Content uploaded by Laurent Bègue
Author content
All content in this area was uploaded by Laurent Bègue on Jun 02, 2023
Content may be subject to copyright.
See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/353572755
Sacrificing Animals in the Name of Scientific Authority: The Relationship
Between Pro-Scientific Mindset and theLethal Use of Animals in
BiomedicalExperimentation
Article · July 2021
DOI: 10.1177/01461672211039413
CITATIONS
3
READS
966
2 authors:
Some of the authors of this publication are also working on these related projects:
Entretien recherche confiance (temporaire) View project
Evolutionary Approach in Binge Drinking View project
Laurent Bègue
Université Grenoble Alpes
183 PUBLICATIONS3,107 CITATIONS
SEE PROFILE
Kevin Vezirian
University of Western Australia
18 PUBLICATIONS228 CITATIONS
SEE PROFILE
All content following this page was uploaded by Laurent Bègue on 29 September 2021.
The user has requested enhancement of the downloaded file.
https://doi.org/10.1177/01461672211039413
Personality and Social
Psychology Bulletin
1 –16
© 2021 by the Society for Personality
and Social Psychology, Inc
Article reuse guidelines:
sagepub.com/journals-permissions
DOI: 10.1177/01461672211039413
journals.sagepub.com/home/pspb
Empirical Research Paper
Introduction
Most people have firm moral inhibitions toward violating
universal moral codes that prohibit unnecessary violence
toward living beings (Graham et al., 2013). However, indi-
viduals are also fully able to disengage from their moral con-
victions when doing so is dictated by specific circumstances
or by what they consider as the pursuit of higher goals
(Bandura, 1999). Scientific inquiry represents a culturally
valued goal that may require the killing of animals in the case
of animal experimentation for biomedical purposes. Since
the enactment of the Nuremberg code, animal experimenta-
tion has become a mandatory phase in most biomedical
development (Annas & Grondin, 1992), and every year,
approximately 125 million animals are used worldwide as
experimental resources (Knight, 2011; Taylor et al., 2008).
They are considered to play a crucial role in medicine, phar-
macology, and biotechnology and are used as models for
human disease, as sources of organs or cells, as well as in
animal experimentation, xeno-transplantation, and cloning
(Monamy, 2017).
This widespread use of animal experimentation raises a
moral paradox. While the Cartesian view of animals merely
as insensitive machines has been widely disproved by scien-
tific studies showing how the cognitive, emotional, and
behavioral lives of animals are complex (de Waal, 2016), and
sometimes similar to human capacities (which justifies their
use as models), in laboratories, animals are ultimately con-
sidered merely as scientific tools. This potentially induces
moral dilemmas for laboratory staff who perform invasive or
painful experiments, and who occasionally feel empathic
reactions or distress (Birke et al., 2007; Filippi et al., 2010;
Westbury & Neumann, 2008), the intensity of which is a
direct function of the phylogenetic closeness of the animals
involved (Miralles et al., 2019; Plous, 1993). For this reason,
when scientists administer a harmful treatment to an animal
during an experimental protocol, they may need to actively
overcome or repress their personal affective reaction. Most
of the time, this emotional disengagement is possible because
1039413PSPXXX10.1177/01461672211039413Personality and Social Psychology BulletinBègue and Vezirian
research-article2021
1Université Grenoble Alpes, Saint-Martin-d’Hères, France
Corresponding Author:
Laurent Bègue, LIP/PC2S, Université Grenoble Alpes, IUF, MSH-Alpes, BP
47, Grenoble Cedex 9, Saint-Martin-d’Heres 38040, France.
Email: Laurent.Begue-Shankland@univ-grenoble-alpes.fr
Sacrificing Animals in the Name of
Scientific Authority: The Relationship
Between Pro-Scientific Mindset and the
Lethal Use of Animals in Biomedical
Experimentation
Laurent Bègue1 and Kevin Vezirian1
Abstract
The present research investigated how scientific authority increases the lethal use of animals in biomedical experimentation.
In two behavioral studies (N = 151 and 150), participants were required to incrementally administer 12 doses of a toxic
chemical to a 53-cm fish (in reality, a biomimetic robot) for research on animal learning. Consistent with the Engaged
Followership Theory on obedience, participants placed in a pro-scientific mindset more severely harmed the laboratory
animal. In a cross-sectional study (N = 351), participants in medical fields endorsed a more pro-scientific attitude than those
in paramedical fields, which mediated their support for animal experimentation. Drawing on a representative European
sample (N = 31,238), we also confirmed the specificity of this link by controlling for potential demographic and ideological
confounds. In a final study (N = 1,598), instrumental harm was shown as mediating the link between a pro-scientific attitude
and support for animal experimentation.
Keywords
conformity, obedience, engaged followership, scientific authority, instrumental harm
Received November 6, 2020; revision accepted July 12, 2021
2 Personality and Social Psychology Bulletin 00(0)
the aim of the behavior is grounded in beliefs regarding the
authority of scientific endeavor (Birke et al., 2007).
Motivated Obedience
The present research starts with the idea that a motivated
obedience toward the authority of science lies at the core of
an individual’s moral disengagement when they are asked to
perform biomedical animal experimentation. By motivated
obedience, we mean that individuals do mainly follow extrin-
sic rules of conduct when they perform lethal operations on
animals for scientific purposes in research institutions.
They do not personally choose to use and sacrifice ani-
mals, or to sometimes expose these animals to painful proce-
dures, but they perform such acts as a part of their professional
requirements. However, even if these potentially problem-
atic behaviors are formally imposed by their role, they are
genuinely and actively involved in consenting, and generally
believe in the utility and the legitimacy of the professional
rules and procedures which they apply. This commitment is
grounded in the cultural authority of science, which repre-
sents a major pillar of the legitimization of animal experi-
mentation (Bauer et al., 2019).
In the context of our research, obedience toward scientific
authority is a specific type of motivated obedience. Motivated
obedience should therefore be contrasted with the “blind obe-
dience” that is generally associated with Milgram’s (1974)
studies on obedience. In the perspective developed in the
present research, obedience is not conceptualized as a passive
(or “agentic”) state involving an “ideological abrogation”
(Milgram, 1974, p. 145) but is, rather, a function of individu-
al’s identification with the scientific enterprise underlying
experimental research (Haslam et al., 2014; Reicher &
Haslam, 2011), and may therefore be more adequately under-
stood as an engaged followership (see Haslam & Reicher,
2017, for review), that is “an active identification with the
scientific project and those leading it” (Haslam et al., 2014, p.
473). We may consider that obedience toward scientific
authority is a specific type of motivated obedience.
In the present study, our aim was to experimentally and
cross-sectionally demonstrate that a pro-scientific mindset
increased attitudinal and behavioral support for painful and
potentially deadly experiments on animals. We also elabo-
rated and tested the hypothesis that a pro-scientific mindset
was related to a preference for a specific form of lay moral
thinking, instrumental harm, defined as a “willingness to
harm and even kill others when this is needed to achieve a
better outcome” (Kahane et al., 2017, p. 135), that would
mediate support for animal experimentation (more on this to
follow).
Scientific Authority and Support for
Animal Experimentation
The way in which scientific authority influences behaviors
was explored 50 years ago in a series of classical experiments
conducted by Stanley Milgram, who observed that the elec-
tric shocks administered to a reluctant victim by laypersons
were maximal in a scientific institution of high prestige (Yale
University) and lowest in a commercial building in Bridgeport
(Milgram, 1974, Experiment 10). In the same vein, Milgram
(1974) showed that obedience was lower when the experi-
menter was presented as an “ordinary man” rather than a sci-
entist (Milgram, 1974, Experiment 13), and considered
potentially important in interpreting his results, in terms of
“the prestige of the scientist, the prestige and worth of the
research” (Reicher et al., 2012, p. 317). While the parallel
between Milgram’s obedience studies and animal use in lab-
oratory experiments seems relevant here, the psychological
process proposed by Milgram (1974) to explain the phenom-
enon of obedience as an “agentic state” in which “responsi-
bility is cast off, and individuals become thoughtless agents
of action” (p. 176) could be argued to be circular and does
not properly capture critical aspects of the phenomenon
(Gibson, 2019; Miller, 1986). Recent studies on obedience
have increasingly focused on the social legitimacy and iden-
tity involved in obedience (Reicher et al., 2014). It is cur-
rently acknowledged that a participant’s willingness to
administer electric shocks cannot be properly explained
purely by blind obedience. Instead, it may be a function of
their active identification with the scientific enterprise under-
lying the experiment, and the perceived legitimacy of the
experimenter’s injunctions (Haslam et al., 2014; Reicher &
Haslam, 2011). According to Engaged Followership Theory,
participants follow the experimenter’s directions because
they believe themselves to be “contributing to a moral, wor-
thy, and progressive cause” (Haslam et al., 2015, p. 60).
