2020RBF - Social Comparison

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 16

The current issue and full text archive of this journal is available on Emerald Insight at:

https://www.emerald.com/insight/1940-5979.htm

Financial risk-taking related to Financial


risk-taking
individual risk preference, social
comparison and competition
Tommy G€arling, Dawei Fang and Martin Holmen
Centre of Finance, University of Gothenburg, G€oteborg, Sweden, and
Received 1 November 2019
Patrik Michaelsen Revised 17 January 2020
Department of Psychology, University of Gothenburg, G€oteborg, Sweden Accepted 17 January 2020

Abstract
Purpose – The purpose of this paper is to investigate how social comparison and motivation to compete
account for elevated risk-taking in fund management corroborated by asset market experiments when
performance depends on rank-based incentives.
Design/methodology/approach – In two laboratory experiments, university students (n1 5 240/n2 5 120)
make choices between risky and certain outcomes of hypothetical sums of money. Both experiments
investigate in which direction risky choices in an individual condition (individual risk preference) are shifted
when participants compare their performance to another participant’s performance (social comparison), being
instructed or not to outperform the other (incentive to compete).
Findings – In the absence of incentives to compete, participants tend to minimize the differences between
expected outcomes to themselves and to the other, but when provided with incentives to compete, they tend to
maximize these differences. An independent additional increase in risk-taking is observed when participants
are provided with incentives to compete.
Originality/value – Original findings include that social comparison does not evoke motivation to compete
unless incentives are offered and that increases in risk-taking depend both on what the other chooses and the
incentives.
Keywords Mutual fund industry, Performance evaluation, Financial risk-taking, Social comparison,
Competition
Paper type Research paper

1. Introduction
It is common that companies in the mutual fund industry seek to improve investment
managers’ performance by offering performance-based compensation (Ma et al., 2019).
Higher salaries are paid to those whose performance ranks highest, and at the same time,
those whose performance ranks lowest run the risk of demotion or even losing their jobs
(Chevalier and Ellison, 1999; Kempf et al., 2009). The rank-based personnel policies are
believed to motivate investment managers to compete for better performance. A drawback of
these policies is that they also tend to induce managerial risk-taking. The existing literature
has identified various conditions under which rank-order tournaments induce managerial
risk-taking, both in actual markets (Brown et al., 1996; Chevalier and Elison, 1997) and in
experimental asset markets (e.g. Fang et al., 2017; Kirchler et al., 2019) [1]. Our aim is to
investigate how social comparison and motivation to compete affect risk-taking.

This research was financially supported by Grant No. 2010-02449 to the Centre for Finance, School of
Business, Economics, and Law, University of Gothenburg, G€oteborg, Sweden, from the Swedish
Governmental Agency for Innovation Systems (VINNOVA), and Grant No. 2015-01713 to Martin
Holmen from the Swedish Research Council. Review of Behavioral Finance
The paper was presented at the Behavioral Finance Working Group conference at Queen Mary © Emerald Publishing Limited
1940-5979
University of London, June 12-13, 2017. We thank a discussant and other participants for comments. DOI 10.1108/RBF-11-2019-0153
RBF G€arling et al. (2019) review research of the relations between financial risk-taking, social
comparison and competition. In the next section, we derive hypotheses based on a selective
summary of the reviewed research. We then report two experiments that investigate how
social comparison and motivation to compete influence risk-taking. In the experiments, the
frequency of choices of a risky option made by participants individually is compared to the
frequency of their choices of the same risky option when information is disclosed about
another participant’s choices. The main results provide qualified support for the hypotheses
that social comparison (information about the other participant’s choices) shifts the
individual risk preference in the direction of minimizing the differences between the expected
outcomes to the participant and to the other, whereas motivation to compete (instructions to
outperform the other participant) shifts the individual risk preference in the direction of
maximizing the differences between expected outcomes. Whether risk-taking increases or
decreases thus depends on the other’s choices and the motivation to compete. An additional
independent increase in risk-taking is observed when participants are motivated to compete.

2. Hypotheses
2.1 Social comparison
An individual risk preference is recognized to be an important determinant of differences in
trading in asset market experiments (Noussair and Tucker, 2013; Palan, 2013; Powell and
Shestakova, 2016). In line with this, we propose Hypothesis H1 that the individual risk
preference is a baseline from which social comparison and motivation to compete may cause
shifts.
Comparing oneself to others provides performance feedback (Garcia et al., 2013, 2020).
Since social comparison frequently evokes the goal of not differing from others (Hodges, 2017;
Karau and Williams, 1993), the performance feedback may reduce motivation to compete.
This is observed in asset markets when investors make the same choices as others do
(Spyrou, 2013). This kind of herding may result from direct influences as when investors
imitate others because they trust the others’ knowledge more than they trust their own
knowledge (Andersson et al., 2009, 2014). Herding may also reflect motivation to cooperate
(Bowden, 2013). In herding investors copy others’ choices believing the outcome is beneficial,
whereas in cooperation others’ choices are copied only if the joint expected outcome is
beneficial.
We propose Hypothesis H2 that information about others’ choices shifts the individual
risk preference in the direction that minimizes the differences between the expected outcomes
to oneself and to others. If the expected outcomes are not known [2], the choices are instead
proposed to be aligned with the others’ choices. Possible motives include reliance on others’
knowledge to reduce uncertainty (Andersson et al., 2009, 2014) or willingness to cooperate
(Balliet et al., 2009; Savikin and Sheremeta, 2013).
Gibbons and Buunk (1999) develop a self-report measure (INCOM) to assess individual
differences in social comparison orientation. INCOM has two correlated subscales, one that
measure the tendency to make social comparison with respect to ability, the other with
respect to opinion [3]. We propose Hypothesis H3 that the stronger the social comparison
orientation with respect to ability, the larger the shifts of the individual risk preference in the
direction of minimizing the differences between the expected outcomes to oneself and to
others, or more frequently aligning choices with the others’ choices.

