation manipulation, and (6) the
control condition.
The COVID-19 health threat and the relaxation manipu-
lations were the same as in the previous experiments using
visuals (see Fig.1a and b), except that the pronoun “you”
was added to the instructions to focus attention on personal
risks (e.g., “you becoming very unwell” or “you lying on
your bed”). The general health threat manipulation was the
same as the COVID-19 health threat manipulation except
that the prompt ended with the phrase “due to a new and
very serious infectious disease and needing emergency
help” rather than the phrase “from COVID-19 and need-
ing emergency help.” To activate perceptions of resource
scarcity, a picture of empty shelves in a supermarket was
displayed (see Fig.1c) together with the prompt “look at
the picture and think about you urgently needing essential
and emergency goods but there being none available…”
For the general scarcity condition, the sentence ended with
“due to a new and very serious economic crisis,” whereas
for the COVID-19 scarcity condition it ended with “due to
COVID-19 related shortages.” The same writing tasks as in
the previous experiments were implemented to complement
the manipulations (see “Sample”). Median response times
in the writing task were 104.5 s for the relaxation manipula-
tion, 111.7 s for the general and 111.1 s for the COVID-19
health threat manipulations, and 115.3 s for the general and
120.9 s for the COVID-19 scarcity threat manipulations.
As in Experiment 3, the control condition measured base-
line rates of threat perceptions without any experimental
manipulation. As in all other experiments, most text boxes
(94.4%) contained three or more words, with an average of
9.72 words per text box.
Next, as in Experiment 3, all participants were prompted
to answer the following questions based on their current
circumstances and state of mind and were given the manipu-
lation checks in the following two screens. The first screen
included the same two questions as in the main threat meas-
ure used in Experiment 3, one on personal health threat and
another on resource scarcity threat perceptions. Since our
modified manipulations were specifically designed to acti-
vate personal threat perceptions, the second screen included
the eight personal threat perception items from the com-
prehensive threat measure in Experiment 3 and excluded
the remaining eight items on societal threat perception (see
Section "Sample").
Using these ten items, we preregistered three main
dependent variables: (1) the personal health threat score
(Cronbach’s α = .861), calculated as the average of the five
items about personal health threat (i.e., one item from the
first screen and four items from the second screen) con-
verted to the percent of maximum possible score (POMP;
Cohen etal., 1999), (2) the personal scarcity threat score
(Cronbach’s α = .782), calculated as the average of five
items about personal resource scarcity threat (i.e., one
item from the first screen and four items from the second
screen) converted to POMP, and (3) the personal threat
score (Cronbach’s α = .842), calculated as the average of
the first two scores.

Behavior Research Methods
Next, for exploratory analysis, participants completed in
two counterbalanced screens (1) the same PANAS and disgust
sensitivity items used in the previous experiments and (2) a
three-item four-option multiple-choice version of the Cog-
nitive Reflection Test (Sirota & Juanchich, 2018). We used
a three-item version to provide a more rigorous evaluation
of our previous test showing no effect of threat on cognitive
performance, which was based on a single item from the test.
Finally, answers to the same survey questions as in Experi-
ment 3 were elicited. Participants were paid a flat fee of £1
for completing the study. In the debriefing, only 2.0% of
participants (53 participants) reported experiencing some
distress during the experiment.
Sample
We estimated our sample size based on testing of H6 using
G*Power 3.1.9.4 (Faul etal., 2009). To detect a small main
effect (f = 0.10) of manipulations in a one-way ANOVA with
six conditions, α = 0.05, and 1− β = 0.99, the required sample
size was calculated to be at least 2682 participants in total.
Sensitivity analysis showed that this sample size allowed
for the detection of an interaction effect size of f = 0.05 or
more with 1− β = 0.99 in a mixed ANOVA for testing H7.
Experiment 4 was conducted on March 3, 2022. A total of
2689 participants (age: M = 38.8, SD = 13.4) were recruited
(control = 458; relaxation = 438; general health = 439;
COVID-19 health = 449; general scarcity = 452; COVID-
19 scarcity = 453).
Hypotheses
H6: The personal threat scores will be higher in the health
and scarcity threat manipulations than in the relaxation
manipulation and the control condition.
H7: The difference in personal threat scores between the
health and scarcity threat manipulations will depend on
the score type (i.e., personal health vs. personal scarcity
threat scores).
H7A: The personal health threat scores will be higher
in the health threat manipulation conditions than in the
relaxation manipulation, the passive control, and the scar-
city threat manipulation conditions.
H7B: The personal scarcity threat scores will be higher
in the scarcity threat manipulation conditions than in
the relaxation manipulation, the passive control, and the
health threat manipulation conditions.