Despite the appeal of the Engaged Followership Theory,
there is still no direct empirical demonstration of this theo-
retical perspective on obedience using Milgram-like behav-
ioral measures. Instead, this perspective has mainly been
based on qualitative studies based on Milgram’s archives
(Haslam et al., 2014, 2015) along with a study based on
immersive digital reality, which showed that participants’
relative identification with the experimenter and their scien-
tific goals was a good predictor of the maximum level of
shock they administered (Haslam et al., 2015; see also
Reicher et al., 2012). However, a test of the Engaged
Followership Theory with Milgram-like strong behavioral
measures has been absent (Gibson, 2019). The present
research therefore aims to fill this gap. It also intends to
extend its scope and provide additional evidence on the spec-
ificity of the link between a scientific mindset and support
for animal experimentation, and to conceptualize and empiri-
cally test the psychological process involved.
Overview of the Present Research
First, after the development and validation of a new realis-
tic and involving biomedical protocol on animal experi-
mentation (Study 1a), we experimentally investigate how
the salience of scientific authority induces individuals to
Bègue and Vezirian 3
administer painful chemical injections to an animal target
(Study 1b). Then, in two cross-sectional studies involving a
sample of medical and paramedical students (Study 2a) and
a large and representative European sample (Study 2b), we
assess the generalizability and specificity of the relationship
between a pro-scientific attitude and support for animal
experimentation. Finally, we further clarify the process by
which a pro-scientific attitude possibly influences an indi-
vidual’s support for animal experimentation (Study 3). We
elaborate and test the hypothesis of a mediating effect of a
core dimension on utilitarianism, instrumental harm, which
is conceptually and empirically distinct from motivated obe-
dience and is defined as a “willingness to harm and even kill
others when this is needed to achieve a better outcome”
(Kahane et al., 2017, p. 135).
Study 1a: Development of a Behavioral
Measure of Animal Objectification
This preliminary study aimed to develop a completely new
experimental protocol in which participants were required to
incrementally administer a noxious chemical substance to an
animal (which they believed to be a fish, although in reality
it was a biomimetic robot) involved in a learning experiment,
leading to the death of the supposed animal. The adminis-
tered substance was explained to the participants to stimulate
learning in the context of a research protocol investigating
Alzheimer’s disease. However, an important side effect of
the drug was its consequences on vital functions at high dos-
ages. To validate this new experimental paradigm, we mea-
sured relevant inter-individual variables involved in
destructive behavior toward animals (Amiot & Bastian,
2015, for review) and expected that individuals’ reluctance
to carry out the task would be contingent on these variables.
The first of these variables was social dominance orienta-
tion (SDO; Sidanius & Pratto, 1999), a pro-hierarchic orien-
tation that represents the value that people place on
non-egalitarian and hierarchically structured relationships
among social groups. In the hierarchy of social categories,
animals represent “the quintessential low status group”
(Dhont et al., 2019, p. 31), which is why attitudes toward
them are reported to be more negative among people with
high SDO (Caviola et al., 2019; Hoffarth et al., 2019).
Another validation measure was speciesism, defined as the
“belief that humans are intrinsically more valuable than indi-
viduals of other species” (Caviola et al., 2019, p. 2).
Speciesism predicts behavioral preferences toward humans
and “superior” animals in relation to allocating money or
time, and is related to the devaluation of the perceived intel-
ligence and suffering capability of animals (Caviola et al.,
2019). We also measured dispositional empathy, which is
inversely related to support for animal experimentation
(Broida et al., 1993), as well as gender and vegetarian or
vegan diet (Amiot & Bastian, 2015; Broida et al., 1993;
Ormandy & Schuppli, 2014).
In summary, we included a selection of relevant measures
to check the validity of our new protocol measuring harm
toward animals. We expected a positive relationship between
SDO, speciesism, and the intensity of harm inflicted on the
animal during the experiment, and a negative relationship
with empathy, vegetarianism, and (female) gender.
Method
Data availability. Materials and data for the five studies are
openly available on OSF.
https://osf.io/6tvgz/
Participants. Because of resource constraints (Lakens,
2021), we were willing to recruit 150 participants to this
pilot study. Using G*Power 3.1 (Faul et al., 2007), we car-
ried out a sensitivity analysis (setting an error alpha rate of
.05, a power of .80, 150 participants, and one predictor). The
analysis indicated that with this design, the minimum effect
we could detect between the drug administration task and
one of our predictors would be a Cohen’s d = .46. In total,
152 participants (95 females, 55 males, and two people who
did not indicate their gender) were recruited. Their ages
ranged from 19 to 71 years, and the sample comprised
45.4% students, 32.3% workers, 14.6% unemployed, and
7.7% who omitted to indicate their activity. The percentage
of vegetarians or vegans was 10.9%. Participants were
recruited through the local press to take part in a 40-min
experiment, which was presented as an observational study
of animal behavior during a learning task. They would be
paid 15 euros (US$16.5) for their participation. They were
first invited to complete an online questionnaire to measure
inter-individual variables and to check their eligibility for
inclusion.1
Measures. SDO was measured using an eight-item scale
(Ho et al., 2015) containing items such as “Some groups of
people are simply not the equals of others” and “It is unjust
to try to make groups equal” (M = 1.93, SD = 0.55, Cron-
bach’s α = .72). To measure speciesism, we used the six-
item Speciesism scale by Caviola et al. (2019), which
includes items such as “Morally, animals always count for
less than humans” and “Humans have the right to use ani-
mals however they want to” (M = 2.09, SD = 0.65, Cron-
bach’s α = .72). We also used the seven-item Empathic
Concern scale by Davis (1983), which includes items such as
“When I see someone being taken advantage of, I feel kind
of protective towards them” and “Other people’s misfor-
tunes do not usually disturb me a great deal” (reverse scored)
(M = 3.88, SD = 0.57, Cronbach’s α = .77). The items from
the three scales were presented as 5-point Likert-type scales.
We imputed the median score of the distribution to sporadic
missing values (less than 5%). Finally, we regarded partici-
pants as vegetarians if they replied positively to the follow-
ing question: “I am vegetarian (or) vegan.”
4 Personality and Social Psychology Bulletin 00(0)
Procedure
Learning protocol. Upon their arrival, participants were
shown a short video containing various items of scientific
and procedural information (Supplemental Material 2). Then,
they were seated at a table in front of a screen. Three meters
from the table was a large aquarium (200 cm × 200 cm × 80
cm, containing 3,000 L of water). Inside the aquarium, what
appeared to be a 53-cm goldfish was moving around. This
was actually not a real fish, but a biomimetic robot (Airo 9
model) covered by very realistic silicone rubber which had
been modeled and painted by the designer Alain Quercia.
The choice of the fish was consistent with the fact that most
of contemporary laboratory research works are based on
rodents and fish. Moreover, fish robots appear more realistic
than mammal robots because of their less complex move-
ments. Participants were instructed to click on the first of
12 buttons, which triggered the following sequence: (a) the
injection via motorized syringe into the water of 10 mL of
the toxic pharmacological substance (which was actually just
water colored yellow), (b) the triggering of a spotlight which
illuminated a specific perimeter of the aquarium for 12 s, and
(c) the emission of bubbles during those 12 s from the floor
of the aquarium. Then, the participants were asked to count
the number of times the fish went behind an orange frame in
a time window of 25 s. The participants were told that the
fish had previously been trained to move toward this frame
when the stimulation (i.e., light and bubbles) was delivered,
and that it was expected that the treatment would increase its
sensitivity to the stimuli. When the time window was over,
they had to click on the next button and so on until the 12th
and final button (see Figure 1).
Below some of the buttons, the expected probability of
the death of the fish was written (Figure 2) as follows: 0%
probability of death (Button 1), 33% (Button 3), 50% (Button
6), 75% (Button 9), and 100% (Button 12). Moreover, the
cardiac pace of the animal was visualized via an oscillo-
scope, which also produced auditory feedback. Contingent to
the number of doses injected, the sine wave and sounds
became increasingly erratic, reflecting cardiac distress. A
short video of the setting is available at https://youtu.be/
exNHKprKNwI.