2.2 Competition
In previous empirical research investigating competition, participants only experience the
presence of others when working on some common task (the “mere physical presence” effect,
e.g. Guerin, 1986), are instructed to outperform physically or virtually present others without Financial
being offered monetary rewards or punishments (see review of psychological research by risk-taking
Murayama and Elliot, 2012), or are incentivized by monetary rewards or punishments
depending on their performance compared to others (as in all-pay auctions, lottery contests or
tournament experiments in economics, see review by Dechenaux et al., 2015).
If financial risk-taking has some competitive context (e.g. trading in asset markets, betting
on soccer games or horse races), people with knowledge may feel more confident and take
more risk even though objectively their knowledge does not increase their probability of
success (Fox and Tversky, 1995; Heath and Tversky, 1991). People may in contests increase
risk-taking because they perceive that the competitors are less skilled or lucky (Fox and
Weber, 2002). G€arling et al. (2019) propose that in risk-taking tournaments social comparison
is susceptible to a motivated-reasoning bias of being more skilled than the competitors which
induces overconfidence resulting in increased risk-taking.
Avoiding risk should in asset markets be preferred to taking risk if a safe option (not
investing) is more likely to be beneficial and the reverse if a risky option (investing) is more
likely to be beneficial. Whether this occurs in risk-taking tournaments is questioned by
observations of elevated risk-taking in experiments conducted by Eriksen and KvalØy (2017)
and Schedlinsky et al. (2016). When participants compete in groups for a prize to the winner,
acting as if risk-taking is believed to be a winning strategy, they choose to invest despite that
the expected value is negative and the dominant strategy is to not invest. Financial risk-
taking also increases for participants lagging behind (Eriksen and KvalØy, 2017;
Schedlinsky et al., 2016). In a similar vein, Schoenberg and Haruvy (2012) show that
upward social comparisons with the highest ranked competitor result in more overpricing of
assets than downward social comparisons with the lowest ranked competitor. Kuziemko et al.
(2014) find that risk-taking is more frequent if participants are ranked last than in other
positions.
Kirchler et al. (2019) conduct an experiment in which financial professionals invest more in
a risky asset when ranks are disclosed even if not payment-related. In the same asset market
experiment with reduced stakes, students invest more in the risky asset only when ranks are
payment-related. Rank information per se is thus sufficient to increase professionals’ risk-
taking, whereas payments need to be added to increase students’ risk-taking. The latter
finding is however inconsistent with the results of an experiment by Dijk et al. (2014) in which
students construct portfolios from stocks with positively (more risky) or negatively (less
risky) skewed return distributions. Information about ranks after each period has similar
effects on risk-taking no matter whether or not payments are associated with the ranks.
The reviewed studies suggest that both monetary and nonmonetary incentives may
increase risk-taking in risk-taking tournaments. We propose Hypothesis H4 that when
comparing oneself to others, monetary or nonmonetary incentives to compete make the
individual risk preference shift in the direction that maximizes the differences between the
expected outcomes to oneself and to other competitors, or by aligning choices away from
the others’ choices.
A meta-analysis by Murayama and Elliot (2012) finds no significant relationship between
performance and competition. In the studies included in the meta-analysis, competition is
operationalized as dispositional competiveness, nonmonetary incentives to compete or
monetary incentives to compete. Only dispositional competitiveness yields a significant
although weak effect. A proposed explanation of the null result is the operation of two
opposing mediating processes, pursuit of performance-approach goals facilitating
performance and pursuit of performance-avoidance goals undermining performance.
Dispositional competitiveness is defined as a desire to win contests. Houston et al. (2002a)
and Harris and Houston (2010) develop a self-report measure referred to as the Competiveness
Index (CI) [4]. Houston et al. (2002b) observe a significant positive correlation between CI and a
RBF measure of need for achievement, thus suggesting that dispositional competitiveness
increases the pursuit of performance-approach goals that shifts the individual risk
preference. But since Hypothesis H4 implies that social comparison provides performance
feedback, we propose Hypothesis H5 that both dispositional competiveness and social
comparison orientation strengthen shifts of the individual risk preference in the direction that
the differences between the expected outcomes to oneself and to others are maximized, or that
choices are more frequently aligned away from others’ choices.

3. Method
In order to investigate whether social comparison and motivation to compete cause shifts
in the individual risk preference as proposed in Hypothesis H1, in one condition the
individual risk preference is measured following Dohmen et al.’s (2011) procedure of
observing choices between a fixed risky monetary option and varying safe monetary
options. The choices in this individual condition are then repeated to investigate shifts
caused by information about others’ choices. The participants are informed about the
choices ostensibly made before them by a single other participant facing the same choices.
Providing information about the choices of a single participant would likely make the
influence maximally strong (Garcia and Tor, 2009; Tor and Garcia, 2010). The other
participant is not physically present to minimize influences of physical presence on risk-
taking (Chou and Nordgren, 2017).
In one between-groups condition referred to as social comparison, following the
measurement of individual risk preference, the participants are presented information
about the other’s choices. The purpose is to investigate whether according to Hypothesis H2
the differences between the expected outcomes are minimized. In another between-groups
condition, referred to as competition, the participants are instructed to outperform the other.
The purpose is to investigate whether according to Hypothesis H4 the differences between
the expected outcomes are maximized.
In Experiment 1 and in one condition of Experiment 2, minimizing the differences between
the expected outcomes is achieved by aligning the choices with the other’s choices and
maximizing the differences by aligning the choices away from the other’s choices. In another
condition of Experiment 2, the reverse pattern of choices in the social comparison and
competition conditions minimizes, respectively, maximizes the differences between the
expected outcomes, thus investigating whether the participants minimize (maximize) the
expected outcomes when they align their choices with (away from) the other’s choices.
Investigating whether risk-taking differs whether the other takes risk or not, the other
participant chooses the safe options in Experiment 1 and the risky option in Experiment 2.
University students recruited as participants are expected to have a stronger motivation
to compete than the general population (Kirchler et al., 2019). To investigate Hypotheses H3
and H5, we use a measure based on INCOM (Gibbons and Buunk, 1999) to assess social
comparison orientation, and a measure based on CI (Harris and Houston, 2010) to assess
dispositional competitiveness.

