Chandrashekar Etal 2024 IRSP Pronin Kugler 2010 Replication Extension Preprint v2

Do people believe they are less predictable than others?
Three replications of Pronin and Kugler (2010)’s Experiment 1
*Subramanya Prasad Chandrashekar

Department of Psychology,
Norwegian University of Science and Technology (NTNU)
prasad.chandrashekar@ntnu.no
ORCID: 0000-0002-8599-9241
*Stephanie Permut
Andlinger Center for Energy and the Environment
Princeton University
sepermut@princeton.edu
*Hallgeir Sjåstad
Norwegian School of Economics
hallgeir.sjastad@nhh.no
ORCID: 0000-0002-8730-1038
** Chelsea (Chi Wing) LO

chelsea1@connect.hku.hk / lochiwingchelsea@gmail.com
Department of Psychology, University of Hong Kong
ORCID: 0009-0008-6052-7759
**Yong Jun KUEH

u3512000@connect.hku.hk / jean_5086@hotmail.com
***Emily Sihui ZHONG (ezhong@connect.hku.hk / zhong_emil@bentley.edu),

***Kai Hin WAN (u3538877@connect.hku.hk / dannywan1222@gmail.com),
***Kai Yi Kelly CHOY (u3529138@connect.hku.hk / ckyk97@gmail.com),
***Man Chung WONG (wong1109@connect.hku.hk / manchungwong@gmail.com),
***Stanley Wei Jian HUGH (u3543385@connect.hku.hk / stanleyhugh2017@u.northwestern.edu),
***Khan TAHIRA (u3537184@connect.hku.hk / t.caan786@gmail.com),
Bo Ley CHENG
boleystudies@gmail.com
^Gilad Feldman
University of Hong Kong
gfeldman@hku.hk / giladfel@gmail.com
ORCID: 0000-0003-2812-6599
* shared first author; ** shared third author; *** shared fifth author ;^ corresponding author
Word: abstract – 173, manuscript - 5074
Pronin and Kugler (2010): Replications and extensions 2
Author Bios:
Subramanya Prasad Chandrashekar is postdoctoral researcher with the Department of
Psychology at the Norwegian University of Science and Technology (NTNU). His research
focuses on moral psychology, lay-beliefs, and judgment and decision-making.
Stephanie Permut is an associate fellow with the Andlinger Center for Energy and the
Environment at Princeton University.
Hallgeir Sjåstad is a professor at the Norwegian School of Economics, Department of Strategy
and Management. His research focuses on judgment and decision-making, and social
psychology.
Chi Wing LO, Yong Jun KUEH, Emily Sihui ZHONG, Kai Hin WAN, Kay Yi Kelly CHOY,
Man Chung WONG, Stanley Wei Jian HUGH, and Khan TAHIRA were students at the
University of Hong Kong in the year 2018.
Bo Ley Cheng was a teaching assistant at the University of Hong Kong psychology department
in the year 2018.
Gilad Feldman is an assistant professor with the University of Hong Kong psychology
department. His research focuses on judgment decision-making.
Declaration of Conflict of Interest:
The author(s) declared no potential conflicts of interests with respect to the authorship and/or
publication of this article.
Financial disclosure/funding:
The author(s) received no financial support for the research and/or authorship of this article.
Acknowledgments or Author Notes:
None.
Authorship declaration
A team comprising of Chi Wing LO, Yong Jun KUEH designed and wrote the pre-registrations
for Study 1b. A team comprising of Emily Sihui ZHONG, Kai Hin WAN, Kay Yi Kelly CHOY,
Man Chung WONG, Stanley Wei Jian HUGH, and Khan TAHIRA designed and wrote the pre-
registrations for Study 1a. Bo Ley CHENG and Gilad Feldman guided the two teams. Gilad
conducted the pre-registrations, and ran data collections. The two teams analyzed and wrote
initial reports of the findings.
Hallgeir and Gilad designed and wrote the pre-registration for Study 2. Gilad conducted Study 2
data collection. Prasad and Stephanie verified teams’ work on Studies 1a and 1b, and analyzed
Study 2.
Hallgeir and Stephanie integrated the three projects to a unified manuscript. Gilad reverified the
three studies, added analyses, and edited the first draft. Prasad verified and extended the analyses
and finalized the manuscript for journal submission, with Gilad’s final editing.
Contributor Roles Taxonomy
Role S.P. Hallgeir Sjåstad Stephanie Chi Wing LO, Yong Jun Bo Ley Gilad
Chandrashekar Permut KUEH, Emily Sihui CHENG Feldman
ZHONG, Kai Hin WAN,
Kay Yi Kelly CHOY, Man
Chung WONG, Stanley Wei
Jian HUGH, and Khan
TAHIRA
Conceptualization X
Pre-registrations X X X
Data curation X
Formal analysis X X X X
Funding acquisition X
Investigation X X
Methodology X X
Pre-registration peer X X X
review / verification
Data analysis peer X X X X X
review / verification
Project administration X X
Resources X
Software X X
Supervision X X X
Validation X X X
Visualization X X X
Writing-original draft X X X X
Writing-review and X X X X
editing
Abstract
Pronin and Kugler (2010) proposed that people believe they have more free will than others. In
their Experiment 1 they showed that American students evaluated their own decisions and life
events as less predictable than similar decisions and life events of close others, presumably
suggesting higher free-will attributions. We conducted three pre-registered replications of this
study, one with undergraduate sample in Hong Kong and two with MTurk participants from the
US using CloudResearch (total N = 1031). In Studies 1a and 1b that mirrored the target article’s
mixed design (self-other between, past-future within), we found support for the original findings
with weaker effects. In Study 2 we contrasted between-subject versus within-subject designs in a
single data collection. We successfully replicated the effects with the between-subject design,
whereas we failed to find support for the effect using the within-subjects design. This suggests
support for the phenomenon in single evaluation mode assessing either the self or the other, but
that people correct for the self-other asymmetry in perceived predictability when the judgment is
made in joint evaluations mode. Materials, data, and code are available on: https://osf.io/ykmqp/
Keywords: Predictability, self-other asymmetry, free will attributions, replications

Do people believe they are less predictable than others?
Three replications of Pronin and Kugler (2010)’s Experiment 1
Do people believe that they have more free will than others? An influential study by
Pronin and Kugler (2010) provided initial evidence for this hypothesis, in which their
Experiment 1 showed that people believe that their own decisions and life outcomes are less
predictable than similar decisions and life outcomes of others. The link the authors made to free
will, was that higher predictability of past and future events would presumably be at odds with
personal agency, at least in laypersons’ views, thereby leaving less room for freedom of choice.
In the current research, we report three pre-registered replications of this experiment
using different samples and variations of study designs to reproduce, replicate, and test the
generalizability of this effect.
‘Free Will’ Research in Psychology
Free will is often associated with the capacity for making choices between alternatives
without internal or external constraints (Feldman et al., 2014; Monroe et al., 2014; Rigoni et al.,
2017). Most laypersons seem to believe in some form of free will (Alquist et al., 2015; Monroe
& Malle, 2010; Nahmias et al., 2005; Nichols, 2004; Sarkissian et al., 2010; Stillman et al.,
2011). In philosophy, there is extensive debate about the very existence of free will, how it
should be conceptualized in formal terms, whether it is compatible with determinism and
biological evolution, and what the implications of free will might be for ethics and society (e.g.,
Dennett & Caruso, 2021). In social psychology and experimental philosophy, the focus has
shifted towards using empirical research methods to study how ordinary people think and feel
about free will, and to examine the causes, consequences, and correlates of people’s free-will
beliefs and attributions. The study by Pronin and Kugler has been one of the first to suggest
interesting asymmetries in the way that people tend to think about concepts related to free will
when reflecting on their own comparing to that of others’ life circumstances.
People readily make inferences about the degree of personal agency involved in a
decision, both when reflecting on their own behavior and on the behaviors of others (Gray et al.,
2012). Previous research documented associations between free-will beliefs with motivation and
work performance (Stillman et al., 2010), psychological well-being and adjustment (Crescioni et
al., 2016), and stricter moral judgments of blame and punishment (Martin et al., 2017; Shariff et
al., 2014). It therefore seems that free-will beliefs are associated with a broad range of factors
that might impact people’s lives. We therefore sought to conduct an independent replication of
what we considered to be one of the first central findings in this literature, suggesting that people
believe that they have more free will than other people by being less predictable a priori.
Early replication attempts of free-will research focused on so-called “priming” activation
experiments, seeking to induce a spill-over effect from encouraging disbelief in free will on a
subsequent and unrelated task. For these methods, rigorous studies have concluded that it is more
difficult than previously thought to create such intrinsic variation in free-will beliefs, where
recent studies have failed to replicate the original effects on the dependent variable as well
(Chartier et al., 2020; Nadelhoffer et al., 2020). A more promising approach seems to be
correlational research on associations with individual differences in free-will beliefs (e.g., Martin
et al., 2017), or experiments regarding free-will attributions that vary one aspect of a
hypothetical choice situation before eliciting agency attributions and related judgments (e.g.,
Everett et al., 2021; Feldman et al., 2016, 2018).