If a participant stated that they were unwilling to con-
tinue, then the experimenter stated as follows: “For our
research, it is important that you go on until 12, but as was
indicated in the consent form, you are allowed to stop at any
time.” If the participant repeated their intention to stop, they
were told by the experimenter, “you want to stop, are you
sure?”
When the experiment was over, the participants were
shown into another room for a filmed semi-directive funnel
interview and to complete a final evaluation survey. Please
see Figures 3 and 4, and Supplemental Material 4.
Post-experimental screening procedure and final sample. Two
independent judges were asked to watch the videos of the
debriefing sessions and assign codes to the participants based
on their responses. Participants who expressed no substantial
doubts about the aim of the experiment were coded 1. If a
participant expressed slight doubts about the meaning of the
experiment, but without threatening the overall credibility of
the cover story, they were coded 2. The judges were instructed
to assign Code 2 to a participant if they were unsure whether
Figure 1. Schematic representation of the experimental setting.
Bègue and Vezirian 5
the participant believed what they were told during the
experiment or not. Finally, participants who expressed clear
doubts about core aspects of the study (e.g., if they suspected
that the substance that they administered or the fish were
fake) were coded 3.
Analysis of the coding indicated a 91.4% initial conver-
gence among judges (139 cases out of 152), and 100% con-
vergence was reached after discussion of the divergent cases.
Ultimately, 73.7% (N = 112) of participants were coded 1;
11.8% were coded 2 (N = 18); and 13.8% (N = 21) were
Figure 2. The numeric device.
Figure 3. The full setting.
6 Personality and Social Psychology Bulletin 00(0)
coded 3. One participant in the latter category who was
included by error in the study (a psychology student, see
exclusion criteria) was excluded from the analysis. Only par-
ticipants coded 1 or 2 were included in the main analysis
(N = 130, 65.4% female, aged 19–71 years, Mage = 31.74
years, SDage = 13.73 years, 10.3% vegetarians).2 Finally,
participants with missing values on gender (N = 2), vegetari-
anism (N = 1), or personality variables (N = 3) were listwise
deleted.
Results
As shown in Figure 5, the shape of the distribution was
bimodal. While 23.1% of participants refused to begin the
task (N = 30), 54.6% fully completed the task through to
the end (N = 71). Between 0.8% and 3.8% of participants
stopped at one of the 11 remaining intermediary modalities.
To analyze the relationships between individual variables
and the number of injections, taking into account the
bimodal distribution, we therefore compared those who
were opposed to pressing the first button (0 injections,
coded 0) to all of the others combined (coded 1). Please see
Figure 5.
Main results. A logistic regression analysis was performed to
estimate the odds ratio (OR) and 95% confidence interval
(95% CI) of every variable in a single block (Table 1). The
results showed that SDO, speciesism, and vegetarianism
were significantly related to the quantity of toxic substance
administered to the target, whereas it was not the case for
age, gender, and empathy. The overall model accounted for
42% of the variance (Nagelkerke’s pseudo R2). Please see
Table 1.
Figure 4. The motorized syringe containing the supposed toxic substance.
0
20
40
60
80
012345678910 11 12
Injections by toxicity level -study 1a
0
10
20
30
40
50
60
012345678910 11 12
Injections by toxicity level -study 1b
Figure 5. Bimodal relationship between harmful injections and
the decision to stop participating in the experiment.
Bègue and Vezirian 7
Ancillary analysis. According to the Engaged Followership
Theory (Reicher & Haslam, 2011), it was expected that par-
ticipants in the experiment would positively rate their par-
ticipation in the experiment (see Haslam et al., 2015).
Overall, the participants were satisfied with the experiment
(M = 3.75, SD = 4.00). Moreover, the more the participants
injected the toxic product into the fish’s tank, the more they
were satisfied (ρ = .32, p < .001) and considered the task to
be easy, interesting, and useful (ρ = .23, .20, and .22, respec-
tively, p < .05). Please see Supplemental Material 4.
Discussion
We developed a new experimental protocol in which partici-
pants were required to incrementally administer a noxious
chemical substance to an animal involved in a learning
experiment, leading to the death of the animal. To check the
validity of this new measure, we analyzed the relationship
between relevant individual factors and the quantity of toxic
substance administered to the laboratory animal. We
observed that the credibility of the overall protocol was
appropriate, as only a low proportion of participants had to
be excluded because of suspicions regarding the experiment.
As we had hypothesized, we observed that the administration
of the lethal substance was positively related to SDO and
speciesism, and that vegetarianism was inversely related to
the completion of the protocol. In conclusion, our behavioral
protocol provided consistent results and seemed fully appro-
priate to behaviorally measure the destructive use of an ani-
mal during a biomedical experiment.
Study 1b: The Influence of a Pro-
Scientific Mindset on Animal Harm
If science represents a cultural authority, the mere priming of
science should increase a participant’s commitment to scien-
tific expectations. In this study, the hypotheses were prereg-
istered (https://osf.io/6tvgz/).3 We hypothesized that the
salience of science would encourage the administration of
the toxic substance to the target fish. We expected that a par-
ticipant’s increased adherence to the scientific goal and the
priming of their support for science would increase their con-
formity to scientific expectations, regardless of the psycho-
logical cost of inducing pain and killing the fish. This
perspective was developed in line with the Engaged
Followership Theory, which supposes that obedience is a
function of identification with the scientific enterprise under-
lying experimental research (Haslam et al., 2014; Reicher &
Haslam, 2011). The same inter-individual measures as in
Study 1a were included, with the same hypothesis.
Method
Participants. Based on the conclusive previous study, we
recruited 150 participants to this study. The sensitivity analy-
sis we ran on G*Power indicated that our design allowed us to
detect a minimal effect of the priming task on the toxic sub-
stance administration of Cohen’s d = .40 (in the t-test compo-
nent, one-tailed, setting an error alpha rate of .05, a power of
.80, and 75 participants in each of the two conditions).
The sample of this study is composed of 150 participants
(98 females, 51 males, and one who did not indicate his or
her gender), with ages ranging from 18 to 65 years. The sam-
ple included 48.7% students, 35.3% workers, 13.3% unem-
ployed, and 2.7% who omitted to indicate their current
activity. Moreover, 16.3% of the participants were vegetari-
ans. We relied on the same recruitment procedure and the
same exclusion criteria that were used in Study 1a.
Measures. The same scales were selected as in Study 1a:
SDO (M = 1.88, SD = 0.64, Cronbach’s α = .68), specie-
sism, (M = 1.90, SD = 0.68, Cronbach’s α = .78), and
empathic concern (M = 3.92, SD = 0.58, Cronbach’s α =
.75). All the items were presented using 5-point Likert-type
scales. The mean score of the distribution was imputed to
missing values (less than 5%).
Procedure
Induction of pro-scientific salience. Following the comple-
tion of the consent form, the participants were randomly
placed either in a science promotion condition or in a sci-
ence critical condition. Following a previously established
Table 1. Logistic Regression Results Contrasting Participants Who Did Not Press the First Button (Coded 0) to All of the Combined
Others (Coded 1), Study 1a.
Variable B SE Exp(B) 95% CI Significance
Age −0.36 0.01 0.96 [0.93, 1.00] .058
Gender −0.46 0.60 0.62 [0.19, 2.04] .441
SDO 1.57 0.57 4.83 [1.55, 15.01] .006
Speciesism 1.48 0.46 4.40 [1.77, 10.93] .001
Vegetarianism −0.68 0.24 0.50 [0.31, 0.81] .006
Empathy −0.02 0.47 0.97 [0.38, 2.47] .965
Constant −1.47 2.52 0.23 / .559
Note. CI = confidence interval; SDO = social dominance orientation.
8 Personality and Social Psychology Bulletin 00(0)
procedure (Haslam et al., 1999; Reicher, Birney, & Haslam,
unpublished, quoted in Haslam & Reicher, 2017), the partici-
pants in the science promotion condition were asked to write
down three things that were important about science, to indi-
cate what they liked about science, and to state what they felt
they had in common with scientists. In the science critical con-
dition, they were asked to list three things they believed to be
problematic about science, what they disliked about science,
and what differentiated them from scientists. The experi-
menter who handled the salience instruction with the partici-
pants (and who remained next to them during the experiment)
was blind to the experimental conditions. To check for the
effectiveness of the manipulation, participants were then
asked to answer a 5-point Likert-type scale composed of five
items adapted from Gonzalez-Franco et al. (2018), including
items such as “I fully share the aims of science,” and “I feel
close links with people to whom science is the most important
thing in their life” (M = 3.63, SD = 0.67, α = .79).
