Relative Contribution of Odour Intensity and Valence to Moral Decisions

Abstract

Meta-analytic evidence showed that the chemical senses affect moral decisions. However, how odours impact on morality is currently unclear. Through a set of three studies, we assess whether and how odour intensity biases moral choices (Study 1a), its psychophysiological responses (Study 1b), as well as the behavioural and psychophysiological effects of odour valence on moral choices (Study 2). Study 1a suggests that the presence of an odour plays a role in shaping moral choice. Study 1b reveals that of two iso-pleasant versions of the same neutral odour, only the one presented sub-threshold (vs. supra-threshold) favours deontological moral choices, those based on the principle of not harming others even when such harm provides benefits. As expected, this odour intensity effect is tracked by skin conductance responses, whereas no difference in cardiac activity - proxy for the valence dimension - is revealed. Study 2 suggests that the same neutral odour presented sub-threshold increases deontological choices even when compared to iso-intense ambiguous odour, perceived as pleasant or unpleasant by half of the participants, respectively. Skin conductance responses, as expected, track odour pleasantness, but cardiac activity fails to do so. Results are discussed in the context of mechanisms alternative to disgust induction underlying moral choices.

Full text

Chemosensory Perception
Relative Contribution
of Odour Intensity and
Valence to Moral Decisions
Cinzia Cecchetto
SISSA – International School for Advanced Studies, Trieste, Italy
Raffaella Ida Rumiati
SISSA – International School for Advanced Studies, Trieste, Italy; ANVUR
– Agenzia Nazionale della Valutazione del sistema Universitario e della
Ricerca, Roma, Italy
Valentina Parma
SISSA – International School for Advanced Studies, Trieste, Italy;
Department of Clinical Neuroscience, Karolinska Institutet, Stockholm,
Sweden; Center for Autism Research, Philadelphia, PA, USA
Abstract
Meta-analytic evidence showed that the chemical senses affect moral decisions. However, how
odours impact on morality is currently unclear. Through a set of three studies, we assess whether
and how odour intensity biases moral choices (Study 1a), its psychophysiological responses (Study
1b), as well as the behavioural and psychophysiological effects of odour valence on moral choices
(Study 2). Study 1a suggests that the presence of an odour plays a role in shaping moral choice.
Study 1b reveals that of two iso-pleasant versions of the same neutral odour, only the one
presented sub-threshold (vs. supra-threshold) favours deontological moral choices, those based
on the principle of not harming others even when such harm provides benefits. As expected, this
odour intensity effect is tracked by skin conductance responses, whereas no difference in cardiac
activity – proxy for the valence dimension – is revealed. Study 2 suggests that the same neutral
odour presented sub-threshold increases deontological choices even when compared to iso-
intense ambiguous odour, perceived as pleasant or unpleasant by half of the participants,
respectively. Skin conductance responses, as expected, track odour pleasantness, but cardiac
activity fails to do so. Results are discussed in the context of mechanisms alternative to disgust
induction underlying moral choices.
Keywords
Odour intensity, odour valence, moral decision-making, psychophysiology
Corresponding author:
Cinzia Cecchetto, SISSA – International School for Advanced Studies, Neuroscience Area, via Bonomea, 265 Trieste, Italy.
Email: ccecchet@sissa.it
Perception
2017, 0(0) 1–28
!The Author(s) 2017
Reprints and permissions:
sagepub.co.uk/journalsPermissions.nav
DOI: 10.1177/0301006616689279
journals.sagepub.com/home/pec

Introduction
The chemical senses have proven to be ideal model systems to disentangle the role of arousal
and valence (Anderson et al., 2003; Winston, Gottfried, Kilner, & Dolan, 2005), the two core
dimensions of affective experiences (e.g., Lang, 1995). Indeed, odours (as well as flavours) can
be described based on their intensity (a proxy for arousal; Bensafi et al., 2002) and – perhaps
more often – based on their hedonic value (or valence; Yeshurun & Sobel, 2010). Beyond the
modulation of subjective experiences, odour intensity and valence induce changes in a variety
of physiological and behavioural processes. For instance, Bensafi et al. (2002) report that skin
conductance variations track odour intensity, heart rate variations track odour pleasantness
and the two indices highly correlate. Furthermore, odour intensity judgements affect cross-
modal perception (i.e., color lightness, Kemp & Gilbert, 1997; sweetness, Stevenson, Boakes,
& Prescott, 1998), whereas odour valence is able to promote mood changes (Chen &
Haviland-Jones, 1999).
The study of the effects of odour intensity and valence has mostly focused on perceptual
issues, and only recently it has moved towards the impact that such olfactory dimensions have
on higher cognitive functions (Mainland, Lundstro
¨m, Reisert, & Lowe, 2014; Yeshurun &
Sobel, 2010). Few attempts have been recently produced in the context of decision-making
processes. In real life situations, intensity-controlled, pleasant ambient odours promote
positive emotions in consumers and bias their decision-making processes towards spending
(Chebat & Michon, 2003). In the lab, Bonini et al. (2011) revealed that male participants
exposed to a disgusting odour increased cooperation during the Ultimatum Game, suggesting
that the disgust felt for poor offers was misattributed to the odour, rather than to the offer
itself. As the meta-analysis by Landy and Goodwin (2015) suggests, the most interesting
attempts at evaluating chemosensory effects on decision processes refer to the moral
domain. Their analysis showed that olfactory (and gustatory) induction of disgust amplifies
its effects on moral behaviour more that the induction of disgust prompted through videos or
images. However, the number of studies inducing disgust through chemosensory stimuli is
disproportionately smaller (3 published studies) than the number of studies including visual
induction of disgust (12 published studies). Despite this asymmetry, the chemosensory
induction of disgust (gustatory and olfactory) produced a stronger amplification effect
(Cohen’s d¼.37) on moral choice than visual induction (Cohen’s d¼.13) or than the total
amplification effect irrespective of the modality (Cohen’s d¼.11). With specific reference to
the olfactory modulation of moral behaviours, Inbar, Pizarro, and Bloom (2012) reported
that a disgusting odour (vs. no odour) lead participants to evaluate gay men more negatively,
whereas both Schnall, Haidt, Clore, and Jordan (2008) and Ugazio, Lamm, and Singer (2012)
used commercially available ‘‘fart spray’’ to prime disgust in participants while they rated
four vignettes requiring either a moral judgment (Schnall et al., 2008) or while they judged the
permissibility of some moral scenarios (Ugazio et al., 2012). In Schnall et al. (2008), authors
assigned participants to one of three olfactory conditions (no vs. four vs. eight sprays
applied). After they completed the task, they were asked to rate how disgusted they felt
and whether they were consciously aware of the presence of the unpleasant odour.
Participants were significantly more disgusted in the strong odour condition than in the
other two conditions, and the more participants experienced olfactory-triggered disgust,
the more severely they judged the proposed vignettes (Schnall et al., 2008). Ugazio et al.
(2012) used the same procedure as in Schnall et al. (2008), but with less intense odour
stimulations (only two sprays applied), which led to the disappearance of disgust-attributed
severity in the moral judgements.
Taking a chemosensory standpoint, some considerations are in order. First, the
comparison of Schnall et al. (2008) results with Ugazio et al. (2012) suggests an interesting
2Perception 0(0)

possible modulation of moral behaviour based on odour intensity, aspect linked to the
awareness of the odour (Smeets, Schifferstein, Boelema, & Lensvelt-Mulders, 2008).
Indeed, Schnall et al. (2008) showed significant modulation of moral behaviour when
participants were exposed to an odour that was clearly perceivable (especially in the
strongest condition – eight sprays), whereas Ugazio et al. (2012) failed to find such effect
with lower intensities of the odour (two sprays). Second, the use of disgusting odours based
on the emotional congruency with moral disgust promoted the testing of negative odours
only (e.g., Schnall et al., 2008; Ugazio et al., 2012). As revealed by Bensafi et al. (2002), the
interdependent nature of odour intensity and odour pleasantness is reflected at the level of
judgments and physiological correlates. Indeed, greater concentrations of one odour,
irrespective of their initial pleasantness judgment, are perceived as less pleasant and are
associated with increased peripheral sympathetic and reduced cardiac parasympathetic
activity. Altogether this evidence calls for an extension of the exploration of the olfactory
effects on moral behaviours that goes beyond negative odours. Moreover, studies that
encompass neutral and pleasant odours are needed in order to disentangle the role played
by valence, especially in the context of decisions modulated by multimodal stimulations.
Third, the evaluation of the odour stimuli, their delivery (e.g., Lundstro
¨m, Gordon, Alden,
Boesveldt, & Albrecht, 2010) and the inclusion/exclusion criteria of participants (Boesveldt
et al., 2011), an aspect generally neglected in the moral literature, should be promoted as to
reduce potential confounding effects.
In order to shed light on the effects that arousal and valence of olfactory stimuli might
have on moral behaviours, we designed three studies in which normosmic participants were
asked to perform a moral decision-making task in the presence of different odour conditions.
As in typical moral decision-making tasks, we present moral dilemmas, hypothetical short
stories that offer two morally conflicting alternatives among which a decision maker is
expected to choose (see also Cecchetto, Rumiati, & Parma, submitted). According to the
most influential theory (the dual process model; e.g., Greene, 2009), moral choices are driven
by the interaction between two competing processing systems, mediated by partially
dissociable neural networks. First, a fast, automatic emotional system is expected to play a
central role in the expression of deontological choices, based on the idea that an individual’s
moral principle should not be infringed, even when the welfare of a greater number of people
is at stake. However, such intuitive moral choice can be overridden by cognitive (executive)
processes which favour, instead, the expression of utilitarian responses grounded on a cost–
benefit analysis tailored to reduce the overall harm across the available options. In other
words, utilitarian choices require more cognitive resources (Greene, Morelli, Lowenberg,
Nystrom, & Cohen, 2008; Van Dillen, van der Wal, & van den Bos, 2012) to contrast
deontological responses, yet the difficulties emerging in the process are associated with
greater negative emotional reactions linked to the sense of responsibility for an immoral
act (e.g., killing a person; Moretto, Ladavas, Mattioli, & Di Pellegrino, 2009).
Additionally, the modulation of the emotional reactions triggered by moral dilemmas can
be associated to the features of the moral dilemmas. The most studied feature refers to the
degree of personal involvement in the immoral act. The popular Trolley dilemma (Foot,
1967) is an example of impersonal dilemmas. In this dilemma, a runaway trolley is about
to run over and kill five people. To save the group of five, the decision maker can hit a switch
that will turn the trolley onto a side track. This act will cause the death of the person standing
on the track. The Footbridge dilemma (Thomson, 1976), instead, is an example of personal
dilemma, in which participants can save the five by pushing a big man off an overpass onto
the track below: The man will die but the five people will survive. In virtue of their stronger
emotional content, personal dilemmas tend to elicit more deontological choices than
Cecchetto et al. 3

