Journal of Personality Assessment

ISSN: 0022-3891 (Print) 1532-7752 (Online) Journal homepage: http://www.tandfonline.com/loi/hjpa20

The Selfishness Questionnaire: Egocentric,

Adaptive, and Pathological Forms of Selfishness

Adrian Raine & Stepheni Uh

To cite this article: Adrian Raine & Stepheni Uh (2018): The Selfishness Questionnaire:
Egocentric, Adaptive, and Pathological Forms of Selfishness, Journal of Personality Assessment,
DOI: 10.1080/00223891.2018.1455692

To link to this article: https://doi.org/10.1080/00223891.2018.1455692

Published online: 19 Apr 2018.

Submit your article to this journal

Article views: 4

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at

http://www.tandfonline.com/action/journalInformation?journalCode=hjpa20
JOURNAL OF PERSONALITY ASSESSMENT
2018, VOL. 0, NO. 0, 1–12
https://doi.org/10.1080/00223891.2018.1455692

The Selfishness Questionnaire: Egocentric, Adaptive, and Pathological Forms

of Selfishness
Adrian Raine,1 and Stepheni Uh2
1
Departments of Criminology, Psychiatry, and Psychology, University of Pennsylvania; 2Center for Neuroscience and Society, University of Pennsylvania

ABSTRACT ARTICLE HISTORY

Despite its importance in society, there is virtually no standardized research on the personality trait of Received 1 September 2017
selfishness, in part due to the absence of an assessment instrument. The central aim of this study was to Revised 9 February 2018
develop a brief (2–3 min) self-report personality measure of selfishness with three main subtypes:
egocentric, adaptive, and pathological. Questionnaires were administered to an undergraduate sample,
with replicability and generalizability tested on a community population. A confirmatory factor analysis
supported the existence of the three hypothesized forms of selfishness. Good internal reliability, test–
retest reliability, convergent validity, discriminant validity, factorial validity, incremental validity, and
criterion validity were documented. Selfishness was associated with reduced levels of mindfulness and
more utilitarian decision making, whereas Tibetan Buddhist monks were less selfish than controls.
Findings from the student sample were replicated in the community sample. To our knowledge, this is the
first systematic examination of the personality construct of selfishness and the first instrument to assess
selfishness and its variants.

Selfishness is widely regarded as an inordinate focus on other forms, those displaying this variant care not just for
one’s own welfare, regardless of the well-being of others. themselves, but also for their family, and at times their
Commonly, selfishness has been invoked in evolutionary friends. Some selfish behaviors could be justified on the
theories of cooperation (Fehr & Gachter, 2002) and mea- grounds that others benefit from such selfishness, even to
sured behaviorally in economic games (Bardsley, 2008). To the extent that giving oneself priority in life can benefit
date, however, there is no standardized instrument for the others. Particularly at an evolutionary level, ensuring one’s
assessment of a selfish personality and no prior examination well-being and survival can be driven by the overriding
of different variants of selfishness. As such, despite being a need to look after one’s family and promote genetic fitness
well-used descriptor of one antisocial feature of human (Dawkins, 2006). For example, if it came down to taking
nature, a direct assessment of selfishness as a personality the last place in a lifeboat in preference to a child, or killing
construct is lacking. Having an instrument that specifically another person to avoid being killed, adaptively selfish peo-
assesses this trait would fill this gap in the personality liter- ple would act affirmatively to ensure that their own family
ature and allow for the future development of a nomologi- would not suffer from their absence. Kin selection theory
cal network of relationships for this construct that would emphasizes the importance of the individual looking after
aid theory development. Because selfishness is a trait com- the interests of family members who are closely genetically
mon to a number of personality disorders, research specifi- related (Hamilton, 1964). From a reciprocal altruism per-
cally on selfishness would facilitate a symptom-based spective (Dawkins, 2006), helping and caring for friends
approach to research the etiology of these clinical disorders who can in turn help one in the future can be taken as a
(Cuthbert, 2014). This article therefore seeks to develop a form of positive selfishness rather than giving selflessness.
brief self-report instrument for assessing degrees of selfish- At a social level, telling white lies is arguably selfish as it
ness in the general population. Specifically, based on princi- benefits the teller by smoothing out social encounters at the
ples from evolutionary psychology, psychopathology, and cost of dishonestly (but beneficially) sparing the listener’s
developmental psychology, we hypothesize the presence of a feelings (Ariely, 2012). As this adaptive form of selfishness
spectrum of selfishness consisting of three basic forms could be theorized as promoting “survival” both in the
(adaptive, pathological, and egocentric) to provide a prelim- extreme and in less harmful social settings, it might be
inary framework for future research. expected to be more prevalent than other forms of selfish-
One variant, adaptive selfishness, is defined as a “softer” ness. Furthermore, because this form of selfishness is some-
form of selfish behavior with an eye to others, and is what more socially positive (i.e., at times including benefits
hypothesized to be the least pathological variant, with a to others), we anticipate that it would be relatively more
basis in evolutionary theory and social psychology. Unlike adaptive and functional. As such, adaptive selfishness was

CONTACT Adrian Raine [email protected] Departments of Criminology, Psychiatry, and Psychology, University of Pennsylvania, Philadelphia, PA 19104.