Pronin and Kugler (2010) used a variant of the latter method and randomly assigned
participants to evaluate free-will related indicators regarding their own lives or the lives of
others. In the case of Experiment 1, they tested people’s attributions of predictability regarding
past and future life events, contrasting evaluations of one’s own life compared to that of other
people.
Self-Other Asymmetries in Human Judgment
There are many documented phenomena of self-other asymmetries. One of those is
“unrealistic optimism”, which is the phenomenon that people tend to believe they are more likely
to experience positive events and life outcomes than is warranted, and to underestimate their
personal risks (Shepperd et al., 2015; Weinstein, 1980). Research on the “better-than-average
effect” has documented that people have a broad tendency to view themselves in a favorable
light when comparing themselves to others (Alicke, 1985; Svenson, 1981; see successful
replication by: Koppel et al., 2023; Ziano et al., 2021), though over time identifying some
nuances and moderators, such as an interaction with difficulty (Kruger, 1999).
Pronin and Kugler (2010) found that people attribute more free will to themselves both
for positive and negative outcomes, which suggested a distinct phenomenon from the classic
form of optimism and better-than-average effects. Drawing on the “actor-observer bias” (Jones
& Nisbett, 1971), Pronin and Kugler (2010) proposed that people might perceive the actions of
others as fixed dispositions of their stable and uncontrollable personalities, whereas people
perceive their own actions as intentional unconstrained responses to changing situations. This
latter perspective presumably leaves more room for free will to operate.
Cognitive neglect vs. Self-enhancement motives

The self-other asymmetry observed in free will attributions may stem from the self-
enhancement motive, wherein individuals tend to perceive themselves more favorably than the
average person across various domains, including intelligence, skill, and positive personality
traits (Alicke, 1985; Alicke et al., 1995; Ziano et al., 2021). Alternatively, this asymmetry could
be partially or entirely influenced by a cognitive bias, where individuals rely on introspective
information, such as feelings and intentions, for self-assessments. Considering these possibilities,
it can be argued that if the self-other asymmetry is primarily driven by self-enhancement
motives, it will persist in within-person comparisons. Conversely, the absence of this asymmetry
in within-person comparisons suggests that cognitive neglect, which individuals can potentially
correct for, plays a significant role in driving the asymmetry. We proposed an extension study to
evaluate this possibility.
Extension: Examining bias, by contrasting within and between subject designs
It remains an open question what research participants themselves would make of their
tendency to claim being less predictable than others. As that experiment was based on a between-
subjects self-other design, where participants either rated the level of predictability in their own
lives or in the lives of close others, the systematic inconsistency between conditions remain
“invisible” to individuals making those ratings. Thus, in addition to examining the replicability
of Pronin and Kugler’s (2010) main finding, in our Study 2 we extended their original design
with a variety of contrasts (1) fully within-subject conditions, 2) a fully between-subject
condition, 3) mixed-design) within a single data collection.
Previous studies suggested that if a given effect is driven by cognitive neglect or
irrational bias, then the effect will tend to be weaker or disappear when tested in a within-
subjects design. This is because the within-subjects structure facilitates direct comparison in joint
evaluation (Caruso et al., 2008; Hsee et al., 1999; Kahneman & Ritov, 1994). Based on this
logic, we reasoned that if we observe that people in a within-subjects condition do not show the
effect, then we might interpret this as cognitive neglect, and less likely to have a motivational
origin. That is, people can recognize the inconsistency when it is made clear by contrasting
conditions against one another in the same experiment (for a recent example of this method, see
Sjåstad et al., 2020).
Studies overview
In the current research, we report three pre-registered replications of Experiment 1 in
Pronin and Kugler (2010). For the target study, they observed very large effects. Our Studies 1a
and 1b were close replications of the original study design, with initial small samples, which we
considered to be pre-tests. In Study 2, we followed-up with a much larger sample, also aiming to
examine the possible role of cognitive neglect versus motivational self-enhancement as
mechanisms. In Study 2 we included an extension contrasting the fully within-subject condition
with the fully between-subject condition, adjusted from the original mixed design. Study 2 also
included a trait achievement measure as a possible individual-difference moderator of the effect.
We summarized the replication hypotheses in Table 1 and the extension hypotheses in Table 2.
We provided an overview of the findings of Experiment 1 of the target article in Table 2.

Table 1
Pronin and Kugler (2010): Summary of hypotheses
Study Hypothesis
Hypothesis 1 People perceive their own past and future decision as less predictable a priori than
those of a roommate.
Hypothesis 2a People perceive their own past decisions as less predictable than the ones of others
Hypothesis 2b People perceive their own future decisions as less predictable than the ones of others.
Hypothesis 3 People perceive future decisions as less predictable than past decisions.
Table 2
Summary of original findings in Pronin and Kugler (2010) Experiment 1
Effect size
df 1 df 2 F p
Cohen’s d with 95% CI
(H1) Across both past
and future outcomes: 1 48 14.46 .0004 -1.08 [-1.67, -0.48]
Self vs. Roommate
(H2a) Across past
outcomes: 1 48 6.40 .010 -0.76 [-1.29, -0.14]
Self vs. Roommate
(H2b) Across future
outcomes: 1 48 8.46 .005 -0.82 [-1.40, -0.25]
Self vs. Roommate
(H3) Future vs. Past 1 48 10.94 .002 -0.94 [-1.52, -0.35]

Open science, preregistrations, and disclosures
Materials, data, and code are available on the Open Science Framework (project:
https://osf.io/ykmqp/; preregistration Study 1a: https://osf.io/jg94e ; preregistration Study 1b:
https://osf.io/yaepf/ ; preregistration Study 2: https://osf.io/sjb8p). All studies, participants,
measures, manipulations, and exclusions conducted for this investigation are reported, and data
collection was completed before hypothesis testing.
Planned Sample
We calculated effects in Pronin and Kugler (2010) using available F-statistics and sample
sizes for each condition with the help of a guide by Jané et al. (2024). The effect size (Cohen’s d)
of 1.1 for the between-subject factor (self vs. other) used as the basis of power analysis.
In study 1a and 1b, we based our power analysis on having 95% (α =0.05) power to
detect a Cohen’s d = 1.08. This resulted in a suggested sample size of 46 across 2 between-
subject conditions. More detailed descriptions of the effect-size calculations and power analyses
can be found in the supplementary materials.
In Study 2, we used the G*Power software to determine the number of participants
needed to detect an effect size of d = 0.32 at 95% power (α =0.05). The effect size of d = 0.32
was based on the results of Study 1b. Our analysis yielded a sample of 426 participants for
studies involving between-subjects effects and a sample of 108 participants for studies involving
within-subject effects. To ensure that both designs were conducted with adequate statistical
power, aiming for a conservative estimate, we took the largest sample size needed to detect a
moderate between-subjects effect and set this as our target sample.

Deviations from Original Studies
We provided detailed information of designs (type of study, sample, variables, exact
wordings) of the original studies in the Methods and Original articles’ results section of the
Supplementary materials. We note several deviations from the original, summarized in Table 3
and Table 4. We recruited participants on MTurk using CloudResearch rather than recruiting
university students and made several adjustments and added extensions.
Table 3
Comparison of the sample in the original study and in the replications
Pronin and Replication Replication Replication

Kugler (2010), Study 1a Study 1b Study 2
Exp. 1
Sample size 50 47 126 1295
Geographic US American Hong Kong, US American US American
origin SAR
Population Undergraduates Undergraduates MTurk/CloudResearch Prolific
participants participants
Gender Not reported 21 females 63 females 770 females
Median age Not reported 20 24 20
Average age Not reported 20.21 25.83 20.40
Standard Not reported 1.02 6.46 1.48
deviation age
Age range Not reported 3 36 26
Medium Not reported Computer Computer (online) Computer
(location) (online) (online)
Compensation Yes, received Yes Yes Yes
candy for
completing the
survey
Year Probably 2009 2018 2018 2019
or 2010
Replications Classification
We categorized Studies 1a and 1b as “very close replications” and Study 2 as a “close
replication” based on the criteria by LeBel et al. (2018), with our classification analysis provided
in Table 4.
Table 4
Evaluation of replication closeness (using criteria by LeBel et al., 2018)
Design facet Replication Details

Effect/hypothesis Same
IV construct Same
DV construct Same
IV operationalization Same
DV Same
operationalization
Population (e.g., age) Different The original study was conducted with a sample of
US undergraduates, whereas the replication effort
involved a sample of undergraduate students from
Hong Kong and US participants from
MTurk/CloudResearch and Prolific.
IV stimuli Some There are minor set of variations in experimental
differences for condition allocations.
Study 2; Same
for Studies 1a
and 1b
DV stimuli Same
Procedural details Similar
Physical settings Different Online data collection in our replication vs filling a

questionnaire
in the target article
Contextual variables Different The original authors conducted their studies before
2010, whereas we conducted the replications in the
years 2018-19.
Replication Studies 1a, 1b: Very close replication; Study 2: Close replication
classification:
Note of alpha tails and p-values interpretations
We note that we did not pre-register whether our alpha is for one-tail or two-tail, with the
pre-registration for Study 1a noting power analysis for two-tail and the pre-registration for Study
1b noting one-tail. In the reported results below we refer to one-tail tests, given that the
hypotheses are well stated and these were replications.
Study 1a: Hong Kong Student sample
Study 1a was our first attempt to replicate the effect, conducted on an undergraduate
student sample in Hong Kong. The experiment was run together with a series of other judgment
and decision-making studies that were designed by the students in the sample and ran together in
randomized order on the designing students. Therefore, given that students were familiar with
other judgment and decision-making phenomena and the relatively small sample, we consider
this study to be a conservative pre-test replication on a well-informed sample.
Method
Participants
We recruited a small sample of 47 participants (16 males, 31 females, Age mean = 20.2,
Age SD = 1.02). A sensitivity analysis indicates that this sample can detect effects of d = 0.97
(one-tail, 95% power), an effect smaller than that in the target article, yet we considered this a
pre-test. Materials were presented in an online survey using the Qualtrics software. We collected
no demographic information in this study.
Procedure
Participants were randomly assigned to one of two conditions: ‘self’ and ‘other’.
Participants in the ‘self’ condition judged the predictability of past and future events in their own
lives whereas participants in the ‘other’ condition judged the predictability of past and future
events in the life of a close other. We prompted participants in the ‘other’ condition to provide
the initials of the person they had in mind before proceeding to provide their predictability
ratings.
We ran a 2 target (self versus other; between-subjects) by 2 time (past versus future;
within-subjects) mixed study design. Participants evaluated three past target events (1 = Not at
all predictable, 7 = Extremely predictable): 1) Chosen university for studies, 2) chosen major for
studies, and 3) failure in dating, and three future events: 1) Career choice, 2) person to marry, 3)
place to live, six target events in total.
Exclusions
We pre-registered an exclusion criteria (failing to complete the entire survey, failing to
adhere to pre-test requirements, indicating low English proficiency and reporting low
seriousness), yet no participants were excluded as a result of these exclusion criteria. The
students who designed the study also participated but their answers were excluded from the
analysis.
Results
We summarized descriptives in Table 5 and Figure 1.
Table 5
Study 1:Predictability descriptives
Target [between] Self (n = 23) Other (n = 24)
Time [within] M SD M SD
Past 4.25 1.39 4.68 0.72
Future 3.01 1.39 4.00 1.08
Overall 3.63 1.02 4.34 1.02