impersonal dilemmas (Greene et al., 2009; Greene et al., 2008; Greene, Nystrom, Engell,
Darley, & Cohen, 2004; Koenigs et al., 2007).
In light of the scarcity of evidence on the effects produced by olfactory stimuli on moral
choices (Inbar et al., 2012; Schnall et al., 2008; Ugazio et al., 2012), we set out to test a series
of hypotheses on whether and how odour intensity and odour valence modulate moral
choices.
In Study 1a and Study 1b, we tested the role of odour intensity across iso-pleasant stimuli
in impacting moral choice. Specifically, we assessed whether the supra-threshold vs. sub-
threshold versions of a neutral odour will comparably modulate moral choice tendencies,
with respect to a no-odour condition. Three hypotheses have been proposed. First, a neutral
odour condition, irrespective of its intensity, might increase deontological responses due to
the activation of limbic areas through olfaction (Zald & Pardo, 1997). Indeed, the olfactory
and limbic systems share common substrates (e.g., amygdala, insula, anterior cyngulate
cortex and orbitofrontal cortex; Soudry, Lemogne, Malinvaud, Consoli, & Bonfils, 2011)
and, in particular, a previous study has showed that odor intensity is associated with
amygdala activation (Anderson et al., 2003). Second, if only the supra-threshold odour
condition produces an increase of deontological responses, compared to the sub-threshold
odour, we could argue that an olfactory stimulus needs to be presented in high concentration
to affect moral behaviour. The pleasantness evaluation should allow us to ascertain whether
or not such higher concentration is responsible for inducing possible unpleasantness/
disgusting reactions. Third, we hypothesize that the eventual selective increases of the
number of deontological responses in sub-threshold condition is caused by insufficient
sensory inputs that, as such, do not trigger the strategic control (Li et al., 2007). In other
words, lower concentrations of an olfactory stimulus, possibly by the effect of lack of
awareness, may be key to the emergence of deontological moral choices.
In Study 1b, we tried to conceptually replicate the above study in an independent group of
participants and, additionally, we collected psychophysiological responses to implicitly assess
the intensity and valence effects of olfactory stimuli during moral decisions. We foresee that,
beyond the arousal induced by the processing of the moral dilemma, odour intensity will be
tracked by skin conductance (SCR) measurements, showing an increased activity in the
supra-threshold odour condition compared to sub-threshold and no odour conditions,
respectively. This effect should be evident for both utilitarian (more arousing) and
deontological choices (less arousing, Moretto et al., 2009). Given that odour valence is
expected to be equivalent between sub- and supra-threshold odour conditions, we do not
foresee modulations at the level of instantaneous heart rate (IHR; Palomba, Sarlo, Angrilli,
Mini, & Stegagno, 2000), an implicit index of valence (Bradley, Codispoti, Cuthbert, & Lang,
2001; Bradley, Greenwald, & Hamm, 1993; Palomba, Angrilli, & Mini, 1997). In line with
previous literature on pleasant and unpleasant stimuli (Palomba et al., 2000), we foresee a
greater deceleration of IHR for deontological rather than utilitarian choices.
In Study 2, we will test the role of valence in iso-intense olfactory stimuli. We will compare
the modulatory effect of neutral, pleasant and unpleasant odours on moral tendencies. Since
it has been shown that an ambiguous odour can be perceived as either pleasant or unpleasant
(de Araujo, Rolls, Velazco, Margot, & Cayeux, 2005), we presented a group of participants
with the same ambiguous odour (butyric acid) and we then split them in two groups, based
on their reported pleasantness evaluation. Following previous studies on morality and
induced affective states (Pasto
¨tter, Gleixner, Neuhauser, & Ba
¨uml, 2013; Schnall et al.,
2008; Valdesolo & DeSteno, 2006; Youssef et al., 2012b), we outline three alternative
hypotheses. The first hypothesis posits that odour valence is not critical to moral choice. If
so, the presence of an odour, irrespective of its valence, may increase deontological choices as
4Perception 0(0)

compared to the no odour condition. The latter two hypotheses suggest that odour valence
plays a role in moral choice. In one case, we hypothesize a generalized effect of valence (as in
Seubert et al., 2010), emerging as an increase of deontological responses for the pleasant and
unpleasant odour, as compared to the neutral and no odour conditions. Alternatively, if
disgust is critical to moral decisions, as suggested elsewhere (e.g., Chapman, Kim,
Susskind, & Anderson, 2009; Rozin, Haidt, & Fincher, 2009), we expect an inverse
relationship between odour pleasantness and the increment in deontological choices (e.g.,
the less pleasant the odour, the more deontological choices). At the psychophysiological level,
we expect SCR to track the effects of odour valence (Bradley, 2009) and also to reflect
decreased arousal in association with deontological moral choices. With respect to the
IHR, we foresee the unpleasant odour to produce the maximal decrease in IHR, as
compared to the neutral and pleasant odour, respectively.
Finally, previous literature has shown that moral choices are modulated also by individual
variability in emotion recognition (Koven, 2011; Patil & Silani, 2014a, 2014b), sensitivity to
disgust (Choe & Min, 2011; Schnall et al., 2008; Ugazio et al., 2012), autistic traits (Buon
et al., 2013; Moran et al., 2011) and anxiety (Starcke, Wolf, Markowitsch, & Brand, 2008;
Youssef et al., 2012a). To control for this variability, we will ask participants to complete
questionnaires that measure these individual traits. Moreover, since our hypothesis is that the
attention that people direct to odours could modulate the effects that odours might have on
moral choices, our participants also performed the questionnaire on the awarness of odour in
the environment (Smeets et al., 2008).
Study 1a – Does Odour Intensity Affect Moral Choice?
Material and Methods
Participants. Seventeen participants were included in Study 1a, based on their eligibility
according to the following exclusion criteria: history of neurological or psychiatric
disorders, being an active smoker, using psychopharmacological substances or other
systemic medications, having experienced a head trauma leading to unconsciousness, score
less than 10 at the Sniffin Sticks Identification test (Hummel, Sekinger, Wolf, Pauli, & Kobal,
1997). Please refer to Table 1 for details on the participants’ characteristics.
The local University Ethics Committee approved all studies, which were in accordance
with the Declaration of Helsinki, and informed written consent was obtained from each
participant.
Self-report questionnaires. The following paper and pencil questionnaires were administered to
assess interpersonal characteristics that have shown to affect moral decisions: Bermond–
Vorst Alexithymia Questionnaire, form B (BVAQ-B; Bermond & Oosterveld, 1994) was
used to assess Alexithymia; Disgust Scale (DS; Rozin, Haidt, McCauley, Dunlop, &
Ashmore, 1999) was used as measure of the individual differences in sensitivity to disgust;
the Trait scale of the State Trait Inventory for Cognitive and Somatic Anxiety (STICSA; Ree,
French, MacLeod, & Locke, 2008) was used to assess anxiety at trait level; the Autism-
Spectrum Quotient (AQ; Baron-Cohen, Wheelwright, Skinner, Martin, & Clubley, 2001)
was used to measure the degree of presence of traits associated with the autism spectrum;
the Odour Awareness Scale (OAS; Smeets et al., 2008) was used to measure individual
differences in awareness of odours in the environment; the State subscale of the State–
Trait Anxiety Inventory (STAI; Spielberger, Gorsuch, & Lushene, 1968). STAI State was
performed on the computer both before and after the moral decision-making task, and a
differential score STAI State post–pre was created and introduced in the analyses.
Cecchetto et al. 5

Odour stimuli. Three odour conditions were used in the task of both studies: no odour, sub-
threshold neutral odour and supra-threshold neutral odour (Cedarwood oil; Sigma-Aldrich,
Italy). This odour was selected on the basis of on a pilot Study (N¼53) showing that its
100% concentration was perceived as a neutral odour (valence rating, M¼5.0; SD ¼2.76 on
a scale ranging from 1 to 10). Detection thresholds were determined using a two-alternative
forced-choice ascending staircase paradigm with seven reversals and no feedback (Doty,
1991; Lundstro
¨m, Boyle, & Jones-Gotman, 2008; Lundstro
¨m, McClintock, & Olsson,
2006). Each trial included one target (a bottle with the diluted odour) and one control
stimulus (a bottle with the diluent only). The odour was presented in ascending
concentrations until the participant discerned correctly the odour in two successive trials,
which triggered a reversal. Mineral oil (Sigma-Aldrich, Italy) was used as diluting agent to
create all the concentrations of the odour threshold test necessary to determine the sub- and
supra-threshold stimulus for each participant. The dilution series was prepared starting from
100% volume to volume (v/v) in liquid phase of cedarwood oil. From there, the odour was
diluted in 16 consecutive dilution steps using a 0.5 volume dilution series (end concentration
0.00305% v/v). Two series of the same odour were prepared and used to allow the odour to
saturate the headspace between potential repetitions of the same dilution step. Each dilution
step and the matched diluent only were delivered using amber 2-oz glass bottles, all visually
identical and containing 10 mL of liquid each. Detection threshold was defined as the
geometric mean of the last four reversals. One step below and above the detection
threshold was considered as odour concentrations for sub- and supra-threshold conditions,
respectively.
Moral dilemmas. Visual stimuli for both studies were 48 moral dilemmas, selected from the
4CONFiDe moral set (Cecchetto et al., submitted). Dilemmas have been revised following
the guidelines proposed by Christensen, Flexas, Calabrese, Gut, & Gomila (2014) and Lotto,
Manfrinati, and Sarlo (2014; e.g., word count, type of moral transgression, the decision
maker’s perspective and the type of question). Each dilemma is characterized by four
Table 1. Summary Table of Demographic Characteristics and Questionnaires of Study 1a and 1b.
Study 1a Study 1b
Mean (SD) Range Mean (SD) Range
No. 17 15
Gender 9 females – 11 females –
Age 23.23 (3.25) 18–32 23.13 (1.62) 20–26
Education 16.18 (1.28) 11–18 16.47 (1.55) 13–18
Threshold test 12.53 (1.69) 8.5–15.5 7.90 (0.90) 6–9.25
Identification test 12.23 (1.52) 10–14 12.87 (1.26) 10–15
BVAQ 42.59 (7.04) 28–53 42.13 (5.09) 34–49
DS 16.17 (5.43) 7.5–25 16.03 (3.38) 10–21.5
STICSA (Trait) 34.18 (6.64) 25–49 36.71 (7.40) 24–50
AQ 17.94 (5.37) 10–30 18.50 (4.45) 11–27
OAS 122.82 (12.75) 101–142 117.07 (17.29) 80–138
STAI (State – PRE) 40.18 (4.36) 27–45 34.33 (7.58) 21–51
STAI (State – POST) 41.00 (4.73) 33–47 36.20 (5.49) 27–47
Note. BVAQ¼Bermond–Vorst Alexithymia Questionnaire; DS ¼Disgust Scale; STICSA ¼State Trait Inventory for
Cognitive and Somatic Anxiety; AQ ¼Autism-Spectrum Quotient; OAS ¼Odour Awareness Scale; STAI (State) ¼State–
Trait Anxiety Inventory; POMS ¼Profile of Mood States.
6Perception 0(0)

conceptual factors: (a) Personal force: defines whether the agent is directly (personal) or
indirectly (impersonal) involved in the production of the harm; (b) Intentionality: defines
whether the harm is produced as a non-desired side effect of the action (accidental), or it is
used as means to save the others (instrumental); (c) Benefit recipient: defines whether the
harm involves the respondent (self-beneficial) or not (other-beneficial); and (d) Evitability:
defines whether the harm could be avoided (avoidable) or not (inevitable). Each dilemma was
presented on two subsequent screens. The first screen described the scenario, in which a
danger threatens to kill a group of people (for instance, ‘‘You are an engineer on the
international space station ISS. A fire breaks out in the cargo bay. The automatic fire
safety system would open the outer door of the cargo bay, letting the oxygen out and
putting out the fire. It only works when the inner portal is sealed, but one mechanic is still
in the cargo bay. He doesn’t have the time to take off the bulky space suit, which will get him
stuck in the inner portal, causing the fire to spread and to kill you all.’’), plus a hypothetical
action that would save these people but harm others (‘‘If you manually close the portal the
emergency system will be activated and it will put out the fire. This will suck the mechanic
into space and you kill him, but you will save yourself and the other ten astronauts.’’). The
second screen presented the question (‘‘Do you put out the fire by manually sealing the inner
portal, which will suck the mechanic into space, so the fire won’t reach you and the ten
astronauts?’’). Participants had to choose between two options: ‘‘I do it,’’ and ‘‘I do not do
it.’’ The first choice is considered to be utilitarian, as it maximises overall utility (i.e., saving
more lives), whereas the second choice was considered deontological. Moral dilemmas were
designed in such a way that the affirmative choice was always the utilitarian option (as in
Christensen et al., 2014). The moral tendency of each participant is based on the relative
frequency of these reciprocal choices. Dilemmas were presented using black font color (font:
Calibri, size: 24) against a white background on a 19-in. computer screen at a viewing
distance of 60 cm. Stimulus presentation was delivered with E-prime 2.0 software
(Psychology Software Tools, Pittsburgh, PA).
Procedure. Upon arrival, each participant was seated in a quiet room and asked to complete
the self-report questionnaires, the odour identification test (Hummel et al., 1997) and the
odour detection threshold test. Afterwards, participants were asked to seat in front of a
computer screen and they performed a practice session (two moral dilemmas) and then the
moral decision-making task. Participants were instructed using an Italian version of the
instructions suggested by Christensen et al. (2014): ‘‘In the following test you will read a
series of short stories about difficult interpersonal situations, similar to those that you could
see on the news every day or may watch in a movie. For each of the difficult situations a
solution will be proposed. You have to decide whether or not you would act as suggested. Do
not linger too much in thinking but try to identify yourself with the characters of the stories.’’
Each trial began with a white screen presented for 4 s followed by a black and green crosses
displayed respectively for 6 s (jittered) and 0.5 s. Successively, the scenario was presented,
displayed until the participants’ click of the mouse, followed by a white screen during which
the odour was released for 3 s. Odours were presented birhinally using a computer-automated
olfactometer delivering odours in a temporally precise, square-shaped manner (Lundstro
¨m
et al., 2010). A low birhinal flow rate of 3.0 L/m (a total of 1.5 L/m per nostril) for a total
duration of 3 s per stimulus was used to prevent irritation of the nasal mucosa over time
(Lo
¨tsch, Ahne, Kunder, Kobal, & Hummel, 1998; Lundstro
¨m et al., 2010). After each odour
presentation, clean air was presented to minimize odour residuals (Parma, Ferraro, Miller,
A
˚hs, & Lundstro
¨m, 2015; Seubert, Gregory, Chamberland, Dessirier, & Lundstro
¨m, 2014),
while the question slide was presented. The two choice options were displayed below the
Cecchetto et al. 7