The development of the overall protocol was almost iden-
tical to that described in Study 1a, with only slight modifica-
tions made to it to improve its realism and impact (please see
Supplemental Material 6).
Post-experimental screening procedure and final sample. All
the participants were coded 1, 2, or 3 by independent judges
following the same criteria used in Study 1a. The procedure
used to achieve the final categorization was slightly differ-
ent from that of the first study, as five judges watched the
whole or some of the videos, and the final code was cho-
sen following the modal categorizations of those judges.
Three participants were excluded because they had to leave
the experiment before the end for independent reasons,
or did not fully understand the instructions. Finally, 70.1%
(N = 103) of participants were coded 1, 17.0% (N = 25)
were coded 2, and 12.9% (N = 19) were coded 3. As in
Study 1a, we only included participants coded 1 or 2
(N = 128, 65.6% females, aged 19–65 years, Mage = 29.67
years, SDage = 12.83 years, 19.2% of vegetarians), and we
excluded all participants coded 3.4 Finally, participants
with missing values for gender (N = 1) and vegetarianism
(N = 3) were listwise deleted.
Results
Manipulation checks and preliminary analysis. Participants in
the pro-science condition scored significantly higher than the
participants in the anti-science on the scale measuring their
attitude toward science (M = 3.51, SD = 0.64 vs. M = 3.75,
SD = 0.66), t(122) = 2.09, d = .38, 95% CI = [0.02, 0.74],
p = .03. The distribution of the injection number was
bimodal, and similar to Study 1a (Figure 5, see also Supple-
mental Material 8). While 28.1% participants refused to
begin the task (N = 36), 44.5% fully completed the task
through to the end (N = 57). Between 0.8% and 6.3% of the
participants stopped at one of the 11 remaining intermediary
modalities. Please see Figure 5.
To analyze the relationships between the individual vari-
ables and the target behavior, we therefore created two
groups, as we did in Study1a: those who stopped before the
first injection (i.e., no injections, coded 0) were therefore
compared with all the others combined (coded 1).
Primary analysis. A logistic regression analysis was per-
formed to estimate the OR and 95% CI of every selected
variable in a single block. The results showed that partici-
pants in the pro-science condition (Code 2) administered a
higher quantity of the toxic substance than participants in the
anti-science condition (Code 1), OR = 2.81, 95% CI =
[1.09, 7.22], p = .03. Furthermore, age was inversely related
to the administration of the toxic substance to the fish (OR =
0.96, 95% CI = [0.93, 1.00], p = .05), and substance admin-
istration was lower among vegetarians (OR = 0.25, 95% CI
= [0.08, 0.74], p = .01). The administration of the toxic sub-
stance was not significantly related with speciesism, SDO, or
empathy. The overall model accounted for 28% of the vari-
ance (Nagelkerke’s pseudo R2). Please see Table 2 and Sup-
plemental Material 7.
Discussion
In this study, we showed that participants who had been
placed in a scientific mindset were more willing to follow
scientific instructions by inflicting pain and harming a target
animal involved in a pharmacological experiment. To our
knowledge, this represents the first direct behavioral confir-
mation of the Engaged Followership Theory, which consid-
ers that obedient behavior is fully motivated, and is based on
an active identification with the scientific enterprise underly-
ing experimental research (Reicher & Haslam, 2011).
However, one may consider that a generalized pro-
authoritarian attitude could have been made salient by scien-
tific priming. While such an interpretation would seem on
the surface to be inconsistent with surveys showing that pro-
scientific attitudes are inversely related to the support of pro-
authoritarian institutions and worldviews (Rutjens et al.,
2018), additional studies should more directly demonstrate
that support for science specifically caused the observed
effect, not support for authority in general. Future research
should also seek to clarify whether or not the observed link
between a pro-science mindset and support for animal exper-
imentation can similarly be observed on a larger scale and in
a more diverse and representative sample.
Study 2: Survey Studies on Pro-
Scientific Attitude and Support for
Animal Experimentation
Study 2a
In the previous study, we experimentally showed that people
who were primed toward scientific authority were more
inclined to administer a lethal substance to an animal during
Bègue and Vezirian 9
biomedical research. If this phenomenon is the consequence
of the cultural authority of science, it should also be observ-
able outside the laboratory and assessable by direct and
explicit verbal measures. To assess the generalizability of our
results, we therefore designed a correlational study on a
larger sample and tested for cross-sectional relationships
between pro-scientific mindset and support for animal exper-
imentation among both medical and pharmacy students
(medical group) and physical therapy and midwifery stu-
dents (paramedical group). The sampling of participants in
contrasting disciplinary fields with unequal support for ani-
mal experimentation enabled us to investigate the structural
relationship between our constructs. We expected to find a
positive association between pro-scientific attitudes and
support for animal experimentation. Moreover, we also
explored how differences in pro-scientific attitudes in these
contrasting samples mediated their attitude toward animal
experimentation.5
Method
Sampling procedure and participants. We benefited from the
context of another independent study to distribute a paper-
and-pencil survey during a university course. The sample
comprised 313 students (73.6% females, aged 19–37 years)
taking first-year medicine (N = 157), pharmacy (N = 67),
physical therapy (N = 52), and midwifery (N = 37) courses.
One participant who did not reply to the gender question
(N = 1) was excluded from the analysis. Missing data (<5%
for every variable) were imputed means.
A sensitivity analysis which we ran on G*Power indicated
(setting an error α rate of .05, a power of .80, 313 partici-
pants) that our design allowed us to detect a minimum
correlation of r = .16 (Cohen’s d = .31), if it existed, on the
link between pro-scientific attitudes and support for animal
experimentation.
Measures. A 5-point Likert-type survey was given to partici-
pants, with options ranging from 1 (Totally disagree) to 5
(Totally agree) for all the scales. Pro-scientific attitude was
measured using two items (“I believe that science can con-
tribute to making the world better” and “I feel at home in the
scientific field,” M = 3.16, SD = 0.62, Cronbach’s α = .60).
Participants’ attitude toward biomedical animal experimen-
tation was also measured via two items (“When it comes to
finding the best treatments for illnesses, doing experiments
on animals is justified” and “Sacrificing animals like mice or
rats for scientific research is normal”). A 5-point Likert-
type survey was proposed, with options ranging from 1
(totally disagree) to 5 (totally agree) (M = 1.95, SD = 0.97,
Cronbach’s α = .82).
Results
As we expected, support for animal experimentation was
significantly related to pro-scientific attitudes (rPearson = .27,
95% CI = [0.16, 0.37], p < .001). The relationships were
similarly observed among males and females. We then rea-
soned that knowledge in the fields of medicine and pharmacy
was more intrinsically related to animal experimentation
than knowledge in the paramedical fields of physical therapy
and midwifery. We therefore expected that medicine and
pharmacy students would be more supportive of animal
experimentation, so we pooled medical and pharmacy stu-
dents and physical and midwifery students into two separate
groups and performed a univariate analysis of variance, with
the two groups as independent variables and attitude toward
animal experimentation as a dependent variable. Gender and
age were also introduced as a control in the model because of
the relevance of these variables in the prior literature on ani-
mal experimentation (Broida et al., 1993; Ormandy &
Schuppli, 2014), as well as the unbalanced distribution of the
gender variable in the two groups (medicine–pharmacy,
n = 224; physical therapy–midwifery, n = 89). The results
showed that animal experimentation was more supported
by males than females (M = 2.40, SD = 1.00 vs. M = 1.79,
SD = 0.91), F(1, 312) = 13.14, p < .001, d = .41, 95% CI
= [0.19, 0.63]. More importantly, the students in the medi-
cine-pharmacy group more strongly supported animal exper-
imentation than those in the physical therapy-midwifery
Table 2. Logistic Regression Results Contrasting Participants Who Did Not Press the First Button (Coded 0) to All of the Combined
Others (Coded 1), Study 1b.