Figure 1
Study 1a: Predictability – Self versus Other
Note. Violin plot of the analysis of Self versus Other in Study 1a. Predictability: 1 = not at all predictable; 7 = extremely predictable. Dependent
variables across the panels: A) Past predictability; B) Future predictability; C) Past and Future predictability.
We conducted an independent t-test and found support for the findings in Pronin and
Kugler (2010). Participants viewed their own decisions as being less predictable than the
decisions of close others (Self: n = 23, M = 3.63, SD = 1.02; other: n = 24, M = 4.34, SD = 0.62;
t(45) = -2.89, p = .003, one-tailed, d = -0.84, 95% CI [-1.46, -0.23]).
Examining past and future separately, future events for self (M = 3.01, SD = 1.39) were
rated as less predictable than others’ (M = 4.00, SD = 1.08; t(45) = -2.72 p = .005, one-tailed, d =
-0.79, 95% CI [-1.41, -0.18]). We did not find support for past events, as self past events (M =
4.25, SD = 1.39) were rated similarly to others’ past events (M = 4.68, SD = 0.72; t(45) = -1.36, p
= .091, one-tailed, d = -0.40, 95% CI [-0.99, 0.20]). Therefore, the effects were stronger for the
future than for the past, although given the small sample we did not detect support for
interaction. The results of the mixed ANOVA analysis are reported in the supplementary section
(see Table S3 and S4).
Consistent with the original findings, there also was a main effect for time, whereby
people perceived the future as less predictable than the past (M = 3.52 (SD = 1.32) vs. 4.47
(1.11); t(92) = -3.78, p < .001, one-tailed, d = -0.78, 95% CI [-1.20, -0.35]).
Study 1b: US Americans on MTurk
Method
Procedure and Exclusions
We recruited an online sample of 136 US American student participants from Amazon
Mechanical Turk using CloudResearch/TurkPrime (Litman et al., 2017). All study materials and
instructions were identical in structure and content to the survey used in Study 1a.
Four participants failed to give consent to partake in the study, and four other participants
indicated that they were not currently American students, and therefore were not allowed to
advance past the consent form of the survey. Two additional participants did not complete the
study. Our final sample consisted of 126 participants (Mage = 25.88, 63 female, 63 male).
Results and discussion
Table 6
Study 1b: Predictability descriptives
Target [between] Self (n = 64) Other (n = 62)

Past 4.56 0.92 4.81 1.22
Future 3.94 1.36 4.33 1.34
Overall 4.24 0.95 4.58 1.15
Figure 2
Study 1b: Self vs Other predictability
Note. Violin plot of the analysis of Self versus Other in Study 1b. Predictability: 1 = not predictable at all; 7 = extremely predictable. Dependent
variables across the panels: A) Past predictability; B) Future predictability; C) Past and Future predictability.
We conducted an independent t-tests and effects were in the expected direction, yet we
failed to meet the evidence criteria we set for concluding support for the findings in Pronin and
Kugler (2010). Participants viewed their own decisions as being less predictable than the
decisions of close others (Self: n = 64, M = 4.25, SD = 0.95; other: n = 62, M = 4.57, SD = 1.16;
t(124) = -1.69, p = .046, one-tailed, d = -0.30, 95% CI [-0.66, 0.05]).
Examining past and future separately, future events for self (M = 3.94, SD = 1.36) were
rated as less predictable than others’ (M = 4.33, SD = 1.34; t(124) = -1.63, p = .053, one-tailed, d
= -0.29, 95% CI [-0.64, 0.06]). We did not find support for past events, as self past events (M =
4.56, SD = 0.92) were rated similarly to others’ past events (M = 4.81, SD = 1.22; t(124) = --
1.29, p = 0.099, one-tailed, d = -0.23, 95% CI [-0.58, 0.12]). Therefore, the effects were stronger
for the future than for the past, although given the small sample we did not detect support for
interaction. Additionally, we found support for main effect for time, whereby people perceived
the future as less predictable than the past (M = 4.69 (SD =1.08) vs. 4.13 (1.36); t(125) = -5.14, p
< 0.001, one-tailed, d = -0.45, 95% CI [-0.70, -0.20]). The results of the mixed ANOVA analysis
are reported in the supplementary section (see Table S5 and S6).
The original effect estimates were in the predicted direction, yet caution against
overinterpretation given the weak effects and just-below threshold one-tail p-values, especially
given that we were not clear in advance about whether we will employ one or two-tail analyses,
although our power analyses referred to one-tail. Still, we provide our data as is. We followed
our pre-registered criteria, and tried to avoid any flexibility in analysis, yet recognize the
challenge with p-values just below .05. We see the reporting of one-tail as reasonable given that
this was a replication with clear directional hypotheses.

Across Study 1a and 1b, we found support for the target study’s findings. Yet we caution
against overinterpretation given the weak effects and just-below threshold one-tail p-values1.
Study 2: US Americans on Prolific
We considered Study 2 as an extension with an aim to test replicability of the original
findings with a set of study design variations in a single data collection effort. We aimed to test
whether the effect would be stronger in the fully between-subjects design compared to the other
two conditions, in that participants would correct for the asymmetry if asked to evaluate both self
and other.
The study participants were randomly assigned to one of the following experimental
design conditions1: 1) fully between-subjects design, 2) fully within-subject design, and 3) a
mixed design. All study materials and instructions were identical in structure and content to the
survey used in previous studies. We focused our analyses below on the first two conditions,
given an oversight in the third condition that we felt limited its contribution, and its analyses are
reported in the supplementary materials.
The study participants in fully between-subjects design were randomly assigned to one of
four conditions in a between-subjects design: 2 (Target: self vs. other) × 2 (Time: Past vs
Future). Whereas the participants in fully within-subjects design answers all four combinations
of Target (self, other) and Time (past, future)2.
1
We note that we included two additional exploratory extensions that in hindsight we felt did not contribute to our
understanding of the phenomenon: 1) a mixed-design condition, and 2) an achievement motivation scale as a
potential predictor/moderator. We decided to report the results of these two extensions in the supplementary
materials.
2
We note that in our pre-registration we referred to one-tail in our power analyses yet we felt that we were not
specific enough about whether we planned to employ a one-tail or two-tail analyses. We provide our data as is. We
followed our pre-registered criteria, and tried to avoid any flexibility in analysis, yet recognize the challenge with p-
Method
Participants
We recruited a total of 1304 online student participants using Prolific participant
recruitment platform. We pre-registered several exclusion criteria for this study: Low English
proficiency, reporting not being serious or if they failed to complete the study. In total, we
excluded 9 participants, which resulted in a sample of 1295 that met our exclusion criteria (525
males, 770 females, mean age = 20.4, mean SD = 1.93).
Procedure
After responding to the consent form, participants completed a 10-item version of the
Achievement Motives Scale (Lang & Fries, 2006). They were then randomly assigned to view
one of three study designs: fully between-subjects (n = 424), fully within-subjects (n = 434) or
the mixed condition (n = 437). We note that we focused our analysis on the fully between and
fully within design, and the analysis of the mixed-subject condition has been moved to the
supplementary materials 3. Materials and measures were exactly same to those used in Studies
1a/1b.
Results
Between-subject Design
values just below .05. We see the reporting of one-tail as reasonable given that this was a replication with clear
directional hypotheses.
3
The mixed condition was planned (based on the pre-registration) to mirror the original study’s mixed allocation
(i.e., self-other is between-subject, past-future is within-subject). The actual presentation of conditions during the
execution of the study made it impossible to conduct a straightforward test (i.e., ANOVA or t-tests) of the original
predictions. 437 of the 1295 participants in the sample were in the excluded condition. An additional analysis (i.e., a
mixed-effect regression analysis) based on the executed mixed-condition is reported in the supplementary materials.
Table 7
Study 2 between design: Predictability descriptives
Target [between] Self Other

Time [between] n M SD n M SD
Past 106 3.61 0.86 108 3.78 0.81
Future 104 2.93 0.97 106 3.38 0.99
Overall 210 3.27 0.97 214 3.58 0.923

Figure 3
Study 2 between-subjects design: Self vs Other predictability
Note. Violin plot of the analysis of Self versus Other in Study between-design. Predictability: 1 = not predictable at all; 7 = extremely
predictable. Dependent variables across the panels: A) Past and Future predictability; B) Past predictability; C) Future predictability.
We conducted an independent t-test and effects were in the expected direction.
Participants viewed their own decisions as being less predictable than the decisions of close
others (Self: n = 210, M = 3.27, SD = 0.97; other: n = 214, M = 3.58, SD = 0.92; t(422) = -3.38, p
< .001, one-tailed, d = -0.33, 95% CI [-0.52, -0.14]). Examining past and future separately, future
events for self (M = 2.93, SD = 0.96) were rated as less predictable than others’ (M = 3.38, SD =
0.99; t(208) = -3.33, p < .001, one-tailed, d = -0.46, 95% CI [-0.74, -0.18]). We did not find
support for past events, as self past events (M = 3.61, SD = 0.86) were rated similarly to others’
past events (M = 3.78, SD = 0.81; t(212) = -1.55, p = 0.061, one-tailed, d = -0.21, 95% CI [-0.48,
0.06]). Therefore, the effects were stronger for the future than for the past. Additionally, we
found support for a main effect for time, whereby people perceived the future as less predictable
than the past (M = 3.70 (SD =0.84) vs. 3.15 (1.00); t(422) = -6.04, p < 0.001, one-tailed, d = -
0.59, 95% CI [-0.78, -0.39]). The results of the 2-way analysis of variance (ANOVA) analysis
are reported in the supplementary section (see Table S7 and S8).
Within-subjects Design
We summarized the descriptives in Table 8 and Figure 4. We conducted paired t-tests as
part of the analyses. In joint evaluation mode, we found no support for participants viewing their
own decisions as being less predictable than the decisions of close others (Self: M = 3.44, SD =
0.96; other: M = 3.44, SD = 0.95; t(433) = -0.01, p = .490, one-tailed, d = -0.00, 95% CI [-0.13,
0.13]). Examining past and future separately, we found no support for differences between future
events for self (M = 3.22, SD = 0.99) and for others’ (M = 3.26, SD = 1.00; t(433) = -0.724, p =
.770, one-tailed, d = 0.04, 95% CI [-0.09, 0.17]). Similarly, we did not find support for past
events, as self past events (M = 3.66, SD = 0.88) were rated similarly to others’ past events (M =
3.62, SD = 0.85; t(433) = -0.821, p = 0.210, one-tailed, d = -0.05, 95% CI [-0.18, 0.09]).
However, we found support for effect for time, people perceived the future (M = 3.24, SD
=0.99) as less predictable than the past (M = 3.64, SD = 0.87; t(433) = -10.30, p < 0.001, one-
tailed, d = -0.53, 95% CI [-0.67, -0.39]). We summarized the results of the within-subjects
ANOVA analyses in the supplementary section (see Table S9 and S10).
Table 8
Study 2 within design: Predictability descriptives
Target [within] Self Other

Past 3.66 0.89 3.62 0.85
Future 3.22 0.99 3.26 1.00
Overall 3.44 0.96 3.44 0.95
Note. n = 434.
Figure 4
Study 2 within-subjects design: Self vs Other predictability
Note. Violin plot of the analysis of Self versus Other in Study 2 within-subject design.
Predictability: 1 = not predictable at all; 7 = extremely predictable.
Discussion
Study 2 involved a larger sample and modified experimental designs to test the
robustness of the findings of the Experiment of Pronin and Kugler (2010). The results based on
the between-subjects design supported the predictions of the original study. However, the results
based on the novel within-subjects and mixed designs failed to support and extend the original
findings, which can shed new light on potential mechanisms. Additionally, there was no support
for the moderation effect of achievement motivation as individual-level factor.