© 2018 Taylor & Francis

2 RAINE AND UH

expected to be less associated with antisocial, narcissistic, 2013), are expected from a discriminant validity standpoint to
and histrionic personality disorders, psychopathy, and be less associated with selfishness.
Machiavellianism than other forms of selfishness. From a construct validity standpoint, we aimed to test sev-
A second variant, egocentric selfishness, is defined as eral other provisional hypotheses. One concerns utilitarianism,
selfishness with a single-minded attentional focus on the which historically has its roots in the work of Malthus (1798),
self. This form is hypothesized as a simpler form of selfish- who argued that an increasing population exponentially bur-
ness in which behaviors are neither ostensibly advantageous dens society’s limited resources. Bentham (1789) similarly
nor disadvantageous to others. It is neither antisocial nor advocated decisions that benefit the greater good of the greater
prosocial in that other individuals are rarely factored into number at the cost of a small minority. Utilitarian moral deci-
one’s need calculus. Such individuals are not concerned sion making has been associated with acting selfishly as mani-
about the needs of either individuals or society, but are sin- fested in variations of the trolley dilemma and the dictator
gle-mindedly centered on themselves. From a sociobiologi- game (Tinghog et al., 2016) as well as with psychopathy (Gao
cal perspective, egoistic incentive theory argues that human & Tang, 2013) and lack of empathy (Gleichgerrcht & Young,
nature is fundamentally selfish (Caporael, Dawes, Orbell, & 2013), two traits provisionally thought to characterize selfish
Van de Kragt, 1989). From a developmental psychology individuals. We consequently hypothesized that selfishness in
perspective, egocentrism as a theoretical concept was most general would be associated with utilitarian moral decision
strongly propagated by Piagetian theory, which drew atten- making.
tion to egocentrism in childhood with a fundamental focus An important conceptual question consists of whether self-
on the self and a failure to take into account others’ per- ishness in general is simply the opposite of altruism. Past
spectives (Piaget, 1951). In the context of adult personality research suggests that altruism and selfishness are related but
development, this lack of others’ perspective taking would different concepts, as altruistic acts are not always selfless. For
be particularly expected to be associated with low levels of example, although individuals might engage in reciprocal altru-
empathy. Lack of perspective taking would be expected in ism, this social reciprocity ultimately maximizes the genetic fit-
turn to be associated with an uncaring, glacial, self-centered ness of the giver and as such can be viewed as selfish (Dawkins,
form of selfishness without reference to others. Based 2006). Altruistic acts can also be used to build an individual’s
broadly on egoistic incentive theory, we hypothesized that reputation in his or her social community, an ultimately selfish
this basic form of selfishness would be characterized by a strategy that has been noted in classic studies of psychopaths
lack of empathy, low altruism, low warmth, and unwilling- (Cleckley, 1976). Furthermore, even in the absence of an audi-
ness to share resources with others. ence, giving to others has been argued to be motivated by a
A third variant, pathological selfishness, is defined as a form “warm glow” reward, a positive feeling associated with activa-
of “hard” selfishness in which others are harmed for self- tion of the ventral striatum, a brain area associated with reward
advancement. It is predicated on theory and research in psy- (Harbaugh, Mayr, & Burghart, 2007). These perspectives sug-
chopathology and is viewed as relatively more antisocial in gest that selfishness and altruism are not polar opposites.
nature. Antisocial behavior, for instance, is frequently self-serv- Although we expect a negative association between selfishness
ing (Koolen, Poorthuis, & van Aken, 2012; van Leeuwen, and altruism, we also anticipate that selfishness will relate to
Rodgers, Gibbs, & Chabrol, 2014), whether it is used to acquire other measures characterized by low altruism (e.g., Machiavel-
resources or as a way to subjugate others. Psychopathy is lianism, psychopathy) independent of its association with altru-
a reward-seeking condition that has been theorized by multiple ism, a test of incremental validity.
researchers to have selfishness at its core, combined with a Finally, we anticipated that higher levels of mindfulness
reckless disregard for the well-being of other people (Patrick, would be associated with lower levels of selfishness. Mindful-
2006). Narcissistic personality disorder is characterized by ego- ness involves approaching one’s experiences and one’s self with
tistical grandiosity, with one diagnostic feature consisting of an attitude of openness and acceptance, and is associated with
taking advantage of other people for one’s own ends (American increased empathy and compassion (Birnie, Speca, & Carlson,
Psychiatric Association, 2013). Histrionic personality disorder 2010). As such, we anticipated a relationship between a mindful
includes self-centered features, including drawing attention to personality and reduced selfishness. In the context of criterion
oneself and craving to be the center of attention (American validity, Buddhist meditation has been associated with a “quiet
Psychiatric Association, 2013). Machiavellianism is a personal- ego” (Wayment, Wiist, Sullivan, & Warren, 2011) and the prac-
ity construct involving the manipulation of others for one’s tice of “mental cultivation” to establish a mindful and compas-
own gain (Wilson, Near, & Miller, 1996). Pathological selfish- sionate awareness of the self (XIV Bstan-‘dzin-rgya, 2012).
ness is conceptualized as incorporating some features of these Compassion and genuine concern for the welfare of others are
externalizing psychopathological conditions, including manip- key foci of the daily practices and studies of Tibetan Buddhist
ulation, exploitation of others, and reward-seeking behavior. monks who incorporate mindfulness practices in their daily
Consequently, we anticipate that these disorders, although lives (XIV Bstan-ʼdzin-rgya, 2012). In collaboration with the
associated with all forms of selfishness, should particularly cor- XIV Dalai Lama and Emory University’s Emory-Tibet Science
relate positively with pathological selfishness, where selfish Initiative (ETSI), we therefore administered our measure to a
actions noticeably harm others. In contrast, anxiety and depres- sample of Tibetan Buddhist monks, hypothesizing that they
sion, disorders that do not contain elements of selfishness as would show reduced selfishness compared to a control group.
defined by the Diagnostic and Statistical Manual of Mental Dis- The overarching aim of this study is to develop a brief self-
orders (5th ed. [DSM–5]; American Psychiatric Association, report measure of selfishness using samples drawn from both
 THE SELFISHNESS QUESTIONNAIRE 3

university and community populations to test generalizability Tibetan buddhist monks

of findings across populations. Psychometric goals were to keep Tibetan Buddhist monks (N D 81) were recruited at the Dre-
the questionnaire relatively simple and brief, establish adequate pung Loseling Monastery in Mundgod, India, for the purpose
internal reliability and test–retest reliability, provide initial con- of criterion validity in collaboration with the ETSI members.
struct validity by documenting convergent, discriminant, crite- Participants were either in their first or second year of a 6-
rion, incremental, and factorial validity, and to replicate year science curriculum established by the XIV Dalai Lama and
findings. A further main goal was to develop items for the three Emory University. The Selfishness Questionnaire (SQ) was
hypothesized subtypes of selfishness, test this a priori model of completed in a classroom at the monastery in two
selfishness using confirmatory factor analysis (CFA), and inves- group sessions for first and second years separately. Items were
tigate differential correlates of these factors. Importantly from a translated and reviewed by ETSI translators and instructions
methodological standpoint, we aimed to test replicability and on the SQ completion were provided orally. Mean age was
generalizability of initial findings from a student sample to a 33.35 years (SD D 4.66), with ethnicity as follows: East Asian
community sample given that it has been argued that the (81.5%) and Indian (18.5%).
majority of findings in psychology do not replicate (Aarts et al.,
2015).
Selfishness questionnaire

Method Questionnaire construction began with the generation of 61

items measuring selfish behaviors and attitudes. Because the
Participants three-factor model was hypothesized a priori, approximately
Student sample equal representation was given to items for the three hypothe-
A total of 453 undergraduate students from the University of sized factors (egocentric, adaptive, pathological). Initial analy-
Pennsylvania participated in this study in two phases. The first ses of these items from the student sample were then
group was used for initial test construction (64.3% female), conducted to select eight items to represent each of the three
with a mean age of 19.93 years (SD D 2.63). Ethnicity was as factors (see later). Item reduction was based on retention of less
follows: White (50.3%), Latino or Hispanic (9.3%), African skewed items, items with better item–total correlations, and
American or Black (8.1%), Asian American (25.3%), and other sampling validity (i.e., obtaining a relatively broad coverage of
(7.0%). Due to questionable data (see later), 14 participants social situations and contexts). These 24 items were then sub-
were dropped from the original sample of 357, leaving a final jected to CFA in both test (student) and replication (MTurk)
sample of 343. samples. Example items from the final questionnaire are as fol-
An additional sample of 117 undergraduates (59.1% female) lows: “I care for myself much more than I care for others”
was similarly recruited at a later phase to increase sample size. (egocentric); “At the end of the day, I care mostly for myself,
These participants had a mean age of 20.04 years (SD D 1.51), my family, and friends who can help me” (adaptive); and “If
with ethnicity as follows: White (44.5%), East Asian (19.1%), I’m honest, there are times when I put myself first, even if it’s
Indian (6.4%), African American or Black (11.8%), Hispanic someone else’s loss” (pathological). To facilitate the availability
(10.0%), and other (8.2%). Seven participants were dropped of this instrument for future research, the questionnaire
due to questionable data (see later), leaving a final sample of together with administration and scoring instructions are given
110. All students were given course credit for their participation in the Appendix.
and human subjects’ approval was obtained from the Univer-
sity of Pennsylvania.
Convergent validity measures
Community sample Antisocial, histrionic, and narcissistic personality
Community participants (N D 402, 51.7% female) were The Personality Diagnostic Questionnare–4C (PDQ–4C) is
recruited through Amazon’s Mechanical Turk (MTurk) and one of the most commonly used self-report measures of per-
provided a replication sample for the CFA. Participants had a sonality disorder symptoms in both clinical and community
mean age of 39.39 years (SD D 12.53), with ethnicity as follows: samples, with acceptable reliability, construct validity, and
White (56.8%), East Asian (2.1%), Indian (30.3%), African cross-cultural validity (Abdin et al., 2011; Bouvard, Vuachet, &
American or Black (6.6%), Hispanic (2.4%), and other (1.8%). Marchand, 2011). In this study, antisocial (manipulation,
Participants were compensated $1.50 for their participation. exploitation, and disregard for the rights of others), histrionic
Due to questionable data (see later), 66 participants were (excessive attention-seeking behavior, an overly dramatic self-
dropped, leaving a final sample of 336. presentation, and discomfort when not the center of attention),
An additional 338 MTurk participants were recruited and narcissistic (heightened grandiosity and an excessive need
(45.9% female) to provide data specifically on the dictator game for admiration) personality disorders were assessed, as they are
as well as moral decision making and mindfulness. Participants clinical disorders that embody facets of pathological selfishness.
had a mean age of 36.81 years (SD D 10.46), with ethnicity as Acceptable levels of reliability, construct validity, and cross-cul-
follows: White (77.2%), East Asian (7.5%), Indian (2.6%), Afri- tural validity have been reported for these measures (Fossati,
can American or Black (7.5%), Hispanic (3.6%), and other Porro, Maffei, & Borroni, 2012; Hopwood et al., 2013; Hyler,
(1.6%). Some participants (n D 31) were dropped due to ques- 1994; Kounou et al., 2015; Wang et al., 2013). In this study,
tionable data (see later), leaving a final sample of 307. internal reliability was acceptable for the community sample:
4 RAINE AND UH

