Staying Positive in a Dystopian Future: A Novel Dissociation Between Personal and Collective Cognition

Abstract

The future of groups of people is a topic of broad interest in society and academia. Nonetheless, relatively little is known about the manner in which people think about the collective future of groups, and whether personal and collective future thinking represent distinct domains of future-oriented cognition. In the present studies (N = 691), we used an adapted future fluency task to demonstrate a novel domain-by-valence interaction between personal and collective future thinking, such that U.S.-based participants were positively biased about their personal future while at the same time being negatively biased about the future of their country. We further present evidence that this valence-based dissociation extends into the distant future, emerges in a non-U.S. (Canadian) sample, depends on the individual’s relation to the group, and has consequences for how people think about the world around them. Taken together, our findings represent the first behavioral evidence of a dissociation between personal and collective future thinking, and suggest that the study of collective future thinking represents a fruitful endeavor for psychological science.

Full text

Journal of Experimental Psychology:
General
Staying Positive in a Dystopian Future: A Novel
Dissociation Between Personal and Collective Cognition
Sushmita Shrikanth, Piotr M. Szpunar, and Karl K. Szpunar
Online First Publication, March 29, 2018. http://dx.doi.org/10.1037/xge0000421
CITATION
Shrikanth, S., Szpunar, P. M., & Szpunar, K. K. (2018, March 29). Staying Positive in a Dystopian
Future: A Novel Dissociation Between Personal and Collective Cognition. Journal of Experimental
Psychology: General. Advance online publication. http://dx.doi.org/10.1037/xge0000421

Staying Positive in a Dystopian Future: A Novel Dissociation Between
Personal and Collective Cognition
Sushmita Shrikanth
University of Illinois at Chicago
Piotr M. Szpunar
University at Albany, State University of New York
Karl K. Szpunar
University of Illinois at Chicago
The future of groups of people is a topic of broad interest in society and academia. Nonetheless, relatively little
is known about the manner in which people think about the collective future of groups, and whether personal
and collective future thinking represent distinct domains of future-oriented cognition. In the present studies
(N⫽691), we used an adapted future fluency task to demonstrate a novel domain-by-valence interaction
between personal and collective future thinking, such that U.S.-based participants were positively biased about
their personal future while at the same time being negatively biased about the future of their country. We
further present evidence that this valence-based dissociation extends into the distant future, emerges in a
non-U.S. (Canadian) sample, depends on the individual’s relation to the group, and has consequences for how
people think about the world around them. Taken together, our findings represent the first behavioral evidence
of a dissociation between personal and collective future thinking, and suggest that the study of collective future
thinking represents a fruitful endeavor for psychological science.
Keywords: future thinking, collective cognition, positivity bias, negativity bias
The future of groups is a topic of much consideration within
political science, communication studies, sociology, and related
disciplines (Szpunar & Szpunar, 2016). Nonetheless, little is cur-
rently known about the manner in which the sociopolitical future
of groups is mentally represented at the level of the individual, and
more importantly, whether or not the personal and collective future
represent distinct domains of future thinking. This is perhaps
somewhat surprising given that cognitive psychologists and neu-
roscientists have long been interested in illuminating the cognitive
and neural mechanisms that support the ability to think about the
future (Schacter et al., 2012; Suddendorf & Corballis, 2007; Sz-
punar, 2010). Much of this attention, however, has focused on the
ability to think about the personal future and not the future of the
collective or sociopolitical context in which individual lives are
embedded (Merck, Topcu, & Hirsh, 2016; Michaelian & Sutton,
2017; Szpunar & Szpunar, 2016).
The few studies that have investigated relations between per-
sonal and collective aspects of future-oriented cognition have done
so in the context of neuropsychological investigations of amnesia.
Specifically, amnesic patients who are unable to remember details
from their past are also often unable to imagine specific experi-
ences that may take place in their future (Hassabis, Kumaran,
Vann, & Maguire, 2007; Race, Keane, & Verfaellie, 2011; but see
Squire et al., 2010), but are nonetheless able to think about the
future of their communities, country, and the world (Andelman,
Hoofien, Goldberg, Aizenstein, & Neufeld, 2010; Klein, Loftus, &
Kihlstrom, 2002). Recent work suggests that elaborating on de-
tailed scenarios relevant to the collective future may be impaired in
amnesia, but the more general ability to identify relevant issues
appears to remain intact (Race, Keane, & Verfaellie, 2013).
While the above-noted findings from neuropsychological pa-
tients suggest that personal and collective future thinking are
indeed dissociable from one another, the study of relations be-
tween these constructs remains largely unexplored in the literature.
One reason for this state of affairs is the lack of viable paradigms
that can demonstrate dissociations between personal and collective
future thinking in healthy cognition. Here, we propose that expec-
tations regarding the personal and collective future may differ
sharply in terms of their emotional tone, and that tests of the
fluency with which associated thoughts come to mind can be used
to reveal such a valence-based dissociation. Specifically, we pre-
dicted that people should expect a positive personal future despite
holding negative expectations for the future of their country.
Country was selected as the unit of collective cognition as this
level of analysis has often been the focus of interest in past
considerations of the topic (e.g., Sani et al., 2007).
Our predictions were motivated by recent work from cognitive
psychology and neuroscience that has convincingly demonstrated
Sushmita Shrikanth, Department of Psychology, University of Illinois at
Chicago; Piotr M. Szpunar, Department of Communication, University at
Albany, State University of New York; Karl K. Szpunar, Department of
Psychology, University of Illinois at Chicago.
We thank Zachary McKenzie for assistance with data collection for
Experiment 5.
Correspondence concerning this article should be addressed to Karl K.
Szpunar, Department of Psychology, University of Illinois at Chicago,
1007 West Harrison Street, M/C 285, Chicago, IL, 60607. E-mail:
szpunar@uic.edu
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
Journal of Experimental Psychology: General
© 2018 American Psychological Association 2018, Vol. 1, No. 2, 000
0096-3445/18/$12.00 http://dx.doi.org/10.1037/xge0000421
1