Variable B SE Exp(B) 95% CI Significance
Age −0.03 0.01 0.96 [0.93, 1.00] .050
Gender −1.04 0.54 0.35 [0.12, 1.03] .057
SDO 0.73 0.50 2.08 [0.78, 5.57] .142
Speciesism 0.41 0.41 1.51 [0.67, 3.42] .316
Vegetarianism −1.37 0.55 0.25 [0.08, 0.74] .013
Empathy 0.06 0.43 1.06 [0.45, 2.49] .878
Science Prime 1.03 0.48 2.81 [1.09, 7.22] .032
Constant 0.33 2.53 1.39 / .896
Note. CI = confidence interval; SDO = social dominance orientation.
10 Personality and Social Psychology Bulletin 00(0)
group (M = 2.06, SD = 0.97 vs. M = 1.66, SD = 0.92),
F(1, 312) = 6.86, p < .01, d = .30, 95% CI = [0.07, 0.52].
The interaction between participants’ gender and their disci-
plinary field was not significant, F(1, 312) = 0.24, p = .62.
To explore the possible mediator status of pro-scientific
attitude, we then analyzed the relationship between the disci-
plinary field and attitude toward animal experimentation. We
performed multiple regressions in the components paths
(Baron & Kenny, 1986) with the “JSmediation” R package
(Yzerbyt et al., 2018) and found that the disciplinary field
represented a significant predictor of attitude toward animal
experimentation (path c), b = 0.390, SE = 0.120, t(311) =
3.24, p < .001, and also of pro-scientific attitude (path a),
b = 0.312, SE = 0.077, t(311) = 4.07, p < .001. These
results indicate that students on medical courses hold signifi-
cantly more positive attitudes toward science and animal
experimentation than students on paramedical courses. Our
analysis also indicated that while controlling for the univer-
sity course, attitudes toward science still significantly
predicted attitudes toward animal experimentation (path b),
b = 0.375, SE = 0.086, t(310) = 4.34, p < .001. Controlling
for the attitude toward science, our model indicated that the
disciplinary field remained a significant predictor of the atti-
tude toward animal experimentation (path c′), b = 0.273,
SE = 0.120, t(310) = 2.28, p = .024, confirming a media-
tion. Upon analyzing the magnitude of this effect using the
Monte Carlo sampling method (5,000 iterations), the indirect
effect was estimated, b = 0.12, 95% CI = [0.05, 0.20].
Please see Figure 6.
Discussion
In this study, we showed that students in medical fields
endorsed more pro-scientific attitudes than students in para-
medical fields, and that this was related to their support for
animal experimentation. This finding suggests that pro-sci-
entific attitudes, albeit obviously distinct from support for
animal experimentation (as indicated by the low correlation
coefficient between the constructs) were linked in a funda-
mental way to this attitude, and that the experimental results
previously observed therefore cannot be reduced to an exper-
imental artifact. Moreover, relying on students in contrasting
Figure 6. Path diagram for c. The total effect of disciplinary field on support for animal experiment c′. The indirect effect of disciplinary
field on support for animal experiments through pro-scientific attitude.
fields, the finding shows that a pro-scientific attitude seemed
to mediate the effect of disciplinary preferences or socializa-
tion (see Guimond & Palmer, 1990) on a relevant attitude.
However, we did not control for some potential confounds
related to ideological variables in this study. Moreover, the
sample of medical and paramedical students was specific,
and not representative of the general population. Study 2b
was consequently devised to assess the solidity of the rela-
tionship between support for scientific authority and support
for animal experimentation, and to rule out the possibility
that this link represents an artifact of ideological overlap.
Study 2b
The aims of this study were twofold. First, we intended to
consolidate the previously observed relationship between
pro-scientific attitude and support for animal experimenta-
tion through a large and representative European sample
(N = 31,238). We also aimed to introduce some necessary
controls to check for the specificity of the observed relation-
ship. Previous studies have shown that attitudes toward
science as well as toward animal experimentation were
related to demographic variables (Crettaz von Rotten, 2008;
Hagelin et al., 2003), and to ideological factors such as
political orientation, authoritarian values, and religion
(Furnham & Heyes, 1993; Hagelin et al., 2003). We there-
fore intended to assess whether or not the observed link
between pro-scientific attitude and support for animal exper-
imentation still holds when five key variables are taken into
account: age, gender, political orientation, religious atten-
dance, and authoritarian values.
Method
Sampling procedure and participants. The Special Eurobarom-
eter (EB) on Science and Technology served as the data
source for this project. This large survey provides represen-
tative information on the populations of the 28 European
Union member states, as well as on Turkey, Switzerland,
Iceland, and Norway (N = 31,238, Mage = 47.35 years,
SDage = 18.23 years, 53.2% female). A multistage random
Bègue and Vezirian 11
(probability) sampling was applied. All the interviews took
place in 2010, face-to-face in people’s homes, and in the
appropriate national language. For the sake of not losing
information, when missing values were below 10% (which
was the case for age, attitude toward science, religious atten-
dance, and attitude toward animal experimentation), we
imputed means for items requiring ordinal or continuous
answers. Listwise deletion was applied for the remaining
variables (gender and political orientation).
We ran a sensitivity analysis with the same design as the
previous study, but the adopted sample size indicated that the
smallest size that we could detect was of r = .01 (Cohen’s
d = .03). Even setting the lower bound to the previous found
effect (r = .16), a post hoc analysis indicated that the power
of this large sample study was estimated to be 1.
Measures
Pro-scientific attitude. Pro-scientific attitude was measured
using eight EB items with a 5-point Likert-type scale, with
options ranging from 1 (totally disagree) to 5 (totally agree).
Example items included the following: “Science should have
no limits to what it is able to investigate”; “The benefits
of science are greater than any harmful effects it may cause”
(M = 3.19, SD = 0.65, Cronbach’s α = .71).
Attitude toward animal experimentation. Attitude toward
the use of animals in biomedical research was measured
using two EB items with a 5-point Likert-type scale, with
options ranging from 1 (totally disagree) to 5 (totally agree).
Example items included the following: “Scientists should be
allowed to do research on animals like mice if it produces
new information about health problems”; “Scientists should
be allowed to experiment on animals like dogs and monkeys
if this can help sort out health problems” (M = 3.39, SD =
1.15, α = .72).
Social and political measures. Political positioning was
measured using the following standard question: “In politi-
cal matters people talk of the ‘left’ and ‘the right’. How
would you place your views on this scale?” (1 = Left, 10 =
Right; M = 5.33, SD = 2.20). We also measured religious
attendance, which is considered as the single most relevant
indicator of religious commitment (Gorsuch & McPherson,
1989). Participants were therefore asked the following ques-
tion: “Apart from weddings or funerals, about how often do
you attend religious services?” (1 = never, 8 = more than
once a week, M = 3.68, SD = 2.21).
Authoritarian-related values. A single-item measure on
value prioritization in relation to opposing the protection of
freedom of speech versus fighting crime and terrorism was
used as a proxy measure for authoritarianism (see Altemeyer,
1998; Hendricks, 2019): “Which of the following do you
think is most important? Protecting freedom of speech
and human rights (1) or fighting crime and terrorism (2)”
(M = 1.45, SD = 0.49).
Results
All the variables were introduced simultaneously in a single
block in a linear multiple regression model.6 As is indicated
in Table 3, pro-scientific attitude was positively related to
support for animal experimentation after controlling for gen-
der, age, religious commitment, political orientation, and
authoritarian-related values. Animal experimentation was
significantly more strongly endorsed among males, older,
religious, right-wing participants, and participants holding
authoritarian-related values. The overall model (adjusted R2)
accounted for 11.1% of variance.
Discussion
In Studies 2a and 2b, we showed that pro-scientific attitude
was cross-sectionally related to attitude toward animal
experimentation, confirming with the survey studies the
results we had experimentally observed in the previous study.
The results confirmed that the relationship between pro-sci-
entific attitude and support for animal experimentation was
not reducible to confounds such as political orientation,
authoritarian-related values, or religious commitment.
Study 3: Testing a Mediation Model
In the previous studies, we showed that a pro-scientific
mindset and support for animal experimentation were caus-
ally and cross-sectionally related to each other. One possible
Table 3. Linear Regression Model Predicting Support for Animal Experiment (Study 2b).
