Moral Preferences in Helping Dilemmas Expressed by Matching and Forced Choice

Judgment and Decision Making, Vol. 15, No. 4, July 2020, pp.
452–475
Moral preferences in helping dilemmas expressed by matching and

forced choice
Arvid Erlandsson∗ Amanda Lindkvist† Kajsa Lundqvist‡ Per A. Andersson§.
Stephan Dickert¶ Paul Slovic‖ Daniel Västfjäll∗∗
Abstract
This paper asks whether moral preferences in eight medical dilemmas change as a function of how preferences are expressed,
and how people choose when they are faced with two equally attractive help projects. In two large-scale studies, participants
first read dilemmas where they “matched” two suggested helping projects (which varied on a single attribute) so that they
became equally attractive. They did this by filling in a missing number (e.g., how many male patients must Project M save
in order to be equally attractive as Project F which can save 100 female patients). Later, the same participants were asked
to choose between the two equally attractive projects. We found robust evidence that people do not choose randomly, but
instead tend to choose projects that help female (vs. male), children (vs. adult), innocent (vs. non-innocent), ingroup (vs.
outgroup) and existing (vs. future) patients, and imply no (vs. some) risk of a harmful side-effect, even when these projects
have been matched as equally attractive as, and save fewer patients than the contrasting project. We also found that some moral
preferences are hidden when expressed with matching but apparent when expressed with forced choice. For example, 88–95%
of the participants expressed that female and male patients are equally valuable when doing the matching task, but over 80%
of them helped female patients in the choice task.
Keywords: moral cognition, expressing moral preferences, helping dilemmas, person trade-offs, prominence effect, medical
decision making
1 Introduction not choose, Sophie saves her son (which means giving away
her daughter to the guards).
In the movie Sophie’s Choice (Pakula, 1979), the main char- This is first and foremost a dreadful moral dilemma,
acter Sophie is forced by Nazi guards to choose between but it also raises questions about Sophie’s preferences re-
saving either her daughter’s or her son’s life. At first So- garding her children. Did Sophie really value both her
phie keeps repeating that she cannot choose, implying that children equally or was it something that made her more
her children are exactly equally valuable to her. Still, when prone to save her son when having to make a choice? One
reminded that both of her children will be shot if she does could argue that when forced to choose between two exactly
equally (un)attractive alternatives, Sophie could (and per-
haps should) choose randomly, but this does not seem to be
This research was financed by a generous grant from the Swedish Science the way most people typically make decisions in the moral
Council (grant number: 2017-01827). We wish to thank the following realm. On the contrary, although coin-tosses are perceived as
persons for help during data collection: Anja Grim, Katrine Svane Bech fair, they are also perceived as inappropriate when resolving
Nielsen, Annica Nilsson, Fanny Plaza, Jacob Andersson, Julia Denkiewicz,
Kalle Kallio Strand, Laura Schmitz, Malin Jakobsson Månsson, Justyna a life-and-death dilemma (Keren & Teigen, 2010).
Svensson and Agnes Andersson. The results from Study 1 was presented at In this paper we ask: (1) whether people’s preferences in
the Society of Judgment and Decision Making (SJDM) conference in New a series of moral dilemmas differ when they are expressed
Orleans in November 2018.
Copyright: © 2020. The authors license this article under the terms of
in two different ways (matching and forced choice); (2) how
the Creative Commons Attribution 3.0 License. people choose when they are faced with two equally attractive
∗ Linköping University, Department of Behavioral Sciences and alternatives.
Learning. Campus Valla, SE-58183, Linköping, Sweden. E-mail:
arvid.erlandsson@liu.se. https://orcid.org/0000-0001-7875-269X.
† Department of Psychology, Göteborg University.
1.1 Helping dilemmas
‡ Department of Psychology, Lund University.
§ Linköping University, Department of Behavioral Sciences and Learn- This paper focuses on a specific form of moral dilemma —
ing helping dilemmas (or person trade-offs; Ubel, Richardson
¶ Queen Mary University London and University of Klagenfurt.
‖ Decision Research, Eugene, Oregon. & Baron, 2002), which occur when a person learns about
∗∗ Linköping University, Department of Behavioral Sciences and Learn- two or more need situations where resources are limited,
ing and Decision Research, Eugene, Oregon. and it is impossible to help everyone in need. In these
452
Judgment and Decision Making, Vol. 15, No. 4, July 2020 Preferences in helping dilemmas 453
situations, people must decide how to allocate help, and in A in order to make it equally attractive as Project B. If you
its extreme form this could mean choosing who will live respond “100 Argentinians” this implies that you have no
and who will die (e.g., which patient will be connected to preference between helping Argentinians or Brazilians (as
the only available respirator). Helping dilemmas are not helping 100 Argentinians is equally good as helping 100
limited to extraordinary situations or to medical decisions, Brazilians according to you). If you respond “150 Argen-
but happen regularly to most of us. For example, we are tinians” this implies that you have a preference for Brazilians
responding to a helping dilemma every time we choose to (because you need a higher number of Argentinians to make
donate to one charity organization but refrain from donating the projects equally good).
to another (Breeze, 2013; Neumayr & Handy, 2019. Also, for
policy makers, politicians, and people working with foreign 1.2.2 Forced choice
aid, choosing how to allocate resources between different
beneficiaries is a vital part of the job (Alesina & Dollar, The most straightforward way to express a moral preference
2000; Bucknall, 2003). in a helping dilemma is arguably to choose between Project
In our studies, we asked participants to imagine that they A and Project B. Importantly, unlike rating, allocation and
have a job that involves evaluating suggested medical projects matching, it is impossible to express indifference between
designed to help patients. Participants were faced with hypo- the alternatives when forced to make a choice. Nevertheless,
thetical dilemmas each consisting of two suggested helping people who have no preference between Argentina and Brazil
projects presented next to each other. In each dilemma, par- could choose randomly, for example by throwing a die or
ticipants communicated whether they believed one of the flipping a coin (Keren & Teigen, 2010; Dwenger, Kübler &
projects was preferable to the other. Weizsäcker, 2012). If people really did so, they would be
equally likely to end up choosing either of the projects (Shah,
Tsuchiya & Wailoo, 2014).1
1.2 Expressing moral preferences in helping
dilemmas 1.3 The Prominence effect: Choosing between
How a person responds to a helping dilemma often reflects two equally attractive helping projects
her moral preference. For example, a foreign aid official who
Matching and forced choice are at the core of a decision
says YES to most suggested helping projects in Argentina
making phenomenon called the Prominence Effect: when
but NO to most (in other ways similar or identical) help-
faced with a trade-off between two or more alternatives,
ing projects in Brazil or Colombia, communicates a moral
people assign a higher weight to the more important (promi-
preference for helping Argentineans.
nent) attribute when making choices than when rating or
Moral preferences in helping dilemmas can be expressed
matching the alternatives. The prominence effect paradigm
in different ways. People can rate the subjective value of each
began with an unexpected finding by Paul Slovic and led
of the helping projects (attractiveness-rating) or distribute
to two influential papers (Slovic, 1975; Tversky, Sattath &
resources between the projects (budget-allocation), but in
Slovic, 1988) where a foundational tenet of decision mak-
this paper we ask participants to express moral preferences
ing – the existence of stable values and preferences – was
by matching and by forced choice. We focus on these ways
questioned. Both these papers demonstrated a systematic in-
because when expressed by the same person, they allow
consistency between preferences expressed using a matching
us to ask how people choose when faced with two equally
task and preferences expressed using a choice task. This in-
attractive alternatives.
consistency was explained by an overweighting of the more
justifiable attributes when making choices and labeled the
1.2.1 Matching Prominence Effect.
In economics, indifference curves are often used to demon- For example, in Slovic (1975), participants first matched
strate the value at which two goods give a consumer equal e.g., pairs of baseball players that differed on two attributes,
satisfaction and utility. Using the same logic, moral pref- so that the two players would be equally valuable for their
erences can be expressed by asking participants to “provide team. This was done by writing how many home-runs Player
their indifference point”, or to equate or “match” helping 1, with a batting average of .287, must hit in a season in or-
der to be equally valuable as Player 2, who had a batting
projects so that they become exactly equally attractive (e.g.,
average of .273 and 26 home runs. In a later session, the
Ubel et al., 2002). To exemplify, imagine that you hear
about a suggested helping Project A that can help Argen- 1A choice task could include a “prefer not to choose” option which makes
tinian patients, and an equally costly and otherwise identical it possible to express no preference. In these situations it is however not
always clear if such a response implies that none of the two helping projects
Project B that can help Brazilian patients. You learn that
will be implemented (as in Sophie’s choice, see e.g., Gordon-Hecker et al.,
Project B will be able to help 100 Brazilians and you are 2017), or if some unspecified other will make the choice (either by their
asked how many Argentinians that must be helped in Project own judgement or randomly).
same participants choose between two players that they had presented in isolation (Hsee & Zhang, 2010). With that said,
matched to be equally valuable. The results showed that par- people are not perfectly number-maximizing even in joint
ticipants did not choose randomly but instead systematically evaluation because the number of people saved is not the
over-selected the player that was superior on the relatively only attribute that humans care about. Most people will,
more prominent attribute, which in this case was batting av- e.g., prefer a helping project that can save their own child
erage. To our knowledge, the prominence effect has not yet over a helping project that can save 2, 10 or even 100,000 un-
been documented in the moral domain. known children. In this paper, we are interested in situational
Participants in our studies saw two helping projects that attributes that can elicit number-overriding preferences in
differed on a single attribute (e.g., only ingroup or outgroup helping dilemmas.
patients can be treated) and had one missing piece of in-
formation on a scope-related attribute (e.g., the number of
1.5 Attributes hypothesized to elicit number-
outgroup patients possible to treat). First, participants had to
match the projects so that they became equally attractive by overriding preferences and to be promi-
filling in the missing piece of information. Later, the same nent in helping dilemmas
participants were forced to make a choice between the two Most studies on preferences in medical helping dilemmas
projects that they had rated as equally attractive. have focused on a single varying attribute (e.g., length or
quality of lives for patients, e.g., Shah, Tsuchiya & Wailoo,
1.4 Number-overriding preferences in helping 2014; Ubel et al., 2002; Nord & Johansen, 2014). We
dilemmas here adopt a much broader perspective and investigate eight
different attributes typically associated with increased help-
This paper focuses on number-overriding preferences in ing, each in a separate dilemma. Besides eliciting number-
helping dilemmas which, in our operationalization, occurs overriding in helping situations, we suspect that these eight
anytime a decision maker prefers a helping project that can attributes might also be prominent.
save fewer over a project that can save more individuals in
need.2 Judgments implying that some lives are valued more
1.5.1 Age of victims
than others, or choices that save fewer lives, are expressions
of a number-overriding moral preference. Please note that People are generally more motivated to help the young than
“number” here refers to the number of lives saved.3 the old (Goodwin & Landy, 2014. One reason for this is that
Neglect of numbers seems to be greater when assessed children are assumed to be more innocent and dependent
in separate evaluation (where participants see and respond than adults, and young children are rarely held responsible
to only one of the alternatives) than when assessed in joint for their own plight (Back & Lips, 1998). Another, more
evaluation helping dilemmas (Hsee & Rottenstreich, 2004; utilitarian, reason is that the anticipated remaining number
Kogut & Ritov, 2005). The main reason for this is that of quality-adjusted life years (QALY) is higher for every
the number of people that can be helped is much easier to saved child than for every saved adult (Goodwin & Landy,
evaluate when presented next to other numbers than when 2014; Bravo Vergel & Sculpher, 2008). Although there are
good consequentialist arguments for preferring to save fewer
2Number-overriding is related to “scope-insensitivity” which refers to
the human inability to adjust ones helping when the amount of good one children rather than more adults, this would still count as a
can do increases or decreases (Dickert & Slovic, 2011; Dickert et al., number-overriding preference using our operationalization.
2015). Although most people say that lives should be valuated according
to a linear function, they do not always act in line with their normative
beliefs (Dickert et al., 2015), and the number of lives one can save or the 1.5.2 Gender of victims
amount of suffering one can prevent for a specified cost, are often very
poor predictors of prosocial behavior (Slovic, 2007). In a dilemma context, Research indicates that women tend to receive more help
preferring to save 95 unknown children rather than 100 unknown children (or than men, and especially so when the helper is a man (e.g.,
expressing indifference) could be thought of as scope-insensitive, because Eagly & Crowley, 1986; Dufwenberg & Muren, 2006; We-
the options differ on only the saved lives-attribute. However, preferring to
save 95 children rather than 100 adults does not necessarily constitute scope- ber, Koehler & Schnauber-Stockmann, 2019. One reason for
insensitivity, because the options differ on two attributes. Nevertheless, this is that helping by males can be used to signal affluence
preferring to help 95 children illustrates that the contrasting attribute (here or kindness toward females (van Vugt & Iredale, 2013; Rai-
children vs. adults) overrides the number of lives-attribute (95 vs. 100). hani & Smith, 2015). Another reason seems to be rooted
3Number-maximizing (preferring the option that saves more lives) vs.
Number-overriding (preferring the option that saves fewer lives) in helping in gender-stereotypic perceptions of victim qualities, imply-
dilemmas can be linked to characteristically utilitarian vs. non-utilitarian ing that women are helped more than men because they are
responding. However, as utilitarians typically care not only about the num- seen as sensitive, kind, non-aggressive and ultimately help-
bers of lives, but also about, e.g., their remaining length and quality, we
less and in need of protection (e.g., Curry, Lee & Rodriguez,
argue that number-maximization and number-overriding are more accu-
rate descriptions. Number-overriding in our operationalization is thus not 2004). A preference is number-overriding if it favors helping
necessarily irrational or non-normative. fewer women rather than more men.
1.5.3 Innocence treme number-overriding meaning that people prefer to help