question. Participants were instructed to make their choice as fast as possible but there was
no time restriction. Following every answer, two rating slides were presented. Participants
rated how arousing and pleasant the odour was on a 10-cm Visual Analog Scale (VAS).
Higher scores indicate greater arousal and pleasantness. Participants were instructed to
answer even if they did not perceive any odour within a 6-s time window/rating. Dilemmas
were presented in three blocks of 16 trials (a total of 48 trials/subject), in which odours were
presented in an event-related manner. Participants were allowed to take a short break at the
end of each block. At the end of the moral decision-making task, participants completed the
second version of the STAI State.
Statistical analysis. Data were analyzed with linear mixed-effects models (LMMs) using R
(version 2.10.1; http://www.r-project.org/) and in particular using lme function (nlme
package; https://cran.r-project.org/web/packages/nlme/nlme.pdf) for continuous variables
and the glmer function (lme4 package; http://cran.r-project.org/web/packages/lme4/index.
html) for the binary variable. To account for individual differences (e.g., some people are
more ‘‘deontological’’ than others), participants were included in the models as random.To
avoid a warning of non-convergence, an optimizer (bobyqa) was applied (Powell, 2009).
Results with and without the optimizer are not significantly different (https://github.com/
lme4/lme4/blob/master/misc/notes/release_notes.md). Estimates on the choice between
utilitarian and deontological responses were based on an adaptive Gaussian Hermite
approximation of the likelihood with 10 integration points. All models included odour
factor, as the main interest of our analysis, and the four conceptual moral factors
(Personal force,Intentionality,Benefit recipient and Evitability) as factors describing our
items. Moreover, for each dependent variable, at the beginning, all self-report
questionnaires, relative to our participants traits, and second-level interactions as fixed
effects was built, were included (Crepaldi et al., 2012; Faraway, 2005; McLean, Sanders, &
Stroup, 1991; Wehling et al., 2016) and then they were progressively removed stepwise until
the deletion of any additional effect caused a significant loss of fit to the model (as tested by
the Akaike Information Criterion (AIC) tests resulting from the use of the generic anova
function). Reduced AIC was used as criterion for model selection because it favours
parsimonious models, also when the sample size is small (Bolker et al., 2009).
Finally, to test whether the odour condition significantly increases the explained variance
of the final model selected, such model was directly contrasted with the same model without
the odor term, included as a main effect or in interaction with the dilemma features. The two
models were compared using the general anova function: In the results section, X
2
and pvalue
are stated for moral choice analysis while L. Ratio and pvalue are reported for continuous
dependent variables (SCR and IHR). Additionally, description of the selected model and the
AIC value per model were reported.
Outliers with respect of reading times and reaction times were removed considering the
outlier-labelling rule (Hoaglin, Iglewicz, & Tukey, 1986). From a starting number of 816
trials, 19 trials were removed because of extremes reading times (>49.27 s; N¼19/816,
2.32%) while 33 trials were removed because of extremely long choice reaction times
(>9.25 s; N¼33/797, 4.14%). Final analyses were run on a sample of 764 trials distributed
over conditions.
Results
Sub- and supra-threshold odour conditions do not differ in arousal ratings. The LMM on arousal
ratings (no-odour: M¼7.22, SD ¼0.07; sub-threshold neutral odour: M¼7.08, SD ¼0.07;
8Perception 0(0)

and supra-threshold neutral odour: M¼7.02, SD ¼0.07; Figure 1(a)) showed no significant
differences across odour conditions (all t<1.3, all p>.2). Please refer to Table 1SI of
Supplemental materials for descriptive data.
The LMM on valence ratings (no-odour: M¼4.76, SD ¼0.06; sub-threshold neutral
odour: M¼4.68, SD ¼0.06; and supra-threshold neutral odour: M¼4.45, SD ¼0.07)
revealed no significant difference between sub- and supra-threshold odour condition and
between sub-threshold and no-odour conditions (all t<0.7, all p>.1; Figure 1(b)).
Considering no-odour condition as reference the model showed a significant difference
between supra-threshold and no-odour conditions [t(745) ¼2.22, p¼.03].
Both sub- and supra-threshold neutral odour increase deontological choices. The best fitting model for
moral choice data (reference factor: utilitarian choice; AIC ¼884.79; X
2
¼30.73, p<.001)
was the one including the odour condition (p<.001), personal force (p<.001), intentionality
(p<.001), benefit recipient (p¼.04) and evitability (p¼.38), the interaction of odour and
each of these three factors, and the scores at AQ (p¼.001), DS (p¼.001) and STAI (p¼.03;
see Table 2 for descriptive data of single parameters). No effect of single odour conditions on
moral choice was retrieved (all z<0.3, all p>.2). However, odour significantly interacted
with the dilemma factor personal force (reference factors: sub-threshold odour and personal).
Post hoc tests (lsmeans function) revealed a selective effect of odour on impersonal dilemmas.
Specifically, the supra-threshold odour condition, as compared to the no-odour condition,
significantly increased the number of deontological choices to impersonal dilemmas (p¼.02).
The sub-threshold odour showed a significant effect in the same direction (sub-threshold vs.
no odour, p¼.04). However, no other contrasts reached the significance level. Additionally,
odour significantly interacted with the dilemma factor intentionality (reference factors: sub-
threshold odour and accidental). Post hoc tests (lsmeans function) showed a selective increase
Figure 1. (a) Distribution of participants’ arousal odour ratings of Study 1a. (b) Distribution of participants’
valence ratings of Study 1a. The black dots represent the value of each participant’s rating, per odour
condition, whereas the box-plot represents the interquartile range of each distribution, with the thick black
horizontal bar corresponding to the median. Each box-plot is surrounded by a violin plot representing the
smoothed distribution of data.
Cecchetto et al. 9

of deontological responses on instrumental dilemmas following the exposure to the supra-
threshold odour (vs. no-odour, p¼.01) and following sub-threshold odour (vs. no-odour,
p¼.001). Finally, the factor odour interacts with the benefit recipient factor (reference
factors: sub-threshold odour and other). Post hoc tests (lsmeans function) showed a
selective increase of deontological responses on other-recipient dilemmas following the
exposure to the sub-threshold odour (vs. no-odour, p¼.04). Irrespective of the odour, the
likelihood of choosing the deontological option increased when dilemmas were personal (vs.
impersonal), instrumental (harm was intentional vs. accidental) and other-benefit (vs. self-
benefit), with reduced autistic traits (i.e., lower AQ scores), increased anxiety (i.e., greater
STAI scores) and increased disgust sensitivity (i.e., greater DS scores).
Discussion
Study 1a was designed to investigate whether moral decision-making is biased by different
intensities of an emotionally neutral olfactory stimulus, either supra-threshold or sub-
threshold intensities. Subjective ratings of arousal and valence did not reveal any
difference between sub- and supra-threshold odour conditions but only between no-odour
and supra-odour conditions for valence ratings. The homogeneity in the arousal ratings
might indicate that sub- and supra-threshold odour conditions were not sufficiently
different at the perceptual level. This might explain the lack of main effect of odour on
deontological response rate. These data seem to support our first hypothesis, stating that
Table 2. Summary of the Best Linear Mixed Effects Model on Moral Choices for Study 1a.
Moral choices
SE z value pvalue exp
95% CI
Fixed effects Lower Upper
Intercept 0.01 0.81 0.02 0.987 1.013 0.206 4.971
No-odour 0.10 0.52 0.19 0.852 1.101 0.401 3.022
Supra-threshold neutral odour 0.34 0.58 0.58 0.563 0.713 0.227 2.243
Personal force (Impersonal)0.81 0.32 2.52 0.012 0.447 0.239 0.835
Intentionality (Instrumental) 1.20 0.33 3.58 <0.001 3.316 1.721 6.392
Benefit recipient (Self)0.83 0.31 2.65 0.008 0.437 0.237 0.806
Evitability (Inevitable) 0.35 0.33 1.08 0.281 0.704 0.372 1.333
AQ 0.15 0.04 3.69 <0.001 0.859 0.792 0.931
DS 0.17 0.04 3.73 <0.001 1.182 1.082 1.290
STAI trait difference 0.06 0.03 2.24 0.025 1.063 1.008 1.121
No-odour Personal Force (Impersonal)1.57 0.60 2.63 0.009 0.209 0.065 0.672
Supra-odour Personal Force (Impersonal) 0.23 0.44 0.52 0.603 1.260 0.527 3.011
No-odour Intentionality (Instrumental)1.19 0.49 2.44 0.015 0.305 0.118 0.791
Supra-odour Intentionality (Instrumental)0.43 0.45 0.95 0.343 0.652 0.269 1.578
No-odour Benefit Recipient (Self) 0.75 0.43 1.72 0.085 2.111 0.902 4.944
Supra-odour Benefit Recipient (Self) 1.17 0.52 2.25 0.024 3.224 1.164 8.927
No-odour Evitability (Inevitable) 0.38 0.45 0.84 0.399 1.457 0.607 3.498
Supra-odour Evitability (Inevitable)0.28 0.57 0.50 0.619 0.752 0.246 2.305
Note. ¼estimate; SE ¼standard error; CI ¼confidence interval; exp ¼exponential of coefficient; Sub-threshold
odour ¼sub-threshold neutral odour condition; Supra-threshold odour ¼supra-threshold neutral odour condition;
DS ¼disgust scale; STAI ¼State subscale of the State–Trait Anxiety Inventory; AQ ¼Autism-Spectrum Quotient.
Significant pvalues are reported in bold. Table shows model with utilitarian choice and sub-threshold odour condition
set as references. Contrast condition from the reference for categorical factors is reported in italic inside bracket.
10 Perception 0(0)