Variable B SE Beta tSignificance 95% CI
Age 0.005 0.000 .078 12.44 .001 [0.004, 0.006]
Gender −0.289 0.014 −.126 −20.15 .001 [−0.31, 0.261]
Pro-scientific attitude 0.498 0.011 .28 45.54 .001 [0.47, 0.519]
Political orientation 0.019 0.003 .037 5.91 .001 [0.013, 0.026]
Authoritarian values 0.056 0.014 −.02 −3.89 .001 [−0.08, 0.020]
Religious attendance 0.009 0.003 .017 2.63 .008 [0.002, 0.015]
Constant 1.98 0.57 34.96 .001 [1.87, 2.09]
Note. CI = confidence interval.
12 Personality and Social Psychology Bulletin 00(0)
process explaining this relationship lies in the instrumental
mindset that may be inherent to the scientific perspective,
and which may numb participants’ attitude to the conse-
quences of experimental practices on animals. Through the
objectification process intrinsic to scientific rationality, most
of the time animals are considered as tools or “models,” and
the expected knowledge outcome is considered sufficient
justification for their use. This logic echoes a core dimension
of lay utilitarianism: instrumental harm (Kahane et al., 2017).
In a recent conceptual analysis of proto-utilitarian tendencies
in moral judgment, Kahane et al. (2017) substantiated a fun-
damental distinction between two dimensions: impartial
benevolence, which is a positive dimension of utilitarianism
involving “instructing moral agents to sacrifice their own
well-being” (p. 134), and instrumental harm, defined as “a
willingness to harm and even kill others when this is needed
to achieve a better outcome” (p. 135). As an individual dif-
ference measure, these authors found that instrumental harm
was inversely related to empathic concern and environmen-
tal protection, and positively linked to psychopathy (Kahane
et al., 2017). It also appears that people who endorse instru-
mental harm (estimated with the trolley dilemma; see Foot,
1967) are more inclined to consider animals merely as things,
with no intrinsic value (Bègue & Laine, 2017).
Our hypothesized relationship between instrumental
harm and support for animal experimentation is also derived
from previous studies on ethical ideologies (Forsyth, 1980),
which have shown that people with low moral idealism
(who believe that an action that could harm an innocent
other may nonetheless be justified to gain some outcomes)
also tend to believe that “scientific concern sometimes jus-
tifies potential harm to participants” (Schlenker & Forsyth,
1977; see also Forsyth & Pope, 1984), and tend to be less
concerned about the use of animals in scientific experi-
ments (Su & Martens, 2018; Wuensch & Poteat, 1998).
Finally, the scientific mindset may numb participants
because of the intrinsic effects of statistical reasoning on
their perceptions of victims. For example, in a previous
study, participants who were induced to follow calculation-
based thought (contrasting with a feeling-based mode of
thought) were reportedly less generous toward an identifi-
able victim (Small et al., 2007).
In the present preregistered study (https://osf.io/6tvgz/),
we expected that instrumental harm would mediate the link
between a pro-scientific attitude and support for animal
experimentation.7
Method
Sampling procedure and participants. To define the sample
size of this study, an a priori calculation on R was carried
out, using the “PowerMediation” package and the follow-
ing guidelines (Perugini et al., 2018). To detect a minimum
effect size of r = .10 in each of the three paths of the media-
tion model, the analysis indicated that a total of 1,550
participants would be needed. The sample of the study
comprised 1,598 participants, 50.3% female, aged 20 to 58
years (Mage = 34.8 years, SDage = 9.55 years). All the par-
ticipants were recruited by the Toluna Panel and were
screened for the following eligibility criteria: aged 18 years
or older and French speaking. We imputed means to miss-
ing data (<5% for every variable).
Measures. A 5-point Likert-type survey was proposed for all
the measures described below, with options ranging from 1
(totally disagree) to 5 (totally agree).
Pro-scientific attitude. Pro-scientific attitude was measured
using the five-item scale adapted from Gonzalez-Franco
et al. (2018) which had previously been used in Study 1b
(M = 3.45, SD = 0.89, Cronbach’s α = .88).
Attitude towards animal experimentation. Attitude toward
animal biomedical experiments was measured using the
same two items as in Studies 2a and 2b, to which two supple-
mentary items were added to increase the scale’s reliability:
“I don’t think that there is anything wrong with using animals
in scientific research,” and “Scientists should be allowed to
perform any animal experiment without constraining ethical
rules if this can benefit medicine” (M = 2.41, SD = 1.08,
Cronbach’s α = .86).
Instrumental harm. We used the Instrumental Harm sub-
scale of the Oxford Utilitarianism Scale (Kahane et al.,
2017), which consists of four items measuring willingness to
cause harm to bring about a greater good (e.g., “Sometimes
it is morally necessary for innocent people to die as collateral
damage—if more people are saved overall,” M = 2.41, SD =
1.04, Cronbach’s α = .86).
Results
Main analysis. We analyzed the relationship between a pro-
scientific attitude and support for animal experimentation
and the mediating effect of instrumental harm by carrying
out a multiple regression analysis on component paths using
the JSmediation R package (Yzerbyt et al., 2018). We found
that the pro-scientific attitude represented a significant pre-
dictor of support for animal experimentation (path c), b =
0.25, SE = 0.03, t(1,596) = 8.39, p < .001, and that it was a
significant predictor of instrumental harm (path a), b = 0.18,
SE = 0.03, t(1,596) = 6.17, p < .001. We also observed that
instrumental harm was a significant predictor of support for
animal experimentation (path b), b = 0.50, SE = 0.02,
t(1,595) = 21.93, p < .001. Controlling for instrumental
harm, we found that pro-scientific attitude remained a sig-
nificant predictor of support for animal experimentation, but
more weakly (path c′), b = 0.16, SE = 0.02, t(1,595) = 6.17,
p < .001, thus confirming a mediation (see Figure 7). By
analyzing the range of the mediation using the Monte Carlo
Bègue and Vezirian 13
sampling method (5,000 iterations), an indirect effect was
estimated, b = 0.08, 95% CI = [0.06, 0.11].
Discussion
In this study, we found that a core dimension of lay utilitari-
anism, instrumental harm (Kahane et al., 2017), mediated the
link between a pro-scientific attitude and support for animal
experimentation. This result suggests that a pro-scientific
attitude was related to the endorsement of a utilitarian arith-
metic, which may facilitate the justification of harm to ani-
mals. To a certain extent, the objectification inherent to
scientific research seems to induce an instrumental view of
the object of science itself, even when this involves a sentient
being.
General Discussion
Across five independent samples, we have shown that a pro-
scientific mindset was related to increased attitudinal and
behavioral support for lethal biomedical experiments on ani-
mals. The overall triangulated evidence from our studies is
consistent with the theoretical prediction that the cultural
authority of science shapes an individual’s attitude toward
the biomedical use of animals. First, as shown in Study 1,
when experimentally primed, a pro-scientific attitude under-
pinned the commission of a painful treatment to, and the kill-
ing of, a sentient animal target. This result, based on a
consequential behavioral measure, represents the first direct
demonstration of the relevance of the Engaged Followership
Theory on obedience (Haslam et al., 2014; Reicher &
Haslam, 2011). According to this perspective, which departs
from the agentic or blind obedience interpretation (Milgram,
1974), an individual’s active identification with the scientific
enterprise underlying research represents a critical cause of
the behavior. In our study, individuals who had been primed
toward the positive features of science by writing down three
things that were important about science, indicating what
they liked about science, and stating what they felt they had
in common with scientists, were more willing to follow sci-
entific instructions by inflicting pain and killing what they
believed to be a target animal involved in a pharmacological
experiment. In his own research, Milgram (1974, p. 159)
himself considered that “the idea of science as a legitimate
social enterprise provides the overarching ideological justifi-
cation for the experiment.” However, he did not develop this
important idea further, and his agentic state perspective
departed from the complex notion of legitimacy which he
incidentally referred to. We believe that the Engaged
Followership Theory introduces a more complex and heuris-
tic understanding of everyday obedience, which should be
understood as a dynamic transaction involving individual
goals and situational features rather than as a passive confor-
mity. As in previous studies carried out in the Milgram para-
digm (e.g., Bègue et al., 2015), we also observed that
individual predictors such as SDO or speciesism were rele-
vant factors able to predict behavioral conformity in Study
1a. However, this was not observed in Study 1b. When we
aggregated samples of Studies 1a and 1b, they were related
to the target behavior.8
The observed link between a pro-scientific attitude and
support for animal experimentation was dealt with in depth
in a cross-sectional study showing that participants in medi-
cal fields endorsed a more pro-scientific attitude than those
in paramedical fields, which mediated their support for ani-
mal experimentation. Moreover, using a large and represen-
tative European sample, the specificity and solidity of this
link was shown by controlling for potential demographic and
ideological confounds.