that memories of the past serve as a basis for forming expectations
of the future (Schacter & Addis, 2007; Schacter et al., 2012; see
also Gilbert & Wilson, 2007). Indeed, prior work has demonstrated
that people tend to both remember (Kahneman, Krueger, Schkade,
Schwarz, & Stone, 2009; Walker & Skowronski, 2009) and expect
more positive than negative experiences in their personal past and
future (for recent reviews, see MacLeod, 2016, 2017). Further
supporting this close relation between the personal past and future,
people who remember their past in a less positive light (e.g., in the
context of depression) also imagine a less positive future (e.g.,
MacLeod & Byrne, 1996).
What about the collective future? To our knowledge, no prior
research has systematically assessed the positive and/or negative
expectations that people hold regarding the future of their country.
However, there does exist relevant work on how people remember
national events. For instance, Liu et al. (2009) asked participants
from 12 nations to list the most important events from their
country’s past. Notably, participant responses from most nations
(including the U.S.)—whether those responses reflected historical
facts or more contemporary occurrences—were dominated by neg-
ative relative to positive events (e.g., wars); that is, a negativity
bias (see also Liu et al., 2005). One possible account for this
phenomenon, particularly in relation to more contemporary events,
is that the primary source through which people learn about
national issues—namely, the media (McCombs, 2006; Norris,
2000)—tends to present information in a manner that is negatively
skewed (e.g., Soroka & McAdams, 2015). Hence, a collective
negativity bias in memory may reflect the type of information that
people have most typically been exposed to in the past. Impor-
tantly, if people tend to remember their nation’s past as relatively
negative, then expectations for the future of their country should
also be negative.
In order to test this hypothesis, we adapted a future fluency
task that has previously been used to assess the accessibility of
positive and negative events in healthy and mood-disordered
populations (MacLeod, Rose, & Williams, 1993). This task
typically requires people to list as many things (i.e., events or
occurrences that are reasonably likely to occur in the future)
that they are excited or worried about in the next week, year,
and 5–10 years (MacLeod & Byrne, 1996; MacLeod, Tata,
Kentish, & Jacobsen, 1997). In the present study, we asked
participants to not only list things that they were excited or
worried about across the specified time periods in relation to
their personal future, but also to list things that they believed
their country may be excited or worried about over those time
frames. Experiments 1 and 2 assessed the generalizability of our
predicted results, that is, a domain-by-valence interaction re-
flecting a personal positivity bias and a collective negativity
bias, in U.S.-based samples. Experiment 3 assessed the extent to
which our hypothesized dissociation extended into the distant
(40 years) future. Experiment 4 assessed the generalizability of
our predicted results to a non-U.S. (Canadian) sample. Exper-
iment 5 assessed whether cognitions about the future of groups
are always negative. Finally, Experiment 6 assessed the impact
of adopting a personal or collective perspective on information
processing in relation to topics (social media and artificial
intelligence) that can have both positive and negative conse-
quences.
Experiment 1
Method
Participants. Forty-eight undergraduate students were re-
cruited through the University of Illinois at Chicago Psychology
Subject Pool (female ⫽33; M
Age
⫽19.1 years). Half of our
participants completed the experiment before and the other half
completed the experiment after the 2016 United States Presidential
election (i.e., either prior to the nomination of Presidential candi-
dates in February 2016 or following the election of a new President
in January 2017). Given that the University of Illinois at Chicago
is a commuter school located in a county that voted nearly 75% in
favor of the Presidential candidate who lost the 2016 general
election (Leip, 2016), this aspect of our design allowed us to assess
whether expectations about the future of the country were influ-
enced by the outcome of the recent election. A power analysis of
a prior study showing valence differences in a personal future
fluency task (MacLeod et al., 1997) revealed that eight participants
would have been needed to give us a statistical power of 80% (with
alpha .05, two-tailed) to detect a within group difference with an
effect size of d⫽1.25 (the estimated effect size). Because we were
interested in identifying a possible interaction between valence and
domain of future thinking (i.e., personal vs. collective), coupled
with assessing the nature of this interaction before and after the
2016 United States Presidential election, we decided a priori that
a larger sample size (N⫽48) than indicated by our power analysis
would be desirable. Written consent was obtained in a manner
approved by the UIC Institutional Review Board. Participants
received course credit for their participation.
Materials and procedure. In order to ascertain cognitions
about the personal and collective future, we adapted the future
fluency task (MacLeod et al., 1997). For the personal future,
participants were asked to list things that they were excited or
worried about in the next week, year, and 5–10 years. Participants
were given 1 min to complete each of the six response prompts
(e.g., list as many things you are excited about in the next year, list
as many things you are worried about in the next year). For the
collective future, participants were asked to list things that they
thought their country (the U.S.) might be excited or worried about
in the next week, year, and 5–10 years. Participants were also
given 1 min to complete each of these six response prompts (e.g.,
list as many things the country is excited about in the next year, list
as many things the country is worried about in the next year). The
personal and collective thinking tasks were presented in a blocked
fashion that was counterbalanced across participants. Within each
block, the six response prompts were presented in random order.
This latter portion of the design deviates from prior work using the
future fluency task, wherein valence (excitement, worry) is typi-
cally counterbalanced and time period (week, year, 5–10 years) is
typically presented in chronological order (MacLeod, 2016, 2017).
Pilot testing revealed that participants had little trouble shifting
between response prompts in a random fashion, and this approach
allowed us to assess whether past observations of a lack of differ-
ence in the number of responses generated across time periods was
possibly the result of the chronological order in which time periods
were presented.
All stimuli were presented using E-Prime 2.0 software on a Dell
desktop computer. Participants entered their responses using a
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
2SHRIKANTH, SZPUNAR, AND SZPUNAR

computer keyboard. Following the 1 min that was allotted for each
response prompt, the computer program automatically advanced to
the next prompt. Finally, after completing the personal and col-
lective future fluency tasks, participants also completed the Iden-
tification With All Humanity Scale (McFarland, Brown, & Webb,
2013), which contained statements regarding one’s affinity toward
their community, country, and the world at large. This scale was
utilized for exploratory purposes, and responses were not analyzed
for the purposes of the present study. The experiment lasted
approximately 25 min. Participants were thanked for their involve-
ment and fully debriefed at the conclusion of the study.
Scoring. As per the scoring procedure described elsewhere
(MacLeod et al., 1997), the total number of responses to each
prompt was summed. Incomplete, incoherent, and repeated re-
sponses were excluded. Additionally, given the focus of the pres-
ent study, we considered the possibility that some collective future
events might be listed in response to personal future prompts (e.g.,
when asked to list things people were worried about in the next
5–10 years, some listed global warming). These responses were
excluded from analysis of personal future prompts in order to
provide a more pure comparison of the personal and collective
future. Nonetheless, for all analyses, inclusion of collective events
that were listed as part of responses to personal future prompts did
not change the observed pattern of results.
Results
All data for this experiment, and additional experiments pre-
sented below, are freely available online through the Open Science
Framework (https://osf.io/nszgu/). We conducted a 2 (Domain:
Personal and Collective) ⫻2 (Valence: Positive and Negative) ⫻
3 (Time: Week, Year, and 5–10 Years) ⫻2 (Order: Personal first
or Collective first) ⫻2 (Election: Preelection or Postelection)
mixed ANOVA, with domain, valence, and time serving as within-
subjects factors and order and election serving as a between-
subjects factors. We begin by focusing our analyses on the pre-
dicted domain-by-valence interaction. The mixed ANOVA
revealed a main effect of domain, F(1, 44) ⫽56.05, p⬍.001,
␩p
2⫽0.56, and a marginally significant effect of valence, F(1,
44) ⫽3.74, p⫽.06, ␩p
2⫽0.078, that were qualified by a
significant domain-by-valence interaction, F(1, 44) ⫽37.07, p⬍
.001, ␩p
2⫽0.457. Pairwise comparisons revealed our predicted
pattern of results. Replicating prior work with healthy adult pop-
ulations (MacLeod, 2016, 2017), participants indicated more
things they were personally excited about (M⫽14.04, SE ⫽0.73)
than they were personally worried about (M⫽10.73, SE ⫽0.60),
t(47) ⫽4.76, p⬍.001, 95% CI ⫽[1.91, 4.71], d⫽0.69.
Conversely, the same participants indicated more things that their
country was worried about (M⫽9.21, SE ⫽0.62) than their
country was excited about (M⫽7.69, SE ⫽0.53), t(47) ⫽3.07,
p⫽.004, 95% CI ⫽[0.52, 2.52], d⫽0.44 (see Figure 1).
The mixed ANOVA further revealed a main effect of time, F(1,
44) ⫽10.42, p⬍.001, ␩p
2⫽0.192. Post hoc comparisons revealed
that participants listed more items 5–10 years into the future (M⫽
3.93, SE ⫽0.20) as compared with 1 week into the future (M⫽
3.29, SE ⫽0.20), t(47) ⫽3.34, p⫽.002, 95% CI ⫽[0.25, 1.02],
d⫽0.48, and 1 year into the future (M⫽3.20, SE ⫽0.17),
t(47) ⫽3.88, p⬍.001, 95% CI ⫽[0.35, 1.11], d⫽0.56. No other
comparisons across time frames were significant. Notably, there
were no effects of order and election, Fs⬍1, and no other
interactions among our factors of interest.
Summary and Discussion
As predicted, we observed a dissociation between personal and
collective future thinking such that people reported more positive
expectations relative to worries about their personal futures and
more worries than positive expectations for the future of their
country. Notably, this pattern of data was not related to the
outcome of a recent election that likely did not favor the prefer-
ences of our sample. In addition, we found that people reported
more positive expectations and worries about the distant as com-
pared with the near future, personal or collective, when order of
time periods was randomized. In order to further establish the
reliability and generalizability of our findings, we carried out a
second experiment that drew participants from the general popu-
lation of the U.S. using Amazon Mechanical Turk, a popular
crowdsourcing platform (Buhrmester, Kwang, & Gosling, 2011).
Experiment 2
Method
Participants. Two-hundred and fifty individuals from across
the U.S. were recruited through Amazon Mechanical Turk (fe-
male ⫽168; M
Age
⫽41.8 years). This large sample size, decided
upon prior to carrying out the experiment, was collected in an
effort to gather enough data to assess whether any observed effects
held across demographic characteristics of our sample (see below).
Seven participants were excluded because they did not properly
follow instructions. Written consent was obtained in a manner
approved by the UIC Institutional Review Board. Participants
received $0.50 for their participation.
Materials and procedure. Materials and procedures were
identical to Experiment 1, with the following exceptions. The
study was programmed using Qualtrics survey software and run
through Amazon Mechanical Turk. Participants completed the
study remotely. The personal and collective future thinking tasks
were also identical to Experiment 1, except that participants were
Figure 1. Mean number of unique things undergraduates indicated they
were excited (positive) or worried about (negative) in relation to their
personal future or the collective future of their country. Error bars represent
95% confidence intervals.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
3
DISSOCIATING THE FUTURE