fewer existing victims rather than more future victims.
An attributional account of helping suggests that the per-
ceived causes of the misfortunes of others influence helping.
External causes (e.g., bad luck) increase the sense of victim 1.5.6 Avoiding sure death
innocence and helping, whereas causes that are seen as in- In helping dilemmas, people often assume that non-helping
ternal to and controllable by the person in need elicit anger implies a certain failure (e.g., death) whereas helping implies
which make helping less likely (Zagefka et al., 2011; Sea- a certain success (e.g. survival). This is an oversimplifica-
cat, Hirschman & Mickelson, 2007; Weiner, 1980). People tion. Instead, estimated outcomes could be better understood
express feeling less compassion, and has less neural activity by providing an estimated probability for success and fail-
in brain areas associated with emotion, when hearing about ure for those who are helped and those who are not helped,
victims who had a role in causing their own suffering (Fehse respectively. To illustrate, a medical doctor can estimate the
et al., 2015), and representations of victims that typically survival-chance to be 10% if a specific patient is not treated
increase empathic concern tend to backfire when perceived in a respirator. The doctor can further estimate that if treated
innocence is low (Kogut, 2011). Number-overriding occurs in a respirator, the survival-chance will increase to 40% for
when people prefer to help fewer “innocent” victims rather the patient. Being treated is no guarantee for survival, but it
than more victims that partially caused their own plight. increases the probability by 30%. We will use the term treat-
ment efficiency to illustrate this increase in survival chance
1.5.4 Group-membership for patients that are treated compared to patients that are not
treated.4
People are more likely to help members of their ingroup Research has, however, shown that it is not only the size
(groups that they identify as members of) than outgroup- of the treatment efficiency that matters but also where on the
members (Dovidio et al., 1997). This tendency has been scale (from 0% to 100%) the survival-chance increase oc-
referred to as intergroup bias, ingroup bias or parochialism curs. According to the prospect theory (Kahneman & Tver-
(Mackie & Smith, 1998; Baron, 2009) and might arise be- sky, 1979; see also Ritov, Baron & Hershey, 1993, and Zhang
cause of a dislike for the outgroup, liking towards the ingroup & Slovic, 2019), we are more sensitive to changes that occur
(Brewer, 1999), or because people believe that they have a closer to the end points and especially changes that involve
greater responsibility to help ingroup-members (Erlandsson, 0% (sure death for non-treated patients) and 100% (sure
Björklund & Bäckström, 2015). There exist many types of survival for treated patients). This could mean that people
ingroups, but in this study we focus on nationality, which prefer projects that can prevent sure deaths (e.g. 0%→40%)
is arguably one of the most salient naturally occurring but or guarantee sure survivals (e.g., 60%→100%) even when
still arbitrary ingroup/outgroup classifications, and mean- these projects are pitted against equally effective projects
ingful because most people identify themselves as citizens that can treat more patients but where the survival chances
of a country or cultural group (Baron, Ritov & Greene, 2013; if untreated/treated are located far from the end points (e.g.,
Levine & Thompson, 2004; Zagefka, Noor & Brown, 2013). 30%→70%). This would imply number-overriding.
Number-overriding occurs when people prefer to help fewer
ingroup-members rather than more outgroup-members.
1.5.7 Avoiding harmful side-effects
1.5.5 Existing (vs. future) lives Saving 20 people by diverting a runaway trolley onto an
empty side-track is the obviously moral thing to do, but
Intertemporal choices are decisions that involve a trade-off saving 50 people by diverting the trolley into a side-track
between costs and benefits occurring at different times and where one person is located seems more problematic despite
the discounted utility model predicts that utilities in the future the net number of lives saved being higher. (Thomson, 1985;
are discounted by their delay (Samuelson, 1937; Chapman Greene, 2008; Bauman et al., 2014). Humans tend to see
& Elstein, 1995; Bischoff & Hansen, 2016). This intertem- harm as a result of an action as worse than harm as a result
poral utility discounting is problematic when people make of an omission (Baron & Ritov, 2004), and we are aversive
decisions regarding themselves (should I benefit the existing to harmful behavior even if it is an unintended side-effect
self or the future self; e.g., retirement savings) but arguably
4It should also be mentioned that the possible outcomes for helped and
even more problematic when making decisions regarding non-helped persons are rarely dichotomous but rather comes on a sliding
others (should I help existing others or future others; e.g., scale (e.g., different degrees of life quality or happiness). The trade-off
Baron & Szymanska, 2011). Discounting may also be one between the length and the quality of lives is often central in health economy
research (e.g. Skedgel, Wailoo & Akehurst, 2015; Nord & Johansen, 2014).
of the major obstacles for combatting climate change as the
Still, in order to not manipulate everything at once, we opted to vary the
primary beneficiaries are future generations (Wade-Benzoni estimated likelihood of survival but not to differentiate between different
& Tost, 2009). Intertemporal discounting can lead to ex- levels of life quality in this study.
of an ultimately prosocial act (Anderson, 2003). Likewise, they have matched to be equally attractive. We could also
some people are hesitant to use vaccines or airbags that are find the opposite, namely dilemmas where participants ex-
overall highly beneficial but in rare cases can cause undesired press no moral preference in the matching task but still chose
harmful side-effects (Ritov & Baron, 1990). This aversion one of the two projects significantly more often in the choice
of incidental harm can lead to number-overriding, meaning task (i.e., choice-dependent number-overriding).
that people prefer less efficient helping projects over more
efficient helping projects that come with a small risk of a
harmful side-effect. 2 Study 1
1.5.8 Causal responsibility Study 1 consisted of two tasks (matching and choice) done
by the same participants but temporally separated by at least
People tend to make appraisals about why a need-situation a month.
occurred and consequently who (if anyone) has the causal
responsibility and is to be held accountable for the problem at
hand (Weiner, 1995). Just as perceived responsibility of the 2.1 Method
victim can decrease helping, so too can perceived personal
responsibility increase it. For example, if a potential helper 2.1.1 The matching task
believes that she caused a specific problem, she is more likely Participants. One thousand and seven Swedish partici-
to perceive herself as having a personal responsibility to help pants (401 male, 596 female, 10 unclassified gender, M age
victims suffering because of that problem, and to anticipate = 24.30 years, SD = 6.75) were recruited by 13 research
feeling guilty if she would not help (Erlandsson, Jungstrand assistants trained to explain the matching task but unaware
& Västfjäll, 2016). Number-overriding occurs when people of the research hypotheses. Participants were approached
prefer projects that can help fewer people who are suffering individually or in small groups at two university campuses
because of a problem they caused over projects that can help during early spring 2017, and (if they agreed to participate)
more people who are suffering from a problem they did not handed a paper-and-pen questionnaire. After completing the
cause. questionnaire, participants received a scratch lottery ticket.
1.6 The current studies

Design. We manipulated the questionnaire using a 2×2
In two large-scale studies, we investigate: (1) Which of the design. Half of the participants were randomly assigned to
eight attributes elicit number-overriding moral preferences fill in a blank box on the number of treated patients attribute
when preferences are expressed with matching and choice (henceforth Number-condition) whereas the other half filled
tasks? (2) Which of the attributes are prominent (relative to in a blank box on the treatment-efficiency attribute (average
the number-attributes)? Recall that the prominent attribute survival-chance for a treated patient; henceforth Efficiency-
systematically drives preferences among equally attractive condition). Further, half of the participants were randomly
options in the choice tasks. If two equally attractive help assigned to fill in the blank on the project presented first
projects are chosen equally often, then there is no prominent in each dilemma (henceforth First-condition), whereas the
attribute in that dilemma. other half filled in the blank on the project presented second
in each dilemma (henceforth Second-condition). See Tables
1.6.1 Hypothesis 1 and 2 and the online supplementary material (OSM 1 and
2) for illustrations of the experimental manipulations in both
Our initial hypothesis for all eight dilemmas is that most studies.
people will express number-overriding preferences in the
matching task (albeit to different degrees for different dilem-
mas), and that they, in the choice-task, will chose the project Procedure and material. Participants were asked to imag-
that is superior on the presumed prominent attribute, even ine that they had a job where they must make decisions about
when that project helps fewer people and the two projects how to distribute resources between medical projects, and
have been matched to be equally attractive. If these results learned that their task, in each dilemma, was to match the
materialize, we have initial support for the notion that the two helping projects so that they became exactly equally at-
prominence effect underlies number-overriding preferences tractive, by writing a number in the blank box (always shaded
in helping dilemmas. in green). “Exactly equally attractive” was explicitly defined
Several other result patterns are possible. For example, as “You would think it was equally good to implement Project
we might find dilemmas where participants express number- 1 as Project 2”.
overriding preferences in the matching task but chose ap- After reading the instruction page, each participant read
proximately 50–50 when presented with two projects that and responded to one version of the test dilemma (see Table
Table 1: The test dilemma in all four conditions in the Study 1 matching task. See OSM 1 and 2 for the exact layout of all
dilemmas in both studies. (Continued on next page.)
Number&First Project 1 Project 2
Who are affected by the disease? Adults Adults
Project cost? 400,000SEK 400,000SEK
In which country will the project be Sweden Sweden
implemented?
Number of ill patients currently in About 1000 patients currently About 1000 patients currently
need of treatment? need treatment need treatment
What is the average chance of 30% chance to survive for each 30% chance to survive for each
surviving the disease for an ill patient patient that is NOT treated patient that is NOT treated
that is NOT treated?
surviving the disease for an ill patient patient that is treated patient that is treated
that is treated?
Number of patients that will be ____ ill patients will be 100 ill patients will be
treated if the project is implemented? treated if the project is treated if the project is
implemented implemented
Number&Second Project 1 Project 2

implemented?
surviving the disease for an ill patient patient that is treated patient that is treated
that is treated?
Number of patients that will be 100 ill patients will be ____ ill patients will be
1 and OSM 1). The test dilemma was included to let par- than 70% in the Efficiency&Second condition) also indicated
ticipants familiarize themselves with the layout and to test non-comprehension.5
their comprehension of the matching task. Participants in After completing the test dilemma, each participant com-
the Number&First condition filled in how many patients that pleted the remainder of the questionnaire individually. The
must be treated in Project 1 (which had a 40% treatment layout of all the dilemmas was identical to the test dilemma,
efficiency) for it to be equally attractive as Project 2 (which but the projects differed so that one attribute was varied at
could treat 100 patients and had a 60% treatment efficiency). the time (always shaded in orange) while all other attributes
A number lower than 100 would indicate that the participant were identical in both projects.
did not comprehend the matching task (as Project 1 then can 5After reading and responding to the test dilemma, participants were
treat less patients AND has a lower treatment efficiency than asked to explain their response to the research assistant. For participants
Project 2). Using the same logic, a response higher than 100 who responded in a way that indicated non-comprehension on the test
dilemma, the assistant spent additional time explaining the matching task
in the Number&Second condition, a response lower than before continuing. Participants could change their response for the test
70% in the Efficiency&First condition, or a number higher dilemma after the explanation, but were not obliged to do so.
Table 1, continued.
Efficiency&First Project 1 Project 2
implemented?
Number of patients that will be 100 ill patients will be 150 ill patients will be
What is the average chance of ____% chance to survive for 70% chance to survive for each
surviving the disease for an ill patient each patient that is treated patient that is treated
that is treated?
Efficiency&Second Project 1 Project 2

implemented?
Number of patients that will be 100 ill patients will be 150 ill patients will be
What is the average chance of 70% chance to survive for each ____% chance to survive for
surviving the disease for an ill patient patient that is treated each patient that is treated
that is treated?
The attributes were presented in a fixed order in all dilem- Inferring moral preferences from the matching task On
mas to make it easier for participants to navigate. A summary each dilemma, we converted participants’ responses on the
of the varying attributes in all dilemmas can be seen in Table matching task to an expression of their moral preference
2 (see also OSM 1). about how to value different lives. For example, in the
After reading and filling in the blank box in each of the age dilemma (see Table 2), a participant in condition Num-
12 dilemmas, participants could report their gender and age. ber&First who writes a number higher than 100 indicates
They were also asked whether they would like to be invited that she thinks that Project A must treat more than 100
to an online follow-up study. Participants who volunteered adults in order to be equally attractive as Project B which
to participate in the follow-up study wrote their contact in- can treat 100 children. This means that she values the life
formation on the last page of the questionnaire.6 of a child as higher than the life of an adult. Conversely, a
participant who writes a number lower than 100 indicates
that she values the life of an adult as higher than the life of a
6This page was later removed, and participants’ contact information was
thus linked to their responses in the study only via an ID-number stored child, whereas a participant who writes exactly 100 indicates
separately. that she values children’s and adult lives equally high. The
Table 2: The four conditions of the dilemmas included in the matching task in Study 1 in the presented order. The letter “X”
indicates that participants filled in this value in order to make the two helping projects equally attractive. (Table continued on
next page.
Dilemma Number&First Number&Second