3.1 Experiment 1: comparison with risk-averse other


According to Hypothesis H1 the risk preference assessed in the individual condition should
influence choices in the social comparison and competition conditions. In the social
comparison condition when the other chooses the safe option, according to Hypothesis H2 the
individual risk preference is expected to shift such that the choices are more frequently
aligned with the other’s choices in an attempt to minimize the differences between the
expected outcomes, thereby decreasing risk-taking compared to the individual condition. In
the competition condition in which participants are asked to outperform the other, according Financial
to Hypothesis H4 the individual risk preference is expected to shift in a way such that choices risk-taking
are more frequently aligned away from the other’s choices in an attempt to maximize the
differences between the expected outcomes, thereby increasing risk-taking compared to the
individual condition.
In risk-taking tournaments, the outcomes are losses or gains compared both to the
contestants’ endowment and the competitors’ outcomes. We refer to the former as absolute
gains and losses and to the latter as relative gains and losses. For instance, an absolute gain is
a relative gain if others gain less and a relative loss if others gain more, whereas an absolute
loss is a relative loss if others loose less and a relative gain if others loose more. An additional
question addressed in Experiment 1 is how the results differ if absolute losses are mixed with
relative losses. Previous research demonstrates that absolute losses receive more attention
than absolute gains (Yechiam and Hochman, 2013), are weighed more heavily in direct
ratings of emotional impact (McGraw et al., 2010), and referred to as loss aversion (Kahneman
and Tversky, 1979), are weighed more heavily in choices between gambles with both absolute
gains and losses (Charpentier et al., 2016). In contrast, Dijk (2017) proposes that relative gains
weigh more than relative losses. Dijk finds in support that a majority of undergraduates are
more risk-taking when comparing themselves to others. In Linde and Sonnemans (2012),
undergraduates choose between low-variance and high-variance gambles with higher, lower
or equal outcomes to a competitor. Consistent with that the reflection effect (Kahneman and
Tversky, 1979) is reversed for relative outcomes compared to absolute outcomes, lower
variance (risk) is preferred when the relative outcome is a loss, whereas higher variance (risk)
is preferred when the relative outcome is a gain. The difference is however not large,
suggesting that individual risk preference has a substantial impact. In addition, no
correlations are found with participants’ ratings of how important it is to compare themselves
to others and to outperform others, respectively.
In the individual, social comparison and competition conditions, participants are
presented gambles with no absolute losses or mixed gambles with both absolute gains
and losses for the same risk-neutral expected values. If absolute losses have larger impacts
than relative losses and absolute gains equal or less impact than relative gains, risk-taking
will decrease for absolute losses in the individual condition but increase in the social
comparison and competition conditions.
Several studies have shown that explicit descriptions of expected outcomes influence
choices (Soman, 2004). A possible caveat is then that implicitly described relative gains and
losses are neglected. To discount this possibility, for approximately half of the participants,
explicit descriptions of the relative outcomes are added.
3.1.1 Participants. Participants are recruited through e-mails sent to volunteering
undergraduates in a pool maintained by the Psychology Department at University of
Gothenburg. Two hundred and forty undergraduates (179 women, median age 5 26 years,
quartile range 5 4 years) from several different study programs (18 pursuing majors in
Psychology and 15 in Economics) reply to the e-mails that direct them to a web page with the
online experiment programmed in Qualtrics (www.Qualtrics.com). Participants are
compensated by a lottery ticket worth the equivalent of approximately US$ 3.50.
3.1.2 Design. An approximately equal number of participants (between 29 and 31) are
randomly assigned to eight different groups in a 2 by 2 by 2 factorial design. The orthogonal
between-group factors are competition vs social comparison, absolute vs no absolute losses
and explicit vs implicit descriptions of the relative outcomes.
3.1.3 Procedure. Those receiving the e-mails are asked to take part in a study of risk-taking
with hypothetical money. They are informed that the study will take about 10 min, that they
are free to quit whenever they wish, that the data will only be used for research purposes, that
their identity will not be disclosed, and that after having answered all questions, they will
RBF electronically receive information about the aggregated results. Additional information
includes what device they may use, that they are required to find a quiet place to individually
do all tasks without being interrupted, and that they are not allowed to use a calculator or
Internet searches. When those choosing to participate access the experimental material on the
web page, the information in the email is partially repeated. Before starting the experiment,
the participants indicate their sex, age, main study program and number of semesters at the
university. More than half of them (135) use smartphones, the remaining (105) stationary
computers, laptops or pads. Median time of participation is 285 sec with a quartile range of
86 sec.
On the following separate pages, in the individual condition the participants make seven
choices between a fixed risky monetary option and a safe monetary option (see translated
instructions in Appendix). In different choices the safe option varies from SEK 300 to SEK
2,100 in steps of SEK 300 presented according to individualized random orders. The risky
option is always a 50% chance of receiving SEK 4,800 or nothing for an approximately half of
the participants assigned to the no absolute loss condition, and a 50% chance of receiving
SEK 6,000 or losing SEK 1,200 (absolute loss) for the remaining participants assigned to the
absolute loss condition. The instructions stress that participants should choose the safe
option (labeled A) or the risky option (labeled B) as they would do if the outcomes are real
money [5]. In comparison, a monthly student scholarship is approximately SEK 10,000.
In four of the treatment groups referred to as the social comparison condition, participants
then make the same seven choices as in the individual condition (with absolute losses or no
absolute losses) informed about the choices made by another participant facing the same
choices before them and to whom they may compare themselves. The safe options are
presented according to individualized random orders. The procedure is identical in the
competition condition to which the remaining four groups are assigned, except that the
participants in this condition are instructed to outperform the other participant.
In two of the groups in the social comparison and competition conditions, respectively,
explicit descriptions of the relative outcomes (the certain outcomes subtracted from the risky
outcomes) are added. For choices of the risky option, the added text reads “. . .SEK X more
than the other receives or SEK Y less than the other receives”. Depending on the amount of the
safe option, X varies for no absolute losses from SEK 4,500 to SEK 2,700 in steps of SEK 300,
for absolute losses from SEK 5,700 to SEK 3,900 in steps of SEK 300. Y varies for no absolute
losses from SEK 300 to SEK 2100 in steps of SEK 300, and for absolute losses from SEK 1,500
to SEK 3,300 in steps of SEK 300.
After the participants finish the experiment, they answer some questions about their
awareness of influence. Of all the participants, 76 (31.7%) participants believe that they are
influenced by the choices made by the other participant. The frequency is significantly higher
in the competition condition (48) than in the social comparison condition (28),
χ 2(1) 5 9.79, p 5 0.002.
Additional questions [6] are asked to assess participants’ social comparison orientation
(SCO) and dispositional competitiveness (DC), respectively. Three scales are chosen with high
loadings on the ability dimension of INCOM (Gibbons and Buunk, 1999) (“I often compare
myself with others with respect to what I have accomplished”, “If I want to find out how well I
have done something, I compare what I have done with how others have done” and “I always
pay a lot of attention to how I do things compared to how others do things”), and three scales
from the CI with high loadings on enjoyment of competition (Harris and Houston, 2010) (“I am a
competitive individual”, “I like to compete”, and “I feel satisfaction from competition against an
opponent”). The statements are rated on 7-point numerical scales ranging from “do not at all
agree” (1) to “fully agree” (7). We also include three scales used by Kirchler et al. (2019) to
measure importance of social comparison (“How important is it what others think about you?”),
importance of competition (“How important is it to be the best at what you do?”) and
importance of financial success (“Social status is primarily defined by financial success”). The Financial
ratings of the first two are obtained on 7-point numerical importance scales ranging from “not risk-taking
important at all” (1) to “very important” (7), and the ratings of the last on a 7-point numerical
scale of agreement ranging from “do not at all agree” (1) to “fully agree” (7). In order to
counterbalance order effects, three scales including one from each set are presented on
separate consecutive pages.
3.1.4 Results. Number of choices of the risky option (coded 1 vs 0 for the safe option) is
calculated for each individual, then submitted to analyses of variance and covariance to test
the significance of the treatment effects. As Model 1 in Table 1 shows, in the individual
condition risk-taking is significantly lower when there are absolute losses (Mean 5 2.10, 95%
CI[1.68, 2.52]) than when there are no absolute losses (Mean 5 3.73, 95% CI[3.31, 4.16]). Model
2 shows that Hypothesis H1 is supported in that the individual risk preference (Mean 5 2.78,
95% CI[2.48, 3.07]) has a significant effect on risk-taking in the social comparison and
competition conditions. When adjusted for the individual risk preference, risk-taking is in
accordance with Hypothesis H2 lower in the social comparison condition (adj. mean 5 2.71,
95% CI[2.45, 2.98]) than in the individual condition but only marginally. Risk-taking is in
accordance with Hypothesis H4 higher in the competition condition (adj. mean 5 2.92, 95% CI
[2.66, 3.19]) than in the individual and social comparison conditions. The difference between
absolute losses (adj. mean 5 2.63, 95% CI[2.36, 2.90] and no absolute losses (adj. mean 5 3.01,
95% CI[2.73, 3.28]) is reduced although still significant. Adding absolute and relative losses
significantly reduces risk-taking more when explicitly described (adj. mean 5 2.58, 95% CI
[2.31, 2.85]) than when implicitly described (adj. mean 5 3.06, 95% CI[2.80, 3.32]).