required to indicate which domain of future thinking they were
about to engage in following the accompanying instructions for the
task. This attention check ensured that participants were following
instructions. Following the completion of the personal and collec-
tive future thinking tasks, participants completed the Identification
With All Humanity Scale, and also a series of questions related to
their news consumption behavior: (a) how many minutes a day
they spent consuming news from various media outlets (e.g., TV,
Internet); (b) what sources they typically relied on (e.g., New York
Times,Fox News); (c) their level of trust in the media; and (d) any
miscellaneous information they wanted to report about their con-
sumption or views on the media. These questionnaires were col-
lected for exploratory purposes and were not analyzed for the
purposes of this study. Finally, participants completed a demo-
graphic survey, which included questions about age, gender, and
political affiliation (variables that were analyzed for the purposes
of this study) and ethnicity, years lived in the U.S., and state of
residency (variables that were not analyzed for the purposes of this
study).
Results
We conducted a 2 (Domain: Personal and Collective) ⫻2
(Valence: Positive and Negative) ⫻3 (Time: Week, Year, and
5–10 Years) ⫻2 (Order: Personal first or Collective first) mixed
ANOVA, with domain, valence, and time serving as within-
subjects factors and order serving as a between-subjects factor. As
above, we begin by focusing our analyses on the predicted domain-
by-valence interaction. The results of the mixed ANOVA revealed
main effects of domain, F(1, 241) ⫽37.66, p⬍.001, ␩p
2⫽0.135,
and valence, F(1, 241) ⫽13.69, p⬍.001, ␩p
2⫽0.054, that were
qualified by a significant domain-by-valence interaction, F(1,
241) ⫽261.41, p⬍.001, ␩p
2⫽0.520. Pairwise comparisons
revealed our predicted pattern of results. Replicating our findings
from Experiment 1, participants indicated more things they were
personally excited about (M⫽10.24, SE ⫽0.31) than they were
personally worried about (M⫽8.06, SE ⫽0.28), t(242) ⫽8.83,
p⬍.001, 95% CI ⫽[1.69, 2.66], d⫽0.57. Conversely, the same
participants indicated more things that their country was worried
about (M⫽9.32, SE ⫽0.30) than their country was excited about
(M⫽5.85, SE ⫽0.24), t(242) ⫽14.08, p⬍.001, 95% CI ⫽
[2.98, 3.95], d⫽0.90.
Further replicating the results of Experiment 1, the mixed
ANOVA revealed a main effect of time, F(1, 241) ⫽58.37, p⬍
.001, ␩p
2⫽0.195. Post hoc comparisons revealed that participants
listed more items 5–10 years into the future (M⫽3.20, SE ⫽0.09)
as compared with 1 week into the future (M⫽2.36, SE ⫽0.08),
t(243) ⫽10.13, p⬍.001, 95% CI ⫽[0.68, 1.00], d⫽0.65, and
1 year into the future (M⫽2.81, SE ⫽0.09), t(243) ⫽5.11, p⬍
.001, 95% CI ⫽[0.24, 0.55], d⫽0.33. Participants also listed
more items 1 year into the future than 1 week into the future,
t(243) ⫽6.11, p⬍.001, 95% CI ⫽[0.30, 0.59], d⫽0.39.
Notably, these effects were further qualified by a three-way
interaction between domain, valence, and time. While the interac-
tion between domain and valence was evident at all three time
frames, smallest F(1, 242) ⫽73.58, p⬍.001, ␩p
2⫽0.233, the
three-way interaction was driven by the fact that participants were
more negatively biased about the collective future, that is, they
indicated more things that elicited worry than excitement, 1 year
from the present (M
diff
⫽1.60) as compared with 1 week from the
present (M
diff
⫽0.74), F(1, 242) ⫽19.30, p⬍.001, ␩p
2⫽0.074.
In contrast to the results obtained in Experiment 1, we observed
an effect of order, F(1, 241) ⫽8.35, p⫽.004, ␩p
2⫽0.034, such
that participants reported more items when asked about the per-
sonal future first (M⫽36.1, SE ⫽1.30) as compared with being
asked about the collective future first (M⫽31.0, SE ⫽1.16). This
effect was qualified by an interaction between domain and order,
F(1, 241) ⫽53.41, p⬍.001, ␩p
2⫽0.181. This interaction was
driven by the fact that participants listed more collective items
when they were asked about the future of the country before (M⫽
18.3, SE ⫽0.69) as compared with after (M⫽12.2, SE ⫽0.58)
being asked about their personal future. Importantly, the interac-
tion between domain and valence, which was of particular interest
to this study, was robustly present regardless of task order, small-
est F(1, 116) ⫽127.66, p⬍.001, ␩p
2⫽0.524.
Finally, in order to further assess the generalizability of the domain-
by-valence interaction, we tested for the presence of the interaction as
a function of key demographic classifications. Specifically, we exam-
ined whether the interaction was present irrespective of the age,
gender, and political affiliation of our participants. With regard to age,
we found a significant domain-by-valence interaction for participants
aged 20 –29 years (N⫽56), 30 –39 years (N⫽69), 40 – 49 years
(N⫽42), 50–59 years (N⫽43), and 60 – 69 years (N⫽27), smallest
F(1, 26) ⫽23.72, p⬍.001, ␩p
2⫽0.477 (see Figure 2). With regard
to gender, we found a significant domain-by-valence interaction for
both women, F(1, 167) ⫽164.40, p⬍.001, ␩p
2⫽0.496, and men,
F(1, 74) ⫽109.20, p⬍.001, ␩p
2⫽0.596 (see Figure 3). Lastly, with
regard to political affiliation, we found a significant domain-by-
valence interaction for Democrats (N⫽89), Republicans (N⫽51),
and Independents (N⫽61), smallest F(1, 50) ⫽45.90, p⬍.001,
␩p
2⫽0.479 (see Figure 4).
Summary and Discussion
The results of Experiments 1 and 2 suggest that U.S.-based
participants are generally positive about their personal future and
negative about the future of their country. One possible concern
regarding the future fluency task is the extent to which participants
actually prospect about the distant future. That is, the most distant
temporal cue asks about events that might take place 5–10 years
Figure 2. Mean number of unique things members of the general public
indicated they were excited (positive) or worried about (negative) in
relation to their personal future or the collective future of their country as
a function of age. Error bars represent 95% confidence intervals.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
4SHRIKANTH, SZPUNAR, AND SZPUNAR