odour intensity does not critically modulate moral choice, perhaps because neutral odours
are not inducing a disgust reaction in the decision maker. If this were completely true, no
odour effect on moral choice should be present. However, results of Study 1a reveal three
odour-dilemma factor interactions. First, both supra-threshold and sub-threshold odour
conditions increase the choice of deontological responses, but such effect is confined to
impersonal and instrumental dilemmas. Second, the sub-threshold odour increases
deontological responses only for dilemmas in which others benefit from the moral choice.
Altogether these data suggest that the odour – irrespective of its intensity – increases
deontological responses to the dilemmas perceived as less arousing (Christensen et al.,
2014; Lotto et al., 2014; Moore, Clark, & Kane, 2008).
In line with previous literature, participants provided more deontological responses when
the moral dilemma was personal (Christensen et al., 2014; Greene et al., 2004; Greene et al.,
2009; Greene et al., 2008; Koenigs et al., 2007; Mendez, Anderson, & Shapira, 2005; Moore
et al., 2008; Moore, Lee, Clark, & Conway, 2011; Moretto et al., 2009), instrumental (Borg,
Hynes, Van Horn, Grafton, & Sinnott-Armstrong, 2006; Christensen et al., 2014; Cushman,
Young, & Hauser, 2006; Greene et al., 2009; Hauser, Cushman, Young, Kang-Xing Jin, &
Mikhail, 2007; Lotto et al., 2014; Moore et al., 2008; Sarlo et al., 2012) and/or self-beneficent
(Christensen et al., 2014; Lotto et al., 2014; Moore et al., 2008).
Moreover, according to the literature, we confirmed that the deontological moral
tendency is reduced in individuals reporting autistic traits (Buon et al., 2013; Moran et al.,
2011; but see Patil, Melsbach, Hennig-Fast, & Silani, 2016), higher anxiety (Starcke et al.,
2008; Youssef et al., 2012a) and increased in those exhibiting higher disgust sensitivity (Choe
& Min, 2011; Schnall et al., 2008; Ugazio et al., 2012, but see Ong, O’Dhaniel, Kwok, & Lim,
2014).
In light of the lack of significant effects between sub- and supra-threshold arousal ratings,
in the next study we will assess the hypotheses tested here in an independent sample of
participants, incrementing the concentration difference between sub- and supra-threshold
odour conditions. Furthermore, we will extend the evaluation of odour-induced arousal
and valence effects to physiological reactions, known implicit indicators of emotional
experiences (e.g., Lang, 1995).
Study 1b – Does Odour Intensity Affect the Psychophysiological
Correlates of Moral Choice?
Material and Methods
Participants. Fifteen participants were included in Study 1b. All screening procedures were the
same as Study 1a. See Table 1 for participants’ characteristics.
Stimuli. The same moral dilemmas, odour conditions, as well as the procedure for establishing
odour concentrations used in Study 1a were used in Study 1b. As to avoid the distribution of
odour thresholds across participants being skewed towards one end of the intensity scale, the
average threshold found for Study 1a was centred on the dilution series scale. In this way, the
dilution series in Study 1b was prepared using 12.5% v/v as dilution starting point. From
there, the odour was diluted in 16 consecutive dilution steps using a 0.5-volume dilution series
(end concentration 0.00038% v/v). To increase the perceptual distance between odour
conditions, two steps below and two above the detection threshold of each participant
were considered as odour concentrations for sub- and supra-threshold conditions,
respectively.
Cecchetto et al. 11

Procedure. The procedure and task was the same as in Study 1a except for the following
improvements. SCR electrodes and the photoplethysmographic probe to measure cardiac
activity were attached 10 min before the beginning of the tasks and 1 min of baseline
was recorded at the beginning of first block. To avoid fatigue effects due to a long
experimental session, we presented the odour ratings at the beginning and at the end of
the moral decision-making task. Each odour presentation was preceded by a green fixation
cross on the screen for 0.5 s followed by a black screen while an odour was presented for 4 s; a
white screen followed for an average interstimulus interval of 6 s (0.12 s). Subsequently, two
questions were asked in random order, namely ‘‘How intense was the odour you just
smelled?’’ and ‘‘How pleasant was the odour you just smelled?’’ Perceptual ratings for
odour intensity and pleasantness were collected on a 10-cm computerized VAS, ranging
from ‘‘not at all’’ to ‘‘very much.’’ Participants were instructed to answer even if they did
not perceive odour. In the moral decision-making task, the scenario was presented for a fixed
time (10 s) plus the time needed by participants to complete reading the slide. Moral decisions
were timed to a maximum of 6 s.
Psychophysiological data acquisition and analysis. SCR and HR were recorded with a PROCOMP
infiniti system (Thought Technology, Montreal, Canada). SCR was recorded using a pair of
prewired 8 mm Ag/AgCl electrodes, attached to the surface of the medial phalanx of the
index and ring fingers of the non-dominant hand as suggested on the PROCOMP infiniti
system manual. Conductive gel was used to reduce impedance. The electrode pair was excited
with a constant voltage of 0.5 V and conductance was recorded using a DC amplifier with a
low-pass filter set at 64 Hz. Such procedures are in accordance with the guidelines by Figner
and Murphy (2011) and Boucsein (2012). A photoplethysmographic probe (3.2 cm/1.8 cm,
photodetector LED type), placed on the middle finger of the non-dominant hand was used to
assess HR in beats per minute (bpm) at a sample rate of 2048 Hz. SCR and HR data were
analysed with Matlab using the same in-house scripts used in our previous study (Cecchetto,
Korb, Rumiati, & Aiello, under review), and partially using the EEGLAB toolbox (http://
sccn.ucsd.edu/eeglab/).
SCR data were filtered with 10 Hz low-pass filters and epoched over the 21 s from the
odour presentation. Two seconds before each odour presentation served as baseline. Peak
amplitude was analysed, defined as the difference in mSiemens between the mean value during
baseline and the peak after stimulus onset. Trials with peak amplitudes below 0.01 mSiemens
were excluded from the analysis. To improve interpretability, peak amplitudes were log-
transformed (Boucsein, 2012).
Heart rate data were filtered with 1 Hz high-pass filter and resampled to 256 Hz. Beat
detection was performed automatically, verified visually and corrected, if necessary.
Frequency was computed as beats per minute (bpm). The 9 s from the odour presentation
were divided in six time windows of 1.5 s. Interbeat intervals were computed, transformed to
heart rate values and averaged for each 1.5 s window. Each time windows was then corrected
by subtracting the 1.5 s before odour presentations to obtain the IHR (Palomba et al., 2000).
Relatively to the trial design, the odour was presented during Time Window 1 and 2, the
question slide appeared at the beginning of the third time window and the amount of time
available to the participants to the answer ended at the end of the sixth time window.
Statistical analysis. For the behavioural and IHR analyses, no reaction times (RT) choice
outliers were identified (720 trials). For the SCR analyses, a subset of data was included.
Specifically, data from one subject were removed due to technical issues. Of the remaining
672 trials, 139 were removed due to SCR peak amplitude lower than 0.01 mSiemens (Roth
12 Perception 0(0)

et al., 2012; N¼139/672, 20.7%, mostly due to habituation issues towards the end of the last
two blocks, Boucsein, 2012, pp. 275–282). Final analyses were run on a sample of 533 trials
distributed over conditions. For all other aspects, analyses are equivalent to those presented
for Study 1a. Please refer to Study 1a statistical analysis section for details.
Results
The supra-threshold odour is more intense– but not more pleasant – than the sub-threshold odour. We
tested whether the three odour conditions had the expected perceptual impact. Since there
were no significant differences between the ratings before and after the moral decision-
making task, data from the two sessions were collapsed. As evident from Figure 2(a), a
LMM on odour intensity ratings (no-odour: M¼2.70, SD ¼0.21; sub-threshold neutral
odour: M¼3.97, SD ¼0.25; and supra-threshold neutral odour: M¼6.97, SD ¼0.20)
revealed a significant difference between odour conditions: The supra-threshold odour
condition was perceived as significantly more intense than both no-odour (p<.0001; the
reference factor was set in no-odour) and sub-threshold odour conditions (p<.0001; the
reference factor was sub-threshold odour); moreover, the sub-threshold odour showed a
significance difference compared to no-odour condition (p¼.02).
For valence ratings (no-odour: M¼4.31, SD ¼1.93; sub-threshold neutral odour:
M¼4.03, SD ¼2.35; supra-threshold neutral odour: M¼3.60, SD ¼1.89), no significant
differences were found (all t<0.6, all p>.5; Figure 2(b)). For the full LMM results, please
refer to Table 2SI of the Supplemental materials.
Deontological choice rate increases during the sub- threshold odour condition. The best fitting model
for moral choice data (reference factor: utilitarian choice; AIC ¼884.88; X
2
¼38.06, p<.001)
resulted being the one including the odour condition (p<.001), personal force (p¼.001),
intentionality (p<.001), benefit recipient (p¼.01) and evitability (p¼.004), the interaction of
odour and each of these three factors, and the scores at AQ (p¼.03; see Table 3 for
descriptive data of single parameters). Four main effects were found to modulate the
moral choice tendency: the odour condition, the personal force factor, the evitability
factor and AQ (reference factors: sub-threshold odour and personal). The likelihood of
choosing deontological responses significantly increased following the exposure to the sub-
threshold odour as compared to the supra-threshold odour condition. Changing no-odour
condition as reference factor reveals an increment of deontological responses following the
exposure to the sub-threshold odour when compared to the no-odour condition,
z(720) ¼2.26, p¼.02. Although the model produced three significant interactions
(Odour Intentionality, Odour Benefit Recipient and Odour Evitability; reference
factors: sub-threshold odour, accidental, other, avoidable), the post hoc contrasts (lsmeans
function) reveal meaningful significant contrasts only for the interaction Odour Evitability:
a selective increase of deontological responses on avoidable dilemmas following the exposure
to the sub-threshold odour compared to no-odour (p¼.02) and compared to supra-threshold
odour conditions (p¼.03). Moreover, as found in Study 1a, the likelihood of choosing
deontological responses increased when the dilemmas were personal (vs. impersonal), when
avoidable (vs. inevitable) and with lower AQ scores.
SCR is greater for sub- and supra-threshold odour conditions compared to clean air. The best fitting
model for amplitude profiles of SCR (AIC ¼312.89; L. Ratio ¼26.82; p<.001) was the one
including the odour factor (p<.001), moral choice (p¼.09) and the interaction between the
two factors, personal force (p¼.06), intentionality (p¼.007), benefit recipient (p¼.09) and
Cecchetto et al. 13

evitability (p¼.34) and the scores at AQ (p¼.04; see Table 3SI of Supplemental materials for
descriptive data of single parameters). The likelihood of presenting greater SCR increases
during supra-threshold odour conditions compared to sub-threshold odour condition
(p¼.03) and to no-odour condition (p¼.03), but no significant difference is evident in
SCR between sub-threshold and no-odour conditions (p¼.97). Moreover, the SCR
amplitude significantly increases for utilitarian choice compared to deontological choice.
The odour significantly interacts with the moral choice (reference factors: sub-threshold
odour conditions and utilitarian choices): Post hoc analysis (lsmeans function) indicates
Figure 2. (a) Distribution of participants’ intensity odour ratings of Study 1b. (b) Distribution of
participants’ valence ratings of Study 1b. (c) SCR per odour conditions of Study 1b. (d) IHR per time window
and odour conditions of Study 1b. ((a) and b)) The black dots represent single data points, whereas the box-
plot represents the interquartile range of each distribution, with the thick black horizontal bar corresponding
to the median. Each box-plot is surrounded by a violin plot representing the smoothed distribution of data.
Significant differences are indicated.
14 Perception 0(0)