antisocial (coefficient a D .82), histrionic (coefficient a D .78), lower social desirability. Internal reliability was good in this
and narcissistic (coefficient a D .84). study for anxiety (coefficient a D .86) and depression
(coefficient a D .88), but relatively low for social desirability
Machiavellianism (coefficient a D .41).
Machiavellianism reflects duplicitous, scheming, and dishonest
behavior often in a way to better serve the interests of the indi-
vidual. In this study, the Machiavellianism-egocentricity facet Criterion validity
of the Psychopathic Personality Inventory–Short Form (PPI–
Dictator game
SF), a brief version of the original PPI (Lilienfeld & Hess,
The dictator game is an experimental economic task gauging
2001), was administered. Scores from the short version corre-
altruistic social preference. The participant (dictator) decides
late well with those from the original PPI scales (Tonnaer,
how much money to give to a recipient anonymously, without
Cima, Sijtsma, Uzieblo, & Lilienfeld, 2013). The community
any negative consequences (Bardsley, 2008). Similar to one ver-
sample showed good internal reliability in this study (coeffi-
sion of the dictator game (Eckel & Grossman, 1996), the recipi-
cient a D .89).
ent for this study was a charity foundation (Save the Children).
Participants in the second community sample were informed
Psychopathy
that they would be given an additional $0.70 and were asked
Psychopathy was assessed using the Triarchic Personality Mea-
how much they would be willing to donate to Save the Children
sure (TriPM), which consists of three phenotypic components:
(participants viewed an online YouTube video from the foun-
boldness, meanness, and disinhibition (Patrick, 2010). This
dation), with assurance that all donations would be sent to the
brief self-report measure has demonstrated strong evidence of
charity at the conclusion of the study.
construct validity (Hall et al., 2014; Sellbom & Phillips, 2013).
Good internal reliability was obtained in this study for the
TriPM (coefficient a D .84).
Additional measures for construct validity (second
community sample only)
Warmth and altruism
Warmth and altruism were hypothesized to show relationships Moral decision making
with reduced selfishness. Affection and genuine like of others Utilitarian moral decision making was assessed in the second
as well as active concern for others were assessed using the community sample using three condensed versions (Glenn,
warmth facet and altruism facet scales, respectively, of the Raine, & Schug, 2009) of high-conflict personal dilemmas:
Revised NEO Personality Inventory (NEO PI–R; Costa & “Crying Baby,” “Footbridge,” and “Sacrifice” (Greene, Morelli,
McCrae, 1992). NEO PI–R facet scales have established reliabil- Lowenberg, Nystrom, & Cohen, 2008; Greene, Sommerville,
ity and validity in both normal and clinical populations, Nystrom, Darley, & Cohen, 2001). Participants were asked to
together with established cross-cultural validity (Trull, Useda, rate how morally appropriate or inappropriate they found utili-
Costa, & McCrae, 1995; Yang et al., 1999). Acceptable internal tarian actions (ones that are harmful but benefit the greater
reliability for both the warmth (coefficient a D .87) and altru- good) on a 7-point Likert scale (Knutson et al., 2010) ranging
ism (coefficient a D .79) facet scales was obtained in this study. from 1 (extremely inappropriate) to 7 (extremely appropriate;
see Supplementary Table S.1).
Affective empathy
The capacity to experience the emotions of how others feel was
Cognitive and affective mindfulness scale–revised
assessed using the 10-item negative affect emotional empathy
The Cognitive and Affective Mindfulness Scale–Revised
subscale of the Cognitive, Affective, and Somatic Empathy Scale
(CAMS–R; Feldman, Hayes, Kumar, Greeson, & Laurenceau,
(CASES; Raine & Chen, 2018). Good internal reliability (coeffi-
2007) is a short 12-item questionnaire assessing mindfulness
cient a D .82) was obtained in this study. Each item (e.g., “See-
(e.g., “I am able to pay close attention to one thing for a long
ing people sad at a funeral would make me feel sad too”) is
period of time”). The CAMS–R is divided into four factors with
answered on a 3-point (rarely, sometimes, often) scale.
three items in each: (a) attention, (b) present focus, (c) aware-
ness, and (d) acceptance (Feldman et al., 2007). Reliability and
Discriminant validity measures validity with constructs including mindfulness, distress, and
emotion regulation have been documented (Feldman et al.,
Anxiety, depression, and social desirability
2007). Good internal reliability was obtained in this study
Both personalities were assessed using the anxiety and depres-
(coefficient a D .87).
sion facet scales of the NEO PI–R (Costa & McCrae, 1992).
High scorers on the anxiety subscale tend to be more apprehen-
sive, fearful, and nervous (Costa & McCrae, 1992). The depres- Data accuracy checks
sion scale assesses the tendency to experience depressive affect To maximize fidelity of the data, the three validity check items
(e.g., guilt, sadness) in normal individuals (Costa & McCrae, from the NEO PI–R (Costa & McCrae, 1992) were adminis-
1992). Social desirability (i.e., people’s desire to make a good tered at the end of the study. Participants were asked to indicate
impression) was assessed using the Too Good scale of the “Yes” or “No” to questions assessing honesty and carelessness
PDQ–4C (Hyler, 1994). An example of a true–false item is (e.g., “Have you answered all the items?”). Participants were
“Sometimes I get upset,” with an affirmative answer reflecting excluded if they did not mark “Yes” to all three questions.
 THE SELFISHNESS QUESTIONNAIRE 5

Statistical methods total (a D .87, range D .33–.52), egocentric (a D .76, range D

.39–.49), adaptive (a D .71, range D .28–.51), pathological (a D
Confirmatory factor analysis
.75, range D .32–.53). Female students scored significantly
The 24 selfishness items (8 egocentric, 8 adaptive, 8 pathologi-
lower on all SQ scores (total, M D 16.98, SD D 8.28, d D .76,
cal) were subjected to CFA conducted on the covariance matrix
p < .001; egocentric, M D 3.67, SD D 2.65, d D .71, p < .001;
using EQS 6 (Bentler, 2000). Due to significant kurtosis for
adaptive, M D 8.21, SD D 3.36, d D .44, p D .027; pathological,
many of the items (Mardia’s normalized estimate for multivari-
M D 4.80, SD D 3.84, d D .82, p < .001) than male students
ate kurtosis was greater than 11 in both samples), the elliptical
(total, M D 23.38, SD D 8.60; egocentric, M D 5.89, SD D 3.77;
estimation method was used to estimate the distribution of
adaptive, M D 9.70, SD D 3.49; pathological, M D 7.80, SD D
covariances in the evaluation of all models (Bentler, 1995;
3.38).
Bentler, Berkane, & Kano, 1991). Two models were evaluated: a
The same items comprising scales developed on the student
one-factor model (general selfishness) and a three-factor model
sample were applied to the MTurk sample to assess indepen-
(egocentric, adaptive, and pathological). In the one-factor
dent replicability. Internal reliabilities were well-replicated,
model, all 24 items were loaded on one latent factor of general
with coefficient alpha and range of item–total correlations as
selfishness. In the three-factor model, the egocentric, adaptive,
follows: total scale (a D .93, range D .28–.70), egocentric (a D
and pathological selfishness items were loaded onto their
.83, range D .41–.65), adaptive (a D .79, range D .45–.62),
respective latent factors. Error terms were set to be uncorrelated
pathological (a D .85, range D .34–.69). As with the student
in all the models.
sample, female community participants also responded signifi-
Three goodness-of-fit indexes were reported to assess the fit
cantly less selfishly (total, M D 15.87, SD D 10.94, d D .41, p <
of the models: chi-square, the root mean square error of
.001; egocentric, M D 4.62, SD D 3.91, d D .37, p < .001; adap-
approximation (RMSEA) index (McDonald, 1989), and the
tive, M D 6.70, SD D 4.22, d D .33, p D .003; pathological, M D
comparative fit index (CFI; Bentler, 1990). High values for CFI
4.56, SD D 4.10, d D .42, p < .001) than male community par-
and RMSEA are indicative of a good-fitting model (Hu &
ticipants on the SQ (total, M D 20.57, SD D 11.76; egocentric,
Bentler, 1999). Direct comparisons were made between the
M D 6.16, SD D 4.43; adaptive, M D 8.07, SD D 4.11; patholog-
one-factor and three-factor models using the chi-square differ-
ical, M D 6.33, SD D 4.36).
ence (Dx2) test (Loehlin, 1992).