from the present, and it is possible that participants do not project
beyond 5 years. In order to address this concern, we set out to
assess whether the observed pattern of data emerges when partic-
ipants are asked to think about the more distant (40 years) future.
Experiment 3
Method
Participants. Fifty 25- to 30-year-olds from across the U.S.
were recruited through Amazon Mechanical Turk (female ⫽31;
M
Age
⫽27.9 years). We selected this age range because it allowed
us to compare the distant personal and collective future. Specifi-
cally, given that the average life expectancy in the U.S. is approx-
imately 78.8 years (Arias, Heron, & Xu, 2017), we reasoned that
25- to 30-year-olds would still expect to be alive 40 years from
now, allowing for meaningful comparisons between the personal
and collective future. We decided to collect data from 50 partici-
pants so as to approximate the average number of participants in
our analyses of age groups in Experiment 2 (M⫽47.4). Written
consent was obtained in a manner approved by the UIC Institu-
tional Review Board. Participants received $0.25 for their partic-
ipation.
Materials and procedure. Materials and procedures were
similar to Experiment 2, with the following exception. Participants
were only asked to list what they or their country were excited or
worried about 40 years from now. As before, information related
to news consumption was collected for exploratory purposes and
was not analyzed for the purposes of this study. Demographic
classifications were also collected, but not analyzed given the
reduced sample size that focused on one specific age group.
Results
We conducted a 2 (Domain: Personal and Collective) ⫻2 (Va-
lence: Positive and Negative) ⫻2 (Order: Personal first or Collective
first) mixed ANOVA, with domain and valence serving as within-
subjects factors and order serving as a between subjects factor. We
begin by focusing our analyses on the predicted domain-by-valence
interaction. The results of the mixed ANOVA revealed no effect of
domain, F(1, 48) ⫽1.12, p⫽.295, and a main effect of valence, F(1,
48) ⫽8.16, p⫽.006, ␩p
2⫽0.145, that were qualified by a significant
domain-by-valence interaction, F(1, 48) ⫽10.50, p⫽.002, ␩p
2⫽
0.180. Interestingly, pairwise comparisons revealed that participants
listed slightly more things they were personally excited about in 40
years (M⫽3.90, SE ⫽0.30) than they were worried about in 40 years
(M⫽3.56, SE ⫽0.33), although this difference was not significant,
t(49) ⫽1.05, p⫽.297, 95% CI ⫽[0.31, 0.99], d⫽0.15. We consider
this lack of significance below. Importantly, the same participants
indicated more things that their country was worried about in 40 years
(M⫽4.68, SE ⫽0.33) than their country was excited about in 40
years (M⫽3.20, SE ⫽0.23), t(49) ⫽3.88, p⬍.001, 95% CI ⫽
[0.71, 2.25], d⫽0.55 (see Figure 5). There was no effect of order,
F⬍1.
Summary and Discussion
The findings of Experiment 3 were notable in two regards. First,
the personal positivity bias was largely attenuated when partici-
Figure 3. Mean number of unique things members of the general public
indicated they were excited (positive) or worried about (negative) in
relation to their personal future or the collective future of their country as
a function of gender. Error bars represent 95% confidence intervals.
Figure 4. Mean number of unique things members of the general public
indicated they were excited (positive) or worried about (negative) in
relation to their personal future or the collective future of their country as
a function of political affiliation. Error bars represent 95% confidence
intervals.
Figure 5. Mean number of things 25- to 30-year-old members of the
general public indicated they were excited (positive) or worried about
(negative) in relation to their distant (40 year) personal future or the distant
(40 year) collective future of their country. Error bars represent 95%
confidence intervals.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
5
DISSOCIATING THE FUTURE

pants prospected about their distant future. We believe that this
aspect of the results can be accounted for by prior findings regard-
ing cultural life scripts (Rubin & Berntsen, 2003). Specifically,
considerable work on cultural life scripts indicates that while
people tend to anticipate positive milestones over the course of
their lives (e.g., getting married, buying a home), the end of life is
typically characterized by more negative expectations (e.g., death
of self and/or loved ones). Indeed, a content analysis of our
participant responses indicated that 32% of participants in the
study explicitly listed worries associated with either their own or a
loved one’s death (in addition, 66% listed worries associated with
declining health). Notably, we observed that the negativity bias
that U.S.-based participants hold about the future of their country
extended into the distant (40 year) future, suggesting that this
negativity bias is not limited in scope (at least not 40 years into the
future).
Having established the robustness of the personal positivity bias
and to a greater extent the collective negativity bias, we set out to
ascertain the extent to which this dissociation between the personal
and collective future extends to participants outside of the U.S.
Experiment 4
Method
Participants. Seventy individuals from across Canada were
recruited through Amazon Mechanical Turk (female ⫽37;
M
Age
⫽34.2 years). One participant was excluded because they
did not properly follow instructions. We note that our initial goal
had been to collect data from 250 Canadians (in order to have
enough participants to examine similarities/differences across de-
mographic categorizations; as with Experiment 2). However, after
sampling data from Amazon’s Mechanical Turk for a period of 2
months (August 1, 2017–October 2, 2017), we were only able to
accrue responses from 70 participants. Nonetheless, a power anal-
ysis using G
ⴱ
Power 3 of our online U.S. sample (Experiment 2)
revealed that only eight participants would be need to demonstrate
the domain-by-valence interaction of interest. Hence, we report
data from the sample we were able to attain, but note that future
studies of Canadian participants using Amazon’s Mechanical Turk
may be limited depending on the sample size required. Written
consent was obtained in a manner approved by the UIC Institu-
tional Review Board. Participants received $0.50 for their partic-
ipation.
Materials and procedures. Materials and procedures were
identical to Experiment 2. As before, information related to news
consumption was collected for exploratory purposes and not ana-
lyzed for the purposes of this study. In addition, demographic
classifications were also collected, but not analyzed given the
limitations noted in collecting a large sample size.
Results
We conducted a 2 (Domain: Personal and Collective) ⫻2
(Valence: Positive and Negative) ⫻3 (Time: Week, Year, and
5–10 Years) ⫻2 (Order: Personal first or Collective first) mixed
ANOVA, with domain, valence, and time serving as within-
subjects factors and order serving as a between-subjects factor. We
begin by focusing our analyses on the predicted domain-by-
valence interaction. The results of the mixed ANOVA revealed a
main effect of domain, F(1, 67) ⫽23.83, p⬍.001, ␩p
2⫽0.195,
and no effect of valence, F⬍1, that were qualified by a significant
domain-by-valence interaction, F(1, 67) ⫽29.85, p⬍.001, ␩p
2⫽
0.308. Pairwise comparisons supported our predicted pattern of
results. Replicating our findings from Experiments 1 and 2 (U.S.
participants), Canadian participants reported more things they
were personally excited about (M⫽7.81, SE ⫽0.50) than they
were personally worried about (M⫽6.49, SE ⫽0.50), t(68) ⫽
2.97, p⫽.004, 95% CI ⫽[0.43, 2.21], d⫽0.36. Conversely, the
same Canadian participants reported more things that their country
was worried about (M⫽6.28, SE ⫽0.42) than their country was
excited about (M⫽4.42, SE ⫽0.30), t(68) ⫽5.22, p⬍.001, 95%
CI ⫽[1.15, 2.56], d⫽0.63 (see Figure 6).
Further replicating the results of Experiments 1 and 2, the mixed
ANOVA revealed a main effect of time, F(1, 67) ⫽18.09, p⬍
.001, ␩p
2⫽0.213. Post hoc comparisons revealed that participants
listed more items 5–10 years into the future (M⫽2.39, SE ⫽0.14)
and 1 year into the future (M⫽2.20, SE ⫽0.14) as compared with
1 week into the future (M⫽1.66, SE ⫽0.11), t(68) ⫽6.29, p⬍
.001, 95% CI ⫽[0.50, 0.97], d⫽0.76 and t(68) ⫽4.63, p⬍.001,
95% CI ⫽[0.31, 0.77], d⫽0.56, respectively. No other compar-
isons across time frames were significant. Finally, there was no
effect of order, F⬍1.
Summary and Discussion
The results of Experiment 4 indicate that the personal positivity
bias and collective negativity bias observed in U.S. participants
extended to a Canadian sample. Next, we set out to assess whether
expectations for the collective future of groups are always nega-
tive. Brewer and Gardiner (1996) proposed an important distinc-
tion between groups with whom people can have tangible experi-
ences (i.e., relational groups) and groups with whom people cannot
have tangible experiences and with whom they hold a more im-
personal sense of membership (i.e., collective groups). A common
example of a relational group is one’s family whereas a common
collective/impersonal group is one’s country or nation. That is, one
can have tangible experiences with their family but not with their
country. Based on this distinction, we hypothesized that because
Figure 6. Mean number of unique things members of the general Cana-
dian public indicated they were excited (positive) or worried about (neg-
ative) in relation to their personal future or the collective future of their
country. Error bars represent 95% confidence intervals.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
6SHRIKANTH, SZPUNAR, AND SZPUNAR