that in utilitarian choices greater SCR is associated to both odour conditions (sub: p<.001;
supra: p¼.007), irrespective of their intensity (Figure 2(c)). Although no significant
differences in post hoc contrasts were revealed for deontological choices, greater SCR is
nominally associated with supra-threshold, but not with sub-threshold and no-odour
conditions. Moreover, the likelihood of showing a greater SCR increases when the
dilemmas presented were accidental. Moreover, the likelihood of presenting greater SCR
increases with higher AQ scores.
IHR decelerates during sub- and supra-threshold odour condition. The best model explaining IHR
activity included (AIC ¼27876.54; L. Ratio ¼10.64, p¼.005), time (p¼.02), odour
conditions (p¼.004), personal force (p¼.93), intentionality (p¼.66), benefit recipient
(p¼.08), evitability (p¼.67), moral choice (p¼.21) and the scores at STAI (p¼.04; see
Table 4SI of Supplemental materials for descriptive data of single parameters). The
likelihood of presenting an decelerated IHR is affected by the odour smelled. In detail, a
significant IHR deceleration is evident when participants smell the sub- and supra-threshold
odour conditions compared to no-odour (p¼.004 and p¼.005, respectively), but no
differences emerge between sub- and supra-threshold odour conditions (p¼.91). Also, a
time effect is present: IHR significantly accelerates compared to odour onset (Time
Window 1) during the second half of the odour stimulation (irrespective of the odour
valence), during the presentation of the question slide (Time Windows 2 to 4; Figure 2(d)).
Finally, there is an effect of anxiety: The greater the STAI score (i.e., the more anxious the
participant), the greater the IHR deceleration.
Table 3. Summary of the Best Linear Mixed Effects Model on Moral Choices for Study 1b.
Moral choices
SE z value pvalue exp
95% CI
Fixed effects with reference to sub-threshold odour Lower Upper
Intercept 2.20 0.88 2.50 0.012 9.012 1.610 50.440
No-odour 1.17 0.52 2.26 0.024 0.310 0.112 0.857
Supra-threshold neutral odour 1.74 0.54 3.25 0.001 0.175 0.061 0.500
Personal force (Impersonal)1.13 0.31 3.60 <0.001 0.324 0.176 0.598
Intentionality (Instrumental) 0.32 0.34 0.95 0.343 1.378 0.710 2.675
Benefit recipient (Self)0.12 0.36 0.33 0.739 0.886 0.434 1.808
Evitability (Inevitable)0.90 0.31 2.87 0.004 0.406 0.219 0.752
AQ 0.10 0.04 2.30 0.022 0.907 0.834 0.986
No-odour Personal Force (Impersonal) 0.80 0.54 1.49 0.136 2.235 0.777 6.430
Supra-odour Personal Force (Impersonal) 0.63 0.46 1.37 0.169 1.877 0.765 4.607
No-odour Intentionality (Instrumental) 0.99 0.48 2.03 0.042 2.679 1.036 6.928
Supra-odour Intentionality (Instrumental) 0.75 0.47 1.59 0.112 2.120 0.840 5.347
No-odour Benefit Recipient (Self)1.46 0.64 2.30 0.022 0.231 0.066 0.807
Supra-odour Benefit Recipient (Self) 0.18 0.51 0.36 0.719 1.200 0.445 3.234
No-odour Evitability (Inevitable) 1.56 0.47 3.32 0.001 4.760 1.897 11.944
Supra-odour Evitability (Inevitable) 1.22 0.45 2.72 0.006 3.392 1.408 8.172
Note. ¼estimate; SE ¼standard error; exp ¼exponential of coefficient; 95% CI ¼confidence interval; sub-threshold
odour ¼sub-threshold neutral odour condition; supra-threshold odour ¼supra-threshold neutral odour condition;
AQ ¼Autism-Spectrum Quotient. Significant pvalues are reported in bold. Table shows model with utilitarian choice
and sub-threshold odour condition set as references. Contrast condition from the reference for categorical factors is
reported in italic inside bracket.
Cecchetto et al. 15

Discussion
The goal of Study 1b was to re-evaluate the hypotheses on the effects of odour intensity on
moral choice tested in Study 1a, by using odour conditions that were more perceptually
discriminable between each other. As the analysis of the odour intensity ratings reveals, in
Study 1b the three odour conditions are parametrically distributed: The supra-threshold
odour condition was significantly more intense than the sub-threshold odour condition,
which itself was significantly more intense than the no-odour condition. Furthermore, the
odour valence ratings resulted homogenous across conditions, indicating that the olfactory
effects on moral choice at the behavioural and psychophysiological level can be attributed to
odour intensity alone.
Deontological choices are more likely to be made in the presence of a sub-threshold odour
as compared to the supra-threshold odour and no odour. This pattern of results supports the
hypothesis that the neutral odour condition impacts on moral choice only when it is
presented sub-threshold. We suggest that this effect might emerge because the elaboration
of irrelevant sensory information – whose sensory signal is minimal – can escape strategic
inhibitory control (e.g., Li, Moallem, Paller, & Gottfried, 2007). As previously shown, odours
– even when presented in very low concentration – can constitute a distractor for the process
of acting, whether acting is considered as a real movement (i.e., Parma, Bulgheroni,
Scaravilli, Tirindelli, & Castiello, 2013; Parma et al., 2012) or, as in the present case,
whether acting is considered as a decision to act. In other words, during moral decision-
making task, the information irrelevant to the decision is processed unconsciously and it
biases the moral choice towards a more emotional tendency, given that cognitive resources
have been taxed by the presence of the irrelevant stimulus (Greene et al., 2008).
One may think that the results of Study 1b are in contrast to the results of Study 1a.
However, we confirm the effects on moral choice that are not dependent on odour variables,
revealing a consistent increase in deontological responses for personal dilemmas as well as the
tendency of participants reporting higher autistic traits to prefer utilitarian choices. With
respect to the odour variables, a clearer definition of the relative intensity of the odour
conditions allowed to show that a disgust reaction is not the only factor able to modulate
moral choices.
The psychophysiological analyses confirm that the behavioural findings are strictly
dependent on differences in intensity of the odours. Indeed, arousal measured through
SCR reveals that arousal is increased following the exposure to the supra-threshold odour
condition with respect to the sub-threshold odour condition. This is in line with the idea that
increments in SCR correspond to the orientation towards newly introduced stimuli (e.g.,
Frith & Allen, 1983). Furthermore, the analysis of SCR is consistent with the differences
in odour intensity reported subjectively. Interestingly, this effect is conserved when
considering deontological choices, whereas SCR does not track odour intensity in
utilitarian choices. Considering that the literature reports that utilitarian choices are
associated with greater SCR as compared to deontological choices (Moretto et al., 2009),
the odour effect is hidden by the ceiling effect of arousal in utilitarian choice. In other words,
the arousal induced by the processing of the utilitarian choice is already high enough not to
leave room for the influence of the odour.
To strengthen the idea that in Study 1b the effect of odour pleasantness was minimal, in
line with the homogeneity of pleasantness ratings, the analysis of IHR showed no significant
differences across sub- and supra-threshold conditions. This is in line with previous evidence
showing that cardiac responses vary with hedonic values and reports of pleasantness (Alaoui-
Ismaı
¨li, Robin, Rada, Dittmar, & Vernet-Maury, 1997; Alaoui-Ismaili, Vernet-Maury,
Dittmar, Delhomme, & Chanel, 1997; Bensafi et al., 2002). The reduction of the cardiac
16 Perception 0(0)

output emerging in the presence of both odour conditions as compared to the no-odour
condition is to be attributed to the attentional processing of the stimuli (Bradley, 2009).
Here, we extend for the first time the study moral behavioural and psychophysiological
effects to the odour intensity of an emotionally neutral stimulus, suggesting that odour
intensity impacts on moral decision-making, by promoting a deontological moral
tendency, as confirmed by the psychophysiological findings. However, the effect of odour
valence on moral decision-making has still to cover the full spectrum of possibilities, ranging
from unpleasant to pleasant odours. In Study 2, we address this issue by considering the effect
of a neutral, pleasant and unpleasant sub-threshold odours in moral decision-making.
Study 2 – Does Odour Valence Affect the Behavioural and
Psychophysiological Correlates of Moral Choice?
Material and Methods
Participants. A total of naı
¨ve 15 participants (Table 4), following the screening procedures of
Study 1a and 1b, were included in Study 2.
Odour stimuli. In Study 2, three odour conditions were presented during the moral decision
making task: no-odour, neutral sub-threshold odour condition (cedarwood oil, Sigma-
Aldrich, Italy) and a solution of butyric acid presented in sub-threshold concentration
(Sigma-Aldrich, Italy). Butyric acid was chosen in virtue of its hedonic ambiguity. Indeed,
it has been reported to be perceived either as a pleasant odour (e.g., parmesan cheese) or an
unpleasant odour (e.g., vomit; Herz, 2006; Herz & von Clef, 2001), while keeping the
chemical properties constant. Mineral oil (Sigma-Aldrich, Italy) was used as diluting agent
for cedarwood oil, while propylene glycol (Sigma-Aldrich, Italy) was used for butyric acid.
All concentrations below are given as volume to volume (v/v) in liquid phase. Odour
Table 4. Summary Table of Demographic Characteristics and Questionnaires of Study 2.
Mean (SD) Range
Gender 13 females –
Age 25.80 (4.95) 19–38
Education 16.93 (2.49) 13–20
Threshold test (Cedarwood Oil) 8.75 (1.12) 6.25–12.25
Threshold test (Butyric Acid) 12.38 (0.34) 11.5–14.5
Identification test 13.27 (1.33) 11–16
BVAQ 46.07 (5.50) 35–52
DS 15.17 (4.16) 7–24
STICSA (Trait) 39.53 (11.09) 23–65
AQ 17.87 (6.01) 8–28
OAS 120.40 (12.29) 103–139
STAI (State – PRE) 35.00 (6.96) 25–52
STAI (State – POST) 37.13 (810.20) 25–59
Note: BVAQ ¼Bermond–Vorst Alexithymia Questionnaire; DS ¼Disgust Scale; STICSA ¼State Trait
Inventory for Cognitive and Somatic Anxiety; AQ ¼Autism-Spectrum Quotient; OAS ¼Odour
Awareness Scale; STAI (State) ¼State–Trait Anxiety Inventory; POMS ¼Profile of Mood States.
Cecchetto et al. 17

concentrations for the cedarwood oil and butyric acid odours were established for each
participant through the same odour detection threshold test used in Study 1b.
Based on the pre-/post-task average valence rating of each participant, the original group
was split in a Pleasant (N¼8; rating above median split, pleasantness >5/10) and an
Unpleasant group (N¼7; rating below median split, <5/10). In case the mean was 5.0,
values of ratings at the end of the study were considered (rating above median split,
namely 5/10, Pleasant group; rating below median split, namely <5/10, Unpleasant group).
Procedure. In Study 2 we replicated the procedure, visual stimuli, psychophysiological
acquisition and statistical analyses used in Study 1b. Only the odour conditions, described
earlier, have partially changed.
Statistical analysis. Behavioural and IHR analyses were performed on a total of 720 trials: No
outliers were identified considering the reaction times. Analyses on SCR data were calculated
on a total of 537 trials, due to SCR peak amplitude lower than 0.01 mSiemens (Roth et al.,
2012; N¼183/720, 25.41%, mostly due to habituation issues towards the end of the last two
blocks, Boucsein, 2012, pp. 275–282). Removed trials were distributed over 15 subjects and
equally over conditions. The analyses are equivalent to those presented for Study 1a and 1b,
except for the addition of Wilcoxon test for the analysis of the ratings across groups.
Results
Odour conditions are perceived as different in intensity and in valence. Given that no significant
differences between the answers before and after the moral decision-making task were
found, the data from the two sessions were collapsed. The LMM on odour intensity
ratings (no-odour: M¼2.77, SD ¼0.17; neutral odour: M¼3.80, SD ¼0.22; pleasant
odour: M¼4.25, SD ¼0.21; unpleasant odour: M¼3.21, SD ¼0.17) showed a main effect
of odour condition (reference factor no-odour): Pleasant odour condition and neutral odour
conditions were rated as more intense than the no-odour condition (respectively: p¼.02 and
p¼.01; Figure 3(a)). Changing the reference factor does not reveal other significant effects.
Wilcoxon tests on odour valence ratings were performed to evaluate whether no-odour and
neutral odour conditions were equivalent across groups. Results showed that no significant
differences were evident for no-odour condition (group Pleasant: M¼4.88, SD ¼0.07; group
Unpleasant: M¼4.29, SD ¼0.33; w¼123, p¼.23, 95%CI: [1.03 10
6
, 2.00]) and for
neutral odour condition (group Pleasant: M¼5.06, SD ¼0.21; group Unpleasant:
M¼3.93, SD ¼0.29; w¼156.5, p¼.06, 95%CI: [3.85 10
5
, 2.99]). As expected, there
was a significant difference in the valence attributed to butyric acid across groups (group
Pleasant: M¼5.38, SD ¼0.14; group Unpleasant: M¼3.36, SD ¼0.31; w¼182.5, p¼.002,
95%CI: 0.99, 3.00]). A LMM on odour valence ratings showed a significant effect of odour:
The unpleasant odour condition was perceived as significantly more unpleasant than the no-
odour condition (p¼.04; see Figure 3(b)). Setting the pleasant odour as reference revealed that
the unpleasant odour condition was perceived as significantly more unpleasant than the
pleasant condition, t(72) ¼2.74, p¼.008. No other significant effects were retrieved.
Please refer to Table 5SI of Supplemental materials for intensity and valence ratings
LMMs’ details.
Deontological responses increased while exposed to the neutral odour but not to pleasant and unpleasant
odour conditions. The best fitting model for moral choice data (reference factor: utilitarian
choice; AIC ¼907.26; X
2
¼53.83, p<.001) resulted being the one including the odour
18 Perception 0(0)