0. Test dilemma [1] X patients, 30%→70% = [2] 100 patients, [1] 100 patients, 30%→70% = [2] X patients,
(Comprehension 30%→90% 30%→90%
check)
1. Age dilemma [A] X adult patients = [B] 100 child patients [A] 100 adult patients = [B] X child patients
2. Gender dilemma [C] X female patients = [D] 100 male patients [C] 100 female patients = [D] X male patients
3. Innocence dilemma [E] X general patients = [F] 100 smoking and [E] 100 general patients = [F] X smoking and
drinking patients drinking patients
4. Comprehension [G] X patients for 400,000 SEK = [H] 100 [G] 100 patients for 400,000 SEK = [H] X
check patients for 600,000 SEK patients for 600,000 SEK
5. Ingroup dilemma [I] X Swedish patients = [J] 100 Canadian [I] 100 Swedish patients = [J] X Canadian
patients patients
6. Group size [K] 1000 in need, possible to treat X patients = [K] 1000 in need, possible to treat 100 patients
dilemma* [L] 200 in need, possible to treat 100 patients = [L] 200 in need, possible to treat X patients
7. Survival chance [M] X patients, 30%→70% = [N] 100 patients [M] 100 patients, 30%→70% = [N] X patients
dilemma 1 0%→40% 0%→40%
8. Survival chance [O] X patients, 30%→70% = [P] 100 patients, [O] 100 patients, 30%→70% = [P] X patients,
dilemma 2* 60%→100% 60%→100%
9. Existence dilemma [Q] X existing patients = [R] 100 future [Q] 100 existing patients = [R] X future
patients patients
10. Personal [S] X patients with a disease you have no [S] 100 patients with a disease you have no
responsibility connection with = [T] 100 patients with a connection with = [T] X patients with a
dilemma disease you partially caused disease you partially caused
11. Attention check [U] X patients, side effects are headache, [U] 100 patients, side effects are headache,
cough and running nose = [V] (100 patients, cough and running nose = [V] (X patients, side
side effects are running nose, cough and effects are running nose, cough and headache
headache
12. Side-effect [X] X patients, no side effects = [Y] 100 [X] 100 patients, no side effects = [Y] X
dilemma patients, small risk of deadly side effect patients, small risk of deadly side effect
opposite is the case for condition Number&Second, and the 2.1.2 The choice task
same logic applies to the Efficiency-conditions.
Participants. We prepared choice task invitations for those
who had participated in (and not been excluded from) the
Excluding participants in the matching task. We re-
matching task, and at that time agreed to be contacted again
moved some participants prior to any analyses using pre-
determined exclusion criteria. Two participants who failed for an online follow-up study (N = 501). Invitations were
to respond to five or more of the helping dilemmas, as well sent out via e-mail to the address they had provided in the
as 44 participants who failed the attention check (i.e., did not end of the matching task. Participants were offered one or
respond with 100 patients or 70%; see Dilemma 11 in Table two electronic scratch lottery ticket for participating in the
2) were excluded. In addition, participants who responded choice task.
in a way that indicated misunderstanding of the matching A presentation of the study was given directly in the e-
task on both comprehension checks (i.e., Dilemmas 0 and mail and the questionnaire was attached as a PDF-file (see
4) were also excluded (n = 21). The reported results for the OSM 1). Participants responded by replying to the invitation
matching task thus include responses from 551 female, 380 e-mail.
male and 9 unclassified participants, M age = 24.26 years, SD After approximately three weeks of data collection (in-
= 6.65. cluding three reminders) 151 participants (88 female, 62
Tabe 2, continued.
Dilemma Efficiency&First Efficiency&Second

0. Test dilemma [1] 100 patients, 30%→X% = [2] 150 patients, [1] 100 patients, 30%→70% = [2] 150
(Comprehension 30%→70% patients, 30%→X%
check)
1. Age dilemma [A] Adult patients, 30%→X% = [B] Child [A] Adult patients, 30%→70% = [B] Child
patients, 30%→70% patients, 30%→X%
2. Gender dilemma [C] Female patients, 30%→X% = [D] Male [C] Female patients, 30%→70% = [D] Male
3. Innocence dilemma [E] General patients, 30%→X% = [F] [E] General patients, 30%→70% = [F]
Smoking and drinking patients, 30%→70% Smoking and drinking patients, 30%→X%
4. Comprehension [G] 400,000 SEK 30%→X% = [H] 600,000 [G] 400,000 SEK 30%→70% = [H] 600,000
check SEK, SEK, 30%→X%
5. Ingroup dilemma [I] Swedish patients, 30%→X% = [J] [I] Swedish patients, 30%→70% = [J]
Canadian patients, 30%→70% Canadian patients, 30%→X%
6. Group size [K] 1000 in need, 30%→X% = [L] 200 in [K] 1000 in need, 30%→70% = [L] 200 in
dilemma* need, 30%→70% need, 30%→X%
7. Survival chance [M] 100 patients, 30%→X% = [N] 100 [M] 100 patients, 30%→70% = [N] 100
dilemma 1 patients, 0%→40% patients, 0%→X%
8. Survival chance [O] 100 patients, 30%→X% = [P] 100 [O] 100 patients, 30%→70% = [P] 100
dilemma 2* patients, 60%→100% patients, 60%→X%
9. Existence dilemma [Q] Existing patients, 30%→X% = [R] Future [Q] Existing patients, 30%→70% = [R] Future
10. Personal [S] Patients with a disease you have no [S] Patients with a disease you have no
responsibility connection with, 30%→X% = [T] Patients connection with, 30%→70% = [T] Patients
dilemma with a disease you partially caused 30%→70% with a disease you partially caused 30%→X%
11. Attention check [U] Side effects are headache, cough and [U] Side effects are headache, cough and
running nose, 30%→X% = [V] Side effects are running nose, 30%→70% = [V] Side effects
running nose, cough and headache, are running nose, cough and headache,
30%→70% 30%→X%
12. Side-effect [X] No side effects, 30%→X% = [Y] Small [X] No side effects, 30%→70% = [Y] Small
dilemma risk of deadly side effect, 30%→70% risk of deadly side effect, 30%→X%
Note 1: “30%→70%” illustrate that the average chance of surviving for each untreated patient is 30% whereas the
average chance of surviving for each treated patient is 70% (i.e., the treatment efficiency is 40%).
Note 2: Characters in brackets denotes the name of the helping projects as shown to participants.
* For reasons explained in the main text, these dilemmas are not included in this manuscript.
male, 1 unclassified, M age = 24.91, SD = 6.84) had com- Procedure and material. For each invited participant, we
pleted the choice task.7 One participant who failed an atten- created a unique questionnaire including 14 help dilemmas
tion check was excluded prior to any analyses. presented similarly to the dilemmas in the matching task.
Four of the dilemmas (0, 4, 7 and 12 in Table 4) were identical
for all participants and represented manipulation or attention
checks. The remaining dilemmas were designed so that the
7The large dropout rate between the two tasks in Study 1 was expected,
as participants in the matching task had to actively opt in to even be invited
two projects that were pitted against each other had been
to the choice task (less than 50% did so). Also, the long retention interval matched to be exactly equally attractive during the matching
between the two tasks likely made participants forget their previous com- task.
mitment when they received the invitation mail. No remarkable selection
bias was found when comparing the matching task responses of those who Participants’ task in each dilemma was to choose which of
completed vs. did not complete the choice task (see Table 3). the two suggested projects to implement. They were asked
to choose the project that they found more attractive and in our strongest indicator of the prominence effect as it include
case they found both projects equally attractive they were only participants who choose between two equally attractive
encouraged to flip a coin, throw a fair die, or use an online projects where one project is superior on the saved lives-
number generator to guide them when making the decision. attribute and the other project is superior on the presumed
Logically, participants who believed that the two projects prominent attribute.
were equally attractive would be equally likely to choose
either of the two projects (because they would choose at 2.2.1 Age dilemma
random). If all participants did this, both projects would
be chosen approximately equally often on the group level. Matching task. 44.0% of the participants valued children
In contrast, our hypothesis in each dilemma was that the higher whereas 24.7% valued adults higher (the remaining
project superior on the presumed prominent attribute would participants valued children and adults equally high).9 The
be chosen more often. mean preference was 0.19 [0.14–0.24] which illustrates a
small preference for helping children (t(936) = 7.34, p <.001,
d = 0.23; one sample t-test with reference value = 0).
2.2 Results
The proportion of participants who, in the matching task, Choice task. Project A which helped adult patients was
expressed each of the three possible preferences in each pitted against Project B which helped child patients. Note
condition of each dilemma are presented in Table 3. The that in this and all other dilemmas reported below, all par-
number of participants who, in the choice task, chose each ticipants made a choice between two projects that they had
of the projects in each dilemma are presented in Table 4. matched to be exactly equally attractive during the matching
We coded matching- and choice-task responses so that task. Despite this, the project helping children was cho-
1 indicated preferences for the project superior on the pre- sen by 80.4% of the participants. The Clopper-Pearson
sumed prominent attribute, 0 indicated no preference (equal 95% confidence interval of the observed proportion was
matching) and −1 indicated a preference for the project infe- [73.1%–86.5%; Fleiss, Levin & Paik, 2003]. Our first z-
rior on the presumed prominent attribute (see Tables 3 and test indicated that this was significantly above a 50–50 dis-
4). tribution (z = 7.39, n = 148, p < .001). Second, 81.6%
When aggregating over all eight focus dilemmas, the mean [67.9%–91.2%] of those who had valued adult and chil-
matching-task preference was 0.13 (SD = 0.37), indicating dren’s lives equally in the matching task chose the project
that projects superior on the presumed prominent attributes that could help children over the project that could help
are preferred when expressing preference with matching, equally many adults (z = 4.42, n = 49, p < .001). Third,
t(939) = 11.00, p < .001, d = 0.35 (one sample t-test with 73.0% [61.4%–82.7%] of the participants who had valued
reference value = 0). The mean choice-task preference was children’s lives higher than adult lives in the matching task
0.59 (SD = 0.32), indicating that, when forced to choose still chose the project that could help fewer children over the
between two equally attractive help projects, people tend to project that could help more adults (z = 3.96, n = 74, p <
choose the project superior on the prominent attribute, t(149) .001).
= 22.86, p <.001, d = 1.84. The rank-order correlation (across
participants) between mean matching and choice preferences
2.2.2 Gender dilemma
was rs = −.34. Additional analyses of aggregated preferences
and their relation to individual differences are presented in Matching task. A large majority (88.3%; 86.8% of fe-
OSM 3. male and 91.0% of male participants) valued female and
To increase readability, the results for each dilemma are male patients equally high, whereas only 6.0% [5.8%] val-
presented separately in the following text. When reporting ued females [males] higher. The mean preference was 0.00
the results from the choice task, we report three one propor-
tion z-tests (which compare the actual distribution against site direction. If the effects in the choice tasks were driven only by these
participants, this would severely undermine our arguments. Likewise, ex-
a 50–50 distribution) for each dilemma. The first z-test in- pressing no preference in the matching task (equal matching) might result
cluded all participants, the second included only participants from laziness, a refusal to consider the conflict, or a kind of default response
who expressed no preference (equal matching) in the match- indicating, e.g., “This decision is too hard for me.”
9In all dilemmas in both studies, we tested if participants’ expressed
ing task, and the third included only participants who ex-
preferences in the matching task differed as a function of which of the
pressed that the project superior on the presumed prominent four experimental conditions they were in (see Tables 3 and 5). In most
attribute was more attractive.8 Crucially, the third z-test is dilemmas, participants’ preferences did not differ, meaning that the attribute
or project they did the matching on, did not change the pattern of preferences.
8This was done because it is conceivable that some participants misun- In some dilemmas, preferences did significantly differ as a function of
derstood the matching task e.g., by thinking “One adult is 75% as important condition but in all but one of these (discussed below), these differences were
as one child, so I respond 75 when asked how many adults are equivalent quantitative rather than qualitative. We therefore aggregated all conditions
to 100 children” and thus mistakenly expressed a preference in the oppo- before the choice-task analyses.
Table 3: The proportion of participants in the Study 1 matching task who valued each of the two helping projects higher
in each condition in each dilemma, and the mean preference for each dilemma. The projects are presented in the order
participants responded to them.
Number Efficiency Total Preference