the tangible experiences that one has with relational groups (e.g.,
family) can be incorporated into their autobiographical past, which
is known to be generally positive (Walker & Skowronski, 2009)
and a basis for simulating the future (Schacter & Addis, 2007),
expectations for relational groups should be positively biased. On
the other hand, we hypothesized that expectations for collective
groups other than country, such as one’s state of residence, should
be based on information that is derived from external sources (e.g.,
history texts and news media), and so exhibit a negativity bias (cf.
Liu et al., 2009). We tested these specific predictions below.
Experiment 5
Method
Participants. Twenty-three undergraduate students were re-
cruited through the University of Illinois at Chicago Psychology
Subject Pool (female ⫽16; M
Age
⫽19.1 years). The sample size
was chosen to reflect the size of our Experiment 1 studies (for
which there was no effect of completing the experiment before or
after the 2016 United States Presidential election). Written consent
was obtained in a manner approved by the UIC Institutional
Review Board. Participants received course credit for their partic-
ipation.
Materials and procedure. The procedure was identical to
that of Experiment 1, with the following exception. Participants
answered questions about the future of their family (e.g., list as
many things your family is excited about in the next year, list as
many things your family is worried about in the next year) and the
future of their state of residence (e.g., list as many things the state
is excited about in the next year, list as many things the state is
worried about in the next year).
Results
In line with our predictions, we found that a positivity bias
emerged for cognitions about the future of an individual’s family,
such that participants indicated more things that their family was
excited about (M⫽9.26, SE ⫽0.76) than their family was worried
about (M⫽7.22, SE ⫽0.73), t(22) ⫽2.83, p⫽.01, 95% CI ⫽
[0.55, 3.54], d⫽0.59. Conversely, the same individuals held a
negativity bias for the future of their state of residence, such that
they indicated more things that their state was worried about (M⫽
7.43, SE ⫽0.72) than their state was excited about (M⫽4.57,
SE ⫽0.66), t(22) ⫽4.03, p⬍.001, 95% CI ⫽[1.39, 4.35], d⫽
0.84.
Summary and Discussion
Across five experiments, we have demonstrated that people tend
to hold a positivity bias about their personal future and a negativity
bias about the future of groups, the latter of which depends on the
nature of one’s relation to the group. Moreover, this valence-based
dissociation extends into the distant future, and emerges in both
U.S. and Canadian samples. Having established the reliability and
generalizability of this domain-by-valence interaction, we set out
to assess the extent to which this phenomenon may have conse-
quences on information processing. That is, if people tend to think
about their personal future in a positive light and the collective
future in a negative light, will the manner in which they think
about the world around them change depending on the perspective
(i.e., personal or collective) they adopt? To test this question, we
conducted a final set of experiments in which participants were
asked to adopt a personal and collective perspective about topics
that can have positive and negative consequences. In Experiment
6a, participants thought about the positive and negative conse-
quences of social media. In Experiment 6b, participants thought
about the positive and negative consequences of artificial intelli-
gence. We hypothesized that participants would think of more
positive than negative consequences, in both experiments, when
adopting a personal perspective, but that they would list more
negative than positive consequences when adopting a collective
perspective.
Experiments 6a and 6b
Method
Participants. Two-hundred and fifty individuals from across
the U.S. were recruited through Amazon Mechanical Turk (fe-
male ⫽160; M
Age
⫽38.9 years). One-hundred and twenty-five
individuals participated in Experiment 6a (consequences of social
media) and 125 participated in Experiment 6b (consequences of
artificial intelligence). Given that there exist no prior studies
assessing the influence of adopting a personal versus collective
perspective on information processing, we reasoned a priori that
the relatively large sample sizes would be appropriate. Seven
participants were excluded from Experiment 6a and six partici-
pants were excluded from Experiment 6b because they did not
properly follow instructions. Written consent was obtained in a
manner approved by the UIC Institutional Review Board. Partic-
ipants received $0.25 for their participation.
Materials and procedure. We asked participants to think
about social media and artificial intelligence because both topics
are commonly discussed in terms of their benefits and drawbacks.
Moreover, we selected two topics in order to ensure that any
significant pattern of results that was obtained was not limited to
one context.
The procedure was based on the future fluency task used in our
previous experiments. Specifically, participants were asked to list
as many positive and negative consequences of social media or
artificial intelligence for themselves and their country, resulting in
a total set of four cues: positive-self, negative-self, positive-
country, and negative-country. The personal and collective (i.e.,
country) cues were counterbalanced across participants such that
half of the participants completed the personal fluency task first,
whereas half of the participants completed the collective fluency
task first. Positive and negative prompts were presented in a
random order within the context of each fluency task. Participants
were given 2 min to provide as many consequences in response to
each cue. As before, information related to news consumption was
collected for exploratory purposes, and was not analyzed for the
purposes of this study. Demographic classifications (e.g., age,
gender, and political affiliation) were also collected. Given that our
earlier analyses did not reveal any meaningful differences across
demographic classifications, these data were also not analyzed for
the purposes of this study.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
7
DISSOCIATING THE FUTURE