Results
Confirmatory tests
The cognitive manipulations were effective (see Fig.4),
with the preregistered one-way ANOVA model showing
significant differences in personal threat scores across the
six conditions (F(5, 2683) = 26.86, p < .001, ηp2 = .048).
Supporting H6, all threat manipulations increased the per-
sonal threat scores above both the relaxation manipulation
and the control conditions, whereas the relaxation manipu-
lation lowered the scores below the control (see Table1).
Supporting H7 (see Fig. 5), the preregistered mixed
ANOVA model indicated a significant interaction effect
between the experimental conditions and the score type
(F(5, 2683) = 44.83, p < .001, ηp2 = .077). Specifically, (H7A)
the personal health threat scores in the health threat manipu-
lations were higher than all other experimental conditions
and (H7B) the personal scarcity threat scores in the scarcity
threat manipulations were higher than all other experimental
conditions (see Table2).
Fig. 4 Personal threat scores for the control, the relaxation manipulation, the COVID-19-specific and the general health threat manipulation, and
the COVID-19-specific and the general resource scarcity threat manipulation conditions. Error bars indicate 95% confidence intervals

Behavior Research Methods
Exploratory analysis
Affect As in Experiment 3, there were small differences
across the experimental conditions in disgust sensitivity
(F(5, 2683) = 7.52, p < .001, ηp2 = .014) and negative affect
(F(5, 2683) = 4.71, p < .001, ηp2 = .009) but not positive
affect (F(5, 2683) = 2.14, p = .058, ηp2 = .004). See Table3
for details.
Cognitive reflection Consistent with the previous experi-
ments, the cognitive manipulations had no effect on cog-
nitive reflection, as measured by performance on the CRT
Fig. 5 Personal health (a) and personal resource scarcity (b) threat
scores for the control, the relaxation manipulation, the COVID-
19-specific and the general health threat manipulation, and the
COVID-19-specific and the general resource scarcity threat manipu-
lation conditions. Error bars indicate 95% confidence intervals
Table 1 Pairwise comparisons of (overall) personal threat scores in Experiment 4
The t-statistics, p-values, and effect sizes (Cohen’s ds) for preregistered two-tailed independent-samples t-tests comparing the personal threat
scores in the cognitive manipulation conditions with the control and the relaxation manipulation conditions
Control Relaxation
t p d t p d
Health threat
COVID-19 3.50 < .001 0.23 8.90 < .001 0.60
General 4.21 < .001 0.28 9.50 < .001 0.64
Scarcity threat
COVID-19 3.45 < .001 0.23 8.65 < .001 0.58
General 3.07 .002 0.20 8.48 < .001 0.57
Relaxation
−5.64 < .001 0.38

Behavior Research Methods
(F(5, 2683) = 0.39, p = .855, ηp2 = .001). See Table3 for
details.
Vaccination As in the previous experiment, we explored
the role of vaccination status in threat perceptions. A total
of 91.1% of participants in Experiment 4 reported having
received at least one dose of the COVID-19 vaccine. As
in Experiment 3, the personal health threat scores in the
control condition were significantly lower for the unvac-
cinated (M = 43.31) than the vaccinated (M = 50.82),
t(456) = −2.76, p = .006, d = 0.43. Similarly, compared to
the relaxation manipulation, the combined effect of the two
health threat conditions on personal health threat perceptions
was stronger for the vaccinated (MHT = 58.13, MR = 43.39;
t(1212) = 13.15, p < .001, d = 0.80) than the unvaccinated
participants (MHT = 49.55, MR = 43.96; t(110) = 1.29,
p = .201, d = 0.26). While the personal scarcity threat
perceptions in the control condition were not significantly
different between the unvaccinated (M = 53.57) and the
vaccinated (M = 51.85; t(456) = 0.602, p = .547, d = 0.09),
when compared to the relaxation manipulation, the com-
bined effect of the two resource scarcity threat conditions on
personal scarcity threat perceptions was stronger for the vac-
cinated (MHT = 58.54, MR = 46.36; t(1225) = 11.04, p < .001,
d = 0.67) than the unvaccinated participants (MHT = 59.66,
MR = 57.15; t(114) = 0.69, p = .494, d = 0.14).