Comparing Replication Findings to Original Findings and Extensions
We compiled a comparison between the effects of the original studies and our
replications in Table 9, along with an interpretation of the findings using the evaluation criteria
established by LeBel et al. (2019) for replication results.
The target study’s main prediction (H1) was that people view their own past and future
decision as less predictable a priori than those of others. Our direct replication samples in Studies
1a and 1b supported the prediction, though with weaker effects and just barely below threshold
in Study 1b. We observed similar patterns for H2a and H2b, yet slightly stronger support for the
hypothesis for future than for the past. Also, we found evidence supporting the prediction (H3)
that individuals perceive the future as less predictable than the past.
In Study 2, we juxtaposed two supplementary designs within a single data collection:
fully between-subject, and fully within-subject. The main effect (H1) replicated in the between-
subject condition, whereas we found no support for the effect in a new within-subject condition.
We also conducted the mini-meta-analyses with a random-effects model method using
the meta package in R (Schwarzer, 2007; see Figure S9-S12 for detailed forest plots in the
Supplementary Materials section). The analysis included effects from Study 1a, Study 1b, and
Study 2. Among effect sizes from Study 2 we only considered effects from between-subject
condition. This analysis provided support for the main prediction (H1) that people view their
own past and future life decisions as less predictable a priori than those of a roommate (d = -0.36
[-0.52, -0.20]).
Table 9
Summary of replication results comparison with the original effect sizes
Sample Predictor Replication Results Original study Replication Interpretation

Study 1a H1: General (Self vs Other) F(1, 45) = 8.33, p = 0.006 -1.08 [-1.67, -0.48] -0.84 [-1.46, -0.23] Signal- consistent
H2a: Past outcomes (Self vs. Other) t(45) = -1.36, p = 0.181 -0.72 [-1.29, -0.14] -0.40 [-0.99, 0.20] No signal- inconsistent
H2b: Future outcomes (Self vs. Other) t(45) = -2.72, p = 0.009 -0.82 [-1.40, -0.25] -0.79 [-1.41, -0.18] Signal- consistent
H3: Future vs Past F(1, 45) = 16.26, p < 0.001 -0.94 [-1.52, -0.35] -0.77 [-1.20, -0.35] Signal- consistent
H1: General (Self vs Other) F(1, 124) = 2.87, p = 0.093 -1.08 [-1.67, -0.48] -0.30 [-0.65, 0.03] No signal- inconsistent
Study 1b H2a: Past outcomes (Self vs Other) t(124) = -1.29, p = 0.198 -0.72 [-1.29, -0.14] -0.23 [-0.58, 0.12] No signal- inconsistent
H2b: Future outcomes (Self vs Other) t(124) = -1.63, p = .106 -0.82 [-1.40, -0.25] -0.29 [-0.64, 0.06] Signal- inconsistent
H3: Future vs Past F (1, 124) = 26.22, p < 0.001 -0.94 [-1.52, -0.35] -0.45 [-0.70, -0.20] Signal- inconsistent (smaller)
Study 2 (between H1: General (Self vs Other) F(1, 420) = 12.59, p < .001 -1.08 [-1.67, -0.48] -0.33 [-0.52, -0.14] Signal- inconsistent (smaller)
condition) H2a: Past outcomes (Self vs Other) t(420) = -1.42, p = .156 -0.72 [-1.29, -0.14] -0.21 [-0.48, 0.06] No signal- inconsistent
H2b: Future outcomes (Self vs Other) t(420) = -3.58, p < .001 -0.82 [-1.40, -0.25] -0.46 [-0.74, -0.18] Signal- inconsistent (smaller)
H3: Future vs Past F(1, 420) = 37.74, p < .001 -0.94 [-1.52, -0.35] -0.59 [-0.78, -0.39] Signal- inconsistent (smaller)
Study 2 (within H1: General (Self vs Other) F(1, 433) = 0.00, p = .988 -1.08 [-1.67, -0.48] 0.00 [-0.13, 0.13] No signal- inconsistent
condition) H2a: Past outcomes (Self vs Other) t(433) = 0.04, p = 0.411 -0.72 [-1.29, -0.14] -0.04 [-0.18, 0.08] No signal- inconsistent
H2b: Future outcomes (Self vs Other) t(433) = -0.04, p = 0.469 -0.82 [-1.40, -0.25] 0.04 [-0.09, 0.17] No signal- inconsistent
H3: Future vs Past F(1, 433) = 105.23, p < .001 -0.94 [-1.52, -0.35] -0.53 [-0.66, -0.39] Signal- inconsistent (smaller)
General Discussion
We aimed to conduct close direct replications of Experiment 1 in Pronin and Kugler
(2010), finding that people believe that their own decisions and life events are less predictable a
priori than similar decisions and life events for other people. We also went beyond the original
findings by adding extensions examining different study designs.
Across three replications we concluded a mostly successful replication of the self-other
asymmetry in predictability ratings, reported in Pronin and Kugler (2010)’s Experiment 1:
People rate themselves as being less predictable than others. The effect sizes were considerably
smaller than that reported in the original study. We found much weaker to no support for the
effect when we ran a novel within-subject design. This suggests that the tendency to believe that
oneself is less predictable than other people only emerged in separate evaluation mode, in which
the participant makes no direct comparison between herself and others. When the response mode
shifts from separate to joint evaluation, people can directly compare their self and other
judgments, and then seem to correct this bias in their judgments.
Our replications differed from the original study in some expected ways. For example,
stimuli used in the original article were targeted at and tested with American undergraduates
more than a decade ago. We used the same materials in an online survey with a more diverse
population in Studies 1b and 2. Direct close replications are never perfectly exact, yet despite the
differences, we believe that the theoretical grounds of the target paper remain relevant and that
the replications indicate a qualified support for the target article’s primary hypothesis.
It might be possible to make the link between the current findings and the phenomenon
that people view their own judgments as less biased than others’. For example, Pronin et al.
(2002) demonstrated a “bias blind-spot” – that people tend to perceive themselves as less
susceptible to biases than others. A recent replication based on a larger and more diverse sample
supported these findings, also showing links between free-will agency beliefs and the bias blind-
spot (Chandrashekar et al., 2021). An interesting direction for future research, would be to
examine whether given the self-other asymmetries regarding the bias blind-spot and agency, and
given the links between the agency beliefs and the bias blind-spot, that there might be a common
mechanism underlying these types of effects.
Our Study 2 also offers an important insight given the comparison of the between-subject
and within-subject design. We found support for an effect in a between-subject design but no
support for the effect using the within-subject design. This experimental contrast has been
commonly used in the judgment and decision-making literature in order to demonstrate that
when comparisons are possible people realize the bias and correct it. For example, Hsee (1998)
provided several curious demonstrations of the “Less Is Better” effect when people evaluate
objects or situations in isolation (separately, in a between-subject design) versus the more
rational “More Is Better” effect when people evaluate those in joint evaluations. His findings
were recently successfully replicated by Vonasch et al. (2023). Therefore, it would seem that in
our current replication, like in Hsee (1998) and Vonasch et al. (2023), the effect seems to appear
when evaluations are done in isolation with no reference point for comparison, yet seems to
dissolve when making joint evaluations.
Perhaps one way to tease apart the variation in the findings across between-persons vs.
within-person comparisons is to investigate the normative understanding of free will, in its
different descriptions. A valuable avenue for future research could involve investigating whether
individuals tend to favor or trust others who exhibit less predictability, and under what
circumstances. Likewise, an intriguing line of inquiry would be to examine whether individuals
express a preference for or trust those who demonstrate a heightened sense of agency.
In replication Study 1a that closely matched the target study, we found strong support for
predictions of the original study (Cohen d = -0.84, 95% CI [-1.46, -0.23]) based on a smaller
student sample (N = 47). In Study 1b with a larger MTurk sample (N = 126) the replication effect
size seemed smaller and just barely detectable signal (Cohen d = -0.30, 95% CI [-0.66, 0.05]),
whereas in Study 2, which had the largest sample, the effect with a between-subject design was
again smaller than the original but this time highly robust, like Study 1a (insert d, CI here). Seen
as a whole, we therefore consider this a mostly successful replication of the main finding in
Pronin and Kugler: Experiment 1 (2010), although the observed effect sizes were smaller. For
future studies aiming to replicate and extend these findings further, for example by looking into
underlying mechanisms and testing potential moderators, we would recommend recruiting a
much larger sample size.