condition (p<.001), personal force (p¼.01), intentionality (p<.001), benefit recipient
(p¼.02) and evitability (p¼.008), the interaction of odour and each of these three factors,
and the scores at STICSA (p¼.02; Table 5). Deontological choices increase during the
neutral odour condition compared to the pleasant (p¼.01), unpleasant (p¼.005) and no-
odour conditions (p<.0001; reference factors: neutral odour and utilitarian choices).
Changing reference in the odour condition factor did not show other significant effects.
The increment in deontological moral choices was also promoted by the odour in
interaction with the intentionality factor (reference factors: accidental). Relevant post hoc
(lsmeans function) tests showed that participants gave more deontological responses when
exposed to the neutral odour condition compared to no odour for accidental (p¼.01)
Figure 3. ((a) and (b)) Intensity and valence odour ratings for Study 2; (c) SCR per odour conditions; (d)
IHR per time windows and odour conditions.((a) and (b)) The black dots represent single data points,
whereas the box-plot represents the interquartile range of each distribution, with the thick black horizontal
bar corresponding to the median. Each box-plot is surrounded by a violin plot representing the smoothed
distribution of data. Significant differences are indicated.
Cecchetto et al. 19

dilemmas. Moreover, the likelihood of choosing deontological answers increased for personal
(vs. impersonal) and avoidable (vs. inevitable) dilemmas as well as for lower STICSA scores.
SCR is greater for pleasant odour conditions. The best model (AIC ¼446.02; L. Ratio ¼11.97,
p¼.06) included the odour condition (p¼.08), moral choice (p¼.46), personal force
(p¼.21), intentionality (p<.001), benefit recipient (p¼.16) and evitability (p¼.27), the
interaction of odour and moral choice (Table 6SI). Participants showed greater SCRs
during the pleasant odour condition compared to the neutral (p¼.03) and no-odour
conditions (p¼.01; setting pleasant odour as reference). The unpleasant odour condition
showed a tendency for greater SCR as compared to no odour (p¼.06). Even though the
interaction between odour condition and moral choice is not significant, nominally, the
pattern (Figure 3(c)) seems to be in line with the arousal elicited by the unpleasant odour
but not by the neutral odour. Moreover, the likelihood that participants showed greater
SCRs increased when dilemmas were accidental (p<.001).
Table 5. Summary of the Best Linear Mixed Effects Model on Moral Choices for Study 2.
Moral choices
SE z value pvalue exp
95%CI
Fixed effects with reference to neutral odour Lower Upper
Intercept 3.06 0.94 3.25 0.001 21.311 3.369 134.809
No-odour 2.00 0.52 3.83 <0.001 0.136 0.049 0.377
Pleasant odour 1.50 0.62 2.42 0.016 0.222 0.066 0.753
Unpleasant odour 1.87 0.68 2.76 0.006 0.154 0.041 0.582
Personal force (Impersonal)0.93 0.32 2.93 0.003 0.396 0.213 0.736
Intentionality (Instrumental) 0.17 0.33 0.51 0.608 1.187 0.617 2.283
Benefit recipient (Self)0.34 0.36 0.94 0.345 0.714 0.354 1.438
Evitability (Inevitable)0.82 0.31 2.65 0.008 0.441 0.240 0.808
STICSA (Trait) 0.05 0.02 2.34 0.019 0.951 0.912 0.992
Odour Personal Force
(No-Odour, Impersonal)
1.00 0.53 1.90 0.058 2.710 0.968 7.589
Odour Personal Force
(Pleasant Odour, Impersonal)
0.17 0.53 0.32 0.748 1.186 0.418 3.367
Odour Personal Force
(Unpleasant Odour, Impersonal)
0.47 0.58 0.81 0.416 1.606 0.514 5.020
Odour Intentionality
(No-Odour, Instrumental)
1.45 0.48 3.01 0.003 4.255 1.659 10.915
Odour Intentionality
(Pleasant Odour, Instrumental)
0.14 0.53 0.27 0.789 1.153 0.407 3.269
Odour Intentionality
(Unpleasant Odour, Instrumental)
0.08 0.57 0.13 0.894 1.079 0.351 3.319
Odour Evitability (No-Odour, Inevitable) 0.95 0.61 1.56 0.118 0.386 0.117 1.271
Odour Evitability
(Pleasant Odour, Inevitable)
0.63 0.58 1.08 0.278 1.873 0.602 5.826
Odour Evitability
(Unpleasant Odour, Inevitable)
1.21 0.62 1.93 0.053 3.337 0.983 11.335
Note: ¼estimate; SE ¼standard error; CI ¼confidence interval; exp ¼exponential of coefficient; STICSA ¼State Trait
Inventory for Cognitive and Somatic Anxiety. Significant pvalues are reported in bold. Table shows model with utilitarian
choice and neutral odour condition set as references. Contrast condition from the reference for categorical factors is
reported in italic inside bracket.
20 Perception 0(0)

IHR is reduced during the neutral odour condition compared to no-odour conditions. The best model
(AIC ¼27077.05; L. Ratio ¼7.95, p¼.04) resulted being the one including time (p<.001),
odour condition (p¼.05), personal force (p¼.53), intentionality (p¼.78), benefit recipient
(p¼.58), evitability (p¼.02) and moral choice (p<.001; Table 7SI; reference factors: neutral
odour). The likelihood to present a reduced IHR is greater for the exposure to neutral odour
condition compared to the no odour condition (p¼.007) while there is only a tendency
between the neutral odour condition compared to the unpleasant condition (p¼.08).
Setting different reference factors does not reveal other significant effects. Moreover, the
likelihood of presenting a reduced IHR increased when participants made deontological
choices and when they answered to inevitable death dilemmas. Furthermore, a time effect
is present. Relevant post hoc contrasts reveal that irrespective of the odour condition, the first
time window, namely the odour onset, shows a significantly decreased IHR as compared to
all following windows (all p<.03; Figure 3(d)).
Discussion
Study 2 was designed to investigate whether odour valance can affect moral choices. Taking
advantage of the peculiar characteristics of ambiguous odours, which could either be perceived
as pleasant or unpleasant stimuli (de Araujo et al., 2005), we keep constant the chemical
stimulation while having participants differentially judge the pleasantness of the odour. As
expected, the butyric acid was perceived as significantly more pleasant by the Pleasant group
and significantly more unpleasant by the Unpleasant group. However, the intensity rating
results show that there were significant differences between the three odour conditions.
Overall, we found that unpleasant and pleasant odours did not bias moral choices while
the neutral odour did. Although this pattern of results does not overlap with any of our
hypotheses, it leans towards the first alternative, suggesting that odour valence has no effect
on moral choice. Specifically, smelling the neutral odour has a greater effect on deontological
tendencies as compared to the pleasant/unpleasant odours and no odour. We advance that
this result may emerge due to a novelty effect specific to the neutral odour (Bradley, 2009).
Indeed, cedarwood oil may be an odour less familiar to our participants as compared to
butyric acid. An additional, though not necessarily contrasting, explanation posits that for
the odour valence effect to emerge on moral choice, higher concentration of the stimulus is
necessary. Here, all odours were presented sub-threshold. As in Ugazio et al. (2012), during
the olfactory condition characterized by lower sprays, the effect on moral behaviour is not
evident. Furthermore, Study 2 confirmed the findings obtained in Study 1b showing that
moral choices were biased towards deontological answers by the neutral odour presented sub-
threshold.
The moral choice analyses confirmed the majority of predictors significantly increasing
deontological choices found in Study 1b, the study in which the odour manipulation
successfully differentiated conditions, as here in Study 2. Indeed, participants gave more
deontological answers when the moral dilemma was described as personal (Christensen
et al., 2014; Greene et al., 2004; Greene et al., 2009; Greene et al., 2008; Koenigs et al.,
2007; Mendez et al., 2005; Moore et al., 2008; Moore et al., 2011; Moretto et al., 2009;
Valdesolo & DeSteno, 2006; Youssef et al., 2012a) and avoidable (Hauser et al., 2007;
Moore et al., 2008; Christensen et al., 2014).
As previously found in Studies 1a and 1b, participants’ characteristics showed also an
effect on moral choices. However, here the test scores significantly predicting deontological
responses were a reduced trait anxiety (STICSA) – as expected from the literature
(Starcke et al., 2008; Youssef et al., 2012a but see Starcke, Polzer, Wolf, & Brand, 2011
Cecchetto et al. 21

for negative result). The lack of a significant effect of autistic traits probably depends on the
distribution of AQ scores across participants, which in this rather small sample is shorter
ranging and more balanced than in Studies 1a and 1b.
In line with the idea that deontological choices are associated with reduced physiological
activation as compared to utilitarian choices (Moretto et al., 2009), and that valenced odours
induce increased arousal as compared to neutral odours (Bensafi et al., 2002), SCR data show a
tendency toward lower arousal associated with the exposure to the neutral odour as compared
to the pleasant odour. Consistently, the behavioural data on moral choice reveal a significant
increment of deontological responses following the neutral odour condition. Specifically,
greater arousal for deontological responses following the pleasant odour condition tends
towards significance when compared to the arousal of deontological responses for the no-
odour condition. The reduced arousal shown in connection with the neutral odour is in line
with the electrodermal activity elicited by neutral novel stimuli, which is reported to be smaller
than that provoked by novel pleasant and unpleasant stimuli (Bradley, 2009).
Moreover, even though we were expecting the cardiac activity to discriminate between
pleasant and unpleasant odour stimulations (Bensafi et al., 2002), we did not found such
result but only a cardiac output reduction during the presentation of the neutral odour. Our
hypothesis to explain such effect is that the negative emotional state induced by the moral
dilemmas and by making moral choices – evident in the generalized acceleration post
dilemma processing – hid the odour valence effect on cardiac activity. We propose that the
effect of the neutral odour only survives in light of its attention-grabbing power, which
promotes cardiac activity reduction.
General Discussion
Our findings systematically evaluate, for the first time, the effects of sub-threshold olfactory
stimuli in the moral domain. Odours presented sub-threshold are shown to affect, for instance,
patterns of EEG activity (Lorig, Herman, Schwartz, & Cain, 1990), mood (Kirk-Smith, Van
Toller, & Dodd, 1983; Zucco, Paolini, & Schaal, 2009) and social preferences (Li et al., 2007).
This capacity, which makes olfactory stimuli relevant for moral decision-making, is related to
the unique anatomical features of the olfactory system. Indeed, it is the only sensory system in
which receptors project directly to the olfactory bulb and to the primary olfactory cortical areas
(Carmichael, Clugnet, & Price, 1994) without the obligatory thalamic relay, that has been
implied in conscious awareness (McAlonan, Cavanaugh, & Wurtz, 2008; Plailly, Howard,
Gitelman, & Gottfried, 2008). This makes olfactory stimuli to be good experimental tools to
investigate sub-threshold information processing (Li et al., 2007).
Besides opening an experimental window on the study of implicit information processing
in the moral domain, the chemical senses are also ideally suited to evaluate the role that
arousal and valence play in moral decision-making (Anderson et al., 2003; Winston et al.,
2005). If, as we have seen, odour intensity matters, with moral choice being biased towards
deontological responses when the odour is presented at a sub-threshold level, Study 2 might
suggest that odour valence – contrary to expectations - did not affect participants’ moral
decisions. We propose that the odour valence effect might have been hidden by the
attentional-grabbing effect of the neutral odour (Bradley, 2009), which may have been
more novel to the participants as the pleasant/unpleasant odour. A further hypothesis is
that the pleasant and unpleasant odours might affect moral choice only when presented in
supra-threshold concentrations.
This lack of an effect of the pleasant/unpleasant odour on moral choice seems to be in
contrast with previous literature on induced pleasant (Eskine, Kacinik, & Prinz, 2011;
22 Perception 0(0)