mean (SD)
Dilemma First Second First Second
0. Test Worse project 15.3% 11.0% 7.0% 7.9% 10.3%
Valued equally 6.0% 8.4% 10.0% 12.4% 9.2%
Better project 78.7% 80.6% 83.0% 79.8% 80.5%
1. Age Adults (−1) 23.6% 28.9% 22.0% 24.1% 24.7% [17.3%]
Valued equally (0) 35.8% 31.0% 30.6% 28.3% 31.4% [33.3%] 0.19 (0.81)
Children (+1) 40.6% 40.2% 47.4% 47.7% 44.0% [49.3%]
2. Gender Female patients (+1) 5.2% 7.5% 6.4% 4.6% 6.0% [6.0%]
Valued equally (0) 88.7% 86.2% 86.3% 92.0% 88.3% [90.1%] 0.00 (0.34)
Male patients (-1) 6.1% 6.3% 7.3% 3.4% 5.8% [4.0%]
3. Innocence General adults (+1) 32.3% 39.7% 41.5% 43.0% 39.1% [40.4%]
Valued equally (0) 29.3% 26.1% 32.9% 28.5% 29.9% [31.1%] 0.08 (0.83)
Smokers & drinkers (-1) 38.4% 31.2% 25.6% 28.5% 30.9% [28.5%]
4. Comprehen- More expensive project 19.8% 11.5% 12.0% 5.6% 12.2%
sion check Valued equally 11.5% 11.9% 17.7% 15.1% 14.0%
Cheaper project 68.8% 76.6% 70.3% 79.4% 73.8%
5. Ingroup Swedish patients (+1) 17.4% 15.9% 18.9% 17.4% 17.4% [22.5%]
Valued equally (0) 73.0% 79.5% 73.4% 77.1% 75.8% [72.8%] 0.11 (0.48)
Canadian patients (-1) 9.6% 4.6% 7.7% 5.5% 6.8% [4.6%]
6. Group size Patients from large group (-1) 8.7% 21.4% 57.9% 55.7% 36.1%
Valued equally (0) 21.0% 27.7% 27.0% 21.1% 24.2%
Patients from small group (+1) 70.3% 50.8% 15.0% 23.2% 39.7%
7. Survival chance 1 30%-project (-1) 50.7% 56.5% 60.1% 64.1% 57.9% [53.0%]
Valued equally (0) 14.0% 15.6% 21.0% 16.5% 16.8% [19.9%] −0.33 (0.85)
0%-project (avoid sure death) (+1) 35.4% 27.8% 18.9% 19.4% 25.3% [27.2%]
8. Survival chance 2 30%-project (-1) 27.0% 32.2% 13.7% 0% 18.2%
Valued equally (0) 12.6% 18.8% 20.5% 10.5% 15.6%
60%-project (+1) 60.4% 49.0% 65.8% 89.5% 66.2%
9. Existence Existing patients (+1) 59.4% 68.2% 73.4% 73.4% 68.7% [72.2%]
Valued equally (0) 16.6% 25.1% 17.2% 16.5% 18.9% [22.5%] 0.56 (0.70)
Future patients (-1) 24.0% 6.7% 9.4% 10.1% 12.5% [5.3%]
10. Personal Unrelated disease (-1) 29.6% 25.7% 24.8% 20.3% 25.1% [23.8%]
responsibility Valued equally (0) 47.8% 46.0% 46.6% 47.9% 47.1% [45.7%] 0.03 (0.73)
Disease they caused (+1) 22.6% 28.3% 28.6% 31.8% 27.9% [30.5%]
11. Attention check Patients helped by Project U 3.6% 0.8% 3.6% 2.0% 2.5%
Valued equally 94.8% 98.0% 94.8% 94.8% 95.6%
Patients helped by Project V 1.6% 1.2% 1.6% 3.2% 1.9%
12. Side-effect No side-effect (+1) 52.2% 52.5% 68.2% 77.6% 62.7% [66.2%]
Valued equally (0) 18.3% 21.8% 18.5% 10.5% 17.3% [19.2%] 0.43 (0.80)
Risk of side-effect (-1) 29.6% 25.6% 13.3% 11.8% 20.0% [14.6%]
Note 1. The scores for the eight focus dilemmas do not include the responses from excluded participants. The scores for
the comprehension/attention checks (Dilemmas 0, 4 and 11) include responses from all participants.
Note 2. Numbers in parentheses indicate whether the preferred project was superior (+1), or inferior (-1) on the presumed
prominent attribute, or if the projects were matched as equally attractive (0).
Note 3. Percentages in brackets illustrate the matching task preferences for those participants who later completed the
choice task.
Table 4: Number of participants in the Study 1 choice task who chose each project as a function of which project they valued
higher in the matching task. The projects are presented in the order participants responded to them.
Dilemma Project chosen to implement