Differential associations with factors Test–retest reliability

All tests of significance were two-tailed. In addition to testing
Two samples of community participants were randomly
the significance of associations between selfishness and the con-
selected to complete the SQ again to assess test–retest reliability
vergent validity measures (e.g., antisocial personality disorder)
at two different time points: 4 weeks and 10 weeks after initial
and discriminant validity measures (e.g., anxiety), the signifi-
administration. Four-week test–retest data were obtained from
cance of the difference between these correlations was calcu-
75 community participants. Test–retest reliabilities were as fol-
lated using the procedure outlined by Lee and Preacher (2013)
lows: total selfishness, r D .84, p < .0001; egocentric, r D .84,
in which each r was converted into a z score using Fisher’s r-to-
p < .0001; adaptive, r D .71, p < .0001; and pathological, r D
z transformation, with Steiger’s Equations 3 and 10 used to
.79, p < .0001. Ten-week test–retest data were also obtained for
compute the asymptotic covariance of the estimates (Steiger,
a different set of 57 community participants. Test–retest reli-
1980).
abilities were as follows: total selfishness, r D .85, p < .0001;
egocentric, r D .79, p < .0001; adaptive, r D .77, p < .0001; and
Incremental validity: Altruism and social desirability
pathological, r D .78, p < .0001.
To assess the extent to which selfishness scales related in the
We tested for any 4- to 10-week differences in the sizes of
predicted manner with convergent validity scales over and
test–retest correlations. No differences were significant (all p
above altruism and social desirability, partial correlations were
values > .41), indicating no significant change in test–retest
run holding the control variable (e.g., altruism) constant for the
reliability over time. We further tested for within-group
total selfish and convergent validity scale (e.g., antisocial per-
changes in mean levels of selfishness scores from the first to the
sonality) relationships, thus comparing the unique variance of
second administration for each group (4 weeks and 10 weeks)
selfishness with the unique variance of antisocial personality
separately using paired t tests. All comparisons were nonsignifi-
free of any shared association with altruism.
cant for both the 4-week group (p > .14) and the 10-week
group (p > .32). Taken together from the three vantage points
Results of test–retest reliability coefficients, change in test–retest reli-
ability over time, and change in mean scores over time, all SQ
Internal reliability and means
measures show a good level of trait stability.
From the original pool of 61 items that had been clustered into
three groups on the basis of face validity, item–total correla-
Confirmatory factor analysis
tions were computed for each of the three putative subscales
using the first student subject pool (N D 343). The eight items The three-factor model of selfishness was initially tested on the
with the highest item–total correlations on their subscale were student sample using CFA (see Supplementary Table S.2). Fit
retained. Internal reliability (coefficient a) for the scales indexes were as follows: x2(249) D 626.49, p < .0001; CFI D
together with range of item–total correlations were as follows: .93; RMSEA D .065, indicating a good-fitting model. The three-
6 RAINE AND UH

factor model was a significantly better fit than a one-factor Table 2. Selfishness Questionnaire convergent validity results for the community
model, Dx2(2) D 75.89, p < .0001 (see Supplemental Table S.3 sample.
for factor loadings and fit indexes for a one-factor model). All Constructs Total selfishness Egocentric Adaptive Pathological
fit indexes were without exception superior for the three-factor a a
Antisocial .556 .498 .459 .557c
model than the one-factor model. All factor loadings were Histrionic .466 .363a .375a .538c
acceptable for egocentric (.42–.56), adaptive (.36–.62), and Machiavellianism .767 .641a .716b .737b
pathological selfishness (.37–.60). Factor intercorrelations were Psychopathy .570 .505a .477a .601c
Altruism ¡.579 ¡.613a ¡.463b ¡.524c
as follows: egocentric with adaptive, r D .52, p < .0001; egocen- Affective empathy ¡.428 ¡.472a ¡.281b ¡.386c
tric with pathological, r D .66, p < .001; adaptive with patho- Warmth ¡.365 ¡.426a ¡.307b ¡.285b
logical, r D .61, p < .001.
Note. N D 336. All correlations are statistically significant (p < .001). For the three
The preceding three-factor solution was well-replicated in factors only, a significant difference (p < .05) in the size of correlation coeffi-
the community sample. Fit indexes were as follows: x2(249) D cients (across rows) is indicated by different superscript (e.g., a, b, c). Coefficients
665.82, p < .0001; CFI D .97; RMSEA D .065, indicating a with the same superscript do not differ significantly.
good-fitting model. The three-factor model was a significantly
better fit than a one-factor model, Dx2(2) D 36.55, p < .0001. across factors, anticipating that all three variants should corre-
All fit indexes were without exception superior for the three- late significantly at some level with construct validity measures.
factor model than the one-factor model. All factor loadings Results of these findings are shown in Table 1 where signifi-
were acceptable for egocentric (.39–.71), adaptive (.29–.69), cant differences (p < .05) between correlations are flagged by
and pathological selfishness (.41–.75). Factor intercorrelations superscripts that differ. Significant differences were found for
were as follows: egocentric with adaptive, r D .71, p < .0001; all external variables. In most cases, all three selfishness factors
egocentric with pathological, r D .75, p < .001; and adaptive differed from each other in the extent to which they were asso-
with pathological, r D .76, p < .001. ciated with these variables. Pathological selfishness was more
strongly related than the other two scales to psychopathy, anti-
social personality, histrionic personality, and Machiavellianism.
Convergent validity Egocentric selfishness was most strongly associated with low
The SQ should relate to other constructs associated with self- altruism, low warmth, and low emotional empathy. In contrast,
ishness if it does indeed measure selfishness. This component adaptive selfishness had the weakest association with these
of construct validity was tested by associating total selfishness traits.
scores with antisocial personality disorder, narcissistic person- The factorial validity findings from the student sample were
ality disorder, histrionic personality disorder, Machiavellian- broadly replicated in the community sample (see Table 2).
ism, psychopathy, affective empathy, and warmth. Correlations Pathological selfishness showed significantly stronger correla-
for the student (N D 343) and community samples (N D 643) tions with psychopathy, narcissism, histrionic personality, and
are shown in Table 1 and Table 2, respectively. All correlations Machiavellianism. Egocentric selfishness again showed signifi-
were significant and in the predicted direction, indicating that cantly stronger correlations with low altruism, low warmth,
the SQ converges with other constructs thought to be associ- and low affective empathy. Prior findings of lower association
ated with selfishness. between adaptive selfishness and convergent validity measures
were confirmed for most traits, except for significantly higher
scores on Machiavellianism compared to egocentric selfishness.
Factorial validity Summarizing these differential relationships, pathological
selfishness was most associated with pathological personality
We sought to examine the extent to which the three factors
traits. Egocentric selfishness was most associated with emo-
derived from the factor analysis had factorial validity in that
tional coldness. Adaptive selfishness was least associated with
they differed from one another by virtue of showing differential
all these traits in the student sample and to a lesser extent in
relations to external variables. We assessed this by testing the
the community sample. Although all three factors are to some
significance of the difference between respective correlations
extent associated with external traits that they should be associ-
ated with if they do indeed measure selfishness, there are also
Table 1. Selfishness Questionnaire convergent validity results for the student
sample. between-factor differences in the extent of these associations,
documenting some degree of factorial validity.
Constructs Total selfishness Egocentric Adaptive Pathological

Antisocial .423 .375b,c .295b .408c

Narcissistic .486 .427a .332b .456a Discriminant validity
Histrionic .307 .226a .209a .340c
Machiavellianism .658 .497a .563a,c .622c We tested whether anxiety and depression would correlate with
Psychopathy .506 .451a .323b .539c selfishness at a significantly lower level than convergent validity
Altruism ¡.539 ¡.576a ¡.324b ¡.472c scales to test discriminant validity. Associations between these
Affective empathy ¡.355 ¡.397a ¡.193b ¡.313c
Warmth ¡.285 ¡.352a ¡.122b ¡.227c traits for the student sample are given in Table 3. Anxiety was
unassociated with selfishness. Depression was associated with
Note. N D 357. All correlations are statistically significant (p < .03). For the three higher selfishness at a low level (r D .155). This association was
factors only, a significant difference (p < .05) in the size of correlation coeffi-
cients (across rows) is indicated by different superscript (e.g., a, b, c). Coefficients significantly weaker than the associations between selfishness
with the same superscript do not differ significantly. and all other convergent validity scales (antisocial personality
 THE SELFISHNESS QUESTIONNAIRE 7