Discussion
Experiment 4 replicated our previous finding that the health
threat manipulation specifically activates personal health
threat perceptions, showed that the novel scarcity threat
manipulation successfully and specifically activates personal
scarcity threat perceptions, and established the validity of
Table 2 Pairwise comparisons of personal health and resource scarcity threat scores in Experiment 4
The table depicts the t-statistics, p-values, and effect sizes (Cohen’s ds) for preregistered two-tailed independent-samples t-tests comparing the
personal health and personal resource scarcity threat scores in the threat manipulations conditions with the other experimental conditions
Health threat scores: Control Relaxation Scarcity threat
COVID-19 General
t p d t p d t p d t p d
Health threat
COVID-19 5.11 < .001 0.34 10.31 < .001 0.69 4.72 < .001 0.31 5.38 < .001 0.36
General 6.92 < .001 0.46 12.03 < .001 0.81 6.48 < .001 0.43 7.18 < .001 0.48
Scarcity threat scores: Control Relaxation Health threat
COVID-19 General
t p d t p d t p d t p d
Scarcity threat
COVID-19 5.58 < .001 0.37 9.35 < .001 0.63 4.93 < .001 0.33 5.48 < .001 0.37
General 5.42 < .001 0.36 9.31 < .001 0.62 4.76 < .001 0.32 5.33 < .001 0.36
Table 3 Affect and cognitive reflection measures in Experiment 4
The means (M) and standard deviations (SD) of disgust sensitivity, negative affect, positive affect, and Cognitive Reflection Test scores across
the conditions in Experiment 4
Disgust Sensitivity Negative Affect Positive Affect Cognitive Reflection
M SD M SD M SD M SD
Health threat
COVID-19 12.82 6.62 16.59 7.62 27.54 7.64 1.45 1.19
General 13.35 6.93 17.22 7.60 27.05 8.20 1.40 1.19
Scarcity threat
COVID-19 14.97 8.23 18.00 8.09 26.85 8.14 1.34 1.20
General 14.26 8.08 17.46 7.59 27.48 7.98 1.42 1.21
Relaxation
12.40 6.01 15.73 6.91 27.68 7.73 1.39 1.20
Control
13.67 7.37 17.06 7.58 28.42 8.25 1.39 1.19

Behavior Research Methods
these manipulations both in the context of the COVID-19
pandemic and in general. Consistent with previous experi-
ments, the influence of the cognitive manipulations on affect
and cognitive reflection were either small or nonsignificant.
Conclusion
We introduced a technique that can separately activate
personal health threat or personal resource scarcity threat
perceptions either in the specific context of the COVID-19
pandemic or in general. We compared these threat manipula-
tions with a passive control and a relaxation manipulation.
The former provides baseline measures of threat percep-
tions at the time of the study in the population under study,
while the latter is useful for assessing the effects of the threat
manipulation when baseline levels of perceived threat are
already high in the population, as it provides a reference
group with relatively weak threat perceptions. Overall,
this technique provides an effective way of manipulating
and assessing health and scarcity threats in experimental
research, whether online or in the laboratory, thereby provid-
ing an alternative to less powerful research designs based on
cross-sectional and correlational studies. The final version
of the experimental materials are available as Qualtrics and
PDF files in the OSF project site (https:// osf. io/ grafm/).
Across four experiments, we found evidence that the
cognitive manipulations introduced here reliably activate
threat perceptions as intended. We did not observe any issues
regarding the psychological safety of the technique among
our Prolific samples, as reports of psychological distress due
to the threat manipulations were rare and manageable by a
debriefing toolkit. However, use of the technique among the
general public without any survey experience or individu-
als with special needs may benefit from added precautions.
Finally, we note that our studies show no important differ-
ences in threat perceptions across different incentivization
schemes (e.g., flat fee, bonus payments), suggesting that
the technique is effective regardless of which schemes are
adopted.
The effect of the threat manipulations was greater for per-
sonal than for public threat perceptions. This is in line with
observed trends that preventive behavior such as vaccination
is motivated more effectively with messages emphasizing
personal than public benefit (Banker & Park, 2020; Milkman
etal., 2021), especially when people perceive themselves to
be at high risk (Isler etal., 2020). Consistent with these find-
ings, the cognitive manipulations were found to be particu-
larly effective among those who reported having received
at least one dose of the COVID-19 vaccine. In contrast, the
unvaccinated participants had lower baseline levels of threat
perceptions and were less affected by the manipulations.
Exploratory analysis further suggested that the techniques
create small but systematic differences in negative affect
and disgust sensitivity between the threat and the relaxation
manipulations, which are indicative of increased threat. In
contrast, none of the four experiments showed an effect of
health or scarcity threat on CRT, failing to support the idea
that people use cognitive resources to suppress thoughts
about these threats, leading to poorer cognitive performance
(Trémolière etal., 2012, 2014; for an alternative perspective
on the effect of scarcity see Isler etal., 2023). While our
null results are based on larger sample sizes when compared
to previous research, one should note that CRT items were
elicited towards the end of the studies, after the initial effec-
tiveness of the manipulations may have waned.