References
Alicke, M. D. (1985). Global self-evaluation as determined by the desirability and controllability
of trait adjectives. Journal of personality and social psychology, 49(6), 1621.
https://doi.org/10.1037/0022-3514.49.6.1621
Alicke, M. D., Klotz, M. L., Breitenbecher, D. L., Yurak, T. J., & Vredenburg, D. S. (1995).
Personal contact, individuation, and the better-than-average effect. Journal of Personality
and Social Psychology, 68(5), 804–825. https://doi.org/10.1037/0022-3514.68.5.804
Alquist, J. L., Ainsworth, S. E., Baumeister, R. F., Daly, M., & Stillman, T. F. (2015). The
making of might-have-beens: Effects of free will belief on counterfactual
thinking. Personality and Social Psychology Bulletin, 41(2), 268-283.
https://doi.org/10.1177/0146167214563673
Caruso, E. M., Gilbert, D. T., & Wilson, T. D. (2008). A wrinkle in time: Asymmetric valuation
of past and future events. Psychological Science, 19(8), 796-801.
https://doi.org/10.1111/j.1467-9280.2008.02159.x
Chandrashekar, S. P., Yeung, S. K., Yau, K. C., Cheung, C. Y., Agarwal, T. K., Wong, C. Y. J.,
Pillai, T., Thirlwell, T. N., Leung, W. N., Tse, C., Li, Y. T., Cheng, B. L., Chan, H. Y. C.
& Feldman, G. (2021). Agency and self-other asymmetries in perceived bias and
shortcomings: Replications of the Bias Blind Spot and link to free will beliefs. Judgment
& Decision Making, 16(6). https://doi.org/10.1017/S1930297500008470
Chartier, C. R., Arnal, J. D., Arrow, H., Bloxsom, N. G., Bonfiglio, D. B. V., Brumbaugh, C. C.,
Corker, K. S., Ebersole, C. R., Garinther, A., Giessner, S. R., Hughes, S., Inzlicht, M.,
Lin, H., Mercier, B., Metzger, M., Rangel, D., Saunders, B., Schmidt, K., Storage, D., &
Tocco, C. (2020). Many Labs 5: Registered Replication of Albarracín et al.(2008),

Experiment 5. Advances in Methods and Practices in Psychological Science, 3(3), 332-
339. https://doi.org/10.1177/2515245920945963
Crescioni, A. W., Baumeister, R. F., Ainsworth, S. E., Ent, M., & Lambert, N. M. (2016).
Subjective correlates and consequences of belief in free will. Philosophical
Psychology, 29(1), 41-63. https://doi.org/10.1080/09515089.2014.996285
Dennett, D. C., & Caruso, G. D. (2021). Just Deserts: Debating Free Will. John Wiley & Sons.
Everett, J. A. C., Clark, C. J., Meindl, P., Luguri, J. B., Earp, B. D., Graham, J., Ditto, P. H., &
Shariff, A. F. (2021). Political differences in free will belief are associated with
differences in moralization. Journal of Personality and Social Psychology, 120(2), 461–
483. https://doi.org/10.1037/pspp0000286
Feldman, G., Baumeister, R. F., & Wong, K. F. E. (2014). Free will is about choosing: The link
between choice and the belief in free will. Journal of Experimental Social
Psychology, 55, 239-245. https://doi.org/10.1016/j.jesp.2014.07.012
Feldman, G., Farh, J. L., & Wong, K. F. E. (2018). Agency beliefs over time and across cultures:
Free will beliefs predict higher job satisfaction. Personality and Social Psychology
Bulletin, 44(3), 304-317. https://doi.org/10.1177/0146167217739261
Feldman, G., Wong, K. F. E., & Baumeister, R. F. (2016). Bad is freer than good: Positive–
negative asymmetry in attributions of free will. Consciousness and Cognition, 42, 26-40.
https://doi.org/10.1016/j.concog.2016.03.005
Gray, K., Young, L., & Waytz, A. (2012). Mind perception is the essence of
morality. Psychological Inquiry, 23(2), 101-124.
https://doi.org/10.1080/1047840X.2012.651387
Hsee, C. K. (1998). Less is better: when low-value options are valued more highly than high-
value options. Journal of Behavioral Decision Making, 11(2), 107–121.
https://doi.org/10.1002/(SICI)1099-0771(199806)11:2%3C107::AID-
BDM292%3E3.0.CO;2-Y
Hsee, C. K., Loewenstein, G. F., Blount, S., & Bazerman, M. H. (1999). Preference reversals
between joint and separate evaluations of options: a review and theoretical
analysis. Psychological Bulletin, 125(5), 576. https://doi.org/10.1037/0033-
2909.125.5.576
Jané, M., Xiao, Q., Yeung, S., Ben-Shachar, M. S., Caldwell, A., Cousineau, D., Dunleavy, D. J.,
Elsherif, M., Johnson, B., Moreau, D., Riesthuis, P., Röseler, L., Steele, J., Vieira, F.,
Zloteanu, M., & Feldman, G. (2024). Guide to Effect Sizes and Confidence Intervals.
http://dx.doi.org/10.17605/OSF.IO/D8C4G
Jones, E. E., & Nisbett, R. E. (1971). The actor and the observer: Divergent perceptions of the
causes of behavior. In E. E. Jones et al. (Eds), Attribution: Perceiving the causes of
behavior. Morristown, N.J.: General Learning Press.
Kahneman, D., & Ritov, I. (1994). Determinants of stated willingness to pay for public goods: A
study in the headline method. Journal of Risk and Uncertainty, 9(1), 5-37.
https://doi.org/10.1007/BF01073401
Korbmacher, M., Kwan, C., & Feldman, G. (2022). Both better and worse than others depending
on difficulty: Replication and extensions of Kruger’s (1999) above and below average
effects. Judgment and Decision Making, 17(2), 449–486.
https://doi.org/10.1017/S1930297500009189
Koppel, L., Andersson, D., Tinghög, G., Västfjäll, D., & Feldman, G. (2023). We are all less
risky and more skillful than our fellow drivers: Successful replication and extension of
Svenson (1981)‎. Meta-Psychology, 7. https://doi.org/10.15626/MP.2021.2932
Kruger, J. (1999). Lake Wobegon be gone! The" below-average effect" and the egocentric nature
of comparative ability judgments. Journal of Personality and Social Psychology, 77(2),
221. https://psycnet.apa.org/doi/10.1037/0022-3514.77.2.221
Lang, J. W., & Fries, S. (2006). A revised 10-item version of the Achievement Motives
Scale. European Journal of Psychological Assessment, 22(3), 216-224.
https://doi.org/10.1027/1015-5759.22.3.216
LeBel, E. P., McCarthy, R. J., Earp, B. D., Elson, M., & Vanpaemel, W. (2018). A unified
framework to quantify the credibility of scientific findings. Advances in Methods and
Practices in Psychological Science, 1(3), 389-402.
https://doi.org/10.1177/2515245918787489
LeBel, E. P., Vanpaemel, W., Cheung, I., & Campbell, L. (2019). A brief guide to evaluate
replications. Meta-Psychology, 3. https://doi.org/10.15626/MP.2018.843
Litman, L., Robinson, J., & Abberbock, T. (2017). TurkPrime. com: A versatile crowdsourcing
data acquisition platform for the behavioral sciences. Behavior Research Methods, 49(2),
433-442. https://doi.org/10.3758/s13428-016-0727-z
Martin, N. D., Rigoni, D., & Vohs, K. D. (2017). Free will beliefs predict attitudes toward
unethical behavior and criminal punishment. Proceedings of the National Academy of
Sciences, 114(28), 7325-7330.

Monroe, A. E., & Malle, B. F. (2010). From uncaused will to conscious choice: The need to
study, not speculate about people’s folk concept of free will. Review of Philosophy and
Psychology, 1(2), 211-224. https://doi.org/10.1007/s13164-009-0010-7
Monroe, A. E., Dillon, K. D., & Malle, B. F. (2014). Bringing free will down to Earth: People’s
psychological concept of free will and its role in moral judgment. Consciousness and
Cognition, 27, 100-108. https://doi.org/10.1016/j.concog.2014.04.011
Nadelhoffer, T., Shepard, J., Crone, D. L., Everett, J. A., Earp, B. D., & Levy, N. (2020). Does
encouraging a belief in determinism increase cheating? Reconsidering the value of
believing in free will. Cognition, 203, 104342.
https://doi.org/10.1016/j.cognition.2020.104342
Nahmias, E., Morris, S., Nadelhoffer, T., & Turner, J. (2005). Surveying freedom: Folk
intuitions about free will and moral responsibility. Philosophical Psychology, 18(5), 561-
584. https://doi.org/10.1080/09515080500264180
Nichols, S. (2004). The folk psychology of free will: Fits and starts. Mind & Language, 19(5),
473-502. https://doi.org/10.1111/j.0268-1064.2004.00269.x
Pronin, E., & Kugler, M. B. (2010). People believe they have more free will than
others. Proceedings of the National Academy of Sciences, 107(52), 22469-22474.
https://doi.org/10.1073/pnas.1012046108
Pronin, E., Lin, D. Y., & Ross, L. (2002). The bias blind spot: Perceptions of bias in self versus
others. Personality and Social Psychology Bulletin, 28(3), 369-381.
https://doi.org/10.1177/0146167202286008
Rigoni, D., Cleeremans, A., & Brass, M. (2017). Causes and consequences of the belief in free
will. In The science of lay theories (pp. 229-242). Springer, Cham.
https://doi.org/10.1007/978-3-319-57306-9_10
Sarkissian, H., Chatterjee, A., De Brigard, F., Knobe, J., Nichols, S., & Sirker, S. (2010). Is
belief in free will a cultural universal?. Mind & Language, 25(3), 346-358.
https://doi.org/10.1111/j.1468-0017.2010.01393.x
Schwarzer, G. (2007). meta: An R package for meta-analysis. R news, 7(3), 40-45.
Shariff, A. F., Greene, J. D., Karremans, J. C., Luguri, J. B., Clark, C. J., Schooler, J. W.,
Baumeister, R. F., & Vohs, K. D. (2014). Free will and punishment: A mechanistic view
of human nature reduces retribution. Psychological Science, 25(8), 1563–
1570. https://doi.org/10.1177/0956797614534693
Shepperd, J. A., Waters, E. A., Weinstein, N. D., & Klein, W. M. (2015). A primer on unrealistic
optimism. Current Directions in Psychological Science, 24(3), 232-237.
https://doi.org/10.1177/0963721414568341
Sjåstad, H., Baumeister, R. F., & Ent, M. (2020). Greener grass or sour grapes? How people
value future goals after initial failure. Journal of Experimental Social Psychology, 88,
103965. https://doi.org/10.1016/j.jesp.2020.103965
Stillman, T. F., Baumeister, R. F., & Mele, A. R. (2011). Free will in everyday life:
Autobiographical accounts of free and unfree actions. Philosophical Psychology, 24(3),
381-394. https://doi.org/10.1080/09515089.2011.556607
Stillman, T. F., Baumeister, R. F., Vohs, K. D., Lambert, N. M., Fincham, F. D., & Brewer, L. E.
(2010). Personal philosophy and personnel achievement: Belief in free will predicts better
job performance. Social Psychological and Personality Science, 1(1), 43-50.
https://doi.org/10.1177/1948550609351600
Svenson, O. (1981). Are we all less risky and more skillful than our fellow drivers? Acta
Psychologica, 47(2), 143–148. https://doi.org/10.1016/0001-6918(81)90005-6
Vonasch, A. J., Hung, W. Y., Leung, W. Y., Nguyen, A. T. B., Chan, S., Cheng, B. L., &
Feldman, G. (2023). ‎“Less Is Better” in Separate Evaluations Versus “More Is Better” in
Joint Evaluations: Mostly ‎Successful Close Replication and Extension of Hsee (1998)‎.
Collabra: Psychology, 9(1). https://doi.org/10.1525/collabra.77859
Weinstein, N. D. (1980). Unrealistic optimism about future life events. Journal of personality
and social psychology, 39(5), 806. https://doi.org/10.1037/0022-3514.39.5.806
Ziano, I., Mok, P. Y., & Feldman, G. (2021). Replication and Extension of Alicke (1985) Better-
Than-Average Effect for Desirable and Controllable Traits. Social Psychological and
Personality Science, 12(6), 1005-1017. https://doi.org/10.1177/1948550620948973