Pasto
¨tter et al., 2013; Valdesolo & DeSteno, 2006) and unpleasant affective states and moral
judgment (Cameron, Payne, & Doris, 2013; Case, Oaten, & Stevenson, 2012; Cheng, Ottati,
& Price, 2013; Eskine et al., 2011; Harle
´& Sanfey, 2010; Horberg, Oveis, Keltner, & Cohen,
2009; Johnson, Cheung, & Donnellan, 2014; Moretti & Di Pellegrino, 2010; Ong et al., 2014;
Schnall et al., 2008; Seidel & Prinz, 2013; Ugazio et al., 2012).
However, some considerations are in order. First, previous studies were focused on moral
judgment, while in our study participants were asked to make moral decisions. Even though
these two labels have been used as synonyms in the past, it has recently been demonstrated
that moral judgment and moral decision-making are two distinct psychological constructs,
both referring to the system of moral norms (Sood & Forehand, 2005; Tassy, Oullier,
Mancini, & Wicker, 2013). While in moral judgment tasks participants are asked to
evaluate the appropriateness or the permissibility of certain actions, in moral decision-
making tasks they are the main characters of a dilemma and, as such, they are responsible
for the action chosen and for its moral consequences. So, it might be possible that the
physical disgust amplification phenomenon theorized for moral judgment, actually it is not
valid or not strong enough in the context of moral decision-making.
A third possibility is that the limitations of the present work might have obscured some
effects. Although participants were carefully selected, the sample size is rather limited. This
may have particularly affected highly variable measures such as IHR, whose variability is
intrinsically important (Palomba et al., 2000). Second, in Study 2 only two odours were
tested, suggesting that conclusions should be generalized with caution, while waiting for
future studies to investigate the effects of wider samples of pleasant and unpleasant odours
on moral choices. Third, dilemmas were designed in a way that the affirmative response was
always the utilitarian option. This could have introduced a bias toward one option. Future
studies should counterbalance the direction of the question, whenever possible.
In conclusion, our study points out that olfactory stimuli affect the processes underlying
moral decisions by incrementing deontological choices and that this effect goes beyond the
ability of the odour to induce disgust.
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) received no financial support for the research, authorship, and/or publication of this
article.
References
Alaoui-Ismaı
¨li, O., Robin, O., Rada, H., Dittmar, A., & Vernet-Maury, E. (1997). Basic emotions
evoked by odorants: Comparison between autonomic responses and self-evaluation. Physiology &
Behavior,62, 713–720.
Alaoui-Ismaili, O., Vernet-Maury, E., Dittmar, A., Delhomme, G., & Chanel, J. (1997). Odor hedonics:
Connection with emotional response estimated by autonomic parameters. Chemical Senses,22,
237–248.
Anderson, A., Christoff, K., Stappen, I., Panitz, D., Ghahremani, D., Glover, G., ...Sobel, N. (2003).
Dissociated neural representations of intensity and valence in human olfaction. Nature Neuroscience,
6, 196–202.
Cecchetto et al. 23

Baron-Cohen, S., Wheelwright, S., Skinner, R., Martin, J., & Clubley, E. (2001). The autism-spectrum
quotient (AQ): Evidence from asperger syndrome/high-functioning autism, malesand females,
scientists and mathematicians. Journal of Autism and Developmental Disorders,31, 5–17.
Bensafi, M., Rouby, C., Farget, V., Bertrand, B., Vigouroux, M., & Holley, A. (2002). Autonomic
nervous system responses to odours: The role of pleasantness and arousal. Chemical Senses,27,
703–709.
Bermond, B., & Oosterveld, P. (1994). Bermond-Vorst Alexithymia Questionnaire; construction,
reliability, validity and uni-dimensionality: Internal Report. Amsterdam: Department of
Psychological Methods, Faculty of Psychology, University of Amsterdam.
Boesveldt, S., Lindau, S. T., McClintock, M. K., Hummel, T., & Lundstro
¨m, J. N. (2011). Gustatory
and olfactory dysfunction in older adults: A national probability study. Rhinology,49, 324.
Bolker, B. M., Brooks, M. E., Clark, C. J., Geange, S. W., Poulsen, J. R., Stevens, M. H. H., & White,
J. S. S. (2009). Generalized linear mixed models: A practical guide for ecology and evolution. Trends
in Ecology & Evolution,24, 127–135.
Bonini, N., Hadjichristidis, C., Mazzocco, K., Dematte
`, M. L., Zampini, M., Sbarbati, A., & Magon, S.
(2011). Pecunia olet: The role of incidental disgust in the ultimatum game. Emotion,11, 965–969.
Borg, J. S., Hynes, C., Van Horn, J., Grafton, S., & Sinnott-Armstrong, W. (2006). Consequences,
action, and intention as factors in moral judgments: An fMRI investigation. Journal of Cognitive
Neuroscience,18, 803–817.
Boucsein, W. (2012). Electrodermal activity. New York: Springer Science & Business Media.
Bradley, M. M. (2009). Natural selective attention: Orienting and emotion. Psychophysiology,46,
1–11.
Bradley, M. M., Codispoti, M., Cuthbert, B. N., & Lang, P. J. (2001). Emotion and motivation I:
Defensive and appetitive reactions in picture processing. Emotion,1, 276.
Bradley, M. M., Greenwald, M. K., & Hamm, A. O. (1993). Affective picture processing. The Structure
of Emotion,25, 48–65.
Buon, M., Dupoux, E., Jacob, P., Chaste, P., Leboyer, M., & Zalla, T. (2013). The role of causal and
intentional judgments in moral reasoning in individuals with high functioning autism. Journal of
Autism and Developmental Disorders,43, 458–470.
Cameron, C. D., Payne, B. K., & Doris, J. M. (2013). Morality in high definition: Emotion
differentiation calibrates the influence of incidental disgust on moral judgments. Journal of
Experimental Social Psychology,49, 719–725.
Carmichael, S. T., Clugnet, M. C., & Price, J. L. (1994). Central olfactory connections in the macaque
monkey. Journal of Comparative Neurology,346, 403–434.
Case, T. I., Oaten, M. J., & Stevenson, R. J. (2012). Disgust and moral judgment. Emotions,
Imagination and Moral Reasoning (pp. 191–214). New York, NY: Psychology Press.
Cecchetto, C., Korb, S., Rumiati, R., & Aiello, M. (Under review). Emotional reactions in moral
decision-making are influenced by empathy and alexithymia.
Cecchetto, C., Rumiati, R., & Parma, V. (Submitted). Promoting cross-culture research on moral
decision-making through the use of standardized, culturally-equivalent dilemmas: The validation
of the 4CONFiDe set.
Chapman, H., Kim, D., Susskind, J., & Anderson, A. (2009). In bad taste: Evidence for the oral origins
of moral disgust. Science (New York, NY),323, 1222–1226.
Chebat, J.-C., & Michon, R. (2003). Impact of ambient odors on mall shoppers’ emotions, cognition,
and spending. Journal of Business Research,56, 529–539.
Chen, D., & Haviland-Jones, J. (1999). Rapid mood change and human odors. Physiology & Behavior,
68, 241–250.
Cheng, J. S., Ottati, V. C., & Price, E. D. (2013). The arousal model of moral condemnation. Journal of
Experimental Social Psychology,49, 1012–1018.
Choe, S. Y., & Min, K.-H. (2011). Who makes utilitarian judgments? The influences of emotions on
utilitarian judgments. Judgment and Decision Making,6, 580–592.
Christensen, J. F., Flexas, A., Calabrese, M., Gut, N. K., & Gomila, A. (2014). Moral judgment
reloaded: A moral dilemma validation study. Emotion Science,5, 607.
24 Perception 0(0)

Cushman, F., Young, L., & Hauser, M. (2006). The role of conscious reasoning and intuition in moral
judgment: Testing three principles of harm. Psychological Science,17, 1082–1089.
Crepaldi, D., Che, W. C., Su, I., & Luzzatti, C. (2012). Lexical-semantic variables affecting picture and
word naming in Chinese: A mixed logit model study in aphasia. Behavioural Neurology,25, 165–184.
de Araujo, I. E., Rolls, E. T., Velazco, M. I., Margot, C., & Cayeux, I. (2005). Cognitive modulation of
olfactory processing. Neuron,46, 671–679.
Doty, R. L. (1991). Psychophysical measurement of odor perception in humans. In D. G. Laing, R.
L. Doty, & W. Breipohl (Eds.), The human sense of smell (pp. 95–134). Berlin: Springer.
Eskine, K., Kacinik, N., & Prinz, J. (2011). A bad taste in the mouth: Gustatory disgust influences
moral judgment. Psychological Science,22, 295–299.
Faraway, J. J. (2005). Extending the linear model with R: Generalized linear, mixed effects and
nonparametric regression models. Boca Raton, FL: CRC Press.
Figner, B., & Murphy, R. O. (2011). Using skin conductance in judgment and decision making research.
In A. Kuehberger, & R. Ranyard (Eds.), A handbook of process tracing methods for decision research
(pp. 163–184). New York, NY: Psychology Press.
Foot, P. (1967). Theories of ethics. New York, NY: Oxford University Press.
Frith, C. D., & Allen, H. A. (1983). The skin conductance orienting response as an index of attention.
Biological Psychology,17, 27–39.
Greene, J., Nystrom, L., Engell, A., Darley, J., & Cohen, J. (2004). The neural bases of cognitive
conflict and control in moral judgment. Neuron,44, 389–400.
Greene, J. D. (2009). Dual-process morality and the personal/impersonal distinction: A reply to
McGuire, Langdon, Coltheart, and Mackenzie. Journal of Experimental Social Psychology,45,
581–584.
Greene, J. D., Cushman, F. A., Stewart, L. E., Lowenberg, K., Nystrom, L. E., & Cohen, J. D. (2009).
Pushing moral buttons: The interaction between personal force and intention in moral judgment.
Cognition,111, 364–371.
Greene, J. D., Morelli, S. A., Lowenberg, K., Nystrom, L. E., & Cohen, J. D. (2008). Cognitive load
selectively interferes with utilitarian moral judgment. Cognition,107, 1144–1154.
Harle
´, K. M., & Sanfey, A. G. (2010). Effects of approach and withdrawal motivation on interactive
economic decisions. Cognition and Emotion,24, 1456–1465.
Hauser, M., Cushman, F., Young, L., Kang-Xing Jin, R., & Mikhail, J. (2007). A dissociation between
moral judgments and justifications. Mind & Language,22, 1–21.
Herz, R. S. (2006). I know what I like: Understanding odor preferences. In J. Drobnick (Ed.), The smell
culture reader (pp. 190–203). New York, NY: Berg.
Herz, R. S., & von Clef, J. (2001). The influence of verbal labeling on the perception of odors: Evidence
for olfactory illusions? Perception,30, 381–391.
Hoaglin, D. C., Iglewicz, B., & Tukey, J. W. (1986). Performance of some resistant rules for outlier
labeling. Journal of the American Statistical Association,81, 991–999.
Horberg, E., Oveis, C., Keltner, D., & Cohen, A. B. (2009). Disgust and the moralization of purity.
Journal of Personality and Social Psychology,97, 963.
Hummel, T., Sekinger, B., Wolf, S., Pauli, E., & Kobal, G. (1997). ‘Sniffin’sticks’: Olfactory
performance assessed by the combined testing of odor identification, odor discrimination and
olfactory threshold. Chemical Senses,22, 39–52.
Inbar, Y., Pizarro, D. A., & Bloom, P. (2012). Disgusting smells cause decreased liking of gay men.
Emotion,12, 23.
Johnson, D. J., Cheung, F., & Donnellan, M. B. (2014). Does cleanliness influence moral judgments? A
direct replication of Schnall, Benton, and Harvey (2008). Social Psychology,45, 209.
Kemp, S. E., & Gilbert, A. N. (1997). Odor intensity and color lightness are correlated sensory
dimensions. The American Journal of Psychology,110, 35.
Kirk-Smith, M., Van Toller, C., & Dodd, G. (1983). Unconscious odour conditioning in human
subjects. Biological Psychology,17, 221–231.
Koenigs, M., Young, L., Adolphs, R., Tranel, D., Cushman, F., Hauser, M., & Damasio, A. (2007).
Damage to the prefrontal cortex increases utilitarian moral judgements. Nature,446, 908–911.
Cecchetto et al. 25