The differences observed in threat perceptions were gen-
erally consistent, but the effect sizes showed large variation,
especially in the first two preliminary experiments. This var-
iation may be due to differences in the location of the threat
perception measures in the experimental protocol, with
the strongest effects observed for measures that were elic-
ited immediately after the cognitive manipulations. These
findings suggest that the manipulations have an immediate
strong effect that diminishes over time, particularly when
there are intervening tasks between the manipulation and the
dependent variable of interest. Hence, dependent variables
that are of primary interest should be elicited immediately
after the manipulation to maximize experimental power.
We advocate that the proper implementation of our tech-
nique requires elicitation of both the passive control and
the relaxation manipulation condition together with any
of the threat manipulation conditions. The elicitation of
the passive control condition is important as it provides
baseline measures of threat perceptions and affect at the
time of data collection. In our research, the passive con-
trol conditions in Experiments 3 and 4 were insightful in
determining whether the effects were driven by threat or
relaxation conditions. Variation in these baseline meas-
ures can be particularly useful for longitudinal studies or
experimental studies completed at different points in time.
If baseline levels of perceived threat are high, then the
threat manipulations may fail to induce even higher lev-
els, such as at the height of health and resource scarcity
threats posed by the COVID-19 pandemic. The elicitation
of the relaxation manipulation is recommended because it
allows for inducing significant differences in threat percep-
tions between two randomly generated groups. In short, the
relaxation manipulation may be necessary in experimen-
tally generating differences in threat perceptions, and the
passive control may be necessary in interpreting the experi-
mental results to determine whether it is the manipulation
or the relaxation manipulation that is driving any reported
experimental effects. Therefore, we highly recommend the

Behavior Research Methods
implementation of both the passive control condition and
the relaxation manipulation together with any of the threat
manipulation techniques, rather than pairing them only
with one or the other.
Although our findings are clear, this research has some
limitations. First, our tests were restricted to online conveni-
ence samples. The technique can be easily implemented in
the laboratory as well, but our findings are yet to be repli-
cated in this context. We expect the controlled environment
of the lab to increase the effectiveness of the manipulations.
Second, althoughthe image of the regular bed in the relaxa-
tion manipulation was originally designed in relation to the
image of the hospital bed in the health threat manipulation,
we usethe relaxation manipulation as a common activecon-
trol condition for all threat types.
The novel technique presented here can be used to acti-
vate perceptions of personal health threat or personal resource
scarcity threat perceptions, both in general and in the context
of the COVID-19 pandemic. Recent studies have documented
correlational findings regarding the psychological and behav-
ioral impact of health and resource scarcity threats due to
the global COVID-19 pandemic. We introduce a reliable,
ethical, and easy-to-use manipulation technique to activate
COVID-19 and general health and scarcity threat perceptions.
Future studies can use the techniques introduced in this study
to experimentally test and expand these correlational find-
ings. Additionally, they can develop novel manipulations for
other psychological phenomena by employing the systematic
perspective utilized here.
Supplementary Information The online version contains supplemen-
tary material available at https:// doi. org/ 10. 3758/ s13428- 024- 02481-6.
Authors’ contributions All authors contributed to the development of
the study concept and the study design. O.I. programmed the experi-
ment, conducted the study, and analyzed the data. O.I. and O.Y. wrote
the manuscript. A.J.M. and S.G. provided critical comments and revi-
sions. All authors approved the final version of the manuscript.
Funding This work was supported by the European Research Council
[grant number ERC-AdG 101020453 PRINCIPLES] and the Think
Forward Initiative (a partnership between ING Bank, Deloitte, Dell
Technologies, Amazon Web Services, IBM, and the Center for Eco-
nomic Policy Research – CEPR). The views and opinions expressed in
this paper are solely those of the authors.
Availability of data and materials Available at the OSF project site:
https:// osf. io/ grafm/
Code availability Available at the OSF project site: https:// osf. io/ grafm/
Declarations
Conflict of interest/Competing interests None.
Ethics approval University of Nottingham and Kadir Has University
provided ethics approvals.
Consent to participate Informed consent was received prior to par-
ticipation.
Consent to publish Not applicable.