Table 3. Selfishness Questionnaire discriminant validity results for students and Table 4. Incremental validation of Selfishness Questionnaire controlling for altru-
community samples. ism and social desirability.
Constructs Total selfishness Egocentric Adaptive Pathological Controlling for Controlling for social
altruism desirability
Students (N D 357)
Anxiety .005 ¡.020 .035 .015 Constructs Students Community Student Community
Depression .155** .113* .130* .153**
Community (N D 336) Antisocial .32*** .37*** .43*** .47***
Anxiety .185** .196** .145** .164** Narcissistic .42*** .48*** .47*** .53***
Depression .331** .279** .321** .324** Histrionic .29*** .41*** .31*** .38***
Machiavellianism .57*** .66*** .65*** .74***
Note. Significant weak associations (p < .05) are seen between depression and the Psychopathy .41*** .43*** .55*** .53***
total and three varieties of selfishness for the student sample. In the community Altruism — — ¡.53*** ¡.60***
sample, significant weak associations are seen between both anxiety and depres- Affective empathy ¡.16** ¡.22*** ¡.36*** ¡.47***
sion and the total and three variances of selfishness. Warmth .08 .05 ¡.24*** ¡.43***

p < .05.
 Note. Relationships between total selfishness scores and convergent validity varia-
p < .01.
bles after controlling for altruism and social desirability in the student and com-
munity sample.

disorder, narcissistic personality disorder, histrionic personality p < .05, two-tailed.

disorder, Machiavellianism, psychopathy, affective empathy, p < .005, two-tailed.

and altruism; all p < .02), with the exception of a trend for p < .0001, two-tailed.
warmth (p D .052, two-tailed).
This pattern of discriminant validity was broadly repli- reexamining the relationship between selfishness scales and
cated in the community replication sample (see Table 3). convergent validity scales after controlling for each of these two
Higher anxiety was associated with higher selfishness at a variables, namely altruism and social desirability.
low level (r D .185). This association was significantly Partial correlations between total selfishness scales and con-
weaker than the associations between selfishness and vergent validity scales are provided in Table 4 for both student
all other convergent validity measures without exception and community replication samples. After controlling for altru-
(p < .002). Although higher depression was associated with ism, selfishness was still associated with all convergent validity
higher selfishness (r D .331), this association was neverthe- scales, with replication across samples. The one exception was
less significantly weaker than the associations between that after controlling for altruism, selfishness was not associated
selfishness and all other convergent validity measures with warmth in the student sample, although robust associa-
(p < .02), except for warmth (p D .49) and affective empa- tions were obtained for this scale in the community sample.
thy (p D .12).

Criterion validity
Incremental validity
Tibetan monks
In addition to its association with altruism (see earlier), high The two MTurk samples were used to form a control group for
selfishness was generally associated with low social desirability, comparison with Tibetan monks as they approximate a normal
with modest correlations for student and community samples, population more closely than university undergraduates
respectively, as follows: total selfishness, r D .09, p < .10; r D (Berinsky, Huber, & Lenz, 2012). Nevertheless, groups differed
.14, p < .02; egocentric, r D .04, p D .32; r D .09, p D .02; adap- on gender, ethnicity, and age, with monks being male, younger,
tive, r D .10, p D .07; r D .17, p D .002; pathological, r D .09, p and more likely to be Asian. To obtain a more balanced sample,
D .09; r D .12, p D .04. In examining incremental validity, we the MTurk control group (N D 108) was restricted to only
assessed the extent to which selfishness provided increased those who were male and Asian, and within the same age range
knowledge on other constructs over and above knowledge pro- as monks (23–45 years). Age was additionally used as a covari-
vided by a related construct (low altruism) or an associated ate in analyses.
response style (low social desirability). This issue also addresses Means and standard deviations for the two groups, together
the conceptual issue of whether selfishness is merely the oppo- with results on analyses of variance and effect sizes, are given in
site of altruism. We addressed these two separate questions by Table 5. Monks scored significantly lower than matched

Table 5. Comparisons on selfishness scales between monks and controls.

Monksa Controlsb

SQ factors M SD M SD df F p d

SQ total 20.68 6.38 25.94 9.63 1,187 18.13 .001 .63

Egocentric 6.94 2.91 8.18 3.61 3,185 6.40 .012 .37
Adaptive 7.91 2.61 9.19 3.49 3,185 7.58 .006 .41
Pathological 5.83 2.91 8.57 3.82 3,185 29.08 .001 .79

Note. SQ D Selfishness Questionnaire.

a
N D 81.
b
N D 108.
8 RAINE AND UH

controls on all scales, with effect sizes ranging from .37 to .79. outcomes over and above altruism, establishing selfishness as
The effect of age was nonsignificant (p D .57). having incremental utility as an independent personality con-
struct in understanding other psychological phenomena. Find-
Dictator game ings from students were cross-validated in a community
The second community sample was used to validate selfishness sample, documenting replicability of findings and evidence of
scores against a known behavioral standard of selfishness: not generalizability to a different population. To our knowledge,
sharing money with others without consequences. Higher total this constitutes the first report of a reliable and valid self-report
selfishness scores were associated with increased hoarding of measure of selfishness, and provides instrumentation for the
financial resources in the dictator game (r D –.27, p < .0001). development of future research on this construct in both clini-
Nonsharing was also associated with egocentric (r D –.32, cal and community populations.
p < .0001), adaptive (r D –.18, p < .003), and pathological self-
ishness (r D –.22, p < .001). Egocentric selfishness was more
Generalizability and social desirability
strongly associated with hoarding than adaptive (p D .002) and
pathological selfishness (p D .018), with no significant differ- Results from the student sample, with some exceptions, were
ence between the latter two (p D .28). Consequently, the SQ well-replicated in the MTurk community sample. Replication
(especially egocentric selfishness) met the behavioral criterion was particularly robust for the CFA, internal reliability, test–
of enhancing self-interest by not sharing resources with others. retest reliability, and convergent validity. The issue of
replication is of significance because the capacity of findings
from psychological research to replicate has been questioned
Construct validity
(Aarts et al., 2015). Replication in our study also speaks to the
Moral decision making issue of the generalizability of findings, as our student sample is
Selfish individuals are more likely to make utilitarian moral a highly select sample even among undergraduate populations,
decisions. High scores favoring utilitarian decision making whereas the MTurk sample is more representative of the gen-
were associated with total (r D .35, d D .75, p < .0001), egocen- eral population. Specifically, although MTurk samples are
tric (r D .35, d D .74, p < .0001), adaptive (r D .29, d D .60, somewhat less representative of national probability samples,
p < .0001), and pathological (r D .32, d D .68, p < .0001) self- they correspond more similarly to the general U.S population
ishness. The three forms of selfishness did not differ signifi- than in-person convenience samples (Berinsky et al., 2012),
cantly (p > .42) in the strength of these relationships. and have a more superior diversity distribution compared to
many in-person samples (Casler, Bickel, & Hackett, 2013). The
Mindfulness associations of selfishness with other constructs did not vary
More selfish individuals are less mindful. Low scores on the greatly across different samples, attesting to the generalizability
CAMS–R were associated with high total selfishness (r D –.30, and replicability of this study’s main findings on the reliability
N D 319, p < .001), egocentric selfishness (r D –.23, p < .001), and validity of this instrument.
adaptive selfishness (r D –.24, p < .001), and pathological self- Although selfishness was associated with antisociality, high
ishness (r D –.31, p < .001). The association of mindfulness selfishness scorers were found to be relatively honest in their
with pathological selfishness was significantly higher than its self-appraisal, as indicated by reduced social desirability in rela-
association with egocentric selfishness (p < .05), and margin- tion to higher and total selfishness. Although high scores on
ally higher than the association with adaptive selfishness social desirability scales assess “faking good” and dissimulation,
(p < .08, two-tailed). they also assess a stable personality trait, with lower scores
reflecting social nonconformity and artfulness (Eysenck &
Eysenck, 1991). It appears that selfish individuals, far from
Discussion
making an attempt to hide this socially undesirable characteris-
The central aim of this study was to develop the first self-report tic, might not be personally concerned about how negatively
measure of selfishness. A CFA supported the existence of ego- they appear to others in society. Furthermore, selfishness was
centric, adaptive, and pathological forms of selfishness. The SQ associated with other scales over and above its association with
was characterized by good internal reliability, test–retest reli- social desirability, documenting that it is not simply a by-prod-
ability, convergent validity, discriminant validity, face validity, uct of the personality trait of social undesirability.
factorial validity, incremental validity, and criterion validity.
Pathological selfishness was more strongly associated with anti-
Optimal selfishness and mindfulness
social, narcissistic, and histrionic personality disorders, psy-
chopathy, and Machiavellianism. Egocentric selfishness, in Selfishness was predominantly associated with negative and
contrast, was more strongly associated with a lack of warmth, malicious outcomes, including psychopathy and Machiavellian-
low altruism, low emotional empathy, and less giving in the ism. Selfishness in society is not entirely unexpected from an
dictator game. Adaptive selfishness was less associated with evolutionary standpoint as self-benefits enhance inclusive fit-
these traits and was more likely to be endorsed. Selfishness was ness (Hamilton, 1964), although we acknowledge that we can-
associated with reduced levels of mindfulness and also not definitively document from our study that adaptive
increased utilitarian moral decision making. Criterion validity selfishness is advantageous from an evolutionary perspective,
was established by reduced selfishness in Tibetan monks, and or is in any way optimal. The term adaptive is used in a relative
less giving in the dictator game. Selfishness predicted to sense to convey a softer form that is somewhat more nuanced
 THE SELFISHNESS QUESTIONNAIRE 9