Koven, N. (2011). Specificity of meta-emotion effects on moral decision-making. Emotion,11,
1255–1261.
Landy, J. F., & Goodwin, G. P. (2015). Does incidental disgust amplify moral judgment? A meta-
analytic review of experimental evidence. Perspectives on Psychological Science,10, 518–536.
Lang, P. J. (1995). The emotion probe: Studies of motivation and attention. American Psychologist,50,
372.
Li, W., Moallem, I., Paller, K., & Gottfried, J. (2007). Subliminal smells can guide social preferences.
Psychological Science,18, 1044–1049.
Lorig, T. S., Herman, K. B., Schwartz, G. E., & Cain, W. S. (1990). EEG activity during administration
of low-concentration odors. Bulletin of the Psychonomic Society,28, 405–408.
Lo
¨tsch, J., Ahne, G., Kunder, J., Kobal, G., & Hummel, T. (1998). Factors affecting pain intensity in a
pain model based upon tonic intranasal stimulation in humans. Inflammation Research,47, 446–450.
Lotto, L., Manfrinati, A., & Sarlo, M. (2014). A new set of moral dilemmas: Norms for moral
acceptability, decision times, and emotional salience. Journal of Behavioral Decision Making,27,
57–65.
Lundstro
¨m, J., Gordon, A., Alden, E., Boesveldt, S., & Albrecht, J. (2010). Methods for building an
inexpensive computer-controlled olfactometer for temporally-precise experiments. International
Journal of Psychophysiology: Official Journal of the International Organization of
Psychophysiology,78, 179–189.
Lundstro
¨m, J. N., Boyle, J. A., & Jones-Gotman, M. (2008). Body position-dependent shift in odor
percept present only for perithreshold odors. Chemical Senses,33, 23–33.
Lundstro
¨m, J. N., McClintock, M. K., & Olsson, M. J. (2006). Effects of reproductive state on olfactory
sensitivity suggest odor specificity. Biological Psychology,71, 244–247.
Mainland, J. D., Lundstro
¨m, J. N., Reisert, J., & Lowe, G. (2014). From molecule to mind: An
integrative perspective on odor intensity. Trends in Neurosciences,37, 443–454.
McAlonan, K., Cavanaugh, J., & Wurtz, R. H. (2008). Guarding the gateway to cortex with attention
in visual thalamus. Nature,456, 391–394.
McLean, R. A., Sanders, W. L., & Stroup, W. W. (1991). A unified approach to mixed linear models.
The American Statistician,45, 54–64.
Mendez, M. F., Anderson, E., & Shapira, J. S. (2005). An investigation of moral judgement in
frontotemporal dementia. Cognitive and Behavioral Neurology,18, 193–197.
Moore, A. B., Clark, B. A., & Kane, M. J. (2008). Who shalt not kill? Individual differences in working
memory capacity, executive control, and moral judgment. Psychological Science,19, 549–557.
Moore, A. B., Lee, N. L., Clark, B. A., & Conway, A. R. (2011). In defense of the personal/impersonal
distinction in moral psychology research: Cross-cultural validation of the dual process model of
moral judgment. Judgment and Decision Making,6, 186.
Moran, J. M., Young, L. L., Saxe, R., Lee, S. M., O’Young, D., Mavros, P. L., & Gabrieli, J. D. (2011).
Impaired theory of mind for moral judgment in high-functioning autism. Proceedings of the National
Academy of Sciences,108, 2688–2692.
Moretti, L., & Di Pellegrino, G. (2010). Disgust selectively modulates reciprocal fairness in economic
interactions. Emotion,10, 169.
Moretto, G., La
`davas, E., Mattioli, F., & Di Pellegrino, G. (2009). A psychophysiological investigation
of moral judgment after ventromedial prefrontal damage. Journal of Cognitive Neuroscience,22,
1888–1899.
Ong, H. H., O’Dhaniel, A., Kwok, K., & Lim, J. (2014). Moral judgment modulation by disgust is bi-
directionally moderated by individual sensitivity. Frontiers in psychology,5, 194.
Palomba, D., Sarlo, M., Angrilli, A., Mini, A., & Stegagno, L. (2000). Cardiac responses associated
with affective processing of unpleasant film stimuli. International Journal of Psychophysiology,36,
45–57.
Palomba, D., Angrilli, A., & Mini, A. (1997). Visual evoked potentials, heart rate responses and
memory to emotional pictorial stimuli. International Journal of Psychophysiology,27, 55–67.
Parma, V., Bulgheroni, M., Scaravilli, T., Tirindelli, R., & Castiello, U. (2013). Implicit olfactory
processing attenuates motor disturbances in idiopathic Parkinson’s disease. Cortex,49, 1241–1251.
26 Perception 0(0)

Parma, V., Ferraro, S., Miller, S. S., A
˚hs, F., & Lundstro
¨m, J. N. (2015). Enhancement of odor
sensitivity following repeated odor and visual fear conditioning. Chemical Senses,40, 497–506.
Parma, V., Straulino, E., Zanatto, D., Cantagallo, A., Tirindelli, R., & Castiello, U. (2012). Implicit
olfactory abilities in traumatic brain injured patients. Journal of Clinical and Experimental
Neuropsychology,34, 977–988.
Pasto
¨tter, B., Gleixner, S., Neuhauser, T., & Ba
¨uml, K.-H. T. (2013). To push or not to push? Affective
influences on moral judgment depend on decision frame. Cognition,126, 373–377.
Patil, I., Melsbach, J., Hennig-Fast, K., & Silani, G. (2016). Divergent roles of autistic and alexithymic
traits in utilitarian moral judgments in adults with autism. Scientific Reports,6, 23637.
Patil, I., & Silani, G. (2014a). Alexithymia increases moral acceptability of accidental harms. Journal of
Cognitive Psychology,26, 597–614.
Patil, I., & Silani, G. (2014b). Reduced empathic concern leads to utilitarian moral judgments in trait
alexithymia. Frontiers in Psychology,5, 501.
Plailly, J., Howard, J. D., Gitelman, D. R., & Gottfried, J. A. (2008). Attention to odor modulates
thalamocortical connectivity in the human brain. The Journal of Neuroscience,28, 5257–5267.
Powell, M. J. (2009). The BOBYQA algorithm for bound constrained optimization without derivatives
(Cambridge NA Report NA2009/06). Cambridge, England: University of Cambridge.
Ree, M. J., French, D., MacLeod, C., & Locke, V. (2008). Distinguishing cognitive and somatic
dimensions of state and trait anxiety: Development and validation of the State-Trait Inventory
for Cognitive and Somatic Anxiety (STICSA). Behavioural and Cognitive Psychotherapy,36,
313–332.
Roth, W. T., Dawson, M. E., & Filion, D. L. (2012). Publication recommendations for electrodermal
measurements. Psychophysiology,49, 1017–1034.
Rozin, P., Haidt, J., & Fincher, K. (2009). From oral to moral. Science,323, 1179–1180.
Rozin, P., Haidt, J., McCauley, C., Dunlop, L., & Ashmore, M. (1999). Individual differences in disgust
sensitivity: Comparisons and evaluations of paper-and-pencil versus behavioral measures. Journal of
Research in Personality,33, 330–351.
Sarlo, M., Lotto, L., Manfrinati, A., Rumiati, R., Gallicchio, G., & Palomba, D. (2012). Temporal
dynamics of cognitive-emotional interplay in moral decision-making. Journal of Cognitive
Neuroscience,24, 1018–1029.
Schnall, S., Haidt, J., Clore, G., & Jordan, A. (2008). Disgust as embodied moral judgment. Personality
& Social Psychology Bulletin,34, 1096–1109.
Seidel, A., & Prinz, J. (2013). Sound morality: Irritating and icky noises amplify judgments in divergent
moral domains. Cognition,127, 1–5.
Seubert, J., Gregory, K. M., Chamberland, J., Dessirier, J.-M., & Lundstro
¨m, J. N. (2014). Odor
valence linearly modulates attractiveness, but not age assessment, of invariant facial features in a
memory-based rating task. PloS One,9, e98347.
Seubert, J., Kellermann, T., Loughead, J., Boers, F., Brensinger, C., Schneider, F., & Habel, U. (2010).
Processing of disgusted faces is facilitated by odor primes: A functional MRI study. NeuroImage,53,
746–756.
Smeets, M. A., Schifferstein, H. N., Boelema, S. R., & Lensvelt-Mulders, G. (2008). The Odor
Awareness Scale: A new scale for measuring positive and negative odor awareness. Chemical
Senses,33, 725–734.
Sood, S., & Forehand, M. (2005). On self-referencing differences in judgment and choice.
Organizational Behavior and Human Decision Processes,98, 144–154.
Soudry, Y., Lemogne, C., Malinvaud, D., Consoli, S. M., & Bonfils, P. (2011). Olfactory system and
emotion: Common substrates. European Annals of Otorhinolaryngology, Head and Neck Diseases,
128, 18–23.
Spielberger, C., Gorsuch, R., & Lushene, R. (1968). Self-Evaluation Questionnaire: STAI Form X-1.
Retrieved from http://hdl.handle.net/10477/1878. Accessed 31 December 2016
Starcke, K., Polzer, C., Wolf, O., & Brand, M. (2011). Does stress alter everyday moral decision-
making? Psychoneuroendocrinology,36, 210–219.
Cecchetto et al. 27

Starcke, K., Wolf, O. T., Markowitsch, H. J., & Brand, M. (2008). Anticipatory stress influences
decision making under explicit risk conditions. Behavioral Neuroscience,122, 1352.
Stevenson, R. J., Boakes, R. A., & Prescott, J. (1998). Changes in odor sweetness resulting from implicit
learning of a simultaneous odor-sweetness association: An example of learned synesthesia. Learning
and Motivation,29, 113–132.
Tassy, S., Oullier, O., Mancini, J., & Wicker, B. (2013). Discrepancies between judgment and choice of
action in moral dilemmas. Frontiers in psychology,4, 250.
Thomson, J. (1976). Aristotle: Ethics. Harmondsworth, England: Penguin.
Ugazio, G., Lamm, C., & Singer, T. (2012). The role of emotions for moral judgments depends on the
type of emotion and moral scenario. Emotion,12, 579.
Valdesolo, P., & DeSteno, D. (2006). Manipulations of emotional context shape moral judgment.
Psychological Science,17, 476–477.
Van Dillen, L. F., van der Wal, R. C., & van den Bos, K. (2012). On the role of attention and emotion
in morality attentional control modulates unrelated disgust in moral judgments. Personality and
Social Psychology Bulletin,38, 1222–1231.
Wehling, E. I., Wollschlaeger, D., Nordin, S., & Lundervold, A. J. (2016). Longitudinal changes in odor
identification performance and neuropsychological measures in aging individuals. Neuropsychology,
30, 87.
Winston, J. S., Gottfried, J. A., Kilner, J. M., & Dolan, R. J. (2005). Integrated neural representations
of odor intensity and affective valence in human amygdala. The Journal of Neuroscience,25,
8903–8907.
Yeshurun, Y., & Sobel, N. (2010). An odor is not worth a thousand words: From multidimensional
odors to unidimensional odor objects. Annual Review of Psychology,61, 219–241.
Youssef, F., Dookeeram, K., Basdeo, V., Francis, E., Doman, M., Mamed, D., ...Legall, G. (2012a).
Stress alters personal moral decision making. Psychoneuroendocrinology,37, 491–498.
Youssef, F. F., Dookeeram, K., Basdeo, V., Francis, E., Doman, M., Mamed, D., ...Ditshotlo, P.
(2012b). Stress alters personal moral decision making. Psychoneuroendocrinology,37, 491–498.
Zald, D. H., & Pardo, J. V. (1997). Emotion, olfaction, and the human amygdala: Amygdala activation
during aversive olfactory stimulation. Proceedings of the National Academy of Sciences,94,
4119–4124.
Zucco, G. M., Paolini, M., & Schaal, B. (2009). Unconscious odour conditioning 25 years later:
Revisiting and extending ‘Kirk-Smith, Van Toller and Dodd’. Learning and Motivation,40, 364–375.
28 Perception 0(0)