Open Access This article is licensed under a Creative Commons Attri-
bution 4.0 International License, which permits use, sharing, adapta-
tion, distribution and reproduction in any medium or format, as long
as you give appropriate credit to the original author(s) and the source,
provide a link to the Creative Commons licence, and indicate if changes
were made. The images or other third party material in this article are
included in the article’s Creative Commons licence, unless indicated
otherwise in a credit line to the material. If material is not included in
the article’s Creative Commons licence and your intended use is not
permitted by statutory regulation or exceeds the permitted use, you will
need to obtain permission directly from the copyright holder. To view a
copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
References
Ackerman, J. M., Hill, S. E., & Murray, D. R. (2018). The behavioral
immune system: Current concerns and future directions. Social
and Personality Psychology Compass, 12(2), e12371. https:// doi.
org/ 10. 1111/ spc3. 12371
Banker, S., & Park, J. (2020). Evaluating prosocial COVID-19 messag-
ing frames: Evidence from a field study on Facebook. Judgment
and Decision Making, 15(6), 1037–1043.
Bonanno, G. A., & Jost, J. T. (2006). Conservative shift among high-
exposure survivors of the September 11th terrorist attacks. Basic
and Applied Social Psychology, 28(4), 311–323. https:// doi. org/
10. 1207/ s1532 4834b asp28 04_4
Brooks, S. K., Webster, R. K., Smith, L. E., Woodland, L., Wessely, S.,
Greenberg, N., & Rubin, G. J. (2020). The psychological impact
of quarantine and how to reduce it: rapid review of the evidence.
The Lancet, 395(10227), 912–920. https:// doi. org/ 10. 1016/ s0140-
6736(20) 30460-8
Camerer, C. F., & Hogarth, R. M. (1999). The effects of financial
incentives in experiments: A review and capital-labor-production
framework. Journal of Risk and Uncertainty, 19, 7–42.
Cappelen, A. W., Falch, R., Sørensen, E. Ø., & Tungodden, B. (2021).
Solidarity and fairness in times of crisis. Journal of Economic
Behavior & Organization, 186, 1–11. https:// doi. org/ 10. 1016/j.
jebo. 2021. 03. 017
Cohen, P., Cohen, J., Aiken, L. S., & West, S. G. (1999). The prob-
lem of units and the circumstance for POMP. Multivariate
Behavioral Research, 34(3), 315–346.
Craig, M. A., & Richeson, J. A. (2014). On the precipice of a “major-
ity-minority” America: Perceived status threat from the racial
demographic shift affects White Americans’ political ideology.
Psychological Science, 25(6), 1189–1197. https:// doi. org/ 10.
1177/ 09567 97614 527113
Curtis, V., De Barra, M., & Aunger, R. (2011). Disgust as an adaptive
system for disease avoidance behaviour. Philosophical Transac-
tions of the Royal Society B: Biological Sciences, 366(1563),
389–401. https:// doi. org/ 10. 1098/ rstb. 2010. 0117
Echebarria-Echabe, A., & Fernández-Guede, E. (2006). Effects of
terrorism on attitudes and ideological orientation. European
Journal of Social Psychology, 36(2), 259–265. https:// doi. org/
10. 1002/ ejsp. 294
Faul, F., Erdfelder, E., Buchner, A., & Lang, A. (2009). Statistical
power analyses using G* Power 3.1: Tests for correlation and
regression analyses. Behavior Research Methods, 41, 1149–1160.

Behavior Research Methods
Faulkner, J., Schaller, M., Park, J. H., & Duncan, L. A. (2004).
Evolved disease-avoidance mechanisms and contemporary
xenophobic attitudes. Group Processes & Intergroup Relations,
7(4), 333–353. https:// doi. org/ 10. 1177/ 13684 30204 046142
Fincher, C. L., & Thornhill, R. (2012). Parasite-stress promotes in-
group assortative sociality: The cases of strong family ties and
heightened religiosity. Behavioral and brain sciences, 35(2),
61–79. https:// doi. org/ 10. 1017/ s0140 525x1 10000 21
Fincher, C. L., Thornhill, R., Murray, D. R., & Schaller, M. (2008).
Pathogen prevalence predicts human cross-cultural variability in
individualism/collectivism. Proceedings of the Royal Society B:
Biological Sciences, 275(1640), 1279–1285. https:// doi. org/ 10.
1098/ rspb. 2008. 0094
Frederick, S. (2005). Cognitive reflection and decision making. Jour-
nal of Economic Perspectives, 19, 25–42.
Greenberg, J., Pyszczynski, T., Solomon, S., Rosenblatt, A., Veeder,
M., Kirkland, S., & Lyon, D. (1990). Evidence for terror man-
agement theory II: The effects of mortality salience on reactions
to those who threaten or bolster the cultural worldview. Journal
of Personality and Social Psychology, 58(2), 308.
Greenberg, J., Simon, L., Pyszczynski, T., Solomon, S., & Chatel, D.
(1992). Terror management and tolerance: Does mortality sali-
ence always intensify negative reactions to others who threaten
one’s worldview? Journal of Personality and Social Psychology,
63(2), 212–220.