Do people believe they are less predictable than others? Three
replications of Pronin and Kugler (2010)’s Experiment 1
Supplementary
Disclosures .............................................................................................................................................. 2
Data collection .................................................................................................................................... 2
Conditions reporting ........................................................................................................................... 2
Variables reporting ............................................................................................................................. 2
Overview of Experimental Design in the Target article ......................................................................... 3
Experimental design........................................................................................................................ 3
Experimental material ..................................................................................................................... 3
Original article’s results .................................................................................................................. 4
Effect size calculations of the target article effects ................................................................................. 4
Power analyses ........................................................................................................................................ 5
Materials and scales related to replication part ....................................................................................... 8
Study Materials: Study 1a & 1b .......................................................................................................... 8
Study Materials: Study 2................................................................................................................... 10
Deviations from Preregistration ............................................................................................................ 13
Figures .............................................................................................................................................. 14
Additional results .................................................................................................................................. 16
Study 1a ............................................................................................................................................ 16
Study 1b ............................................................................................................................................ 16
Study 2 .............................................................................................................................................. 17
Study 2: Mixed design: Past versus future separately for self and for other ................................. 18
Study 2 Extension: Achievement Motives ..................................................................................... 21
Introduction ................................................................................................................................... 21
Results............................................................................................................................................... 21
Summary and Mini-meta analysis .................................................................................................... 23
Mini-meta-analyses including the original study .......................................................................... 23
Mini-meta analyses excluding the original study.......................................................................... 25
Appendix A .......................................................................................................................................... 27
References ............................................................................................................................................ 28
Pronin and Kugler (2010) Experiment 1: Supplementary 2
Disclosures
Data collection
Data collection was completed before analyzing the data.
Conditions reporting
We report all the conditions we collected.
Variables reporting
All variables collected for this study are reported and included in the provided data.
Overview of Experimental Design in the Target article

Experimental design
Original study authors ran 2(between) x 2 (within) mixed-design ANOVA comparing self-
other difference and past- future difference in the predictability of personal action.
Experimental material
Other/close friend
Please think about one of your roommates (or suitemates) at Princeton. Pick one person
only, and indicate his/her initials here: ____. If you do not have a roommate or suitemate,
choose one you previously had, and indicate his/her initials here: ____.
Looking back…
1. Think about your roommate’s decision to attend Princeton. How easy would it
have been to predict that he/she would end up going to Princeton? (Not at all
predictable = 1; Extremely predictable = 7)
2. Think about a person whom your roommate recently dated (but is not dating any
longer). How easy would it have been to predict that things would not last
between the two of them? (Not at all predictable = 1; Extremely predictable = 7)
3. Think about your roommate’s choice of what to major in. How easy would it have
been to predict that he/she would end up choosing that major? (Not at all
predictable = 1; Extremely predictable = 7)
Looking forward…
1. Think about what your roommate will end up doing for his/her career. How easy
would it be to predict what his/her career will be? (Not at all predictable = 1;
Extremely predictable = 7)
2. Think about the person whom your roommate will marry. How easy would it be to
predict who this person will be? (Not at all predictable = 1; Extremely predictable = 7)
3. Think about your roommate’s decision about where in the country he/she will live
during the decade after he/she graduates. How easy would it be to predict where
he/she will live? (Not at all predictable = 1; Extremely predictable = 7)
Self
Looking back…
1. Think about your decision to attend Princeton. How easy would it have been to
predict that you would end up going to Princeton?
2. Think about a person whom you recently dated (but are not dating any longer).
How easy would it have been to predict that things would not last between the two
of you?
3. Think about your choice of what to major in. How easy would it have been to
predict that you would end up choosing that major?
Looking forward…
1. Think about what you will end up doing for your career. How easy would it be to
predict what your career will be?
2. Think about the person whom you will marry. How easy would it be to predict
who this person will be?
3. Think about your decision about where in the country you will live during the
decade after you graduate. How easy would it be to predict where you will live?
Original article’s results

We calculated the effect sizes based on the results of the target study. See Appendix A for
sample calculation of Cohen’s d (95% CI intervals).
Prediction 1: Participant perceived their own outcome as less predictable a priori than those
of a roommate (the main effect of a priori predictability of self and others’ past and future
decision)
- Mean = 3.86(self) vs. 4.81(other), (on a 7-point scale ranging from 1 = not at all
predictable to 7 = extremely predictable);
- Standard Deviation = Not Reported
- The Statiscal Test: two-way ANOVA: F(1, 48) = 14.46
- The reported degree of freedom = 48
- The reported p-value = 0.0004.
- Cohen’s d = -1.08 [-1.68, -0.47]
Prediction 2: The simple effect for self-other difference for past outcomes
predictable to 7 = extremely predictable)
- The Statistical Test: two-way ANOVA: F(1, 48) = 6.40
- The reported p-value = 0.01
- Cohen’s d = -0.72 [-1.30, -0.13]
Prediction 3: The simple effect for self-other difference for future outcomes
- Cohen’s d = -0.82 [-1.41, -0.23]
Prediction 3: The differences for future vs past outcome predictabilities

- Cohen’s d = -0.94 [-1.53, -0.34]
Effect size calculations of the target article effects
Effect sizes were not reported in the target article. We thus computed these whenever
possible (when there was sufficient information).
Table S1. Summary of target findings in Pronin and Kugler (2010) Experiment 1
Cohen's d with 95% Partial eta-square
df 1 df 2 F p
CI with 95% CI
(H1) Across both past and future
outcomes: 1 48 14.46 0.0004 -1.08 [-1.68, -0.47] 0.23 [0.08, 1.00]
Self vs. Roommate
(H2a) Across both past outcomes:
1 48 6.40 0.010 -0.72 [-1.30, -0.13] 0.12 [0.01, 1.00]
Self vs. Roommate
(H2b) Across both future
outcomes: 1 48 8.46 0.005 -0.82 [-1.41, -0.23] 0.15 [0.03, 1.00]
Self vs. Roommate
(H3) Future vs. Past 1 48 10.94 0.002 -0.94 [-1.53, -0.34] 0.08 [0.00, 1.00]
Power analyses
Steps for power analysis (Pronin & Kugler, 2010):

These are the results reported in the target study:
Participants perceived their own outcomes as less predictable a priori than those of a
roommate (M = 3.86 vs. 4.81, on a 7-point scale ranging from 1 = not at all predictable to 7
= extremely predictable); F (1, 48) = 14.46, P = 0.0004. This self–other difference was
significant for both past outcomes (M = 4.32 vs. 5.16), F (1, 48) = 6.40, P = 0.01 and future
outcomes (M = 3.40 vs. 4.47), F (1, 48) = 8.46, P = 0.005. Consistent with previous
theorizing (1), there also was a main effect for time, whereby people perceived the future as
less predictable than the past (M = 3.93 vs. 4.74); F (1, 48) = 10.94, P = 0.002. There was no
interaction between past/future and self/other (F < 1).
Here is a summarised form of the statistics of the results:

● General: Self vs other (M = 3.86 vs 4.81, SD not provided), F (1, 48) = 14.46, p = 0.0004
● Past outcomes: Self vs other (M = 4.32 vs 5.16, SD not provided), F (1, 48) = 6.40, p = 0.01
● Future outcomes: Self vs Other (M = 3.40 vs 4.47, SD not provided), F (1, 48) = 8.46, p =
0.005
● Future vs Past (M = 3.93 vs 4.74, SD not provided), F (1, 48) = 10.94, p = 0.002
Standard deviations were not provided so we took the F statistic and the overall number of subjects to
extract the effect size in Cohen’s d. As the target study did not state how were the subjects divided
into the two conditions, we assumed that the total number of subjects was divided equally into the
conditions. We used the effect size calculator on http://www.uv.es/~friasnav/TEdatospublicados.xls -
‘How to Calculate Effect Sizes from Published Research: A Simplified Spreadsheet’, by Will
Thalheimer and Samantha Cook.
For the general main effect (self vs other), the F statistic = 14.46, and the overall number of subjects
was 50 with 25 in each condition, we calculated a Cohen’s d of 1.1, a very large effect.
For the effect on past outcomes (self vs other), the F statistic = 6.40, and the overall number of
subjects was 50 with 25 in each condition, we calculated a Cohen’s d of .73, a medium effect.
For the effect on future outcomes (self vs other), the F statistic = 8.46, and the overall number of
subjects was 50 with 25 in each condition, we calculated a Cohen’s d of .84, a large effect.
For the effect of time (past vs future), the F statistic = 10.94, and the overall number of subjects was
50 with 25 in each condition, we calculated a Cohen’s d of .95, a large effect.
1. We calculated the effect size (Cohen’s d).

2. Used the Cohen’s d for the power analysis with GPower (Faul et al., 2007).
Gpower captures
G*Power output
Power analysis Analysis: A priori: Compute required sample size
output for Study Input: Tail(s) = Two
1a and Study 1b Effect size d = 1.1
α err prob = 0.05
Power (1-β err prob) = 0.95
Allocation ratio N2/N1 = 1
Output: Noncentrality parameter δ = 3.7302815
Critical t = 2.0153676
Df = 44
Sample size group 1 = 23
Total sample size = 46
Actual power = 0.9542682
Considering the effect size of the general main effect which is the most
crucial to the experiment, d = 1.1. Hence, a sample size of 23
participants in each condition is required for two-tail power .95 alpha
.05. Therefore, a sample of 46 participants is required.
Cohen’s d = 0.32 based on the results of the Study 1b

Power analysis
G*Power output
output for Study
2 t tests - Means: Difference between two independent means (two
groups)
Analysis: A priori: Compute required sample size
Input: Tail(s) = One
Effect size d = 0.32

α err prob = 0.05
Power (1-β err prob) = 0.95
Allocation ratio N2/N1 = 1
Output: Noncentrality parameter δ = 3.3023628
Critical t = 1.6484553
Df = 424
Total sample size = 426
Actual power = 0.9507565
Materials and scales related to replication part
Study Materials: Study 1a & 1b

OSF Link to the Qualtrics survey file (in word doc format):
Study 1a: https://osf.io/qmhy
Study 1b: https://osf.io/n6gsy
Introduction to evaluations of others