To clarify the psychological process involved, we then
introduced a core dimension of utilitarianism, instrumental
harm. Defined as “a willingness to harm and even kill others
when this is needed to achieve a better outcome” (Kahane
et al., 2017, p. 135), this dimension of lay utilitarianism was
conceived as an intermediary psychological variable con-
necting an individual’s pro-scientific attitude and support for
animal experimentation. We observed that this hypothesized
mediation was supported by our data in a final study. This
result was consistent with the idea that through the scientific
process, animals are transformed into commodities to be
used to obtain data (Lynch, 1988).
In conclusion, by introducing a new obedience paradigm
involving an animal victim, this study refreshes the classical
investigation of obedience and provides a psychological
Figure 7. Path diagram for c. The total effect of pro-scientific attitude on support for animal experiment c′. The indirect effect of
pro-scientific attitude on support for animal experiment through instrumental harm.
14 Personality and Social Psychology Bulletin 00(0)
view on the contemporary dilemma of animal experimenta-
tion. It suggests that the authority assigned by laypeople to
science is a powerful source of legitimacy that may in some
instances overtake their core human moral principles in rela-
tion to how to treat animals.
Acknowledgments
The authors wish to thank Jérôme Maisonnasse, Véronique
Aubergé, and Alain Quercia for their help at various stages of the
research program.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect
to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support
for the research, authorship, and/or publication of this article: The
study was funded in part by University Institute of France (IUF).
ORCID iD
Laurent Bègue https://orcid.org/0000-0001-8078-5802
Supplemental Material
Supplemental material is available online with this article.
Notes
1. Please refer to Supplemental Material 1 regarding the exclusion
criteria.
2. Analyses were also performed which included Code 3 partici-
pants, which did not produce substantially changed results. See
also Supplemental Material 5.
3. Some final analyses differ from the original plan, but the hypoth-
eses remain the same.
4. Analyses were also performed which included Code 3 partici-
pants, which did not produce substantially different results.
5. The same participants also answered a survey 6 months later, but
the data are not presented here.
6. Eurobarometer data are collected according to a very systematic
protocol across countries. No cross-national differences were
hypothesized, and analyses were therefore performed without
standardizing within nations.
7. Although the tested hypothesis remained identical, the analy-
sis presented here departs from the preregistration plan. Our
initial aim was to test a mediation model by manipulating pro-
scientific priming through an online experiment. As predicted,
participants in the pro-scientific prime condition endorsed sig-
nificantly more instrumental harm (p < .05). However, their
support for animal experimentation was not significantly higher,
whereas manipulation check items still suggested that the induc-
tion was efficient (see, however, Sigall & Mills, 1998, regard-
ing the intrinsic limitations of manipulation check items). We
believe that these unexpected results were possibly due to the
online procedure, which may have impacted the quality of the
experimental induction, as the data may be less consistent in
online experiments than in lab-based experiments for the pro-
duction task we relied on (see Finley & Penningroth, 2015). We
therefore tested our mediation effect using the five manipula-
tion checks items as our independent variable, which was then
invoked and not provoked.
8. Following cumulative science recommendations (see
Asendorpf et al., 2013; Funder et al., 2014), we aggregated
the two datasets from our experimental studies (which was
made possible as they used the same protocol) to assess the
relationships between SDO, speciesism, empathy, vegetari-
anism, age, gender, and the injection of the toxic substance.
As in Studies 1a and 1b, only participants coded “1” and “2”
by the judges were included, amounting to a total of 258
participants. A logistic regression analysis was performed
following an identical procedure to Studies 1a and 1b. The
following variables were related to toxic substance adminis-
tration: SDO (odds ratio [OR] = 3.04, 95% confidence inter-
val [CI] = [1.45, 6.37], p = .003), speciesism (OR = 2.40,
95% CI = [1.36, 4.22], p = .002), and vegetarianism (OR
= 0.21, 95% CI = [0.09, 0.50], p < .001). Age was also
inversely related to toxic substance administration (OR =
0.96, 95% CI = [0.94, 0.99], p = .01), whereas gender and
empathy remained unrelated (OR = 0.50, 95% CI = [0.22,
1.12], p = .09 and OR = 1.07, 95% CI = [0.56, 2.03], p =
.82). The overall model accounted for 34% of the variance
(Nagelkerke’s pseudo R2).
References
Altemeyer, B. (1998). The other “authoritarian personality.” In M.
P. Zanna (Ed.), Advances in experimental social psychology
(Vol. 30, pp. 47–92). Academic Press.
Amiot, C. E., & Bastian, B. (2015). Toward a psychology of
human–animal relations. Psychological Bulletin, 141(1), 6–47.
Annas, J., & Grondin, M. A. (1992). The Nazi doctors and the
Nuremberg code: Human rights in human experimentation.
Oxford University Press.
Asendorpf, J. B., Conner, M., de Fruyt, F., de Houwer, J., Denissen,
J. J. A., Fiedler, K., Fiedler, S., Funder, D. C., Kliegl, R., Nosek,
B. A., Perugini, M., Roberts, B. W., Schmitt, M., van Aken, M.
A. G., Weber, H., & Wicherts, J. M. (2013). Recommendations
for increasing replicability in psychology. European Journal of
Personality, 27, 108–119.
Bandura, A. (1999). Moral disengagement in the perpetration of
inhumanities. Personality and Social Psychology Review, 3(3),
193–209.
Baron, R. M., & Kenny, D. A. (1986). The moderator–mediator vari-
able distinction in social psychological research: Conceptual,
strategic, and statistical considerations. Journal of Personality
and Social Psychology, 51(6), 1173–1182.
Bauer, M. W., Pansegrau, P., & Shulka, R. (2019). Image, per-
ception and cultural identity of science. In M.W. Bauer, P.
Pansegrau, & R. Shukla (Eds.), The cultural authority of
science. Comparing across Europe, Asia, Africa, and the
Americas (pp. 3–21). Routledge.
Bègue, L., Beauvois, J. L., Courbet, D., Oberlé, D., Lepage, J., &
Duke, A. (2015). Personality predicts obedience in a Milgram
paradigm. Journal of Personality, 83, 299–312.
Bègue, L., & Laine, P. J. (2017). Moral utilitarianism and attitudes
toward animals. Ethics and Behavior, 27(3), 173–178.
Birke, L. I., Birke, L., Arluke, A., & Michael, M. (2007). The sacri-
fice: How scientific experiments transform animals and people.
Purdue University Press.
Bègue and Vezirian 15
Broida, J., Tingley, L., Kimball, R., & Miele, J. (1993). Personality
differences between pro and antivivisectionists. Society and
Animals, 1, 129–144.
Caviola, L., Everett, J. A. C., & Faber, N. S. (2019). The moral
standing of animals: Towards a psychology of speciesism.
Journal of Personality and Social Psychology, 116(6),
1011–1029.
Crettaz von Rotten, F. (2008). Mapping perceptions of animal
experimentation: Trend and explanatory factors. Social Science
Quarterly, 89(2), 537–549.
Davis, M. H. (1983). Measuring individual differences in empa-
thy: Evidence for a multidimensional approach. Journal of
Personality and Social Psychology, 44(1), 113–126.
de Waal, F. (2016). Are we smart enough to know how smart ani-
mals are? W.W. Norton.
Dhont, K., Hodson, G., Leite, A. C., & Salmen, A. (2020). The psy-
chology of speciesism. In K. Dhont & G. Hodson (Eds.), Why
we love and exploit animals: Bridging insights from academia
and advocacy (pp. 29–49). Routledge.
Faul, F., Erdfelder, E., Lang, A. G., & Buchner, A. (2007). G*Power
3: A flexible statistical power analysis program for the social,
behavioral, and biomedical sciences. Behavior Research
Methods, 39(2), 175–191.
Filippi, M., Riccitelli, G., Falini, A., Di Salle, F., Vuilleumier, P.,
Comi, G., & Rocca, M. A. (2010). The brain functional net-
works associated to human and animal suffering differ among
omnivores, vegetarians and vegans. PLOS ONE, 5(5), Article
e10847.