Greenberg, J., Pyszczynski, T., Solomon, S., Simon, L., & Breus, M.
(1994). Role of consciousness and accessibility of death-related
thoughts in mortality salience effects. Journal of Personality
and Social Psychology, 67(4), 627–637.
Greenberg, J., Schimel, J., Martens, A., Solomon, S., & Pyszcznyski,
T. (2001). Sympathy for the devil: Evidence that reminding
Whites of their mortality promotes more favorable reactions to
White racists. Motivation and Emotion, 25(2), 113–133.
Helzer, E. G., & Pizarro, D. A. (2011). Dirty liberals! Reminders
of physical cleanliness influence moral and political attitudes.
Psychological Science, 22(4), 517–522. https:// doi. org/ 10. 1177/
09567 97611 402514
Hensel, L., Witte, M., Caria, A. S., Fetzer, T., Fiorin, S., Götz, F.
M., Gomez, M., Haushofer, J., Ivchenko, A., & Kraft-Todd, G.
(2022). Global Behaviors, Perceptions, and the Emergence of
Social Norms at the Onset of the COVID-19 Pandemic. Journal
of Economic Behavior & Organization, 193, 473–496.
Hertwig, R., & Ortmann, A. (2001). Experimental practices in eco-
nomics: A methodological challenge for psychologists? Behav-
ioral and Brain Sciences, 24, 383–403.
Horita, Y., & Takezawa, M. (2018). Cultural differences in strength of
conformity explained through pathogen stress: A statistical test
using hierarchical Bayesian estimation. Frontiers in Psychology,
9, 399830. https:// doi. org/ 10. 3389/ fpsyg. 2018. 01921
Inbar, Y., Pizarro, D. A., Knobe, J., & Bloom, P. (2009). Disgust sen-
sitivity predicts intuitive disapproval of gays. Emotion, 9(3), 435.
https:// doi. org/ 10. 1037/ a0015 960
Isler, O., Isler, B., Kopsacheilis, O., & Ferguson, E. (2020). Limits of
the social-benefit motive among high-risk patients: A field experi-
ment on influenza vaccination behaviour. BMC Public Health,
20(1), 1–9.
Isler, O., Yilmaz, O., & Dulleck, U. (2023). Scarcity improves eco-
nomic valuations when cognitively salient. Journal of Economic
Behavior & Organization, 209, 362–371.
Isler, O., Günaydın, G., Doğruyol, B., Dulleck, U., & Yilmaz, O.
(2024, August 11). Does scarcity impede cognitive function? A
comprehensive study with alternative resource scarcity perception
manipulations. https:// doi. org/ 10. 17605/ OSF. IO/ MKZAN
Janoff-Bulman, R., & Usoof-Thowfeek, R. (2009). Shifting moralities:
Post-9/11 responses to shattered national assumptions. The Impact
of 9/11 on Psychology and Education: The Day that Changed
Everything? (pp. 81–96). Springer.
Jones, A., & Fitness, J. (2008). Moral hypervigilance: The influence
of disgust sensitivity in the moral domain. Emotion, 8(5), 613.
https:// doi. org/ 10. 1037/ a0013 435
Jost, J. T., Stern, C., Rule, N. O., & Sterling, J. (2017). The politics of
fear: Is there an ideological asymmetry in existential motivation?
Social Cognition, 35(4), 324–353. https:// doi. org/ 10. 1521/ soco.
2017. 35.4. 324
Karwowski, M., Groyecka-Bernard, A., Kowal, M., & Sorokowski,
P. (2020a). Does thinking about coronavirus impact insight and
analytical reasoning? Thinking Skills and Creativity, 38, 100715.
https:// doi. org/ 10. 1016/j. tsc. 2020. 100715
Karwowski, M., Kowal, M., Groyecka, A., Białek, M., Lebuda, I.,
Sorokowska, A., & Sorokowski, P. (2020b). When in danger,
Turn right: Does Covid-19 threat promote social conservatism
and right-wing presidential candidates? Human Ethology, 35(1),
37–48. https:// doi. org/ 10. 22330/ he/ 35/ 037- 048
Krosch, A. R., & Amodio, D. M. (2014). Economic scarcity alters the per-
ception of race. Proceedings of the National Academy of Sciences,
111(25), 9079–9084. https:// doi. org/ 10. 1073/ pnas. 14044 48111
Landau, M. J., Solomon, S., Greenberg, J., Cohen, F., Pyszczynski, T.,
Arndt, J., ..., Cook, A. (2004). Deliver us from evil: The effects
of mortality salience and reminders of 9/11 on support for Presi-
dent George W. Bush. Personality and Social Psychology Bulletin,
30(9), 1136–1150. https:// doi. org/ 10. 1177/ 01461 67204 267988
Mani, A., Mullainathan, S., Shafir, E., & Zhao, J. (2013). Poverty
impedes cognitive function. Science, 341(6149), 976–980.