0. Test Worse project [1] Better project [2]
1 145
1. Age Adults [A] (-1) Children [B] (+1)
Adults (-1) 0 25
Valued equally (0) 9 40
Children (+1) 20 54
2. Gender Females [C] (+1) Males [D] (-1)
Male patients (-1) 6 0
Female patients (+1) 7 3
3. Innocence General adults [E] (+1) Smokers & drinkers [F] (-1)
Smokers & drinkers (-1) 40 2
General adults (+1) 41 18
4. Manipulation check Better and cheaper [G] Worse and more expensive [H]
147 1
5. Ingroup Ingroup Swedes [I] (+1) Outgroup Canadians [J] (-1)
Canadian patients (-1) 7 0
Swedish patients (+1) 26 8
8. Survival chance 1 30%-project [M] (-1) 0%-project (avoid sure death) [N] (+1)
30%-project (-1) 10 69
0%-project (+1) 14 27
10. Existence Existing patients [Q] (+1) Future patients [R] (-1)
Future patients (-1) 8 0
Existing patients (+1) 81 28
11. Personal responsibility Unrelated disease [S] (-1) Disease they caused [T] (+1)
Unrelated disease (-1) 0 36
Disease they caused (+1) 28 18
12. Manipulation check Identical Project [U] Identical Project [V]
89 57
13. Side-effect Without side-effect [X] (+1) With side-effect [Y] (-1)
Risk of side-effect (-1) 21 0
No side-effect (+1) 58 41
Note 1. Dilemmas 0, 4, and 12 were identical for all participants. In all other dilemmas, participants saw two projects
that they previously had matched to be equally attractive.
Note 2. Characters in brackets denote the name of the projects as shown to participants.
prominent attribute, or if the projects were equally preferred (0).
Note 4. The group size dilemma (Nr 6) and the Survival chance dilemma 2 (Nr 9) are not included in this manuscript,
but included in the raw data file.
Note 5. The attention check (Nr 7) looked like the other dilemmas but the text in the row shaded in orange was: “This is
not a real question but a way to test if you are paying attention. Show that you are paying attention by writing [code] in
the response box” (see OSM 1).
[−0.02–0.02] implying no preference for helper either gender (z = 9.04, n = 108, p < .001). Third, 76.5% [58.9%–89.3%]
(t(938) = 0.19, p = .849, d < 0.01). This indicates a clear gen- of those who had valued ingroup lives higher in the matching
eral preference for valuing female and male patients equally task, chose the project that could help fewer ingroup patients
when expressing preferences with a matching task. over the project that could help more outgroup patients (z =
3.09, n = 34, p = .002).
Choice task. The project helping female patients was cho-
sen by 84.7% [77.9%–90.1%] of the participants (z = 8.50,
2.2.5 Patient group size dilemma.
n = 150, p < .001). Second, among the majority who valued
female and male lives equally in the matching task, 84.4% Due to highly varying preferences as a function of which
[77.2%–90.1%] chose the project helping females (z = 7.99, attribute participants did the matching on (number of treated
n =135, p < .001).10 Third, 70% [34.75%–93.33%] of the or treatment-efficiency, see Table 3), we have opted to lift
few who valued females higher in the matching task, still this dilemma from this manuscript.11
chose to help fewer females rather than more males (z =
1.27, n =10, p = .206).
2.2.6 Survival chance dilemma 1
2.2.3 Innocence dilemma Matching task. 57.9% valued untreated patients with a
Matching task. 39.1% valued “innocent” general patients 30% survival chance higher, whereas 25.3% valued un-
higher whereas 30.9% valued “non-innocent” smokers & treated patients with no chance of surviving higher. Against
drinkers higher and 29.9% valued innocent and non-innocent expectations, the mean preference was −0.33 [−0.38– −0.27]
patients equally high. The mean preference was 0.08 which illustrates a small preference for saving untreated pa-
[0.03–0.14] which illustrates a very small preference for tients with a 30% survival chance rather than patients that
helping innocent patients (t(934) = 3.02, p =.003, d = 0.10). will unavoidably die if left untreated (t(935) = 11.69, p <.001,
d = 0.39).
Choice task. The project helping innocent patients was
chosen by 83.0% [75.9%–88.7%] of the participants (z = Choice task. The helping project that could avoid a sure
8.00, n = 147, p < .001). Second, 89.1% [76.4%–96.4%] death was chosen by 79.3% [71.9%–85.5%] of the partic-
of those who had valued innocent patients and non-innocent ipants, (z = 7.18, n = 150, p < .001). Second, 76.7%
patients equally in the matching task chose the project help- [57.8%–90.1%] of those who had valued patients with both
ing innocent patients (z = 5.30, n = 46, p < .001). Third, diseases equally in the matching task chose the project that
69.5% [56.1%–80.8%] of those who valued innocent pa- could avoid a sure death for treated patients (z = 2.93, n = 30,
tients higher in the matching task, still chose the project that p = .003). Third, 65.9% [49.5%–80.0%] of those who had
could help fewer innocent patients over the project that could valued patients suffering from a disease with a 0% chance of
help more non-innocent patients (z = 3.00, n = 59, p = .003). survival if not treated higher in the matching task still chose
the project that could help fewer patients (but avoid a sure
2.2.4 Ingroup dilemma death for those treated) over the project that could help more
patients (z = 2.04, n = 41, p = .042).
Matching task. A large majority (75.8%) valued ingroup
and outgroup patients equally high whereas 17.4% valued
ingroup patients higher and 6.8% valued outgroup patients 2.2.7 Survival chance dilemma 2
higher. The mean preference was 0.11 [0.07–0.14] which
illustrates a small preference for helping ingroup patients During data collection, we realized that this dilemma was
(t(937) = 6.72, p <.001, d = 0.23). flawed. Specifically, it was logically impossible to express a
preference for Project O (30% survival chance if untreated)
in the Efficiency&Second condition. For this reason, we
Choice task. The project helping ingroup patients was
have lifted this dilemma from this manuscript.
chosen by 89.9% [83.9%–94.2%] of the participants (z =
9.74, n = 149, p < .001). Second, 93.5% [87.1%–97.3%] of 11The expressed preferences varied substantially in this dilemma, but this
those who had valued ingroup and outgroup lives equally in was primarily a function of participants’ preferences changing as a function
the matching task chose the project helping ingroup patients of which condition they read 𝜒 2 [6, n =937] = 235.17, p < .001. A majority
(over 50%) of the participants in the Number-conditions valued patients
10Among those who had valued lives equally in the matching task, both from the small patient group higher whereas a majority in the Efficiency-
female (97.4% [90.9%–99.7%], z = 8.26, n = 76, p < .001) and to a lesser conditions valued patients from the large patient-group higher. This finding
extent male participants (67.2% [53.6%–79.0%], z = 2.62, n = 58, p = .009) is interesting (and will be discussed in a separate manuscript) but poses a
tended to choose the project that could help female patients over the project problem for the choice-task as preferences in this dilemma seem strongly
that could help equally many male patients. influenced by which dimension one does the matching.
2.2.8 Existence dilemma. p < .001). Second, 93.1% [77.2%–99.2%] of those who had
valued lives equally in the matching task chose the project
Matching task. 68.7% valued existing patients higher
without a side-effect (z = 4.64, n = 29, p < .001). Third,
whereas 12.5% valued future patients higher. The mean
58.6% [48.3%–68.4%] of those who had valued the lives
preference was 0.56 [0.52–0.61] which illustrates a strong
of patients that could be treated without risk of side-effect
preference for helping existing patients (t(937) = 24.43, p
higher in the matching task, chose to help fewer patients that
<.001, d = 0.80).
could be treated without any risk of side-effect over helping
more patients that could be treated with a risk of side-effect
Choice task. The project helping existing patients was (z = 1.71, n = 99, p = .087).
chosen by 81.3% [74.1%–87.2%] of the participants (z =
7.67, n = 150, p < .001). Second, everyone (100%) who
2.2.11 Attention and manipulation checks in the choice
had valued existing and future victims equally in the match-
task.
ing task, chose the project that could help existing patients.
Third, 74.3% [65.0%–82.2%] of those who had valued exist- Dilemma 7 in the choice task was an attention check (one
ing lives higher in the matching task, still chose the project participant failed this) whereas Dilemmas 0 and 4 were ma-
that could help fewer existing patients over the project that nipulation checks (see Table 4). Against expectations, par-
could help more future patients (z = 5.07, n = 109, p < .001). ticipants did not seem to choose randomly when choosing
between two identical helping projects (U and V) in Dilemma
2.2.9 Personal responsibility dilemma 12. Project U was chosen by 61.0% [52.6%–69.0%] of the
participants (z = 2.66, n = 146, p = .008). We discuss this
Matching task. 47.1% valued patients suffering from an finding in the general discussion.
unrelated disease and patients suffering from a disease that
the participant ostensibly caused equally high whereas 27.9%
valued patients suffering from the participant-caused disease 2.3 Summary of Study 1
higher and 25.1% valued patients suffering from an unrelated Our initial hypothesis in all dilemmas was that we would find
disease higher. The mean preference was 0.03 [−0.02–0.07] group-level number-overriding preferences in the matching
which illustrates no preference for either project, t(936) = task, and that people later still would choose the project
1.17 (p =.243, d = 0.04). that was superior on the varying (supposed prominent) at-
tribute disproportionally often, even when that project could
Choice task. The project helping patients with a dis- save fewer lives. This hypothesis was supported in the Age,
ease that participants caused was chosen by 67.1% Innocence, Ingroup, Existence and Side-effect dilemmas, al-
[58.9%–74.6%] of the participants, (z = 4.18, n = 149, p though the matching task preferences differed much between
< .001). Second, 68.7% [56.2%–79.5%] of those who had dilemmas.
valued patients suffering from the two diseases equally in the The alternative “choice-dependent number-overriding”
matching task chose the project that helped patients suffering hypothesis predicted that most people would express no
from the participant-caused disease (z = 3.06, n = 67, p = preference in in the matching task, but that they, in the
.002). However, only 39.1% [25.1%–54.6%] of those who choice task, would prefer the project superior on the sup-
had valued patients suffering from the participant-caused posed prominent attribute rather than choosing at random.
disease higher in the matching task, chose to help fewer pa- We found support for this hypothesis in the Gender dilemma.
tients suffering from the participant-caused disease over the It also received partial support in the Personal responsibil-
project that could help more patients with the disease that ity dilemma. There, participants disproportionally preferred
the participant did not cause (z = 1.48, n = 46, p = .139). the project that helped patients whose plight they were re-
sponsible for when the two projects could help equally many
2.2.10 Side-effect dilemma (i.e., had been matched as equally good), but not when the
opposing project could help more patients. We did however
Matching task. 62.7% valued patients that could be note that the personal-responsibility manipulation was diffi-
treated without any risk for a side-effect whereas 20.0% cult to convey in a concise way with this paradigm, and we
valued patients that could be treated with a risk for a side- therefore dropped it for Study 2.
effect higher. The mean preference was 0.43 [0.37–0.48] In the Survival chance dilemma, we found preferences in
which illustrates a medium preference for avoiding causing opposing directions in the matching and choice tasks. The
incidental harm (t(937) = 16.25, p <.001, d = 0.54). matching task revealed preferences in the opposite direction
of what we predicted in that people valued patients who
Choice task . The no side-effect project was chosen by would die if untreated less than patients who had a 30%
71.1% [63.1%–78.2%] of the participants (z = 5.15, n = 149, chance to survive if untreated. In the choice task however,
participants responded as predicted and chose the project dilemma and this response was one of three comprehension
helping patients who would die if untreated more frequently, checks in the matching task.
even when this meant that less people would be saved. After the tutorial, participants read ten dilemmas (in ran-
domized order) and in each they matched the two projects
to become equally attractive by writing how many patients
3 Study 2 must be treated in one project to make it equally attractive
as the project it was pitted against (which could always treat
Although Study 1 provided strong support for the promi- 100 ill patients). To make it easier for participants to com-
nence effect in several of the included helping dilemmas, it municate in case they believed it was impossible to make the
suffered from some methodological drawbacks. For exam- two projects equally attractive, we told them to indicate this
ple, the fixed order of the dilemmas as well as of the projects by writing the number 0 (zero).
in each dilemma could have influenced the results (Ubel et
al., 2002; Carney & Banaji, 2012). We preregistered and
conducted Study 2 as a well-powered internal replication in Comprehension checks in the matching tasks. Three
order to test the robustness of the obtained results and at the comprehension checks were included in the matching task
same time controlling for several of the potential problems (Dilemmas M1–3, see Table 5 and OSM 2). In line with
in Study 1.12 The presentation as well as the information in preregistered criteria, we excluded participants who did not
in the helping dilemmas were identical or very similar to the comprehend the matching task on two or three of these dilem-
ones used in Study 1 (see OSM 1 and 2). mas.
3.1 Method
3.1.4 Procedure and material in the choice task
3.1.1 Participants
Unlike Study 1, participants completed the choice task right
Six hundred and five US participants recruited through Ama- after they completed the matching task. We explicitly stated
zon Mechanical Turk completed an online questionnaire. that in case they believed that the two contrasted projects
Participants were payed $2. were equally attractive to them, they should choose randomly.
To make this alternative even more accessible, we provided
3.1.2 Design participants with an online number generator obtained from
www.random.org.
In Study 2, all participants did the matching task on the num- Participants then read 14 dilemmas (randomized order),
ber of treated patients-attribute (not on treatment-efficiency). and in all dilemmas they had to write the name of one of
In order to control for possible order effects that potentially the two proposed helping projects. In 8 of the 14 dilem-
could have confounded the results in Study 1, the dilemmas in mas, participants were faced with two helping projects that
the matching and choice tasks were presented in an order ran- they previously had matched to be exactly equally attractive.
domized for each participant in Study 2. We also varied the
The remaining dilemmas were either comprehension checks,
order of two projects, i.e., half of the participants compared
manipulation checks or an attention check (see Table 6).
Project A (e.g., adults) against Project B (children) whereas
the other half compared Project B (children) against Project
A (adults). As in Study 1, half of the participants filled in the Attention and comprehension checks in the choice task.
blank on the project presented first (First-condition) whereas We embedded the same attention check as used in Study 1 in
the other half filled in the blank on the project presented last the choice task dilemmas. As preregistered, participants who
(Second-condition). For each helping dilemma, participants did not pass this check were excluded prior to any analyses.
were randomly assigned to read one of four versions of the There were four comprehension checks in the choice task
dilemma (AB/BA × First/Second). (see Table 6 and OSM 2). As preregistered, we excluded
participants who responded in a way indicating that they did
3.1.3 Procedure and material in the matching task not comprehend two or more of these.
In total, we excluded 121 participants for failing the atten-
After reading an instruction page, participants saw a tutorial
tion check task and an additional 49 participants for missing
about how to match the two projects. Participants were
more than one comprehension check in either the matching
shown a test dilemma like the one used in Study 1 and read
task or the choice task. This left us with 435 participants (219
several paragraphs with explanatory text. In the end of the
female, 215 male and 1 unclassified, M age = 37.40 years (SD
tutorial we asked participants to match the projects in the test
= 10.43) which was more than the 400 we deemed necessary
12https://osf.io/jrg38/?view_only=107c07abcc054caf97002bc2ed7400ee. in the preregistration.
Table 5: The proportion of participants in the Study 2 matching task who valued each of the two helping projects higher in
each condition, and the mean preference for each dilemma.
AB BA Total Preference
mean (SD)
Dilemma First Second First Second
Test Dilemma M1 Worse project 11.9% 11.9%
Valued equally 8.6% 8.6%
Better project 76.5% 76.5%
Age dilemma [2.5%] Adults (-1) 9.6% 19.1% 20.2% 13.7% 15.8%
Valued equally (0) 57.7% 60.6% 52.4% 57.8% 56.8% 0.12 (0.65)
Children (+1) 32.7% 20.2% 27.4% 28.4% 27.4%
Gender dilemma [0.9%] Female patients (+1) 2.7% 4.5% 1.9% 3.8% 3.2%
Valued equally (0) 97.3% 94.5% 93.2% 94.3% 94.9% 0.01 (0.23)
Male patients (-1) 0% 0.9% 4.9% 1.9% 1.9%
Innocence dilemma [4.4%] Runners & dieters (+1) 24.0% 39.4% 35.7% 28.7% 31.7%
Valued equally (0) 45.2% 42.4% 50.0% 47.0% 46.2% 0.10 (0.73)
Smokers & drinkers (-1) 30.8% 18.2% 14.3% 24.3% 22.1%
Ingroup dilemma [1.4%] American patients (+1) 13.0% 12.1% 17.8% 11.5% 12.1%
Valued equally (0) 80.0% 82.8% 80.4% 82.4% 81.4% 0.06 (0.43)
German patients (-1) 7.0% 5.1% 1.9% 12.0% 6.5%
Survival chance 30%-project (-1) 56.4% 68.8% 65.7% 44.9% 59.3%
dilemma [7.4%] Valued equally (0) 13.8% 12.5% 17.2% 18.4% 15.4% −0.34 (0.86)
0%-project (avoid sure death) (+1) 29.8% 18.8% 17.2% 36.7% 25.3%
Existence dilemma [13.8%] Existing patients (+1) 67.3% 82.5% 90.2% 61.8% 74.9%
Valued equally (0) 18.7% 11.3% 8.5% 20.2% 14.9% 0.65 (0.66)
Future patients (-1) 14.0% 6.2% 1.2% 18.0% 10.1%
Side-effect dilemma [3.9%] No side-effect (+1) 47.7% 67.0% 74.5% 44.2% 58.1%
Valued equally (0) 32.7% 21.1% 16.3% 32.7% 25.8% 0.42 (0.75)
Risk of side-effect (-1) 19.6% 11.9% 9.2% 23.1% 16.0%
Comprehension check M2 More expensive project 20.5% 10.3% 9.7% 22.6% 16.0%
Preferred equally 29.5% 20.5% 29.2% 25.8% 26.2%
Cheaper project 50.0% 69.2% 61.1% 51.6% 57.8%
Comprehension check M3 Patients in Project U 6.8% 11.6% 6.0% 10.4% 8.8%
Valued equally 90.2% 81.3% 87.3% 81.8% 84.9%
Patients in Project V 3.0% 7.1% 6.7% 7.8% 6.3%
Note 1: The scores for the seven focus dilemmas does not include the responses from excluded participants. The scores
for the comprehension checks (M1–3) include responses from all participants.
Note 2. Numbers in parentheses in the first column indicate whether the preferred project was superior (+1), or inferior
(-1) on the presumed prominent attribute, or if the projects were matched as equally attractive (0).
Note 3: Percentages in brackets show the number of participants who believed it was impossible to match the projects so
that they became equally attractive.
Table 6: Number of participants in the Study 2 choice task who chose each project as a function of which project they valued
higher in the matching task.
Dilemma Project chosen to implement