than the other two forms, which either callously make use of centered at the midpoint of the 0, 1, 2 response scale, and such
others (pathological) or have a single-minded focus on the self uncentered responses in questionnaire data can result in Type
(egocentric). Adaptive selfishness could ultimately prove to be II error inflation (Holtzman & Donnellan, 2017).
a maladaptive trait, albeit at a lower level than other forms of
selfishness. Furthermore, we should clarify that the term patho-
Future directions
logical selfishness is a shorthand label for a form of selfishness
that harms others and is more related to psychopathology. It There are several directions in which future research on selfish-
has been suggested that if one had only 2 seconds to explain ness could usefully contribute incremental knowledge. From a
human nature, one might say “self-interest” (Haidt, 2007). developmental perspective, one issue concerns whether pro-
Clearly, at some level, looking after oneself is an essential com- spective longitudinal research, by teasing out the temporal
ponent of life, and this raises the question of whether extremely ordering of variables, can help establish whether selfishness
low levels of selfishness could be maladaptive. Future research predisposes individuals to some clinical disorders, or alterna-
could examine to what extent, and in which outcome areas, an tively whether the development of psychopathological condi-
“optimum” level of selfishness might exist. tions results in increased selfishness. From a psychopathology
An important question at a clinical level concerns how perspective, it would be of value to explore links between self-
excessive selfishness that harms others can be attenuated. Find- ishness and other disorders, including borderline personality
ings of this study provide initial suggestions. Criterion validity disorder. At a psychometric level, this study at least provides a
for the scale was documented by showing that Tibetan monks basis for building alternative selfishness questionnaires with a
who endorse compassion and practice mindfulness were less different conceptualization, item wording, and factor structure.
selfish than controls. Relatedly, selfishness was also associated At a sampling level, we anticipate that the scale could have util-
with reduced compassion and increased mindfulness, particu- ity in researching individual differences in the general popula-
larly with respect to pathological selfishness. There is evidence tion, as well as pathological selfishness in clinical samples.
that a mindful self-compassion intervention experimentally At a mechanistic level, an unresolved issue concerns the
increases compassion (Neff & Germer, 2013), and that even genetic, neural, cognitive, social, and macrosocial processes
kindergarten children can become less selfish compared to con- that give rise to selfishness. This study was not designed to
trols following mindfulness-based kindness training (Flook, speak to the important issue of causal processes, but inclu-
Goldberg, Pinger, & Davidson, 2015). Mindfulness training has sion of the SQ in future brain imaging, molecular genetic,
also been reported to reduce antisocial behaviors (Shonin, Van social, clinical, and cognitive studies could begin to address
Gordon, Slade, & Griffiths, 2013; Wupperman et al., 2012), this question. For example, is the reward circuit consisting
behaviors that we find to be associated with selfishness. Future of the ventral striatum, orbitofrontal cortex, and anterior
intervention studies could therefore address whether selfishness cingulate, together with the dorsolateral prefrontal cortex,
could be attenuated by loving-kindness meditation programs, amygdala, hippocampus, and thalamus that regulate this
particularly for pathological selfishness, which was more associ- system (Haber & Knutson, 2010) more easily activated in
ated with reduced mindfulness compared to other variants excessively selfish individuals?
(Logie & Frewen, 2015).
Conclusions
Limitations
Selfishness as a concept has been frequently invoked in the
Several limitations of this study need to be recognized. First and study of human behavior, particularly in the area of clinical
foremost, the samples used to develop the scale consisted of psychology where selfishness is a trait purported to be present
undergraduates and online (MTurk) samples that do not in several clinical conditions. In the context of personality
include children, adolescents, and the elderly. Second, this assessment, the lack of any instrument to specifically quantify
study only provides a very beginning investigation of the selfishness and its variants is unexpected. The SQ represents an
nomological network of relationships that encapsulate the self- initial instrument for the assessment of three forms of selfish-
ishness construct, a network that needs to be extended to solid- ness—egocentric, adaptive, and pathological—that might have
ify its construct validity (Cronbach & Meehl, 1955). Third, utility in further investigating a construct that is well-known to
although we a priori conceptualized a three-factor structure to everyone, but is virtually unresearched in a standardized man-
selfishness that was confirmed and replicated, other conceptu- ner. The future promise of such research, at least at the clinical
alizations of selfishness could be evaluated in future research level, lies in the potential to parse more finely a trait that is
using a different instrument. Fourth, the wording of some items believed to characterize narcissism, psychopathy, Machiavel-
could be improved, as, for example, for adaptive selfishness, lianism, and histrionic personality, and yet is not specified as
which contained two items on nonhurtful lying (e.g., “white such in their formal DSM definitions. Whether selfishness
lies”), yielding an overlap between this construct and telling lies should be included as one of the defining traits in these clinical
that can benefit others as well as oneself. Fifth, although dis- disorders in future DSM revisions remains to be seen.
criminant validity was documented for both student and com-
munity samples, correlations were somewhat larger for the
community sample (see Table 3). Sixth, Tibetan monks differ Acknowledgments
from controls in ways other than mindfulness that we were not We thank Emory-Tibet Science Initiative and Drepung Loseling Monas-
fully able to control. Seventh, selfishness scores were not tery for assistance in the criterion validity component of this study, and
10 RAINE AND UH