Mikulincer, M., Gillath, O., & Shaver, P. R. (2002). Activation of the
attachment system in adulthood: Threat-related primes increase
the accessibility of mental representations of attachment figures.
Journal of Personality and Social Psychology, 83(4), 881–895.
Milkman, K. L., Patel, M. S., Gandhi, L., Graci, H. N., Gromet, D. M.,
Ho, H., Kay, J. S., Lee, T. W., Akinola, M., Beshears, J., Bogard,
J. E., Buttenheim, A., Chabris, C. F., Chapman, G. B., Choi, J. J.,
Dai, H., Fox, C. R., Goren, A., Hilchey, M. D., ..., Duckworth, A.
L. (2021). A megastudy of text-based nudges encouraging patients
to get vaccinated at an upcoming doctor’s appointment. Proceed-
ings of the National Academy of Sciences, 118(20), e2101165118.
https:// doi. org/ 10. 1073/ pnas. 21011 65118
Mullainathan, S., & Shafir, E. (2013). Scarcity: Why having too little
means so much. Time Books.
Murray, D. R., & Schaller, M. (2012). Threat (s) and conformity decon-
structed: Perceived threat of infectious disease and its implications
for conformist attitudes and behavior. European Journal of Social
Psychology, 42(2), 180–188. https:// doi. org/ 10. 1002/ ejsp. 863
Murray, D. R., Schaller, M., & Suedfeld, P. (2013). Pathogens and poli-
tics: Further evidence that parasite prevalence predicts authoritari-
anism. PLoS One, 8(5), e62275. https:// doi. org/ 10. 1371/ journ al.
pone. 00622 75
Nail, P. R., McGregor, I., Drinkwater, A. E., Steele, G. M., & Thomp-
son, A. W. (2009). Threat causes liberals to think like conserva-
tives. Journal of Experimental Social Psychology. https:// doi. org/
10. 1016/j. jesp. 2009. 04. 013
Navarrete, C. D., & Fessler, D. M. (2006). Disease avoidance and eth-
nocentrism: The effects of disease vulnerability and disgust sen-
sitivity on intergroup attitudes. Evolution and Human Behavior,
27(4), 270–282.
Nelson, L. D., & Morrison, E. L. (2005). The symptoms of resource
scarcity: Judgments of food and finances influence preferences for
potential partners. Psychological Science, 16(2), 167–173.
O’Donnell, M., Dev, A. S., Antonoplis, S., Baum, S. M., Benedetti,
A. H., Brown, N. D., ..., Nelson, L. D. (2021). Empirical audit
and review and an assessment of evidentiary value in research on
the psychological consequences of scarcity. Proceedings of the

Behavior Research Methods
national academy of sciences, 118(44), e2103313118. https:// doi.
org/ 10. 1073/ pnas. 21033 13118
Oaten, M., Stevenson, R. J., & Case, T. I. (2009). Disgust as a disease-
avoidance mechanism. Psychological Bulletin, 135(2), 303.
Roux, C., Goldsmith, K., & Bonezzi, A. (2015). On the Psychology
of scarcity: When reminders of resource scarcity promote selfish
(and Generous) behavior. Journal of Consumer Research. https://
doi. org/ 10. 1093/ jcr/ ucv048
Schaller, M. (2011). The behavioural immune system and the psychol-
ogy of human sociality. Philosophical Transactions of the Royal
Society B: Biological Sciences, 366(1583), 3418–3426. https://
doi. org/ 10. 1098/ rstb. 2011. 0029
Schaller, M., Miller, G. E., Gervais, W. M., Yager, S., & Chen, E.
(2010). Mere visual perception of other people’s disease symp-
toms facilitates a more aggressive immune response. Psychologi-
cal Science, 21(5), 649–652.
Shah, A. K., Mullainathan, S., & Shafir, E. (2012). Some consequences
of having too little. Science, 338, 682–685. https:// doi. org/ 10.