Age Adults [A] (-1) Children [B] (+1)
Adults (-1) 1 66
Children (+1) 20 96
Gender Females [C] (+1) Males [D] (-1)
Male patients (-1) 8 0
Female patients (+1) 9 5
Innocence Runners & dieters [E] (+1) Smokers & drinkers [F] (-1)
Smokers & drinkers (-1) 90 2
Runners & dieters (+1) 97 35
Ingroup Ingroup Americans [I] (+1) Outgroup Germans [J] (-1)
German patients (-1) 28 0
American patients (+1) 38 14
Survival chance 30%-project [M] (-1) 0%-project [N] (+1)
30%-project (-1) 123 116
0%-project (+1) 64 38
Existence Existing patients [Q] (+1) Future patients [R] (-1)
Future patients (-1) 37 1
Existing patients (+1) 231 50
Side-effect Without side-effect [X] (+1) With side-effect [Y] (-1)
Risk of side-effect (-1) 67 0
No side-effect (+1) 144 99
Manipulation check Identical Project U Identical Project V
204 231
Comprehension check C1 Better and cheaper Worse and more expensive
546 59
Comprehension check C2 More efficient Less efficient
547 58
Comprehension check C3 Treating 100 patients Treating 3500 patients
92 513
Comprehension check C4 Treating 2300 patients Treating 100 patients
540 65
Note 1: The scores for the eight focus dilemmas do not include the responses from excluded participants. The scores for
the comprehension checks (C1–4) include responses from all participants.
Note 2. Characters in brackets denote the name of the projects as shown to participants.
prominent attribute, or if the projects were equally preferred (0).
3.2 Results Choice. The results from Study 1 were replicated as 82.1%
[78.2%–85.6%] chose the project that treated female pa-
We coded matching- and choice-task responses in the same
tients (z = 13.33, n = 431, p < .001). Second, 82.4%
way as in Study 1 (−1, 0 and +1, see Tables 5 and 6).
[78.4%–86.0%] of those who had valued male and female
When aggregating over all seven focus dilemmas, the mean
lives equally high in the matching task chose the project that
matching-task preference was 0.13 (SD = 0.31), indicating
treated 100 females over the project that treated 100 males (z
that projects that are superior on the presumed prominent
= 13.11, n = 409, p < .001).13 Third, 60% [32.3%–83.66%]
attributes are preferred when expressing preference with
of the few who valued female lives higher in the matching
matching (t(434) = 9.03, p < .001, d = 0.42). The mean
task, still chose to help fewer females rather than more males
choice-task preference was 0.56 (SD = 0.32), indicating that,
(z = 0.77, n = 15, p = .439).
when forced to choose between two equally attractive help
projects, people tend to choose the project superior on the
prominent attribute (t(434) = 36.34, p <.001, d = 1.75). 3.2.3 Innocence dilemma.
The rank-order correlation across participants between mean Matching. 46.2% valued innocent and non-innocent pa-
matching and choice preferences was rs = −.20. Additional tients equally high whereas 31.7% valued innocent patients
analyses about aggregated preferences and their relation to higher and 22.1% valued non-innocent patients higher. The
individual differences are presented in OSM 3. mean preference was 0.10 [0.03–0.17] indicating a small
As in Study 1, we present the results for each dilemma sep- preference for helping innocent patients (t(415) = 2.69, p =
arately. Again, three one-proportion z-tests for each choice-
.007, d = 0.14).14
task were conducted to compare the observed proportion
against 50–50: (1) Including all participants (2) Including
only participants who expressed no preference in the match- Choice. The results from Study 1 were replicated as 83.4%
ing task (3) including only participants who expressed a pref- [79.5%–86.9%] chose the project that treated innocent pa-
erence for the project superior on the presumed prominent tients, (z = 13.63, n = 416, p < .001). Second, 83.3%
attribute in the matching task. [77.3%–88.3%] of those who had valued innocent and non-
innocent patients equally high in the matching task chose
the project that treated 100 innocent patients over the project
3.2.1 Age dilemma
that treated 100 non-innocent patients (z = 9.23, n = 192, p <
Matching. 56.8% valued adult and child patients equally .001). Third, 73.5% [65.1%–80.8%] of those who valued in-
high whereas 27.4% valued children higher and 15.8% val- nocent patients higher in the matching task chose the project
ued adults higher (see Table 5). The mean preference was helping fewer innocent patients over the project helping more
0.12 [0.05–0.18] which illustrates a small preference for non-innocent patients (z = 5.40, n = 132, p < .001).
helping children (t(423) = 3.68, p <.001, d = 0.18).
3.2.4 Ingroup dilemma
Choice. The results from Study 1 were replicated as 88.2%
[CI95: 84.7%–91.1%] chose the project that treated child Matching. A large majority (81.4%) valued ingroup (US)
patients, and this was significantly above a 50–50 distribution and outgroup (German) patients equally high whereas 12.1%
(z = 15.73, n = 424, p < .001; see Table 6). Second, 88.0% valued ingroup patients higher and 6.5% valued outgroup
[83.2%–91.8%] of those who had valued adult and child patients higher. The mean preference was 0.06 [0.02–0.10],
lives equally high in the matching task chose the project that indicating a very small preference for helping ingroup pa-
helped 100 children over the project that helped 100 adults (z tients (t(428) = 2.70, p = .007, d = 0.14).
= 11.80, n = 241, p < .001). Third, 82.8% [74.7%–89.2%] of
those who valued children higher in the matching task chose Choice. The results from Study 1 were replicated, as 88.3%
the project helping fewer children over the project helping [84.9%–91.2%] chose the project treating ingroup patients (z
more adults (z = 7.07, n = 116, p < .001). = 15.87, n = 429, p < .001). Second, 89.7% [86.0%–92.7%]
of those who had valued ingroup and outgroup patients
3.2.2 Gender dilemma
13Among those who had valued lives equally in the matching task, both
Matching. A large majority (94.9%; 94.0% for female and female (91.2% [86.4%–94.7%], z = 11.77, n =204, p < .001) and to a lesser
extent male participants (74.0% [67.4%–79.9%], z = 6.86, n =204, p < .001)
95.8% for male participants) valued male and female patients tended to choose the project that could help female patients over the project
equally high in the matching task, whereas only 3.2% [1.9%] that could help equally many male patients.
valued female [male] patients higher. The mean preference 14It should be noted that this manipulation was stronger than the inno-
cence manipulation in Study 1. In Study 2, Project E which helped “in-
was 0.01 [−0.01–0.04] meaning that participants expressed
nocent” patients who exercised regularly and ate nutritious food was pitted
no preference for helping females over males or vice versa against Project F which helped “non-innocent” patients who ate unhealthy,
(t(430) = 1.28, p = .201, d = 0.04). smoke and drank alcohol (see OSM 1 and 2.
equally high in the matching task chose the project that higher. The mean preference was 0.42 [0.35–0.49] indicat-
treated 100 ingroup patients over the project that treated ing a medium preference for the help project without any
100 outgroup patients (z = 14.83, n = 349, p < .001). Third, side-effect (t(417) = 11.44, p <.001, d = 0.56).
73.1% [59.0%–84.4%] of those who valued ingroup patients
higher in the matching task, still chose the project helping Choice. The results from Study 1 were replicated as 75.6%
fewer ingroup patients over the project helping more out- [71.2%–79.6%] chose the project that could treat patients
group patients (z = 3.33, n = 52, p < .001). without any risk of side effect (z = 10.47, n = 418, p < .001).
Second, 97.2% [92.1%–99.4%] of those who had valued
3.2.5 Survival chance dilemma patients in the without and with side effect projects equally
high in the matching task, chose the project that could treat
Matching. 59.3% valued patients from the 30→70%-
100 patients without any risk of side effect (z = 9.81, n = 108,
group higher whereas 25.3% valued patients from the
p < .001). Third, 59.3% [52.8%–65.5%] of those who valued
0→40%-group higher. The mean preference was -0.34
patients in the no side effect project higher in the matching
[−0.42– −0.26], indicating a small-medium preference for
task chose the project helping fewer patients without risk
the 30→70% project, t(402) = −7.97 p <.001, d = 0.40).
of side effect over the project helping more patients with a
small risk of side effect (z = 2.90, n = 243, p = .004).
Choice. The results from Study 1 were not replicated in the
choice task. Only 46.2% [41.3%–51.2%] chose the project
3.2.8 Manipulation check
that treated patients from the 0→40%-group and this was not
significantly different from a 50–50 distribution (z = 1.53, This dilemma, which pitted Project U against identical
n = 403, p = .127). Likewise, 48.5% [42.0%–55.0%] of Project V, is identical to the one used in Study 1 but we
those who had valued patients from the 30%→70%-group here varied the dilemma so that half of the participants read
higher and 51.6% [38.6%–64.5%] of those who had valued Project U first whereas the other half read Project V first. Un-
patients from the two projects equally high in the matching like Study 1, we found that the two identical projects where
task, chose the project that treated patients in the 0→40%- chosen equally often (46.9% chose Project U, z = 1.29, n =
group (z = 0.46, n = 239, p = .643 and z = 0.25, n = 62, p = 435, p = .196).
.801 respectively). We address these diverging results in the
general discussion.
4 General discussion
3.2.6 Existence dilemma
This research examined how preferences in moral dilemmas
Matching. 74.9% valued existing patients higher whereas are influenced by the way we express them and how we
10.1% valued future patients higher in the matching task. make choices when faced with two equally attractive helping
The mean preference was 0.65 [0.58–0.71], indicating a large projects. There are several novel findings to discuss.
preference for helping existing patients (t(374) = 19.09, p For at least five of the included dilemmas (Age, Innocence,
<.001, d = 0.98). Ingroup, Existence and Side-effect), we found the hypothe-
sized pattern of results in both studies. To varying degrees,
Choice. The results from Study 1 were replicated as 86.1% participants expressed number-overriding moral preferences
[82.2%–89.4%] chose the project that treated existing pa- in the matching task (e.g., they generally matched the projects
tients (z = 13.98, n = 375, p < .001). Second, 98.2% so that more than 100 future patients had to be helped in
[90.4%–100%] of those who had valued existing and future order to be equally attractive as a project helping 100 ex-
patients equally high in the matching task chose the project isting patients). In the choice-task however, most of these
that treated 100 existing patients over the project that treated participants still chose the projects helping existing, young,
100 future patients (z = 7.21, n = 56, p < .001). Third, 82.2% innocent, ingroup patients without any risk of a harmful side-
[77.2%–86.5%] of those who valued existing patients higher effect, and they did so even when these projects could help
in the matching task chose the project helping fewer existing fewer patients and had been matched as equally attractive
patients over the project helping more future patients (z = as the helping project it was pitted against. These results
10.80, n = 281, p < .001). support the notion the prominence effect underlies number-
overriding preferences in these helping dilemmas.
A possible mechanism for these results could be the
3.2.7 Side-effect dilemma
anchoring-and-adjustment heuristic (Tversky & Kahneman,
Matching. 58.1% valued patients that could be treated 1974). The number in the cell contrasted against the blank
without any risk of side effect higher whereas 16.0% val- box (e.g., 100 treated patients) could be understood as an
ued patients that could only be treated with a side effect-risk anchor, and participants might then adjust their estimates to
better fit with their moral preference. However, adjustments Importantly, the results obtained in Study 1 and Study
tend to be insufficient because people adjust only until reach- 2 correspond well to each other despite being collected in
ing the lower boundary of the range of plausible estimates two different samples (undergraduate students in Sweden
(Epley & Gilovich, 2006). If one’s “real” indifference point in Study 1 and MTurk-workers in USA in Study 2), using
is closer to the middle of the plausible range, this could two different methods (paper and pen in Study 1, online in
explain why people do not choose at random in the choice Study 2), and having different retention intervals between
task. the matching and the choice task (one month later in Study
Two dilemmas (Gender and Ingroup) clearly stuck out 1, right afterwards in Study 2). This, together with the
regarding the popularity of equal matches. To illustrate, 88– rather clear effect on most of the dilemmas, suggest that the
95% matched men and women as equally valuable whereas obtained results are robust and generalizable.
75–82% did so for outgroup and ingroup patients (all other The first take-home message of this article is that people
dilemmas had 14–57% equal matches). However, when do not choose at random when two equally attractive helping
these “indifferent” people were later forced to choose, 82– projects are pitted against each other. Instead, they choose
85% helped 100 females rather than 100 males, and 89–94% the project that is superior on the more prominent attribute,
helped 100 ingroup rather than 100 outgroup patients. These and children, female, innocent, ingroup and existing vic-
results are consistent with the “choice-dependent number-
tims as well as absence of harmful side-effects are all more
overriding” hypothesis, and remarkable because they show
prominent attributes than the number of individuals possible
that participants’ preferences in helping dilemmas dramat-
to save. The second take-home message is that some moral
ically change as a function of how we ask them to express
preferences are hidden when it is possible to express indif-
their preferences.
ference but become revealed when we are forced to make a
Although speculative, our suggested underlying mecha-
choice.
nism for choice-dependent number-overriding is people’s
desire to express justifiable, politically correct, and status-
enhancing preferences (Slovic, 1975; Grubbs, et al., 2019).
For some dilemmas (e.g., Age, Existence and Side-effect), 4.1 Are people ever indifferent in the choice
it is relatively easy and uncontroversial to justify why some task?
lives are valued higher than others. In contrast, the moral
Is it impossible to be indifferent when faced with a forced
preference that all lives are equally valuable becomes espe-
cially pronounced when women are pitted against men and choice in a helping dilemma? Although we explicitly told
as people of different background and ethnicity are pitted participants that they could use a coin flip or a fair die toss
against each other. In these situations, expressing anything (and even provided a random number generator to partici-
else than a preference for equal value would likely be seen as pants in Study 2), we did not measure how many used it.
socially controversial and possibly upsetting, so most partic- To further test whether it is possible to be indifferent when
ipants express neutrality (equal value) in the matching task. making a choice, we included one dilemma where two iden-
In a choice situation however, there is no way to express tical helping projects were pitted against each other. We
these neutral preferences. Then, rather than choosing ran- were initially concerned when Project U was chosen dispro-
domly (which one would do if one truthfully valued men portionally more often than identical Project V in Study 1,
and women and ingroup and outgroup equally), people opt but upon reflection, this was attributed to a limitation in the
for the alternative that is relatively easier to justify, namely methodology. In Study 1, the order of the presented projects
helping women and ingroup patients. This explanation cor- in each dilemma was fixed so that Project U was always pre-
responds nicely to research where decisions in a trade-off sented first. According to the “first is best” heuristic (Carney
game changed dramatically when varying which of the op- & Banaji, 2012) people tend to prefer the first when faced
tions that was framed as the moral one (Capraro & Rand, with two equally attractive alternatives. In Study 2 we var-
2018). Rather than having specific and stable moral pref- ied the order of the helping projects in all dilemmas and as
erences, it seems like most people are primarily motivated predicted Projects U and V were now chosen equally often.
to express what they think is generally considered to be the
most socially accepted moral attitudes.15
15One could further argue that this ease-of-justification explanation pri-
4.2 Limitations
marily applies for different people in the different dilemmas. Specifically,
choosing to help women rather than men seems easier to justify for polit- In this section, we first discuss the dilemma where the results
ically left-leaning people (to compensate for the inherent societal inequal- did not turn out as expected, and then consider issues related
ities between men and women), but choosing to help ingroup rather than to the matching task, as well as possible solutions.
outgroup patients seems easier to justify for politically right-leaning par-
ticipants (to avoid signaling disloyalty). Future studies should test this by
including individual difference measures such as political orientation and concern for political correctness (Strauts & Blanton, 2015).
4.2.1 The survival chance dilemma At the same time we argue that we did much to mitigate
the influence of possible misunderstandings. For example,
This dilemma stuck out in two ways. First, in line with
we had participants justify their responses in a test dilemma
prospect theory, we expected that helping projects that could
(in Study 1), or included a tutorial (in Study 2) to explain
avoid a sure death for treated patients would be preferred
the matching task. We also included several comprehension
when pitted against a helping project that could increase
checks and excluded participants who consistently responded
the chance of survival quantitatively. In the matching task,
in ways which indicated misunderstanding of the matching
we found the opposite pattern in both studies meaning that
task. Most importantly, the preference for the project su-
most participants valued, e.g., 30%→70% patients higher
perior on the prominent attribute was, for most dilemmas,
than 0%→40% patients (which would unavoidably die if not
found not only among those who matched in ways suggesting
treated). One possible explanation for this is a general affect
misunderstanding or among those who matched equally, but
heuristic (Slovic et al., 2002) which would predict that par-
also among participants who matched in the “predicted” di-
ticipants feel a negative affect towards the project were all
rection. This suggests that prominent attributes loom larger
untreated patients, and more than half of all treated patients
in choice than in matching.
will die. This negative affect elicits a general preference for
Even so, there are alternatives to the traditional match-
the opposing project that in comparison seems more promis-
ing task. Indifference point can be inferred from sev-
ing.
eral choices presented in quick succession (e.g., Dolan &
Second, most participants in Study 1 preferred the project
Tsuchiya, 2011). This would mean asking participants mul-
that could avoid a certain death in the choice task. Even
tiple times if they would choose Project A (X adults) or B
65.9% of those who choose between saving more people in
(100 children) while increasing or decreasing the value of
the 30%→70% project and fewer people in the 0%→40%
X. A participant’s indifference point is the value of X where
project chose the project that could avoid a sure death for
they switch to the other project. (This method may, however,
treated patients. This was in line with predictions and would
be treated more like the choice task.) Another alternative
suggest that avoiding a sure death is a prominent attribute.
way to test how people choose between two equally attrac-
To our surprise, Study 2 did not generate the same results
tive options could be to simply ask them to imagine that two
and the projects were chosen about equally often. A provi-
helping projects are equally attractive to them, and then ask
sional explanation is that the affect heuristic in the matching
them to choose between these two. Although it remains to
task influenced participants in the choice task in Study 2
be tested, we suspect that we would find the same pattern in
but not in Study 1. One reason could be the retention in-
the choice task also if we inferred participants indifference
terval between matching and choice. The negative affect
elicited by the 0%→40% project in the matching task was points through several choices (at least if there were some
no longer present when participants in Study 1 completed time between the two tasks) or if we asked participants to
the choice task a month later, but it could have remained “assume” that the two projects were equally attractive.
for participants in Study 2 who completed the choice task Relatedly, one difference between this paper and the sem-
some minutes later. Another reason could be that the Study 1 inal paper by Tversky, Sattath and Slovic (1988) is that
choice task tested undergraduate students who responded to whereas they had the two options differing on two continuous
our online invitation (presumably more conscientious than variables, most of our dilemmas had options which varied
those who did not respond) whereas Study 2 tested experi- on one categorical variable (e.g., children vs. adults or in-
enced MTurk-workers (presumably quicker and less deliber- group vs. outgroup) and one numerical variable (e.g., number
ate in their responding). Past research has shown that people of patients possible to treat), and the matching was always
experiencing time-pressure have different decision making done on the numerical variable. This raises the question of
processes than those who do not (e.g., Payne, Bettman & whether the observed effects are results of the content of the
Johnson, 1988). presumed prominent attributes or of them being expressed
categorically. We note that some categorical attributes can
be expressed numerically and that it, in future studies, might
4.2.2 The matching task
be possible to have participants make the matching task also
One could argue that expressing preferences using the match- on the presumed prominent attribute (e.g., 50 ten-year old
ing task is counterintuitive as your task is to eliminate your patients = 100 X-year old patients?).
preference rather than expressing it. In addition, the match-
ing task is complex and more cognitively demanding not only 4.3 Conclusion
compared to the choice task, but also compared to other non-
binary ways of expressing preferences such as attractiveness- This study investigated how moral preferences in different
ratings or budget allocations (see footnote 8). We agree medical helping dilemmas change as a function of how pref-
with these concerns and realize that some of the participants erences are expressed, and how people choose between two
might not completely have understood the matching task. equally attractive helping projects. We found that, when
faced with two helping projects that had been rated as Bischoff, C., & Hansen, J. (2016). Influencing support of
equally attractive, a significant majority chose the projects charitable objectives in the near and distant future: De-
that helped children (vs. adult), female (vs. male), innocent lay discounting and the moderating influence of construal
(vs. smokers & drinkers), ingroup (vs. outgroup) and exist- level. Social Influence, 11(4), 217–229. http://dx.doi.org/
ing (vs. future) patients and implied no (vs. some) risk of 10.1080/15534510.2016.1232204.
harmful side-effect. These projects were chosen more often Bravo Vergel, Y., & Sculpher, M. (2008). Quality-adjusted
even when they could help fewer patients than the oppos- life years. Practical Neurology, 8(3), 175–182. http://dx.
ing project, and this implies that these attributes influence doi.org/10.1136/pn.2007.140186.
preferences more when expressed with forced choice than Breeze, B. (2013). How donors choose charities: The role of
when expressed with matching. This study is the first to personal taste and experiences in giving decisions. Volun-
suggest the prominence effect as an underlying mechanism tary Sector Review, 4(2), 165–183. http://dx.doi.org/10.
for number-overriding in helping dilemmas, and that some 1332/204080513x667792.
moral preferences that are hidden when it is possible to ex- Brewer, M. B. (1999). The psychology of prejudice: Ingroup
press indifference become revealed when people are forced love and outgroup hate? Journal of Social Issues, 55(3),
to choose. 429–444. http://dx.doi.org/10.1111/0022-4537.00126.
Bucknall, T. (2003). The clinical landscape of critical care:
nurses’ decision-making. Journal of Advanced Nursing,
References 43(3), 310–319. http://dx.doi.org/10.1046/j.1365-2648.
2003.02714.x.
Alesina, A., & Dollar, D. (2000). Who gives foreign aid Capraro, V., & Rand, D. G. (2018). Do the right thing:
to whom and why? Journal of Economic Growth, 5(1), Experimental evidence that preferences for moral behav-
33–63. http://dx.doi.org/10.1023/A:1009874203400. ior, rather than equity or efficiency per se, drive human
Anderson, C. J. (2003). The psychology of doing nothing: prosociality. Judgment & Decision Making, 13(1). http://
Forms of decision avoidance result from reason and emo- journal.sjdm.org/17/171107/jdm171107.html.
tion. Psychological Bulletin, 129(1), 139–166. http://dx. Carney, D. R., & Banaji, M. R. (2012). First is best. PloS
doi.org/10.1037/0033-2909.129.1.139. one, 7(6), e35088. http://dx.doi.org/10.1371/journal.
Back, S., & Lips, H. M. (1998). Child sexual abuse: Victim pone.0035088.
age, victim gender, and observer gender as factors con- Chapman, G. B., & Elstein, A. S. (1995). Valuing the fu-
tributing to attributions of responsibility. Child Abuse & ture: Temporal discounting of health and money. Medical
Neglect, 22(12), 1239–1252. http://dx.doi.org/10.1016/ Decision Making, 15(4), 373-386. http://dx.doi.org/10.
S0145-2134(98)00098-2. 1177/0272989x9501500408.
Baron, J. (2009). Parochialism as a result of cognitive biases. Curry, T. R., Lee, G., & Rodriguez, S. F. (2004). Does victim
In R. Goodman, D. Jinks, & A. K. Woods (Eds.), Un- gender increase sentence severity? Further explorations
derstanding social action, promoting human rights. Ox- of gender dynamics and sentencing outcomes. Crime &
ford University Press. http://dx.doi.org/10.1093/acprof: Delinquency, 50(3), 319–343. http://dx.doi.org/10.1177/
oso/9780195371895.001.0001. 0011128703256265.
Baron, J., & Ritov, I. (2004). Omission bias, individual Dickert, S., & Slovic, P. (2011). Unstable values in lifesaving
differences, and normality. Organizational Behavior and decisions. Frontiers in Psychology, 2(294). http://dx.doi.
Human Decision Processes, 94(2), 74–85. http://dx.doi. org/10.3389/fpsyg.2011.00294.
org/10.1016/j.obhdp.2004.03.003. Dickert, S., Västfjäll, D., Kleber, J., & Slovic, P. (2015).
Baron, J., Ritov, I., & Greene, J. D. (2013). The duty to sup- Scope insensitivity: The limits of intuitive valuation of
port nationalistic policies. Journal of Behavioral Decision human lives in public policy. Journal of Applied Research
Making, 26(2), 128–138. http://dx.doi.org/10.1002/bdm. in Memory and Cognition, 4(3), 248–255. http://dx.doi.
768. org/10.1016/j.jarmac.2014.09.002.
Baron, J., & Szymanska, E. (2011). Heuristics and biases Dolan, P., & Tsuchiya, A. (2011). Determining the pa-
in charity. In D. M. Oppenheimer & C. Y. Olivola (Eds.), rameters in a social welfare function using stated pref-
The science of giving: Experimental approaches to the erence data: An application to health. Applied Eco-
study of charity. (pp. 215–235). Psychology Press. nomics, 43(18), 2241–2250. http://dx.doi.org/10.1080/
Bauman, C. W., McGraw, A. P., Bartels, D. M., & Warren, 00036840903166244.
C. (2014). Revisiting external validity: Concerns about Dovidio, J. F., Gaertner, S. L., Validzic, A., Matoka, K.,
trolley problems and other sacrificial dilemmas in moral Johnson, B., & Frazier, S. (1997). Extending the ben-
psychology. Social and Personality Psychology Compass, efits of recategorization: Evaluations, self-disclosure,
8(9), 536–554. http://dx.doi.org/10.1111/spc3.12131. and helping. Journal of Experimental Social Psychol-
ogy, 33(4), 401–420. http://dx.doi.org/10.1006/jesp. 14(10), e0223749. http://dx.doi.org/10.1371/journal.