3.2 Experiment 2: comparison with risk-taking other


The same choices are made in Experiment 2 as in Experiment 1 except that the other
participant chooses the risky option in the social comparison and competition conditions. In
an otherwise comparable condition to Experiment 1, the outcomes are the same to the
participants and the other. According to Hypothesis H2, the participants are expected in the
social comparison condition to align their choices with the other’s choices such that risk-
taking increases compared to the individual risk preference. According to Hypothesis H4, the
participants are expected in the competition condition to align their choices away from the
other’s choices such that risk-taking decreases compared to the individual risk preference.
In order to investigate whether the differences between the expected outcomes to the
participant and the other are minimized in the social comparison condition vs maximized in
the competition condition, for another group of the participants the expected risky outcome is
not the same but reversed. If the participant chooses the risky option, with equal probability

Model 1 Model 2
Df MS F p Df MS F p

Constant 1 1544.42 326.24 <0.001 1 57.24 30.50 <0.001


Individual risk preference 1 549.60 292.81 <0.001
Gender 1 12.70 2.68 0.103 1 0.03 0.01 0.906
Competition vs social comparison 1 2.72 1.45 0.230
Absolute losses vs no absolute losses 1 160.92 33.99 <0.001 1 7.55 4.02 0.046
Explicit vs implicit descriptions 1 13.42 7.15 0.008
Error 237 4.73 234 1.888
Adj R2 0.13, F(2, 237) 5 18.25, p < 0.001 0.61, F(5, 234) 5 76.73, p < 0.001
Note(s): Analysing the number of choices of the risky option (ranging from 0 to 7) in individual condition
(Model 1) and analysis of covariance of the number of choices of the risky option in social comparison and
competition conditions with the number of choices of the risky option in individual condition (individual risk Table 1.
preference) as covariate (Model 2) (Experiment 1) Analysis of variance
RBF either the participant receives the gain outcome and the other nothing or the other receives the
gain outcome and the participant nothing. Choosing the safe option results in a certain
absolute gain and with equal probability a relative loss or gain depending on the other’s
outcome. According to Hypothesis H2, risk-taking is expected to decrease in the social
comparison condition since this minimizes the differences between the expected outcomes.
According to Hypotheses H4, risk-taking is expected to increase in the competition condition
since this maximizes the differences between expected outcomes. The expected choices are
thus according to the hypotheses reversed compared to the condition in which the outcomes
are the same.
3.2.1 Participants. Another 120 undergraduates (96 women, median 5 29 years, quartile
range 5 4 years) is recruited through e-mails sent to the pool of volunteering undergraduates
maintained by the Department of Psychology. Those recruited are enrolled in several
different study programs. Ten participants are majors in Psychology and eight in Economics.
Participants are directed to a web page to do the online experiment programmed in Qualtrics.
They are compensated by a lottery ticket worth the equivalent of approximately US$ 3.50.
3.2.2 Design. An equal number of participants are randomly assigned to four different
groups in a 2 by 2 factorial design. The orthogonal between-group factors are social
comparison vs competition and same vs reversed outcomes.
3.2.3 Procedure. A majority of participants (66) uses smartphones, the remaining
participants (54) stationary computers, laptops or pads. Median time for participation is 306
sec with a quartile range of 86 seconds.
The procedure is the same as in Experiment 1 in that choices in the individual condition
precede choices in the social comparison and competition conditions. The choices are the
same except that in the reversed-outcome condition, the participants either gain when the
other loses or the reverse. The instructions read (cf. Appendix): “If you choose B, the chance is
50% that you receive the larger sum of money and the other participant nothing or that you
receive nothing and the other participant the larger sum of money”. There are no absolute
losses and the descriptions of the relative outcomes are implicit.
Participants answer the same additional questions as in Experiment 1 after making all
choices. Twenty six (21.7%) participants believe they are influenced by the other’s choices, a
significantly higher number in the competition condition (20) than in the social comparison
condition (6), χ 2(1) 5 9.62, p 5 0.002.
3.2.4 Results. Number of choices of the risky option (coded 1 vs 0 for the safe option) is
calculated for each individual. The results of an analysis of variance on the number of risky
choices in the social comparison and competition conditions (see Table 2) show in support of
Hypothesis H1 that risk-taking in the individual condition (mean 5 3.69, 95% CI[3.33, 4.05]) is
significant. After adjustments for the individual risk preference, risk-taking in the