Results
We conducted a 2 (Domain: Personal and Collective) ⫻2
(Valence: Positive and Negative) ⫻2 (Order: Personal first or
Collective first) ⫻2 (Experiment: Social Media or Artificial
Intelligence) mixed ANOVA, with domain and valence serving as
within-subjects factors and order and experiment serving as
between-subjects factors. We begin by focusing our analyses on
the predicted domain-by-valence interaction, collapsing across Ex-
periments 6a and 6b. The results of the mixed ANOVA revealed
no main effect of domain, F⬍1, and a main effect of valence, F(1,
233) ⫽10.29, p⫽.002, ␩p
2⫽0.042, that were qualified by a
significant domain-by-valence interaction, F(1, 233) ⫽71.22, p⬍
.001, ␩p
2⫽0.234. Pairwise comparisons supported our predicted
pattern of results. Participants listed more positive (M⫽3.19,
SE ⫽0.12) than negative (M⫽2.42, SE ⫽0.09) consequences
when adopting a personal perspective, t(236) ⫽7.09, p⬍.001,
95% CI ⫽[0.55, 0.98], d⫽0.46. Conversely, the same partici-
pants listed more negative (M⫽2.88, SE ⫽0.10) than positive
(M⫽2.62, SE ⫽0.09) consequences when adopting a collective
perspective, t(236) ⫽2.86, p⫽.005, 95% CI ⫽[0.08, 0.45], d⫽
0.19.
Notably, these effects were qualified by a three-way interaction
between domain, valence, and experiment, F(1, 233) ⫽6.13, p⫽
.014, ␩p
2⫽0.026. While the interaction between domain and
valence was evident in both Experiments 6a and 6b, smallest F(1,
118) ⫽23.50, p⬍.001, ␩p
2⫽0.166, the three-way interaction was
driven by the fact that participants listed more personal positive
consequences of social media (M⫽3.86, SE ⫽0.19) than personal
positive consequences of artificial intelligence (M⫽2.51, SE ⫽
0.14), t(235) ⫽5.80, p⬍.001, 95% CI ⫽[0.89, 1.81], d⫽0.75,
and more personal negative consequences of social media (M⫽
2.83, SE ⫽0.15) than personal negative consequences of artificial
intelligence (M⫽2.01, SE ⫽0.11), t(235) ⫽4.54, p⬍.001, 95%
CI ⫽[0.47, 1.18], d⫽0.59, while not exhibiting any difference in
the number of collective positive or negative consequences across
experiments, largest t(235) ⫽1.32, p⫽.189 (see Figure 7).
Summary and Discussion
In Experiments 6a and 6b, we set out to assess whether people
may think differently about the consequences of topics that can
have both positive and negative implications as a function of
whether they adopt a personal or collective perspective. Indeed, we
found that people were more likely to think of positive conse-
quences when adopting a personal perspective, and slightly but
reliably more likely to think of negative consequences when adopt-
ing a collective perspective. We interpret these findings as dem-
onstrating that the personal positivity and collective negativity
biases identified in Experiments 1 through 5 may have important
implications for how people think about the world.
General Discussion
People are generally optimistic about their future (Newby-Clark
& Ross, 2003; Weinstein, 1980). Here we show that this optimism
bias persists in the face of a negative outlook for the country in
which people reside. To our knowledge, our findings represent the
first behavioral demonstration of a dissociation between personal
and collective future thinking in healthy cognition. This domain-
by-valence interaction appears to be robust. We observed the
interaction in college undergraduates; people drawn from the gen-
eral public who varied in terms of age, gender, and political
affiliation; in the distant (40 years) future; and in a non-U.S.
(Canadian) sample. Importantly, we further highlighted that the
collective negativity bias tends to only emerge under conditions
wherein people cannot have tangible experiences with the group in
question (e.g., state of residence, country). Finally, we showed that
these personal positivity and collective negativity biases have
implications for how people think about the world. Specifically,
people were more likely to think of positive than negative conse-
quences of social media and artificial intelligence when adopting a
personal perspective, whereas they were more likely to think of
negative than positive consequences when adopting a collective
perspective. Next, we provided additional discussion regarding the
possible basis of the observed domain-by-valence interaction be-
tween personal and collective cognition, and concluded by con-
sidering further the practical implications of this phenomenon.
We interpret our findings as reflecting the impact of past expe-
rience on expectations for the future. People tend to experience and
subsequently remember more positive than negative experiences
from their personal past (Kahneman et al., 2009; see also MacLeod
& Byrne, 1996; MacLeod et al., 1997), and this positive personal
past likely biases people to expect that their future will also be
positive. Conversely, people tend to be exposed to negatively
skewed media coverage of national events (Soroka & McAdams,
2015), and also remember more negative than positive national
events (Liu et al., 2005, 2009). Accordingly, this negative collec-
tive past likely biases people to expect that the future of their
country will also be negative. However, precisely how positive
personal and negative collective past experiences impact expecta-
tions of the future remains to be worked out.
One possibility is that relatively accessible past instances related
to a positive personal past or a negative collective past serve as a
basis for the construction of possible future occurrences (cf. Tver-
sky & Kahneman, 1973). Indeed, considerable evidence suggests
that this is the case in the context of expectations related to the
personal future (e.g., Gilbert, & Wilson, 2007; Pillemer, 2003).
Figure 7. Mean number of positive and negative consequences of social
media (Experiment 6a) and artificial intelligence (Experiment 6b) listed by
members of the general public in relation to their personal future or the
collective future of the country. Error bars represent 95% confidence
intervals.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
8SHRIKANTH, SZPUNAR, AND SZPUNAR