First, please think about your closest friend or your roommate at your university.
Pick one person only and indicate his/her first name initials here (do NOT write full name or
any identifying details, to ensure full anonymity):
Past events in others' lives
Read the following statements about certain past events. Think about how predictable those
events were in your friend/roommate's life and rate them on a scale from 1 (not at all
predictable) to 7 (extremely predictable).
1. Think about your friend/roommate’s decision to attend your university. How easy would
it have been to predict that he/she would end up going to this university? (1= Not at all
predictable; 7 = Extremely predictable)
2. Think about a person whom your friend/roommate recently dated (but is not dating any
longer). How easy would it have been to predict that things would not last between the
two of them? (1= Not at all predictable; 7 = Extremely predictable)
3. Think about your friend/roommate's choice of what to major in. How easy would it have
been to predict that he/she would end up choosing that major? (1= Not at all predictable;
7 = Extremely predictable)
Future events in others' lives

Read the following statements about certain future events. Think about how predictable those
events are in your friend/roommate's life and rate them on a scale from 1 (not at all
1. Think about what your friend/roommate will end up doing for his/her career. How easy
would it be to predict what his/her career would be? (1= Not at all predictable; 7 =
Extremely predictable)
2. Think about the person whom your friend/roommate will marry. How easy would it be to
predict who this person would be? (1= Not at all predictable; 7 = Extremely predictable)
3. Think about your friend/roommate's decision about where he/she will live during the
decade after he/she graduates. How easy would it be to predict where he/she would live?
(1= Not at all predictable; 7 = Extremely predictable)
Introduction to evaluations of self

In this section you're asked to evaluate events in your own life.
Read the following statements about past events in your life. Think about how predictable
those events are and rate them on a scale from 1 (not at all predictable) to 7 (extremely
predictable).
1. Think about your decision to attend your current university. How easy would it have been
to predict that you would end up going to this university? (1= Not at all predictable; 7 =
2. Think about a person whom you recently dated (but are not dating any longer). How easy
would it have been to predict that things would not last between the two of you? (1= Not
at all predictable; 7 = Extremely predictable)
3. Think about your choice of what to major in. How easy would it have been to predict that
you would end up choosing that major? (1= Not at all predictable; 7 = Extremely
predictable)
Future events in your own life

Read the following statements about future events in your life. Think about how predictable
predictable).
1. Think about what you will end up doing for your career. How easy would it be to predict
what your career would be? (1= Not at all predictable; 7 = Extremely predictable)
2. Think about the person whom you will marry. How easy would it be to predict who this
person would be? (1= Not at all predictable; 7 = Extremely predictable)
3. Think about your decision about where you will live during the decade after you
graduate. How easy would it be to predict where you would live? (1= Not at all
Study Materials: Study 2
OSF Link to the Qualtrics survey file (in word doc format):
Study 2: https://osf.io/vx4pa
Introduction: Students only
This experiment is only for current American undergraduate students.
This experiment has 6 questions presented in one page asking for you evaluations of
behaviors. There are no right or wrong answers, answer to the best of your understanding.
Are you currently an American undergraduate student?
 Yes, I am currently an undergraduate student (1)
 No, I am not an undergraduate student (2)
What year of your undergraduate studies are you currently at?
 First year (1)
 Second year (2)
 Third year (3)
 Forth year+ (4)
 Not currently enrolled (5)
Achievement motivations
The following questions are meant to understand your motivations for achievement.
Please indicate your agreement with each of the following statements: (1= Strongly
disagree; 4 = Strongly agree)
1. I like situations, in which I can find out how capable I am.
2. When I am confronted with a problem, which I can possibly solve, I am enticed to start
working onit immediately.
3. I enjoy situations, in which I can make use of my abilities.
4. I am appealed by situations allowing me to test my abilities.
5. I am attracted by tasks, in which I can test my abilities.
6. I am afraid of failing in somewhat difficult situations, when a lot depends on me.
7. I feel uneasy to do something if I am not sure of succeeding.
8. Even if nobody would notice my failure, I’m afraid of tasks, which I’m not able to solve.
9. Even if nobody is watching, I feel quite anxious in new situations.
10. If I do not understand a problem immediately I start feeling anxious.
Introduction to evaluations of others

First, please think about your closest friend or your roommate at your university.
Pick one person only and indicate his/her first name initials here (do NOT write full name or
any identifying details, to ensure full anonymity):
Past events in others' lives
Read the following statements about certain past events. Think about how predictable those
events were in your friend/roommate's life and rate them on a scale from 1 (not at all
4. Think about your friend/roommate’s decision to attend your university. How easy would
it have been to predict that he/she would end up going to this university? (1= Not at all
5. Think about a person whom your friend/roommate recently dated (but is not dating any
longer). How easy would it have been to predict that things would not last between the
two of them? (1= Not at all predictable; 7 = Extremely predictable)
6. Think about your friend/roommate's choice of what to major in. How easy would it have
been to predict that he/she would end up choosing that major? (1= Not at all predictable;
7 = Extremely predictable)
Future events in others' lives

Read the following statements about certain future events. Think about how predictable those
events are in your friend/roommate's life and rate them on a scale from 1 (not at all
4. Think about what your friend/roommate will end up doing for his/her career. How easy
would it be to predict what his/her career would be? (1= Not at all predictable; 7 =
5. Think about the person whom your friend/roommate will marry. How easy would it be to
predict who this person would be? (1= Not at all predictable; 7 = Extremely predictable)
6. Think about your friend/roommate's decision about where he/she will live during the
decade after he/she graduates. How easy would it be to predict where he/she would live?
(1= Not at all predictable; 7 = Extremely predictable)
Introduction to evaluations of self

In this section you're asked to evaluate events in your own life.
Read the following statements about past events in your life. Think about how predictable
predictable).
4. Think about your decision to attend your current university. How easy would it have been
to predict that you would end up going to this university? (1= Not at all predictable; 7 =
5. Think about a person whom you recently dated (but are not dating any longer). How easy
would it have been to predict that things would not last between the two of you? (1= Not
at all predictable; 7 = Extremely predictable)
6. Think about your choice of what to major in. How easy would it have been to predict that
you would end up choosing that major? (1= Not at all predictable; 7 = Extremely
predictable)
Future events in your own life

Read the following statements about future events in your life. Think about how predictable
predictable).
4. Think about what you will end up doing for your career. How easy would it be to predict
what your career would be? (1= Not at all predictable; 7 = Extremely predictable)
5. Think about the person whom you will marry. How easy would it be to predict who this
person would be? (1= Not at all predictable; 7 = Extremely predictable)
6. Think about your decision about where you will live during the decade after you
graduate. How easy would it be to predict where you would live? (1= Not at all
Deviations from Preregistration

Table S2. Pre-registration plan versus final report
Components in Location of Deviatio Description of deviation Rationale Impact of deviation Date/time of decision for
preregistration preregistered ns for on results deviation + stage
decision/plan deviation
Study design Study 1a: no / / / /

https://osf.io/j5ms4
Measured variables no / / / /
Study 1b:
Exclusion criteria https://osf.io/xwhu3 no / / / /
https://osf.io/tfwrg
IV no / / / /
DV Study 2: no / / / /
https://osf.io/7ucjv
Data analysis Yes We do not deviate in our partial eta- none June 2023
analysis but report the squared is
effect using a partial eta- more apt
squared effect size effect size
measure. Power analysis measure
was based on cohen’d for anova
effect size measure. analysis
Additional results
Study 1a
Table S3. Study 1a: Results of mixed ANOVA
p-
Effect df MSE F 𝜼𝟐𝒑 (95% CI)
value
Self vs. Other 1, 45 1.42 8.33 .006 0.16 [0.03, 1.00]
Future vs. Past 1, 45 1.32 16.26 <.001 0.27 [0.10, 1.00]
(Self vs. Other) x (Future
1, 45 1.32 1.35 .251 0.03 [0.00, 1.00]
vs. Past)
Table S4. Study 1a: Results of post hoc comparisons based on mixed ANOVA
Contrast estimate SE df T p-value Cohen’s d, 95% CI

Future
Self - Other -0.99 0.36 45 -2.72 0.0092 -0.79 [-1.40, -0.18]
Past
Self - Other -0.43 0.32 45 -1.36 0.1817 -0.40 [-0.99, -0.20]
Study 1b
Table S5. Study 1b: Results of mixed ANOVA
p-
Effect df MSE F 𝜼𝟐𝒑 with (95% CI)
value
Self vs. Other 1, 124 2.25 2.87 .093 0.02 [0.00, 1.00]
Past v. Future 1, 124 0.74 26.22 <.001 0.17 [0.08, 1.00]
(Self vs. Other) x (Past vs.
1, 124 0.74 0.42 .516 0.00 [0.00, 1.00]
Future)
Table S6. Study 1b: Results of post hoc comparisons based on mixed ANOVA
Future
Self - Other -0.39 0.24 124 -1.626 0.106 -0.29 [-0.64, -0.06]
Past
Self - Other -0.25 0.19 124 -1.29 0.198 -0.23 [-0.58, -0.12]
Study 2
Table S7. Study 2: Results of ANOVA based on responses from between-subjects design
p-
Effect df MSE F 𝜼𝟐𝒑 with (95% CI)
value
Self vs. Other 1, 420 0.82 12.59 <.001 0.03 [0.01, 1.00]
Past v. Future 1, 420 0.82 37.74 <.001 0.08 [0.05, 1.00]
1, 420 0.82 2.40 .122 0.00 [0.00, 1.00]
Future)
Table S8. Study 2: Results of post hoc comparisons based on ANOVA from between-subjects
design

Future
Self - Other -0.45 0.12 420 -3.59 <.001 -0.46 [-0.74, -0.18]
Past
Self - Other -0.18 0.12 420 -1.42 0.156 -0.21 [-0.48, -0.06]
Table S9. Study 2: Results of ANOVA based on responses from within-subjects design
Effect df MSE F p-value 𝜼𝟐𝒑 with (95% CI)
Self vs. Other 1, 433 0.66 0.00 .988 0.00 [0.00, 1.00]
Past v. Future 1, 433 0.68 105.23 <.001 0.20 [0.14, 1.00]