Df MS F p

Constant 1 20.25 13.38 <0.001


Individual risk preference 1 251.48 166.12 <0.001
Gender 1 0.01 0.01 0.922
Competition vs social comparison 1 20.84 13.77 <0.001
Same vs reversed outcomes 1 2.72 1.80 0.182
Competition X Same outcomes vs reversed outcomes 1 9.47 6.26 0.014
Error 114 1.51
Adj R2 0.60, F(5, 114) 5 37.29, p < 0.001
Note(s): Analysing the number of choices of the risky option (ranging from 0 to 7) in social comparison and
Table 2. competition conditions with the number of choices of the risky option in individual condition (individual risk
Analysis of variance preference) as covariate (Experiment 2)
competition condition is significantly more frequent (adj. mean 5 4.12, 95% CI[3.78, 4.47]) Financial
than in the social comparison condition (adj. mean 5 3.28, 95% CI[2.90, 3.66]). The interaction risk-taking
between comparison vs competition and reversed outcome vs same outcome is also
significant. Figure 1 plots the adjusted means (filled circles) corresponding to the interaction.
When the outcomes are the same, the difference is not significant between the adjusted mean
in the competition condition (adj. mean 5 3.99, 95% CI[3.52, 4.46]) and the adjusted mean in
the social comparison condition (adj mean 5 3.71, 95% CI[3.21, 4.21]), │t(56)│ < 1. When the
outcomes are reversed, the difference is larger and significant (adj. mean 5 4.25, 95% CI[3.79,
4.72] vs adj. mean 5 2.85, 95% CI[2.37, 3.32]), t(56) 5 4.56, p < 0.001. In the social comparison
condition the adjusted mean difference between reversed and same outcomes is significant,
t(56) 5 2.82, p 5 0.007, but not in the competition condition, │t(56)│ < 1. After adjusting for
also the main effect of competition (open circles in Figure 1), in support of Hypotheses H2 and
H4 a cross-over interaction is revealed showing that risk-taking is more frequent in the
competition condition than in the social comparison condition for reversed outcomes to the
participant and the other but less frequent for same outcomes to the participant and the other.

3.3 Social comparison orientation and dispositional competitiveness


Based on the results of both experiments we investigate Hypothesis H3 proposing that social
comparison orientation is positively correlated with choices minimizing the difference
between the expected outcomes in the social comparison conditions, and Hypothesis H5 that
dispositional competitiveness and social comparison orientation are both positively
correlated with choices maximizing the difference between the expected outcomes in the
competition conditions.
A principal component analysis is first performed on the ratings in Experiments 1 and 2
of Social Comparison Orientation (SCO), Dispositional Competitiveness (DC), and the
additional three ratings of importance of social comparison, competition and financial
success. Three components are extracted with eigenvalues of 3.65, 1.86 and 0.96,
respectively. An oblique rotation results in positive loadings higher than 0.80 for the three

7 7
Same Outcomes Reversed Outcomes
6 6

5
Adjusted Mean Risk Taking

4 4

3 3

2 2

1 1
Figure 1.
Adjusted mean choices
of risky option in
0 0
experiment 2
Social Comparison Competition Social Comparison Competition
RBF SCO ratings on the first component, for the three DC ratings on the second component
(correlating r 5 0.26 with the first component), and for the rating of importance of financial
success (IFS) on the third component (correlating r 5 0.07 with both the first and second
components). Indexes of the first two components are constructed by averaging the ratings
with high loadings (SCO: mean 5 4.34, 95% CI[4.19, 4.50]), α 5 0.86; DC: mean 5 3.95, 95%
CI[3.78, 4.12]), α 5 0.89). The SCO and DC indexes correlate 0.32 (p < 0.001), IFS (M 5 3.39,
95% CI[3.19, 3.52]) correlates r 5 0.21 (p < 0.001) with SCO and r 5 0.12 (p 5 0.028) with DC,
the rating of importance of social comparison correlates r 5 0.45 (p < 0.001) with SCO and
r 5 0.03 (p 5 0.596) with DC, and the rating of importance of competition correlates r 5 0.23
(p < 0.001) with the rating of importance of social comparison, r 5 0.35 (p < 0.001) with DC,
and r 5 0.38 (p < 0.001) with SCO. Adding the importance ratings to the SCO and DC
indexes decreases αs. Analyses of variance performed separately on the SCO index, DC
index and IFS rating yield no significant effects (p > 0.05) of the experimental treatments,
thus failing to show any reverse carry over.
In order to make possible to use all the data in the same correlation analyses, choices of the
risky option are reverse coded in Experiment 2 when the outcomes to the participant and the
other are the same. The choices then always maximize the differences between the expected
outcomes. Table 3 shows that when controlling for the individual risk preference, in
agreement with Hypothesis H3 the correlation with the SCO index is negative in the social
comparison conditions, and in agreement with Hypothesis H5 both the correlations with the
SCO index and the DI index are positive in the competition conditions. However, the only
significant correlation is between the rating of IFS and the maximal differences between the
expected outcomes in the competition condition.

4. Conclusions and discussion


We tested the following hypotheses in the experiments: (H1) individuals’ risk preference is
affected by the information about another’s choices; (H2) if comparisons are made with
another’s choices and there is no incentives to compete, the individual risk preference shifts in
the direction that minimizes differences in the expected outcomes compared to the other’s
choices of a risky or safe option and (H4) if there are incentives to compete, the individual risk
preference shifts in the direction that maximizes the differences in the expected outcomes
compared to the other’s choices of a risky or safe option.
The results of both experiments support Hypothesis H1 in that the individual risk
preference influences risk-taking when the information about the other participant’s choices
is disclosed. This finding supports that contextual influences on risk-taking is partly
counteracted by an individual risk preference correlated with relatively stable socioeconomic
factors (e.g. Dohmen et al., 2011). The observation that information about the other
participant’s choices does not have substantial impacts should be viewed in this light. In fact,

Social comparison condition (n 5 184) Competition condition (n 5 176)