1997.1327. pone.0223749.
Dufwenberg, M., & Muren, A. (2006). Generosity, Hsee, C. K., & Rottenstreich, Y. (2004). Music, pandas,
anonymity, gender. Journal of Economic Behavior & and muggers: On the affective psychology of value. Jour-
Organization, 61(1), 42–49. http://dx.doi.org/10.1016/j. nal of Experimental Psychology: General, 133(1), 23–30.
jebo.2004.11.007. http://dx.doi.org/0.1037/0096-3445.133.1.23.
Dwenger, N. and Kübler, D. F. and Weizsäcker, G. (2012). Hsee, C. K., & Zhang, J. (2010). General evaluability theory.
Flipping a coin: Theory and evidence. Working paper of Perspectives on Psychological Science, 5(4), 343–355.
the Max Planck institute for tax law and public finance http://dx.doi.org/10.1177/1745691610374586.
No. 2012–14. https://ssrn.com/abstract=2192422. Kahneman, D., & Tversky, A. (1979). Prospect theory: An
Eagly, A. H., & Crowley, M. (1986). Gender and helping be- analysis of decision under risk. Econometrica: Journal
havior: A meta-analytic review of the social psychological of the Econometric Society, 263–291. http://dx.doi.org/
literature. Psychological Bulletin, 100(3), 283–308. 10.2307/1914185.
Erlandsson, A., Björklund, F., & Bäckström, M. (2015). Keren, G., & Teigen, K. H. (2010). Decisions by coin
Emotional reactions, perceived impact and perceived re- toss: Inappropriate but fair. Judgment and Decision
sponsibility mediate the identifiable victim effect, pro- Making, 5(2), 83–101. http://journal.sjdm.org/10/10203/
portion dominance effect and in-group effect respectively. jdm10203.html.
Organizational Behavior and Human Decision Processes, Kogut, T. (2011). Someone to blame: When identifying a
127(0), 1–14. http://dx.doi.org/10.1016/j.obhdp.2014.11. victim decreases helping. Journal of Experimental Social
00. Psychology, 47(4), 748–755. http://dx.doi.org/10.1016/j.
Erlandsson, A., Jungstrand, A., & Västfjäll, D. (2016). An- jesp.2011.02.011.
ticipated guilt for not helping and anticipated warm glow Kogut, T., & Ritov, I. (2005). The singularity effect of
identified victims in separate and joint evaluations. Or-
for helping are differently impacted by personal responsi-
ganizational Behavior and Human Decision Processes,
bility to help. Frontiers in Psychology, 7(1475). http://dx.
97(2), 106–116. http://dx.doi.org/10.1016/j.obhdp.2005.
doi.org/10.3389/fpsyg.2016.01475.
02.003.
Epley, N., & Gilovich, T. (2006). The anchoring-and-
Levine, M., & Thompson, K. (2004). Identity, place, and
adjustment heuristic: Why the adjustments are insuffi-
bystander intervention: Social categories and helping af-
cient. Psychological Science, 17(4), 311–318. http://dx.
ter natural disasters. The Journal of Social Psychology,
doi.org/10.1111/j.1467-9280.2006.01704.x.
144(3), 229–245. http://dx.doi.org/10.3200/socp.144.3.
Fehse, K., Silveira, S., Elvers, K., & Blautzik, J. (2015).
229-245.
Compassion, guilt and innocence: An fMRI study of re-
Mackie, D. M., & Smith, E. R. (1998). Intergroup relations:
sponses to victims who are responsible for their fate. So- Insights from a theoretically integrative approach. Psy-
cial Neuroscience, 10(3), 243–252. http://dx.doi.org/10. chological Review, 105(3), 499–529. http://dx.doi.org/10.
1080/17470919.2014.980587. 1037/0033-295X.105.3.499.
Fleiss, J. L., Levin, B., & Paik, M. C. (2003). Statistical Neumayr, M., & Handy, F. (2019). Charitable giving: What
methods for rates and proportions (3 ed.). Hoboken: John influences donors’ choice among different causes? Volun-
Wiley & Sons. tas, 30(4), 783–799. http://dx.doi.org/10.1007/s11266-
Greene, J. D. (2008). The secret joke of Kant’s soul. In W. 017-9843-3.
Sinnott-Armstrong (Ed.), Moral psychology, Vol 3: The Nord, E., & Johansen, R. (2014). Concerns for severity
neuroscience of morality: Emotion, brain disorders, and in priority setting in health care: A review of trade-off
development. (pp. 35–80). MIT Press. data in preference studies and implications for societal
Goodwin, G. P., & Landy, J. F. (2014). Valuing different hu- willingness to pay for a QALY. Health Policy, 116(2),
man lives. Journal of Experimental Psychology: General, 281–288. http://dx.doi.org/10.1016/j.healthpol.2014.02.
143(2), 778–803. http://dx.doi.org/10.1037/a0032796. 009.
Gordon-Hecker, T., Rosensaft-Eshel, D., Pittarello, A., Pakula, A. J. (Director). (1979). Sophies choice [Film]. ITC
Shalvi, S., & Bereby-Meyer, Y. (2017). Not taking re- Entertainment; Universal Pictures.
sponsibility: Equity trumps efficiency in allocation de- Payne, J. W., Bettman, J. R., & Johnson, E. J. (1988). Adap-
cisions. Journal of Experimental Psychology: General, tive strategy selection in decision making. Journal of
146(6), 771–775. http://dx.doi.org/10.1037/xge0000273. Experimental Psychology: Learning, Memory, and Cog-
Grubbs, J. B., Warmke, B., Tosi, J., James, A. S., & nition, 14(3), 534–552. http://dx.doi.org/10.1037/0278-
Campbell, W. K. (2019). Moral grandstanding in pub- 7393.14.3.534.
lic discourse: Status-seeking motives as a potential ex- Raihani, N. J., & Smith, S. (2015). Competitive helping in
planatory mechanism in predicting conflict. PloS one, online giving. Current Biology, 25(9), 1183–1186. http://
dx.doi.org/10.1016/j.cub.2015.02.042. Tversky, A., & Kahneman, D. (1974). Judgment under