1, 433 0.47 1.43 .233 0.00 [0.00, 1.00]
Future)
Table S10. Study 2: Results of post hoc comparisons based on ANOVA from within-subjects
design
Future
Self - Other 0.04 0.04 433 0.82 .412 0.05 [-0.09, 0.18]
Past
Self - Other -0.04 0.05 433 -0.72 .469 -0.04 [-0.17, -0.09]
Study 2: Mixed design: Past versus future separately for self and for other
We also added a condition with a design in which participants rated the predictability
of past or future events across each of the self and other’s reference point. The participants in
the mixed condition were assigned to one of the four scenario combinations: 1) Self-past and
Other-past; 2) Self-past and Other-future; 3) Self-future and Other-past; 4) Self-future and
Other-future. The order of the presentations within the combinations was randomized. We
summarized descriptive in Table 5 and Figure 5. The means were very similar to that of the
within-subject condition, with no differences between self and other.
Table S11. Study 2 mix design: Predictability descriptives

Target [Between] Self Other
Past 3.78 0.75 3.68 0.77
Future 3.22 0.94 3.22 0.95
Note. N = 437.
Figure S4. Study 2 mix design: Self-other predictability separately for past and future
choices
Given the nature of mixed experimental design, we conducted a mixed effects
regression analysis. We regressed predictability ratings on target frame (self vs. other), time
factor (past vs future), and interaction between reference frame and time factor as fixed
effects predictors, and participants’ ID as the random intercept. We constructed the linear
mixed model (using the lme4 package in R; Bates et al., 2015) in R. The p values were
calculated using the Satterthwaite approximation for the mixed-effects regressions
(Kuznetsova et al., 2017). The results of the analysis are presented in the Table #. There was
no support for target frame as a predictor. However, we found support for the view that future
events (M= 3.22; SD= 0.94) are rated less predictable than the past events (M= 3.73; SD=
0.76; estimate = -0.57, 95% CI [-0.73 – -0.41], p <.001).
Table S12. Study 2: Results of mixed effect regression analysis

DV: Predictability
Predictors Estimates CI p
(Intercept) 3.78 3.67 – 3.90 <0.001
Target frame (Other vs. Self) -0.10 -0.25 – 0.06 0.222
Time (Future vs. Past) -0.57 -0.73 – -0.41 <0.001
Target frame × Time 0.09 -0.13 – 0.32 0.420
Random Effects
σ2 0.65
τ00 ParticipantID 0.08
ICC 0.11
N ParticipantID 437
Observations 874
Marginal R2 / Conditional R2 0.086 / 0.191
Study 2 Extension: Achievement Motives

Introduction
Our Study 2 also included a second extension in which we measured individual differences in
achievement motivation (Lang & Fries, 2006) as a potential trait moderator. Since a high need for
achievement is positively associated with work ethic and performance, it could well be the case that
the ego-centric effect in predictability judgments is more pronounced for participants with high rather
than low achievement motivation. Such a result would provide initial correlational evidence that the
effect is associated with motivated beliefs.
Table S13
Summary of extension hypotheses
Study Hypothesis
Hypothesis 4 People with high (vs. low) achievement motivation will have a stronger
tendency to think that they have more free will than others (i.e., less
predictable)
Hypothesis 5 People with high (vs. low) achievement motivation will show a greater future
vs. past discrepancy.
Results
Our Study 2 involved experimental designs in which the Target (self vs. other) and Time (past vs
future) varied as a between or within subject factors. Although, we presented the results
independently within each designed, they are less helpful in summarizing the results across the
sample. Therefore, we conducted a linear mixed-effects analysis to draw summary conclusions. We
pooled the responses across experimental conditions to conduct a linear mixed-effects analysis. Linear
mixed-effects models (LMEMs) have several advantages. More importantly, mixed effect analysis is
appropriate for our extension hypothesis that proposed the moderating role of the achievement motive
measured at the individual level (Lang & Fries, 2006). The results of the analysis are presented in
Table 10.
We found no support for differences between self and other. The results of interaction as test of
Hypothesis 4 and 5. The results suggest no support the prediction (H4) that people with high (vs. low)
achievement motivation will have a stronger tendency to think view outcomes are less predictable
than for others (estimate = -0.10, 95% [-0.28, 0.08]), p = .291). Similarly, we found no support for the
prediction (H5) that people with high (vs. low) achievement motivation will show a greater future vs.
past discrepancy (estimate = 0.01, 95% [-0.17, 0.20]), p = .896). Besides, results held support for the
view that future events as more predictable than participants who rated future events.
Table S14
Study 2: Results of mixed effect regression analysis
Predictability
Predictors Estimates CI p
(Intercept) 3.34 2.96 – 3.72 <0.001

Perspective (Other vs. Self) 0.23 -0.26 – 0.72 0.365
Time (Past vs. Future) -0.54 -1.04 – -0.04 0.034
Achievement Motivation Scale (AMS) 0.13 -0.01 – 0.28 0.067
Perspective × Time 0.48 -0.23 – 1.19 0.183
Perspective × AMS -0.10 -0.28 – 0.08 0.291
Time × AMS 0.01 -0.17 – 0.20 0.896
Perspective × Time × AMS -0.14 -0.40 – 0.12 0.290
Random Effects
σ2 0.61
τ00 ParticipantID 0.22
ICC 0.26
N ParticipantID 1295
Observations 3034
Marginal R2 / Conditional R2 0.062 / 0.308
Summary and Mini-meta analysis
Mini-meta-analyses including the original study
Forest plots of the mini meta-analyses. CI = confidence interval. ES = effect size (Cohen's d).
Figure S5. Testing H1 (People would view their own past and future decision as less
predictable a priori than those of a roommate.)
Figure S6. Testing H2a (People will consider their own past decisions as less predictable than
the ones of others.)
Figure S7. Testing H2b (People will consider their own future decisions as less predictable
than the ones of others.)
Figure S8. Testing H3 (People would view future decisions as less predictable than past
decisions.)
Mini-meta analyses excluding the original study
Note: Forest plots of the mini meta-analyses. CI = confidence interval. ES = effect size
(Cohen's d).
Figure S9. Testing H1 (People would view their own past and future decision as less
predictable a priori than those of a roommate.)
Figure S10. Testing H2a (People will consider their own past decisions as less predictable
Figure S11. Testing H2b (People will consider their own future decisions as less predictable
Figure S12. Testing H3 (People would view future decisions as less predictable than past
decisions.)
Appendix A
Effect size and 95 % Confidence Intervals Calculation
Using: http://kylehamilton.net/shiny/MAVIS/
MAVIS: Meta Analysis via Shiny
● Cohen’ d = 1.08
● 95% CI of the main effect of self vs others = [0.47, 1.68]
References
Pronin, E., & Kugler, M. B. (2010). People believe they have more free will than
others. Proceedings of the National Academy of Sciences, 107(52), 22469-22474.
LeBel, E. P., McCarthy, R. J., Earp, B. D., Elson, M., & Vanpaemel, W. (2018). A unified
framework to quantify the credibility of scientific findings. Advances in Methods and
Practices in Psychological Science, 1, 389-402.
LeBel, E. P., Vanpaemel, W., Cheung, I., & Campbell, L. (2019). A brief guide to evaluate
replications. Meta-Psychology, 3.

Chandrashekar Etal 2024 IRSP Pronin Kugler 2010 Replication Extension Preprint v2

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Chandrashekar Etal 2024 IRSP Pronin Kugler 2010 Replication Extension Preprint v2

Uploaded by

Copyright:

Available Formats

Do people believe they are less predictable than others?

Three replications of Pronin and Kugler (2010)’s Experiment 1

*Subramanya Prasad Chandrashekar

** Chelsea (Chi Wing) LO

**Yong Jun KUEH

***Emily Sihui ZHONG (ezhong@connect.hku.hk / zhong_emil@bentley.edu),

Contributor Roles Taxonomy

suggesting higher free-will attributions. We conducted three pre-registered replications of this

with weaker effects. In Study 2 we contrasted between-subject versus within-subject designs in a

Keywords: Predictability, self-other asymmetry, free will attributions, replications

Do people believe they are less predictable than others?

Three replications of Pronin and Kugler (2010)’s Experiment 1

In the current research, we report three pre-registered replications of this experiment

generalizability of this effect.

‘Free Will’ Research in Psychology

should be conceptualized in formal terms, whether it is compatible with determinism and

when reflecting on their own comparing to that of others’ life circumstances.

Early replication attempts of free-will research focused on so-called “priming” activation

Everett et al., 2021; Feldman et al., 2016, 2018).

Self-Other Asymmetries in Human Judgment

There are many documented phenomena of self-other asymmetries. One of those is

nuances and moderators, such as an interaction with difficulty (Kruger, 1999).

Cognitive neglect vs. Self-enhancement motives

be partially or entirely influenced by a cognitive bias, where individuals rely on introspective

it can be argued that if the self-other asymmetry is primarily driven by self-enhancement

evaluate this possibility.

Extension: Examining bias, by contrasting within and between subject designs

with a variety of contrasts (1) fully within-subject conditions, 2) a fully between-subject

condition, 3) mixed-design) within a single data collection.

Previous studies suggested that if a given effect is driven by cognitive neglect or

Sjåstad et al., 2020).

In the current research, we report three pre-registered replications of Experiment 1 in

examine the possible role of cognitive neglect versus motivational self-enhancement as

mechanisms. In Study 2 we included an extension contrasting the fully within-subject condition

included a trait achievement measure as a possible individual-difference moderator of the effect.

We provided an overview of the findings of Experiment 1 of the target article in Table 2.

Pronin and Kugler (2010): Summary of hypotheses

Summary of original findings in Pronin and Kugler (2010) Experiment 1

(H3) Future vs. Past 1 48 10.94 .002 -0.94 [-1.52, -0.35]

Open science, preregistrations, and disclosures

https://osf.io/ykmqp/; preregistration Study 1a: https://osf.io/jg94e ; preregistration Study 1b:

https://osf.io/yaepf/ ; preregistration Study 2: https://osf.io/sjb8p). All studies, participants,

collection was completed before hypothesis testing.

can be found in the supplementary materials.

In Study 2, we used the G*Power software to determine the number of participants

moderate between-subjects effect and set this as our target sample.

Deviations from Original Studies

We provided detailed information of designs (type of study, sample, variables, exact

university students and made several adjustments and added extensions.

Comparison of the sample in the original study and in the replications

Pronin and Replication Replication Replication

We categorized Studies 1a and 1b as “very close replications” and Study 2 as a “close

Evaluation of replication closeness (using criteria by LeBel et al., 2018)

Design facet Replication Details

Physical settings Different Online data collection in our replication vs filling a

Note of alpha tails and p-values interpretations

hypotheses are well stated and these were replications.

Study 1a: Hong Kong Student sample

this study to be a conservative pre-test replication on a well-informed sample.

no demographic information in this study.

place to live, six target events in total.

We pre-registered an exclusion criteria (failing to complete the entire survey, failing to

We summarized descriptives in Table 5 and Figure 1.