SCO index DC index IFS rating SCO index DC index IFS rating

Social comparison 0.05 (0.267) 0.06 (0.225) 0.01 (0.453) 0.02 (0.377) 0.05 (0.259) 0.17 (0.014)
or competition
condition
Table 3. Note(s): Correlations between the number of risky choices maximizing differences between expected
Product moment outcomes in experiments 1 and 2, social comparison orientation (SCO) index, dispositional competitiveness (DC)
correlations (one-tailed index and rating of influence of financial success (IFS) in social comparison and competition conditions.
p values in (Adjustments are made for the correlations with the number of choices of the risky option in the individual
parentheses) condition and the correlations with the other indexes are partialled out, and adjustments)
a minority of the participants retrospectively report that they are influenced by the other’s Financial
choices, fewer in the social comparison than in the competition condition. The results for both risk-taking
experiments displayed in Figure 2 substantiate the important role of the individual risk
preference in that close to half of the choices when the other’s choices are disclosed (in
Experiment 1 choices of the safe option and in Experiment 2 choices of the risky option with
the same or reversed outcomes) do not differ from those made when the other’s choices are not
disclosed. This is consistent with the finding in Linde and Sonnemans (2012, 2015) that
individual risk preference is strongly affected by the information about others’ choices.
Figure 2 shows that consistent with Hypotheses H2 and H4, we observe that more
frequently in the competition condition than in the social comparison condition participants
shift the individual risk preference in the direction that maximizes the differences in the
expected outcomes compared to the other’s expected outcomes. These differences in both
experiments are however only statistically significant in Experiment 2. The results of
Experiment 1 are similar compared to the results in the condition of Experiment 2 in which
the outcomes are the same to the participant and the other, thus suggesting that it makes little
difference whether the other takes risk or not. However, when in Experiment 2 the risky
outcomes are reversed, the choices are no longer aligned with the other’s choices in the social
comparison condition or away from the other’s choices in the competition condition. In the
former condition, the safe options tend to be chosen instead of the same risky option as the
other chooses. When it is not possible to attain the same outcomes, the participants thus tend
to prefer a possible relative loss to foregoing an absolute gain. This is additional evidence for
that social comparison alone does not evoke motivation to compete. The main effect of
competition in Experiment 2, which accounts for the lower percent of choices of maximizing
the differences between the expected outcomes in Figure 2, suggests that when motivated to
compete, participants increase risk-taking. This is thus observed in our choice experiments as
in the trading-like experiments conducted by Eriksen and KvalØy (2017) and Schedlinsky
et al. (2016).
It may be argued that the weak effect of social comparison is the result of inattention to the
relative outcomes. Although this possibility cannot by fully ruled out, the results for explicit

Differing from No Social comparison: 44.0%, [37.0%, 47.0%]


individual choices? Competition: 47.2%, [39.8%, 54.6%]

Yes

Maximizing
differences No
between the Social comparison: 66.0%, [56.9%, 75.1%]
expected Competition: 43.0%, [32.9%, 53.1%]
outcomes?

Figure 2.
Yes
Percent choices and
Social comparison: 34.0%, [24.9%, 43.1%] 95% confidence
Competition: 57.0%, [46.9%, 67.1%]
intervals
RBF descriptions in Experiment 1 suggest that relative outcomes are attended. Explicitly
described outcomes still reduce risk-taking when absolute and relative outcomes are added.
This is an indication of that explicit vs implicit descriptions play some role, as has been found
in other choice experiments (Soman, 2004).
Another finding is the demonstration of the role of absolute losses compared to losses
relative to the other’s outcomes. The results of Experiment 1 show in accordance with loss
aversion in Prospect Theory (Kahneman and Tversky, 1979) that risk-taking decreases when
there are absolute losses, implying that absolute losses are weighed more than absolute gains.
The decrease in risk-taking is slightly larger than in the individual condition when the other’s
choices are disclosed, failing to support that relative gains are weighed more than relative
losses (Dijk, 2017; Linde and Sonnemans, 2012). It also highlights that in the experiments by
Eriksen and KvalØy (2017) and Schedlinsky et al. (2016), normal returns may have an effect
independent of the relative outcomes. For instance, Schedlinsky et al. (2016) show that the
outcomes’ expected values influences risk-taking.
In analyzing the results of Experiments 1 and 2 together, SCO and DC influence
differences between the expected outcomes to the participant and to the other in the
hypothesized direction (Hypotheses H3 and H5) but fail to reach significance. Only the
correlation with importance of financial success is significant. It is conceivable that
the results are different than in previous studies (e.g. Garcia et al., 2013, 2020; Gibbons and
Buunk, 1999) because in contrast to these studies, financial risk-taking is investigated. The
correlation with the importance of financial success may reflect individual differences in the
desire to participate in contests about money.
In concluding, our research shows that an important influence on increased risk-taking is
incentives that increase motivation to compete. Only information about another’s choices is
not sufficient. Yet, in Experiment 1 risk-taking in the competition condition is not as high as
expected. The reason may be that increasing risk-taking also increases the possibility of
losing, augmented by the fact that losses are both absolute and relative. Thus, fear of losing
may also play a role. In Experiment 2 when the other takes risk and the outcomes are
reversed, the possibility to lose relative to the other is the same whether increasing or
decreasing risk-taking, whereas increasing risk-taking may result in a higher relative gain. In
line with G€arling et al. (2019), that the other chooses risk may in the competition condition
boost motivation to compete due to overconfidence if social comparison is biased in the
direction of being more skilled or lucky than the other. Although the results in the competition
condition of Experiment 2 are similar to the results of Eriksen and KvalØy (2017) and
Schedlinsky et al. (2016), the explanation by G€arling et al. differs from theirs.
The results of this research cannot be directly generalized to how social comparison and
motivation to compete affect actual fund management but call for future research that moves
toward such generalizations. One research topic is to identify factors that induce asset market
participants with tournament incentives to cooperate instead of to compete (cf. Fang et al.,
2017; Savikin and Sheremeta, 2013). The role of overconfidence caused by rank-order
incentives is another topic worth investigating (G€arling et al., 2019). Of several available
definitions and methods of measuring overconfidence (Glaser et al., 2007), studies should use
measures of beliefs of how one’s own skill compares to others’ skill.

Notes
1. A vast number of rank-order tournament experiments (for a review, see Dechenaux et al., 2015)
investigate optimal effort choices by observing contestants choosing between outcome distributions
with different means associated with costly effort. As noted by Nieken and Sliwka (2010), choices
between outcome distributions with different variances (e.g. risky and safe outcomes) have been
investigated less frequently. We primarily focus on such risk-taking tournaments, which arguably
are the most relevant for applications to mutual fund managements.
2. In asset market experiments with rank-based incentives, the expected outcomes to oneself and Financial
competitors may not be sufficiently transparent to be known. Traders may then align their choices
with (away from) others’ choices to minimize (maximize) the difference between the expected outcomes. risk-taking
3. In INCOM agreement ratings are made of 11 statements. The instructions read: “Most people
compare themselves from time to time with others. . . . There is nothing particularly good or bad
about this type of comparison . . . We would like to find out how often you compare yourself to other
people. . . .” An example of statements about self-other comparisons of ability is “I always pay a lot of
attention to how I do things compared to how others do things”. Psychometric analyses reported in
Gibbons and Buunk (1999) yielded satisfactory reliability (αs ranging from 0.77 to 0.85), temporal
stability (test-retest rs ranging from 0.60 to 0.72) and convergent and discriminant validity.
4. CI consists of agreement ratings of 14 statements forming two orthogonal subscales labeled
Enjoyment of Competition (e.g. “I enjoy competing against an opponent”) and Contentiousness (e.g. “I
try to avoid arguments”). Both subscales have high reliability (αs of 0.93 and 0.82 and test-retest rs of
0.85 and 0.78).
5. In a review of 74 studies, Camerer and Hogarth (1999) are unable to demonstrate substantial effects
of real monetary incentives on average performance in judgment and decision-making tasks.
Another study shows that large fictitious monetary outcomes (as in the present experiments) may be
stronger incentives than small real monetary outcomes (K€ uhberger et al., 2002).
6. In order to increase the statistical power, the results for these questions combined with the results for
the same questions in Experiment 2 are reported below in a separate section.