With regard to the collective future, participant responses in our
studies certainly indicated that recently acquired information was
being used to guide expectations as to what the country should be
worried about in the future (i.e., topical news media items related
to government leadership, immigration, and so on). Nonetheless,
studies are needed to more directly demonstrate the role of acces-
sible past instances in guiding expectations of the collective future.
One possibly useful approach would be to require participants to
complete fluency tasks that probe accessible representations re-
garding the collective past and future (cf. MacLeod et al., 1997),
and to compare and contrast the content that is evoked in the
context of these tasks.
It will also be important for future work to hash out the repre-
sentational nature of past instances that may serve as a basis for
generating expectations of the future. Specifically, are these in-
stances based on recollections of specific experiences (i.e., epi-
sodic) or general knowledge that has been extracted from past
experience (i.e., semantic)? Prior work from neuropsychology
(Andelman et al., 2010; Klein et al., 2002; Race et al., 2013) and
neuroimaging (Abraham, Schubotz, & von Cramon, 2008) has
spurred conclusions that episodic memory may enable thoughts
about the personal future, whereas as semantic memory may play
a more central role in enabling thoughts about the collective future
(e.g., Klein et al., 2002). Although our data cannot speak directly
to this system-based distinction, it is certainly possible that mem-
ories of personal experiences— our proposed basis of expectations
for the personal future—are more likely to be sourced from epi-
sodic than semantic knowledge as compared with memories for
information learned through the media— our proposed basis of
expectations for the collective future.
1
More work is needed to
directly test this assertion. For instance, additional studies employ-
ing methods that require participants to more fully elaborate on
what they or their country are excited and worried about should
help to further disentangle the relative contributions of episodic
and semantic memory to personal and collective cognitions about
the future. Another advantage of such an approach would be to
further enhance the ability of the experimenter to determine the
importance and perhaps validity of the responses that participants
offer in relation to the personal and collective future. For instance,
when a participant states that they are excited about traveling or
that the country is worried about war with another country, it
would help to have more detailed descriptions, and perhaps rat-
ings, to aid in gauging how likely it is that participants think that
these events are to occur, and—if likely— how excited or worried
they or the country might be.
In addition to the possible role that past instances may play in
guiding the generation of expectations for the future, we propose
that extensive experience thinking about the personal past as
positive and the collective past as negative may give rise to more
general schema-based biases that guide expectations of the future.
The results of Experiments 6a and 6b lend some support to this
conjecture. Participants in those experiments were presented with
topics that are commonly associated with benefits and drawbacks
(i.e., social media and artificial intelligence). Notably, participants
listed more positive than negative consequences when adopting a
personal perspective, but more negative than positive conse-
quences when adopting a collective perspective, suggesting that
distinct schemas were being adopted depending on context (i.e.,
personal vs. collective). Clearly, more work is needed to distin-
guish between the possible influence of accessible instances in
memory and more general schema-based processes as candidate
mechanisms that may give rise to a positive view of the personal
future and a negative view of the collective future. For instance,
one possibly interesting avenue of research in relation to collective
cognition might be to teach participants about the history and
current events associated with a country that they are relatively
unfamiliar with, and to manipulate the extent to which the histor-
ical and topical information is characterized as positive and/or
negative. To the extent that schema-based processing impacts
cognitions about the future of countries, people should expect a
negative collective future regardless of the amount of positive
historical and topical content they are exposed to in the context of
such an experiment.
The influence of episodic and/or semantic instances in memory
and more schema-based processes in guiding positive personal and
negative collective biases are likely not mutually exclusive. Cog-
nitions about the future, whether personal or collective, are likely
to be based on several interacting factors (i.e., the future is mul-
tiply determined). Indeed, some of our data speak directly to this
point. For instance, in Experiment 3 we found that the personal
positivity bias was largely attenuated when people projected into
their distant future. This pattern of results likely reflects the fact
that people hold more general expectations about milestones that
they may achieve over the course of their lifetime (Rubin &
Berntsen, 2003), and that later life tends to be characterized by
more negative general milestones (e.g., death) than earlier life.
2
While the collective future appeared to be more stable over time
(i.e., the negativity bias was persistent), it is certainly possible that
expectations of the near and distant future of groups differ in ways
that may require more detail-oriented response criteria to illumi-
nate.
Finally, we conclude our discussion by highlighting that it will
be important to assess the generalizability of our findings in
additional cultural contexts (Markus & Kitayama, 1991). For in-
stance, Liu et al. (2009) found that people from some countries
(e.g., China and Hungary) nominated more positive than negative
events as best representing the history of their country. One in-
triguing prediction that can be made on the basis of our findings is
that people in China and Hungary may, by virtue of having
positive memories of their nation’s past, also expect the future of
their respective countries to be positive. Related to this point, it
will be important for studies of this sort to more clearly pinpoint
the source of collective biases (whether negative or positive).
While we hypothesize that media exposure likely plays an impor-
tant role, more work is needed to determine the manner in which
media exposure gives rise to collective cognitions. One possibly
fruitful approach may be the adoption of longitudinal studies that
1
Of course, information learned from the media is done so in the context
of personal experiences. Our point is that people may be less likely to
remember the context in which they learn this information as compared
with experiences that are relevant to the personal future (i.e., those related
to considerations about one’s social relationships, family, education/career,
and so on).
2
It is noteworthy to point out that although 25- to 30-year-olds did not
exhibit a positivity bias about their distant (40 year) future (Experiment 3),
older adults in that age range (i.e., 60-year-olds) did nonetheless exhibit a
strong positivity bias extending into the following week, year, and 5–10
years (Experiment 2).
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
9
DISSOCIATING THE FUTURE

track changes in media coverage and collective cognition over
time. In addition to media coverage, we suspect that other sources
of collective knowledge, such as history texts (e.g., Liu et al.,
2005, 2009) and group interaction (e.g., Choi, Kensinger, & Ra-
jaram, 2017), among others, will also be important to consider.
Context of the Research
The work presented here represents a novel extension of our
contributions to the study of the cognitive and neural mechanisms
that support the ability to simulate or imagine the future (Schacter,
Benoit, & Szpunar, 2017; Szpunar, Spreng, & Schacter, 2014).
Specifically, whereas much of the extant research on future-
oriented cognition has focused on how people simulate the per-
sonal future, relatively little attention has been devoted to how
people think about the future of groups. Our findings provide
evidence for a valence-based dissociation between personal and
collective future thinking that has implications for information
processing. Notwithstanding, we acknowledge that the present
findings may pose more questions than they necessarily answer.
For instance, in addition to the need to more systematically assess
the generality of this phenomenon across cultural contexts, it will
also be important to assess what variables might moderate the
personal positivity and collective negativity biases within cultural
contexts. Do constructs such as self-enhancement, group identifi-
cation, and perceived control impact the pattern of results? Each of
these future directions highlights the current dearth of research on
collective future thinking and its relation to personal future think-
ing, and also its potential as a novel direction in psychological
research.
References
Abraham, A., Schubotz, R. I., & von Cramon, D. Y. (2008). Thinking
about the future versus the past in personal and non-personal contexts.
Brain Research, 1233, 106 –119. http://dx.doi.org/10.1016/j.brainres
.2008.07.084
Andelman, F., Hoofien, D., Goldberg, I., Aizenstein, O., & Neufeld, M. Y.
(2010). Bilateral hippocampal lesion and a selective impairment of the
ability for mental time travel. Neurocase, 16, 426 – 435. http://dx.doi
.org/10.1080/13554791003623318
Arias, E., Heron, M., & Xu, J. (2017). United States life tables, 2013:
National Vital Statistics Report. Hyattsville, MD: National Center for
Health Statistics.
Brewer, M. B., & Gardiner, W. (1996). Who is this “We”? Levels of
collective identity and self representations. Journal of Personality and
Social Psychology, 71, 83–93. http://dx.doi.org/10.1037/0022-3514.71
.1.83
Buhrmester, M., Kwang, T., & Gosling, S. (2011). Amazon’s Mechanical
Turk: A new source of inexpensive, yet high-quality, data? Perspectives
on Psychological Science, 6, 3–5. http://dx.doi.org/10.1177/17456
91610393980
Choi, H.-Y., Kensinger, E. A., & Rajaram, S. (2017). Mnemonic transmis-
sion, social contagion, and emergence of collective memory: Infleunce
of emotional valence, group structure, and information distribution.
Journal of Experimental Psychology: General, 146, 1247–1265. http://
dx.doi.org/10.1037/xge0000327
Gilbert, D. T., & Wilson, T. D. (2007). Prospection: Experiencing the
future. Science, 317, 1351–1354. http://dx.doi.org/10.1126/science
.1144161
Hassabis, D., Kumaran, D., Vann, S. D., & Maguire, E. A. (2007). Patients
with hippocampal amnesia cannot imagine new experiences. Proceed-
ings of the National Academy of Sciences of the United States of
America, 104, 1726 –1731. http://dx.doi.org/10.1073/pnas.0610561104
Kahneman, D., Krueger, A. B., Schkade, D. A., Schwarz, N., & Stone,
A. A. (2009). A survey method for characterizing life experience: The
day reconstruction method. Science, 306, 1776 –1780. http://dx.doi.org/
10.1126/science.1103572
Klein, S. B., Loftus, J., & Kihlstrom, J. (2002). Memory and temporal
experience: The effects of episodic memory loss on an amnesic patient’s
ability to remember the past and imagine the future. Social Cognition,
20, 353–379. http://dx.doi.org/10.1521/soco.20.5.353.21125
Leip, D. (2016). Atlas of U.S. Presidential elections. Retrieved from
https://uselectionatlas.org/RESULTS/
Liu, J. H., Goldstein-Hawes, R., Hilton, D. J., Huang, L. L., Gastardo-
Conaco, C., Dresler-Hawke, E.,...Hidaka, Y. (2005). Social represen-
tations of events and people in world history across twelve cultures.
Journal of Cross-Cultural Psychology, 36, 171–191. http://dx.doi.org/
10.1177/0022022104272900
Liu, J. H., Paez, D., Slawuta, P., Cabecinhas, R., Techio, E., Kokdemir,
D.,...Zlobina, A. (2009). Representing world history in the 21st
century: The impact of 9/11, the Iraq war, and the nation-state on
dynamics of collective remembering. Journal of Cross-Cultural Psy-
chology, 40, 667– 692. http://dx.doi.org/10.1177/0022022109335557
MacLeod, A. K. (2016). Prospection, well-being and memory. Memory
Studies, 9, 266 –274. http://dx.doi.org/10.1177/1750698016645233
MacLeod, A. K. (2017). Prospection: Well-being and mental health. Ox-
ford, UK: Oxford University Press.
MacLeod, A. K., & Byrne, A. (1996). Anxiety, depression, and the antic-
ipation of future positive and negative experiences. Journal of Abnormal
Psychology, 105, 286 –289. http://dx.doi.org/10.1037/0021-843X.105.2
.286
MacLeod, A. K., Rose, G. S., & Williams, J. M. G. (1993). Components of
hopelessness about the future in parasuicide. Cognitive Therapy and
Research, 17, 441– 455. http://dx.doi.org/10.1007/BF01173056
MacLeod, A. K., Tata, P., Kentish, J., & Jacobsen, H. (1997). Retrospec-
tive and prospective cognitions in anxiety and depression. Cognition and
Emotion, 11, 467– 479. http://dx.doi.org/10.1080/026999397379881
Markus, H. R., & Kitayama, S. (1991). Culture and self: Implications for
cognition, emotion, and motivation. Psychological Review, 98, 224 –
253. http://dx.doi.org/10.1037/0033-295X.98.2.224
McCombs, M. (2006). A look at agenda-setting: Past, present and future.
Journalism Studies, 6, 543–557. http://dx.doi.org/10.1080/14616700
500250438
McFarland, S., Brown, D., & Webb, M. (2013). Identification with all
humanity as a moral concept and psychological construct. Current
Directions in Psychological Science, 22, 194 –198. http://dx.doi.org/10
.1177/0963721412471346
Merck, C., Topcu, W., & Hirsh, W. (2016). Collective mental time travel:
Creating a shared future through our shared past. Memory Studies, 9,
284 –294. http://dx.doi.org/10.1177/1750698016645236
Michaelian, K., & Sutton, J. (2017). Collective mental time travel: Re-
membering the past and imagining the future together. Synthese. Ad-
vance online publication. http://dx.doi.org/10.1007/s11229-017-1449-1
Newby-Clark, I. R., & Ross, M. (2003). Conceiving the past and future.
Personality and Social Psychology Bulletin, 29, 807– 818. http://dx.doi
.org/10.1177/0146167203029007001
Norris, P. (2000). A virtuous circle: Political communications in post-
industrial societies. New York, NY: Cambridge University Press. http://
dx.doi.org/10.1017/CBO9780511609343
Pillemer, D. B. (2003). Directive functions of autobiographical memory:
The guiding power of the specific episode. Memory, 11, 193–202.
http://dx.doi.org/10.1080/741938208
Race, E., Keane, M. M., & Verfaellie, M. (2011). Medial temporal lobe
damage causes deficits in episodic memory and episodic future thinking
not attributable to deficits in narrative construction. The Journal of
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
10 SHRIKANTH, SZPUNAR, AND SZPUNAR