Finley, A., & Penningroth, S. (2015). Online versus in-lab: Pros
and cons of an online prospective memory experiment. In
A. M. Columbus (Ed.), Advances in psychology research
(pp. 135–161). Nova Science.
Foot, P. (1967). The problem of abortion and the doctrine of double
effect. Oxford Review, 5, 5–15.
Forsyth, D. R. (1980). A taxonomy of ethical ideologies. Journal of
Personality and Social Psychology, 39(1), 175–184.
Forsyth, D. R., & Pope, W. R. (1984). Ethical ideology and judg-
ments of social psychological research: Multidimensional
analysis. Journal of Personality and Social Psychology, 46(6),
1365–1375.
Funder, D. C., Levine, J. M., Mackie, D. M., Morf, C. C., Sansone,
C., Vazire, S., & West, S. G. (2014). Improving the depend-
ability of research in personality and social psychology:
Recommendations for research and educational practice.
Personality and Social Psychology Review, 18(1), 3–12.
Furnham, A., & Heyes, C. (1993). Psychology students’ beliefs
about animals and animal experimentation. Personality and
Individual Differences, 15(1), 1–10.
Gibson, S. (2019). Arguing, obeying and defying: A rhetorical
perspective on Stanley Milgram’s obedience experiments.
Cambridge University Press.
Gonzalez-Franco, M., Slater, M., Birney, M. E., Swapp, D.,
Haslam, S. A., & Reicher, S. D. (2018). Participant concerns
for the Learner in a Virtual Reality replication of the Milgram
obedience study. PLOS ONE, 13(12), Article e0209704.
Gorsuch, R. L., & McPherson, S. E. (1989). Intrinsic/Extrinsic
measurement: I/E-Revised and single-item scales. Journal for
the Scientific Study of Religion, 28, 348–354.
Graham, J., Haidt, J., Koleva, S., Motyl, M., Iyer, R., Wojcik, S., &
Ditto, P. H. (2013). Moral foundations theory: The pragmatic
validity of moral pluralism. Advances in Experimental Social
Psychology, 47, 55–130.
Guimond, S., & Palmer, D. L. (1990). Type of academic training
and causal attributions for social problems. European Journal
of Social Psychology, 20, 61–75.
Hagelin, J., Hau, J., & Carlsson, H. E. (2003). The refining influ-
ence of ethics committees on animal experimentation in
Sweden. Laboratory Animals, 37(1), 10–18.
Haslam, S. A., Oakes, P. J., Reynolds, K. J., & Turner, J. C. (1999).
Social identity salience and the emergence of stereotype con-
sensus. Personality and Social Psychology Bulletin, 25(7),
809–818.
Haslam, S. A., & Reicher, S. D. (2017). 50 years of “obedience
to authority”: From blind obedience to engaged followership.
Annual Review of Law and Social Science, 13, 59–78.
Haslam, S. A., Reicher, S. D., & Birney, M. E. (2014). Nothing
by mere authority: Evidence that in an experimental analogue
of the Milgram paradigm participants are motivated not by
orders but by appeals to science. Journal of Social Issues, 70,
473–488.
Haslam, S. A., Reicher, S. D., Millard, K., & McDonald, R.
(2015). “Happy to have been of service”: The Yale archive
as a window into the engaged followership of participants in
Milgram’s “obedience” experiments. British Journal of Social
Psychology, 54(1), 55–83.
Hendricks, M. J. (2019). Toward the development of a Terrorist
Threat Scale. Peace and Conflict: Journal of Peace Psychology,
25(2), 148–151.
Ho, A. K., Sidanius, J., Kteily, N., Sheehy-Skeffington, J., Pratto,
F., Henkel, K. E., Foels, R., & Stewart, A. L. (2015). The
nature of social dominance orientation: Theorizing and mea-
suring preferences for intergroup inequality using the new
SDO₇ scale. Journal of Personality and Social Psychology,
109(6), 1003–1028.
Hoffarth, M. R., Azevedo, F., & Jost, J. T. (2019). Political con-
servatism and the exploitation of non-human animals: An
application of System Justification Theory. Group Processes
& Intergroup Relations, 22, 858–878.
Kahane, G., Everett, J. A., Earp, B. D., Caviola, L., Faber, N. S.,
Crockett, M. J., & Savulescu, J. (2017). Beyond sacrificial
harm: A two-dimensional model of utilitarian psychology.
Psychological Review, 125, 131–164.
Knight, A. (2011). The costs and benefits of animal experimenta-
tion. Palgrave Mc Millan.
Lakens, D. (2021, January 4). Sample size justification. https://doi.
org/10.31234/osf.io/9d3yf
Lynch, M. (1988). Sacrifice and the transformation of the animal
body into a scientific object: Laboratory culture and ritual prac-
tice in the neurosciences. Social Studies of Science, 18, 265–289.
Milgram, S. (1974). Obedience to Authority: An experimental view.
Harper & Row.
Miller, A. G. (1986). The obedience experiments: A case study of
controversy in social sciences. Praeger.
Miralles, A., Raymond, M., & Lecointre, G. (2019). Empathy and
compassion toward other species decrease with evolutionary
divergence time. Scientific Reports, 9, 19555.
16 Personality and Social Psychology Bulletin 00(0)
Monamy, V. (2017). Animal experimentation. A guide to the issue
(3rd ed.). Cambridge.
Ormandy, E. H., & Schuppli, C. A. (2014). Public attitudes toward
animal research: A review. Animals, 4, 391–408.
Perugini, M., Gallucci, M., & Costantini, G. (2018). A practical
primer to power analysis for simple experimental designs.
International Review of Social Psychology, 31(1), 20. http://
doi.org/10.5334/irsp.181
Plous, S. (1993). Psychological mechanisms in the human use of
animals. Journal of Social Issues, 49, 11–52.
Reicher, S. D., & Haslam, S. A. (2011). After shock? Towards a
social identity explanation of the Milgram “obedience” studies.
British Journal of Social Psychology, 50, 163–169.
Reicher, S. D., Haslam, S. A., & Miller, A. G. (2014). What makes a
person a perpetrator? The intellectual, moral, and methodologi-
cal arguments for revisiting Milgram’s research on the influ-
ence of authority. Journal of Social Issues, 70(3), 393–408.
Reicher, S. D., Haslam, S. A., & Smith, J. (2012). Working
toward the experimenter: Reconceptualizing obedience within
the Milgram paradigm as identification-based followership.
Perspectives on Psychological Science, 7, 315–324.
Rutjens, B. T., Heine, S. J., Sutton, R. M., & van Harreveld,
F. (2018). Attitudes towards science. In J. M. Olson (Ed.),
Advances in experimental social psychology: Volume 57
(pp. 125–165). Elsevier Academic Press.
Schlenker, B., & Forsyth, D. (1977). On the ethics of psychologi-
cal research. Journal of Experimental Social Psychology, 13,
369–396.
Sidanius, J., & Pratto, F. (1999). Social dominance: An inter-
group theory of social hierarchy and oppression. Cambridge
University Press.
Sigall, H., & Mills, J. (1998). Measures of independent variables
and mediators are useful in social psychology experiments: But
are they necessary? Personality and Social Psychology Review,
2, 218–226.
Small, D. A., Loewenstein, G., & Slovic, P. (2007). Sympathy and
callousness: The impact of deliberative thought on donations
to identifiable and statistical victims. Organizational Behavior
and Human Decision Processes, 102, 2143–2153.
Su, B., & Martens, P. (2018). How ethical ideologies related to pub-
lic attitudes toward animals: The Dutch case. Anthrozöos, 31,
2179–2194.
Taylor, K., Gordon, N., Langley, G., & Higgins, W. (2008).
Estimates for worldwide laboratory animal use in 2005.
Alternatives to Laboratory Animals, 36(3), 327–342.
Westbury, H., & Neumann, D. L. (2008). Empathy-related
responses to moving film stimuli depicting human and non-
human animal targets in negative circumstances. Biological
Psychology, 78(1), 66–74.
Wuensch, K. L., & Poteat, G. M. (1998). Evaluating the morality of
animal research: Effects of ethical ideology, gender, and pur-
pose. Journal of Social Behavior and Personality, 13, 139–150.
Yzerbyt, V., Muller, D., Batailler, C., & Judd, C. M. (2018). New
recommendations for testing indirect effects in mediational
models: The need to report and test component paths. Journal
of Personality and Social Psychology, 115(6), 929–943.
View publication stats