References
Andersson, M., G€arling, T. and Hedesstr€om, M. (2014), “A social-psychological perspective on herding
in stock markets”, Journal of Behavioral Finance, Vol. 15 No. 3, pp. 226-234.
Andersson, M., Hedesstr€om, M. and G€arling, T. (2009), “Social influence on predictions of simulated
stock prices”, Journal of Behavioral Decision Making, Vol. 22 No. 3, pp. 271-279.
Balliet, D., Parks, C.D. and Joireman, J. (2009), “Social value orientation and cooperation in social
dilemmas: a meta-analysis”, Group Processes andand Intergroup Relations, Vol. 12 No. 4,
pp. 533-547.
Bowden, M.P. (2013), “Social learning, herd behavior and information cascades: a review of the recent
developments in relation to their criticism”, Journal of Economic Research, Vol. 18 No. 3,
pp. 205-236.
Brown, K.C., Harlow, W.V. and Starks, L.T. (1996), “Of tournaments and temptations: an analysis of
managerial incentives in the mutual fund industry”, Journal of Finance, Vol. 51 No. 1, pp. 85-110.
Camerer, C.F. and Hogarth, R.M. (1999), “The effects of financial incentives in experiments: a
review and capital-labor-production framework”, Journal of Risk and Uncertainty, Vol. 19
Nos 1-3, pp. 7-42.
Charpentier, C.J., De Neve, J.-E., Li, X., Roiser, J.P. and Sharot, T. (2016), “Models of affective decision
making: how do feelings predict choice?”, Psychological Science, Vol. 27 No. 6, pp. 763-775.
Chevalier, J. and Ellison, G. (1997), “Risk taking by mutual funds as a response to incentives”, Journal
of Political Economy, Vol. 105 No. 6, pp. 1167-1200.
Chevalier, J. and Ellison, G. (1999), “Career concerns of mutual fund managers”, Quarterly Journal of
Economics, Vol. 114 No. 2, pp. 389-432.
Chou, E.Y. and Nordgren, L.F. (2017), “Safety in numbers: why the mere physical presence of others
affects risk-taking behaviors”, Journal of Behavioral Decision Making, Vol. 30 No. 3, pp. 671-682.
Dechenaux, E., Kovenock, D. and Sheremeta, R.M. (2015), “A survey of experimental research on
contests, all-pay auctions and tournaments”, Experimental Economics, Vol. 18 No. 4, pp. 609-669.
Dijk, O. (2017), “For whom does social comparison induce risk-taking?”, Theory and Decision, Vol. 82
No. 4, pp. 519-541.
RBF Dijk, O., Holmen, M. and Kirchler, M. (2014), “Rank matters: the impact of social competition on
portfolio choice”, European Economic Review, Vol. 66 No. 3, pp. 97-110.
upp, J. and Wagner, G.G. (2011), “Individual risk
Dohmen, T., Falk, A., Hursman, D., Sunde, U., Sch€
attitudes: measurement, determinants, and behavioral consequences”, Journal of the European
Economic Association, Vol. 9 No. 3, pp. 522-550.
Eriksen, K.W. and KvalØy, O. (2017), “No guts, no glory: an experiment on excessive risk taking”,
Review of Finance, Vol. 21 No. 3, pp. 1327-1351.
Fang, D., Holmen, M., Kleinlercher, D. and Kirchler, M. (2017), “The impact of different tournament
incentives on asset markets: theory and experiment”, Journal of Economic Dynamics and
Control, Vol. 75, pp. 1-27.
Fox, C.R. and Tversky, A. (1995), “Ambiguity aversion and comparative ignorance”, Quarterly Journal
of Economics, Vol. 110 No. 3, pp. 879-895.
Fox, C.R. and Weber, M. (2002), “Ambiguity aversion, comparative ignorance, and decision context”,
Organizational Behavior and Human Decision Processes, Vol. 88 No. 1, pp. 476-498.
Garcia, S.M. and Tor, A. (2009), “The N-effect: more competitors, the less competition”, Psychological
Science, Vol. 21 No. 5, pp. 871-877.
Garcia, S.M., Reese, Z.A. and Tor, A. (2020), “Social comparison before, during, and after the
competition”, in Suls, J., Collins, R. and Wheeler, L. (Eds), Social Comparison, Judgment, and
Behavior, Oxford University Press, New York, NY, doi: 10.1093/oso/9780190629113.
001.0001.
Garcia, S.M., Tor, A. and Schiff, T.M. (2013), “The psychology of competition: a social comparison
perspective”, Perspectives on Psychological Science, Vol. 8 No. 6, pp. 634-650.
G€arling, T., Fang, D. and Holmen, M. (2019), “Review of behavioral explanations of how rank-based
incentives influence risk taking by investment managers in mutual fund companies”, Review of
Behavioral Finance, Ahead-of-print, doi: 10.1108/RBF-01-2019-0013.
Gibbons, F.X. and Buunk, B.P. (1999), “Individual differences in social comparison: development of a
scale of social comparison orientation”, Journal of Personality and Social Psychology, Vol. 76
No. 1, pp. 129-142.
Glaser, M., N€oth, M. and Weber, M. (2007), “Behavioral finance”, in Koehler, D.J. and Harvey, N.
(Eds), Blackwell Handbook of Judgment and Decision Making, Blackwell, Oxford,
pp. 527-546.
Guerin, B. (1986), “Mere presence effects in humans: a review”, Journal of Experimental Social
Psychology, Vol. 22 No. 1, pp. 38-77.
Harris, P.B. and Houston, J.M. (2010), “A reliability analysis of the revised competitiveness index”,
Psychological Reports, Vol. 106 No. 3, pp. 870-874.
Heath, C. and Tversky, A. (1991), “Preference and belief: ambiguity and competence in choice under
uncertainty”, Journal of Risk and Uncertainty, Vol. 4 No. 1, pp. 5-28.
Hodges, B.H. (2017), “Conformity and divergence in interactions, groups, and culture”, in Harkins, S.G.,
Williams, K.D. and Burger, J.M. (Eds), The Oxford Handbook of Social Influence, Oxford
University Press, New York, NY, pp. 87-105.
Houston, J.M., Harris, P.B., Mcintire, S. and Francis, D. (2002a), “Revising the competitiveness index”,
Psychological Reports, Vol. 90 No. 1, pp. 31-34.
Houston, J.M., Mcintire, S., Kinnie, I. and Terry, C. (2002b), “A factor analysis of scales measuring
competitiveness”, Educational and Psychological Measurement, Vol. 62 No. 2, pp. 284-298.
Kahneman, D. and Tversky, A. (1979), “Prospect Theory: an analysis of decision under risk”,
Econometrica, Vol. 47 No. 2, pp. 263-291.
Karau, S.J. and Williams, K.D. (1993), “Social loafing: a meta-analytic review and theoretical
integration”, Journal of Personality and Social Psychology, Vol. 65 No. 4, pp. 681-706.
Kempf, A., Ruenzi, S. and Thiele, T. (2009), “Employment risk, compensation incentives, and Financial
managerial risk taking: evidence from the mutual fund industry”, Journal of Financial
Economics, Vol. 92 No. 1, pp. 92-108. risk-taking
Kirchler, M., Lindner, F. and Weitzel, U. (2019), “Rankings and risk-taking in the finance industry”,
Journal of Finance, Vol. 73 No. 5, pp. 2271-2302.
uhberger, A., Schulte-Mecklenbeck, M. and Perner, J. (2002), “Framing decisions: hypothetical
K€
and real”, Organizational Behavior and Human Decision Processes, Vol. 89 No. 2,
pp. 1162-1175.
Kuziemko, I., Buell, R.W., Reich, T. and Norton, M.I. (2014), “‘Last-place aversion’: evidence and
redistributive implications”, Quarterly Journal of Economics, Vol. 129 No. 1, pp. 105-149.
Linde, J. and Sonnemans, J. (2012), “Social comparison and risky choices”, Journal of Risk and
Uncertainty, Vol. 44 No. 1, pp. 45-72.
Linde, J. and Sonnemans, J. (2015), “Decisions under risk in a social and individual context: the limits
of social preferences?”, Journal of Behavioral and Experimental Economics, Vol. 56 No. 3,
pp. 62-71.
Ma, L., Tang, Y. and Gomez, J.-P. (2019), “Portfolio manager compensation in the US mutual fund
industry”, Journal of Finance, Vol. 74 No. 2, pp. 587-638.
McGraw, A.P., Larsen, J.T., Kahneman, D. and Schkade, D. (2010), “Comparing gains and losses”,
Psychological Science, Vol. 21 No. 10, pp. 1438-1445.
Murayama, K. and Elliot, A.J. (2012), “The competition-performance relation: a meta-analytic review
and test of the opposing processes model of competition and performance”, Psychological
Bulletin, Vol. 136 No. 6, pp. 1035-1070.
Nieken, P. and Sliwka, D. (2010), “Risk-taking tournaments: theory and experimental evidence”,
Journal of Economic Psychology, Vol. 31 No. 3, pp. 254-268.
Noussair, C.N. and Tucker, S. (2013), “Experimental research on asset pricing”, Journal of Economic
Surveys, Vol. 27 No. 3, pp. 554-569.
Palan, S. (2013), “A review of bubbles and crashes in experimental asset markets”, Journal of
Economic Surveys, Vol. 27 No. 3, pp. 570-588.
Powell, O. and Shestakova, N. (2016), “Experimental asset markets: a survey of recent developments”,
Journal of Behavioral and Experimental Finance, Vol. 12 No. 1, pp. 14-22.
Savikin, A.C. and Sheremeta, R.M. (2013), “Simultaneous decision-making in competitive and
cooperative games”, Economy Inquiry, Vol. 51 No. 2, pp. 1311-1323.
Schedlinsky, I., Sommer, F. and Wohrman, A. (2016), “Risk-taking in tournaments: an experimental
analysis”, Journal of Business Economics, Vol. 86 No. 8, pp. 837-866.
Schoenberg, E.J. and Haruvy, E. (2012), “Relative performance information in asset markets: an
experimental approach”, Journal of Economic Psychology, Vol. 33 No. 6, pp. 1143-1155.
Soman, D. (2004), “Framing, loss aversion, and mental accounting”, in Koehler, D.J. and Harvey, N.
(Eds), Blackwell Handbook of Judgment and Decision Making, Blackwell, Oxford,
pp. 379-398.
Spyrou, S. (2013), “Herding in financial markets: a review of the literature”, Review of Behavioral
Finance, Vol. 5 No. 2, pp. 371-398.
Tor, A. and Garcia, S.M. (2010), “The N-effect: beyond probability judgments”, Psychological Science,
Vol. 20 No. 7, pp. 748-877.
Yechiam, E. and Hochman, G. (2013), “Losses as modulators of attention: review and analyses
of the unique effects of losses over gains”, Psychological Bulletin, Vol. 136 No. 2,
pp. 407-518.
RBF
Appendix.
Instructions to experiment 1