also Sonam Choepel and Tsondue Samphel for translation and administra- Eysenck, H. J., & Eysenck, S. (1991). Eysenck personality scales. London,
tion of the Selfishness Questionnaire from English into Tibetan for data UK: Hodder & Stoughton.
collection. Fehr, E., & Gachter, S. (2002). Altruistic punishment in humans. Nature,
415, 137–140. doi:10.1038/415137a
Feldman, G., Hayes, A., Kumar, S., Greeson, J., & Laurenceau, J. P. (2007).
Mindfulness and emotion regulation: The development and initial vali-
Funding dation of the Cognitive and Affective Mindfulness Scale–Revised
This work was supported by the GVR Khodadad Family Foundation, grant (CAMS–R). Journal of Psychopathology and Behavioral Assessment, 29,
no. 001. 177–190. doi:10.1007/s10862-006-9035-8
Flook, L., Goldberg, S. B., Pinger, L., & Davidson, R. J. (2015). Promoting
prosocial behavior and self-regulatory skills in preschool children
References through a mindfulness-based kindness curriculum. Developmental Psy-
chology, 51, 44–51. doi:10.1037/a0038256
Aarts, A. A., Anderson, J. E., Anderson, C. J., Attridge, P. R., Attwood, A., Fossati, A., Porro, F. V., Maffei, C., & Borroni, S. (2012). Are the DSM–IV
Axt, J., … Penuliar, M. (2015). Estimating the reproducibility of psy- personality disorders related to mindfulness? An Italian study on clini-
chological science. Science, 349, aac4716. doi:10.1126/science.aac4716 cal participants. Journal of Clinical Psychology, 68, 672–683.
Abdin, E., Koh, K. G. W. W., Subramaniam, M., Guo, M. E., Leo, T., Teo, doi:10.1002/jclp.21848
C., … Chong, S. A. (2011). Validity of the Personality Diagnostic Ques- Gao, Y., & Tang, S. (2013). Psychopathic personality and utilitarian moral
tionnaire–4 (PDQ–4) among mentally ill prison inmates in Singapore. judgment in college students. Journal of Criminal Justice, 41, 342–349.
Journal of Personality Disorders, 25, 834–841. doi:10.1521/ doi:10.1016/j.jcrimjus.2013.06.012
pedi.2011.25.6.834 Gleichgerrcht, E., & Young, L. (2013). Low levels of empathic concern pre-
American Psychiatric Association. (2013). Diagnostic and statistical man- dict utilitarian moral judgment. PLoS ONE, 8, e60418. doi: 10.1371/
ual of mental disorders (5th ed.). Washington, DC: Author. journal.pone.0060418
Ariely, D. (2012). The honest truth about dishonesty: How we lie to every- Glenn, A. L., Raine, A., & Schug, R. A. (2009). The neural correlates of
one, especially ourselves. New York, NY: HarperCollins. moral decision-making in psychopathy. Molecular Psychiatry, 14, 5–6.
Bardsley, N. (2008). Dictator game giving: Altruism or artefact? Experi- doi:10.1038/mp.2008.104
mental Economics, 11, 122–133. doi:10.1007/s10683-007-9172-2 Greene, J. D., Morelli, S. A., Lowenberg, K., Nystrom, L. E., & Cohen,
Bentham, J. (1789). An introduction to the principles of morals and legisla- J. D. (2008). Cognitive load selectively interferes with utilitarian
tion. Oxford, UK: Clarendon. moral judgement. Cognition, 107, 1144–1154. doi:10.1016/j.
Bentler, P. M. (1990). Comparative fit indexes in structural models. Psy- cognition.2007.11.004
chological Bulletin, 107, 238–246. doi:10.1037/0033-2909.107.2.238 Greene, J. D., Sommerville, R. B., Nystrom, L. E., Darley, J. M., & Cohen, J.
Bentler, P. M. (1995). EQS Structural Equations program manual. Encino, D. (2001). An fMRI investigation of emotional engagement in moral
CA: Multivariate Software. judgment. Science, 293, 2105–2108. doi:10.1126/science.1062872
Bentler, P. M. (2000). EQS 6 Structural Equation program manual. Encino, Haber, S. N., & Knutson, B. (2010). The reward circuit: Linking primate
CA: Multivariate Software. anatomy and human imaging. Neuropsychopharmacology, 35, 4–26.
Bentler, P. M., Berkane, M., & Kano, Y. (1991). Covariance structure analy- doi:10.1038/npp.2009.129
sis under a simple kurtosis model. In E. M. Keramidas (Ed.), Comput- Haidt, J. (2007). The new synthesis in moral psychology. Science, 316, 998–
ing science and statistics (pp. 463–465). Fairfax Station, VA: Interface 1002. doi:10.1126/science.1137651
Foundation of North America. Hall, J. R., Drislane, L. E., Patrick, C. J., Morano, M., Lilienfeld, S. O., &
Berinsky, A. J., Huber, G. A., & Lenz, G. S. (2012). Evaluating online labor Poythress, N. G. (2014). Development and validation of triarchic con-
markets for experimental research: Amazon.com’s Mechanical Turk. struct scales from the Psychopathic Personality Inventory. Psychologi-
Political Analysis, 20, 351–368. doi:10.1093/pan/mpr057 cal Assessment, 26, 447–461. doi:10.1037/a0035665
Birnie, K., Speca, M., & Carlson, L. E. (2010). Exploring self-compassion Hamilton, W. D. (1964). The genetical evolution of social behaviour: I.
and empathy in the context of mindfulness-based stress reduction Journal of Theoretical Biology, 7, 1–16. doi:10.1016/0022-5193(64)
(MBSR). Stress and Health, 26, 359–371. doi:10.1002/smi.1305 90038-4
Bouvard, M., Vuachet, M., & Marchand, C. (2011). Examination of the Harbaugh, W. T., Mayr, U., & Burghart, D. R. (2007). Neural responses to
screening properties of the Personality Diagnostic Questionnaire–4C taxation and voluntary giving reveal motives for charitable donations.
(PDQ–4C) in a non-clinical sample. Clinical Neuropsychiatry, 8, 151–158. Science, 316, 1622–1625. doi:10.1126/science.1140738
Caporael, L. R., Dawes, R. M., Orbell, J. M., & Van de Kragt, A. J. (1989). Holtzman, N. S., & Donnellan, M. B. (2017). A simulator of the degree to
Selfishness examined: Cooperation in the absence of egoistic incentives. which random responding leads to biases in the correlations between
Behavioral and Brain Sciences, 12, 683–699. doi:10.1017/ two individual differences. Personality and Individual Differences, 114,
S0140525X00025292 187–192. doi:10.1016/j.paid.2017.04.013
Casler, K., Bickel, L., & Hackett, E. (2013). Separate but equal? A compari- Hopwood, C. J., Donnellan, M. B., Ackerman, R. A., Thomas, K. M.,
son of participants and data gathered via Amazon’s MTurk, social Morey, L. C., & Skodol, A. E. (2013). The validity of the Personality
media, and face-to-face behavioral testing. Computers in Human Diagnostic Questionnaire–4 narcissistic personality disorder scale for
Behavior, 29, 2156–2160. doi:10.1016/j.chb.2013.05.009 assessing pathological grandiosity. Journal of Personality Assessment,
Cleckley, H. (1976). The mask of sanity (5th ed.). St. Louis, MO: Mosby. 95, 274–283. doi:10.1080/00223891.2012.732637
Costa, P. T., & McCrae, R. R. (1992). Revised NEO Personality Inventory Hu, L., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance
(NEO PI–R) and NEO Five-Factor Inventory (NEO–FFI) professional structure analysis: Conventional criteria versus new alternatives. Struc-
manual. Odessa, FL: Psychological Assessment Resources. tural Equation Modeling, 6, 1–55. doi:10.1080/10705519909540118
Cronbach, L., & Meehl, P. (1955). Construct validity in psychological tests. Hyler, S. E. (1994). PDQ–4 and PDQ–4C instructions for use. New York,
Psychological Bulletin, 52, 281–302. doi:10.1037/h0040957 NY: New York State Psychiatric Institute.
Cuthbert, B. N. (2014). The RDoC framework: Facilitating transition from Knutson, K. M., Krueger, F., Koenigs, M., Hawley, A., Escobedo, J. R., &
ICD/DSM to dimensional approaches that integrate neuroscience and Vasudeva, V., … Grafman, J. (2010). Behavioral norms for condensed
psychopathology. World Psychiatry, 13, 28–35. doi:10.1002/wps.20087 moral vignettes. Social Cognitive and Affective Neuroscience, 5, 378–
Dawkins, R. (2006). The selfish gene (30th anniversary ed.). Oxford, UK: 384. doi:10.1093/scan/nsq005
Oxford University Press. Koolen, S., Poorthuis, A., & van Aken, M. A. G. (2012). Cognitive distor-
Eckel, C. C., & Grossman, P. J. (1996). Altruism in anonymous dictator tions and self-regulatory personality traits associated with proactive
games. Games and Economic Behavior, 16, 181–191. doi:10.1006/ and reactive aggression in early adolescence. Cognitive Therapy and
game.1996.0081 Research, 36, 776–787. doi:10.1007/s10608-011-9407-6
 THE SELFISHNESS QUESTIONNAIRE 11