Neuroscience, 31, 10262–10269. http://dx.doi.org/10.1523/JNEURO-
SCI.1145-11.2011
Race, E., Keane, M. M., & Verfaellie, M. (2013). Losing sight of the
future: Impaired semantic prospection following medial temporal lobe
lesions. Hippocampus, 23, 268 –277. http://dx.doi.org/10.1002/hipo
.22084
Rubin, D. C., & Berntsen, D. (2003). Life scripts help to maintain auto-
biographical memories of highly positive, but not highly negative,
events. Memory & Cognition, 31, 1–14. http://dx.doi.org/10.3758/
BF03196077
Sani, F., Bowe, M., Herrera, M., Manna, C., Cossa, T., Miao, X., & Zhou,
Y. (2007). Perceived collective continuity: Seeing groups as entities that
move through time. European Journal of Social Psychology, 37, 1118 –
1134. http://dx.doi.org/10.1002/ejsp.430
Schacter, D. L., & Addis, D. R. (2007). The cognitive neuroscience of
constructive memory: Remembering the past and imagining the future.
Philosophical Transactions of the Royal Society of London Series B,
Biological Sciences, 362, 773–786. http://dx.doi.org/10.1098/rstb.2007
.2087
Schacter, D. L., Addis, D. R., Hassabis, D., Martin, V. C., Spreng, R. N.,
& Szpunar, K. K. (2012). The future of memory: Remembering, imag-
ining, and the brain. Neuron, 76, 677– 694. http://dx.doi.org/10.1016/j
.neuron.2012.11.001
Schacter, D. L., Benoit, R. G., & Szpunar, K. K. (2017). Episodic future
thinking: Mechanisms and functions. Current Opinion in Behavioral
Sciences, 17, 41–50. http://dx.doi.org/10.1016/j.cobeha.2017.06.002
Soroka, S., & McAdams, S. (2015). News, politics, and negativity. Polit-
ical Communication, 32, 1–22. http://dx.doi.org/10.1080/10584609
.2014.881942
Squire, L. R., van der Horst, A. S., McDuff, S. G. R., Frascino, J. C.,
Hopkins, R. O., & Mauldin, K. N. (2010). The role of the hippocampus
in remembering the past and imagining the future. Proceedings of the
National Academy of Sciences of the United States of America, 107,
19044 –19048. http://dx.doi.org/10.1073/pnas.1014391107
Suddendorf, T., & Corballis, M. C. (2007). The evolution of foresight:
What is mental time travel, and is it unique to humans? Behavioral and
Brain Sciences, 30, 299 –313. http://dx.doi.org/10.1017/S0140525X
07001975
Szpunar, K. K. (2010). Episodic future thought: An emerging concept.
Perspectives on Psychological Science, 5, 142–162. http://dx.doi.org/10
.1177/1745691610362350
Szpunar, K. K., Spreng, R. N., & Schacter, D. L. (2014). A taxonomy of
prospection: Introducing an organizational framework for future-
oriented cognition. Proceedings of the National Academy of Sciences of
the United States of America, 111, 18414 –18421. http://dx.doi.org/10
.1073/pnas.1417144111
Szpunar, P. M., & Szpunar, K. K. (2016). Collective future thought:
Concept, function, and implications for collective memory studies.
Memory Studies, 9, 376 –389. http://dx.doi.org/10.1177/175069801
5615660
Tversky, A., & Kahneman, D. (1973). Availability: A heuristic for judging
frequency and probability. Cognitive Psychology, 5, 207–232. http://dx
.doi.org/10.1016/0010-0285(73)90033-9
Walker, W. R., & Skowronski, J. J. (2009). The fading affect bias: But
what the hell is it for? Applied Cognitive Psychology, 23, 1122–1136.
http://dx.doi.org/10.1002/acp.1614
Weinstein, N. D. (1980). Unrealistic optimism about future life events.
Journal of Personality and Social Psychology, 39, 806 – 820. http://dx
.doi.org/10.1037/0022-3514.39.5.806
Received June 5, 2017
Revision received January 31, 2018
Accepted February 4, 2018 䡲
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
11
DISSOCIATING THE FUTURE