Individual condition/no absolute losses and absolute losses


Below we ask you to make a number of choices between a specified sum of money (A) and a 50% chance
of receiving a larger sum of money or /nothing/losing a smaller sum/ (B). If you choose A, you receive the
specified sum of money. If you instead choose B, you have a 50% chance of receiving the larger sum or
nothing. Which one you receive is determined after all choices are made. Make your choice as you expect
you will do if the money is real.
Click on your choice of A or B
(A) SEK 300/SEK 600/SEK 900/SEK 1,200/SEK 1,500/SEK 1,800/SEK 2,100;
(B) A 50% chance to receive/SEK 4,800 or nothing/SEK 6,000 or losing SEK 1,200/.

Social comparison and competition conditions/no absolute losses and absolute losses
Imagine that you again can make the same choices between a specified sum of money (A) and a 50%
chance of receiving a larger sum of money or /nothing/losing a smaller sum/ (B). /You are informed about
the choices made before you by another participant to whom you may compare yourself. This
participant always chooses A. /Another participant before you always chooses A when making these
choices. Since it is a competition, you should try to outperform this participant. /If you choose A, you
both receive the specified sum of money. If you instead choose B, you have a 50% chance of receiving the
larger sum or nothing. Which one you receive is determined after all choices are made. Make your choice
as you expect you will do if the money is real.
The other participant chooses A. Click on your choice of A or B.
(A) SEK 300/SEK 600/SEK 900/SEK 1,200/SEK 1,500/SEK 1,800/SEK 2,100;
(B) A 50% chance to receive SEK 4,800 or nothing/SEK 6,000 or losing SEK 1,200/.

Corresponding author
Tommy G€arling can be contacted at: tommy.garling@psy.gu.se

For instructions on how to order reprints of this article, please visit our website:
www.emeraldgrouppublishing.com/licensing/reprints.htm
Or contact us for further details: permissions@emeraldinsight.com

You might also like