Ritov, I., & Baron, J. (1990). Reluctance to vaccinate: Omis- uncertainty: Heuristics and biases. Science, 185(4157),
sion bias and ambiguity. Journal of Behavioral Decision 1124–1131. http://dx.doi.org/10.1126/science.185.4157.
Making, 3(4), 263–277. http://dx.doi.org/10.1002/bdm. 1124.
3960030404. Tversky, A., Sattath, S., & Slovic, P. (1988). Contingent
Ritov, I., Baron, J., & Hershey, J. C. (1993). Framing effects weighting in judgment and choice. Psychological Re-
in the evaluation of multiple risk reduction. Journal of view, 95(3), 371–384. http://dx.doi.org/10.1037/0033-
Risk and Uncertainty, 6(2), 145–159. http://dx.doi.org/ 295x.95.3.371.
10.1007/BF01065355. Ubel, P. A., Richardson, J., & Baron, J. (2002). Exploring
Samuelson, P. A. (1937). A note on measurement of utility. the role of order effects in person trade-off elicitations.
The Review of Economic Studies, 4(2), 155–161. http:// Health Policy, 61(2), 189–199. http://dx.doi.org/10.1016/
dx.doi.org/10.2307/2967612. S0168-8510(01)00238-X.
Seacat, J. D., Hirschman, R., & Mickelson, K. D. (2007). Wade-Benzoni, K. A., & Tost, L. P. (2009). The egoism and
Attributions of HIV onset controllability, emotional reac- altruism of intergenerational behavior. Personality and
tions, and helping intentions: Implicit effects of victim Social Psychology Review, 13(3), 165–193. http://dx.doi.
sexual orientation. Journal of Applied Social Psychol- org/10.1177/1088868309339317.
ogy, 37(7), 1442–1461. http://dx.doi.org/10.1111/j.1559- van Vugt, M., & Iredale, W. (2013). Men behaving nicely:
1816.2007.00220.x. Public goods as peacock tails. British Journal of Psy-
Shah, K. K., Tsuchiya, A., & Wailoo, A. J. (2014). Valuing chology, 104(1), 3–13. http://dx.doi.org/10.1111/j.2044-
health at the end of life: an empirical study of public 8295.2011.02093.x.
preferences. The European Journal of Health Economics, Weber, M., Koehler, C., & Schnauber-Stockmann, A. (2019).
15(4), 389–399. http://dx.doi.org/10.1007/s10198-013- Why should I help you? Man up! Bystanders’ gen-
0482-3. der stereotypic perceptions of a cyberbullying incident.
Skedgel, C. D., Wailoo, A. J., & Akehurst, R. L. (2015). Deviant Behavior, 40(5), 585–601. http://dx.doi.org/10.
Choosing vs. allocating: Discrete choice experiments 1080/01639625.2018.1431183.
and constant-sum paired comparisons for the elicitation of Weiner, B. (1980). A cognitive (attribution)-emotion-action
societal preferences. Health Expectations, 18(5), 1227– model of motivated behavior: An analysis of judgments of
1240. http://dx.doi.org/10.1111/hex.12098. help-giving. Journal of Personality and Social Psychol-
Slovic, P. (1975). Choice between equally valued alterna- ogy, 39(2), 186–200. http://dx.doi.org/10.1037/0022-
tives. Journal of Experimental Psychology: Human Per- 3514.39.2.186.
ception and Performance, 1(3), 280–287. http://dx.doi. Weiner, B. (1995). Inferences of responsibility and social
org/10.1037/0096-1523.1.3.280. motivation. In M. P. Zanna (Ed.), Advances in experimen-
Slovic, P. (2007). “If I look at the mass I will never act”: tal social psychology. Academic Press. http://dx.doi.org/
Psychic numbing and genocide. Judgment and Deci- 10.1016/S0065-2601(08)60402-5.
sion Making, 2(2), 79–95. http://journal.sjdm.org/7303a/ Zagefka, H., Noor, M., & Brown, R. (2013). Familiar-
jdm7303a.htm. ity breeds compassion: Knowledge of disaster areas and
Slovic, P., Finucane, M., Peters, E., & MacGregor, D. G. willingness to donate money to disaster victims. Applied
(2002). The affect heuristic. In T. Gilovich, D. Griffin, Psychology, 62(4), 640–654. http://dx.doi.org/10.1111/j.
& D. Kahneman (Eds.), Heuristics and biases: The psy- 1464-0597.2012.00501.x.
chology of intuitive judgment. (pp. 397–420). Cambridge Zagefka, H., Noor, M., Brown, R., de Moura, G. R., &
University Press. Hopthrow, T. (2011). Donating to disaster victims: Re-
Strauts, E., & Blanton, H. (2015). That’s not funny: Instru- sponses to natural and humanly caused events. European
ment validation of the concern for political correctness Journal of Social Psychology, 41(3), 353–363. http://dx.
scale. Personality and Individual Differences, 80, 32–40. doi.org/10.1002/ejsp.781.
http://dx.doi.org/10.1016/j.paid.2015.02.012. Zhang, Y., & Slovic, P. (2019). Much ado about nothing:
Thomson, J. J. (1985). The trolley problem. Yale Law The zero effect in life-saving decisions. Journal of Be-
Journal, 94(6), 1395–1415. havioral Decision Making, 32(1), 30–37. http://dx.doi.
org/10.1002/bdm.2089.

Moral Preferences in Helping Dilemmas Expressed by Matching and Forced Choice

Uploaded by

Document Informationclick to expand document information

Copyright:

Available Formats

Moral Preferences in Helping Dilemmas Expressed by Matching and Forced Choice

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Moral Preferences in Helping Dilemmas Expressed by Matching and Forced Choice

Uploaded by

Copyright:

Available Formats

Judgment and Decision Making, Vol. 15, No. 4, July 2020, pp.