1126/ scien ce. 122242
Shook, N. J., Thomas, R., & Ford, C. G. (2019). Testing the relation
between disgust and general avoidance behavior. Personality and
Individual Differences, 150, 109457. https:// doi. org/ 10. 1016/j.
paid. 2019. 05. 063
Sibley, C. G., Osborne, D., & Duckitt, J. (2012). Personality and
political orientation: Meta-analysis and test of a threat-constraint
model. Journal of Research in Personality, 46(6), 664–677.
https:// doi. org/ 10. 1016/j. jrp. 2012. 08. 002
Sirota, M., & Juanchich, M. (2018). Effect of response format on
cognitive reflection: Validating a two- and four-option multiple
choice question version of the Cognitive Reflection Test. Behavior
Research Methods. https:// doi. org/ 10. 3758/ s13428- 018- 1029-4
Spiller, S. A. (2011). Opportunity cost consideration. Journal of Con-
sumer Research, 38(4), 595–610. https:// doi. org/ 10. 1086/ 660045
Starmer, C., & Sugden, R. (1991). Does the random-lottery incentive
system elicit true preferences? An experimental investigation. The
American Economic Review, 81, 971–978.
Terrizzi, J. A., Jr., Shook, N. J., & McDaniel, M. A. (2013). The behav-
ioral immune system and social conservatism: A meta-analysis.
Evolution and Human Behavior, 34(2), 99–108.
Thornhill, R., Fincher, C. L., & Aran, D. (2009). Parasites, democ-
ratization, and the liberalization of values across contemporary
countries. Biological Reviews, 84(1), 113–131. https:// doi. or g/ 10.
1111/j. 1469- 185x. 2008. 00062.x
Trémolière, B., Neys, W. D., & Bonnefon, J. F. (2012). Mortality sali-
ence and morality: Thinking about death makes people less utili-
tarian. Cognition. https:// doi. org/ 10. 1016/j. cogni tion. 2012. 05. 011
Trémolière, B., De Neys, W., & Bonnefon, J. F. (2014). The grim
reasoner: Analytical reasoning under mortality salience. Think-
ing and Reasoning. https:// doi. org/ 10. 1080/ 13546 783. 2013.
823888
Tybur, J., Frankenhuis, W., & Pollet, T. (2014). Behavioral immune
system methods: Surveying the present to shape the future. Evo-
lutionary Behavioral Sciences, 8(4), 274–283. https:// doi. org/ 10.
1037/ ebs00 00017
Ullrich, J., & Cohrs, J. C. (2007). Terrorism salience increases system
justification: Experimental evidence. Social Justice Research, 20,
117–139. https:// doi. org/ 10. 1007/ s11211- 007- 0035-y
Van Bavel, J. J., Cichocka, A., Capraro, V., Sjåstad, H., Nezlek, J. B.,
Pavlović, T., Alfano, M., Gelfand, M. J., Azevedo, F., Birtel, M.
D., Cislak, A., Lockwood, P. L., Ross, R. M., Abts, K., Agadul-
lina, E., Aruta, J. J. B., Besharati, S. N., Bor, A., Choma, B. L.,
..., Boggio, P. S. (2022). National identity predicts public health
support during a global pandemic. Nature Communications, 13(1),
517. https:// doi. org/ 10. 1038/ s41467- 021- 27668-9
Van Leeuwen, F., Jaeger, B., Sleegers, W. W., & Petersen, M. B.
(2023). Do experimental manipulations of pathogen avoidance
motivations influence conformity?. Personality and Social Psy-
chology Bulletin. https:// doi. org/ 10. 1177/ 01461 67223 11606 55
Van de Vyver, J., Houston, D. M., Abrams, D., & Vasiljevic, M. (2016).
Boosting belligerence: How the July 7, 2005, London bombings
affected liberals’ moral foundations and prejudice. Psychol Sci,
27(2), 169–177. https:// doi. org/ 10. 1177/ 09567 97615 615584
Van der Toorn, J., Feinberg, M., Jost, J. T., Kay, A. C., Tyler, T. R.,
Willer, R., & Wilmuth, C. (2015). A sense of powerlessness
fosters system justification: Implications for the legitimation of
authority, hierarchy, and government. Political Psychology, 36(1),
93–110. https:// doi. org/ 10. 1111/ pops. 12183
Varma, M. M., Chen, D., Lin, X. L., Aknin, L. B., & Hu, X. (2020).
Prosocial behavior promotes positive emotion during the COVID-
19 pandemic. Emotion, 23(2), 538–553. https:// doi. org/ 10. 1037/
emo00 01077
Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and
validation of brief measures of positive and negative affect: The
PANAS scales. J Pers Soc Psychol, 54(6), 1063–1070. https:// doi.
org/ 10. 1037/ 0022- 3514. 54.6. 1063
Wu, B. P., & Chang, L. (2012). The social impact of pathogen threat:
How disease salience influences conformity. Personality and Indi-
vidual Differences, 53(1), 50–54.
Publisher's Note Springer Nature remains neutral with regard to
jurisdictional claims in published maps and institutional affiliations.