Kounou, K. B., Foli, A. A. D., Djassoa, G., Ametepe, L. K., Rieu, J., & Buddhist practitioners. Journal of Happiness Studies, 12, 575–589.
Mathur, A., … Schmitt, L. (2015). Childhood maltreatment and per- doi:10.1007/s10902-010-9218-6
sonality disorders in patients with a major depressive disorder: A com- Wilson, D. S., Near, D., & Miller, R. R. (1996). Machiavellianism: A synthe-
parative study between France and Togo. Transcultural Psychiatry, 52, sis of the evolutionary and psychological literatures. Psychological Bul-
681–699. doi:10.1177/1363461515572001 letin, 119, 285–299. doi:10.1037/0033-2909.119.2.285
Lee, I. A., & Preacher, K. J. (2013). Calculation for the test of the difference Wupperman, P., Marlatt, G. A., Cunningham, A., Bowen, S., Berking, M.,
between two dependent correlations with one variable in common Mulvihill-Rivera, N., & Easton, C. (2012). Mindfulness and modifica-
[Computer software]. tion therapy for behavioral dysregulation: Results from a pilot study
Lilienfeld, S. O., & Hess, T. H. (2001). Psychopathic personality traits and targeting alcohol use and aggression in women. Journal of Clinical
somatization: Sex differences and the mediating role of negative emo- Psychology, 68, 50–66. doi:10.1002/jclp.20830
tionality. Journal of Psychopathology and Behavioral Assessment, 23, XIV Bstan-ʼdzin-rgya, D. L. (2012). Beyond religion: Ethics for a whole
11–24. doi:10.1023/A:1011035306061 world. New York, NY: Random House.
Loehlin, J. C. (1992). Latent variable models: An introduction to factor, Yang, J., McCrae, R. R., Costa, P. T., Dai, X. Y., Yao, S. Q., Cai, T. S., &
path, and structural analysis. Hillsdale, NJ: Erlbaum. Gao, B. (1999). Cross-cultural personality assessment in psychiatric
Logie, K., & Frewen, P. (2015). Self/other referential processing following populations: The NEO-PI–R in the People’s Republic of China. Psy-
mindfulness and loving-kindness meditation. Mindfulness, 6, 778–787. chological Assessment, 11, 359–368. doi:10.1037/1040-3590.11.3.359
doi:10.1007/s12671-014-0317-z
Malthus, T. R. (1798). An essay on the principle of population. London, UK:
J. Johnson in St Paul’s Church-Yard. Appendix
McDonald, R. P. (1989). An index of goodness-of-fit based on non-
centrality. Journal of Classification, 6, 97–103. doi:10.1007/
BF01908590 The Selfishness Questionnaire (SQ)
Neff, K. D., & Germer, C. K. (2013). A pilot study and randomized con-
trolled trial of the mindful self-compassion program. Journal of Clinical
Psychology, 69, 28–44. doi:10.1002/jclp.21923 Instructions to participant
Patrick, C. J. (2006). Handbook of psychopathy. New York, NY: Guilford.
Patrick, C. J. (2010). Operationalizing the triarchic conceptualization of psy- We can’t always be charitable to others, and there are times
chopathy: Preliminary description of brief scales for assessment of bold- when you have to look after your own self-interests. Answer
ness, meanness, and disinhibition. Tallahassee, FL: Florida State the following questions as honestly as you can by indicating
University.
whether you: Disagree (0), Neither Agree nor Disagree (1), or
Piaget, J. (1951). The child’s conception of the world (Vol. 213). Lanham,
MD: Rowman & Littlefield. Agree (2) with each statement.
Raine, A., & Chen, F. R. (2018). Cognitive, Affective, and Somatic Empa-
thy Scale (CASES) for children. Journal of Clinical Child and Adoles- 1. I have no problem telling “white lies” if it will help me
cent Psychology, 47, 24–37. doi:10.1080/15374416.2017.1295383 achieve my goals.
Sellbom, M., & Phillips, T. T. (2013). An examination of the triarchic con-
2. I’m not too concerned about what is best for society in
ceptualization of psychopathy in incarcerated and nonincarcerated
samples. Journal of Abnormal Psychology, 122, 208–214. doi:10.1037/ general.
a0029306 3. Now and again I’ve manipulated my friends to gain an
Shonin, E., Van Gordon, W., Slade, K., & Griffiths, M. D. (2013). Mindful- advantage.
ness and other Buddhist-derived interventions in correctional settings: 4. At the end of the day I care mostly for myself, my family,
A systematic review. Aggression and Violent Behavior, 18, 365–372.
and friends who can help me.
doi:10.1016/j.avb.2013.01.002
Steiger, J. H. (1980). Tests for comparing elements of a correlation 5. I’ve occasionally put others down to achieve my goals.
matrix. Psychological Bulletin, 87, 245–251. doi:10.1037/0033- 6. I don’t give to charities.
2909.87.2.245 7. Even if it meant giving my kids an unfair advantage over
Tinghog, G., Andersson, D., Bonn, C., Johannesson, M., Kirchler, M., Kop- others, I’d do it for them.
pel, L., & V€astfj€all, D. (2016). Intuition and moral decision-making—
The effect of time pressure and cognitive load on moral judgment and
8. Sometimes you need to take advantage of other people
altruistic behavior. PLoS ONE, 11, e0164012.doi: 10.1371/journal. before they take advantage of you.
pone.0164012 9. I’m not always honest because honesty can end up
Tonnaer, F., Cima, M., Sijtsma, K., Uzieblo, K., & Lilienfeld, S. O. (2013). harming myself and others.
Screening for psychopathy: Validation of the Psychopathic Personality 10. When it comes to helping myself or helping others, I
Inventory–Short Form with reference scores. Journal of Psychopathol-
tend to help myself.
ogy and Behavioral Assessment, 35, 153–161. doi:10.1007/s10862-012-
9333-2 11. It’s not nice to exploit others, but there are times when
Trull, T. J., Useda, J. D., Costa, P. T., & McCrae, R. R. (1995). Comparison you simply need to.
of the MMPI–2 Personality Psychopathology Five (PSY–5), the NEO– 12. If there was only one space left on a lifeboat that a child
PI, and the NEO PI–R. Psychological Assessment, 7, 508–516. needed, I’d honestly have to take it for myself and my
doi:10.1037/1040-3590.7.4.508
family.
van Leeuwen, N., Rodgers, R. F., Gibbs, J. C., & Chabrol, H. (2014). Cal-
lous-unemotional traits and antisocial behavior among adolescents: 13. Quite often in life, it is more important to receive than to
The role of self-serving cognitions. Journal of Abnormal Child Psychol- give.
ogy, 42, 229–237. doi:10.1007/s10802-013-9779-z 14. I know I love rewards in life, even if there is a cost to
Wang, Y. P., Zhu, X. Z., Cai, L., Wang, Q., Wang, M. C., Yi, J. Y., & Yao, S. others.
(2013). Screening cluster A and cluster B personality disorders in Chi-
15. It’s better to save for a rainy day than to give to charities
nese high school students. BMC Psychiatry, 13, 116.doi: 10.1186/1471-
244X-13-116 where money can be misspent.
Wayment, H. A., Wiist, B., Sullivan, B. M., & Warren, M. A. (2011). Doing 16. If I’m honest, there are times when I put myself first,
and being: Mindfulness, health, and quiet ego characteristics among even if it’s someone else’s loss.
12 RAINE AND UH

17. If the choice was between killing someone or being 24. I mostly help those around me who will help me
killed, I’d kill. later.
18. I care for myself much more than I care for others.
19. I have sometimes dumped friends that I don’t need
Scoring instructions
anymore.
20. I sometimes lie to others for my own good, and theirs too. Each item has a score ranging from 0 to 2. Add scores as
21. Even when I see people in need, I don’t feel the urge to follows:
help them. Egocentric Selfishness: Questions 2, 6, 10, 13, 15, 18, 21, 23.
22. I go out of the way to exploit situations for my own Adaptive Selfishness: Questions 1, 4, 7, 9, 12, 17, 20, 24.
advantage. Pathological Selfishness: Questions 3, 5, 8, 11, 14, 16, 19, 22.
23. At the end of the day, I have to admit that I’m quite a Total Selfishness: add Egocentric C Adaptive C Pathological
selfish person. Selfishness scores