ArticlePDF Available

Folk Beliefs About Where Manipulation Outside of Awareness Occurs, and How Much Awareness and Free Choice Is Still Maintained

Authors:

Abstract and Figures

To examine folk beliefs on manipulative techniques targeting the unconscious and their effects on free choice and level of awareness, we collected judgments based exclusively on examples participants volunteered themselves. Our sample (N = 961) consisted of respondents from six continents with 46 different nationalities and residing in 27 different countries. Participants were first asked to report (via free text) a personal experience in which they suspected that unconscious manipulation had taken place. After this, they rated their experience across a number of dimensions (e.g., level of awareness of manipulation, success of manipulation, level of free choice, level of concern). Consistent with previous findings, participants thought of marketing as the most common context in which unconscious manipulation takes place (45% of all participants) followed by research (11%)—typically psychological studies, therapy (2%)—typically hypnosis, media (11%)—including entertainment, news media, social media, and politics (3%). In addition, free choice, awareness, and most other ratings were not reliably predicted by the context of the example volunteered, or by group level differences (place of residence, age, gender, religiosity, political affiliation, education), suggesting near universally shared beliefs in the way unconscious manipulation is being conceptualized. As observed in previous work, we also find that, irrespective of context, participants believe that even if manipulated, they retain some degree of free choice. We focus the discussion on new insights that the data set revealed with respect to other domains of influence and manipulation.
Content may be subject to copyright.
Psychology of Consciousness: Theory, Research,
and Practice
Folk Beliefs About Where Manipulation Outside of Awareness Occurs, and
How Much Awareness and Free Choice Is Still Maintained
Magda Osman and Christos Bechlivanidis
Online First Publication, November 2, 2023. https://dx.doi.org/10.1037/cns0000379
CITATION
Osman, M., & Bechlivanidis, C. (2023, November 2). Folk Beliefs About Where Manipulation Outside of Awareness Occurs,
and How Much Awareness and Free Choice Is Still Maintained. Psychology of Consciousness: Theory, Research, and
Practice. Advance online publication. https://dx.doi.org/10.1037/cns0000379
BRIEF REPORT
Folk Beliefs About Where Manipulation Outside
of Awareness Occurs, and How Much Awareness
and Free Choice Is Still Maintained
Magda Osman
1
and Christos Bechlivanidis
2
1
Centre for Science and Policy, University of Cambridge
2
Department of Experimental Psychology, University College London
To examine folk beliefs on manipulative techniques targeting the unconscious and their
effects on free choice and level of awareness, we collected judgments based exclusively
on examples participants volunteered themselves. Our sample (N=961) consisted of
respondents from six continents with 46 different nationalities and residing in 27 different
countries. Participants were rst asked to report (via free text) a personal experience in
which they suspected that unconscious manipulation had taken place. After this, they rated
their experience across a number of dimensions (e.g., level of awareness of manipulation,
success of manipulation, level of free choice, level of concern). Consistent with previous
ndings, participants thought of marketing as the most common context in which
unconscious manipulation takes place (45% of all participants) followed by research
(11%)typically psychological studies, therapy (2%)typically hypnosis, media (11%)
including entertainment, news media, social media, and politics (3%). In addition, free
choice, awareness, and most other ratings were not reliably predicted by the context of the
example volunteered, or by group level differences (place of residence, age, gender,
religiosity, political afliation, education), suggesting near universally shared beliefs in the
way unconscious manipulation is being conceptualized. As observed in previous work, we
also nd that, irrespective of context, participants believe that even if manipulated, they
retain some degree of free choice. We focus the discussion on new insights that the data set
revealed with respect to other domains of inuence and manipulation.
Keywords: folk beliefs, unconscious, manipulation, awareness, free choice
If it is the case that our ability to freely choose
in day-to-day contexts is subverted by manip-
ulative techniques that target us outside our
awareness, we would expect that free choice
should be undermined. On this basis, a simple
prediction is that the more we suspect that
unconscious manipulation takes place in some
context, the less free choice we would believe
Magda Osman https://orcid.org/0000-0003-1480-6657
No external funding sources contributed to this study. The
authors have no known conicts of interest to disclose.
All the raw anonymized data collected and analyzed for
this study will be made available in a weblink to the published
article at https://osf.io/3xvwq/.
Magda Osman developed the methods, materials, ran the
study, and conducted the data collection. The coding of the
data was carried out independently by Magda Osman and
Christos Bechlivanidis, and all analyses presented in the
article were conducted by Christos Bechlivanidis. The
writing up of the article was prepared by Magda Osman and
reviewed and revised by Christos Bechlivanidis.
Open Access funding provided by Centre for Science and
Policy, University of Cambridge: This work is licensed
under a Creative Commons Attribution 4.0 International
License (CC BY 4.0; http://creativecommons.org/licenses/
by/4.0). This license permits copying and redistributing the
work in any medium or format, as well as adapting the
material for any purpose, even commercially.
Correspondence concerning this article should be
addressed to Magda Osman, Centre for Science and
Policy, University of Cambridge, 10 Trumpington Street,
Cambridge CB2 1QA, United Kingdom. Email: m.osman@
jbs.cam.ac.uk
1
Psychology of Consciousness:
Theory, Research, and Practice
© 2023 The Author(s)
ISSN: 2326-5523 https://doi.org/10.1037/cns0000379
to be left with. A more puzzling question is how
we can assess the role of unconscious inu-
ences in our decision-making behavior, given
that, by denition, we are not aware of them.
A solution would be to associate certain contexts
or scenarios with mechanisms that override our
conscious deliberative faculties (e.g., uncon-
scious priming) and assume that in such cases
we cannot condently believe the choices we
make to be (exclusively) our own. However,
the nding from several studies examining
peoples folk beliefs (Osman, 2020;Osman &
Bechlivanidis, 2021,2022) is that the relation-
ship between the judged level of manipulation
and the level of free choice is weak. The studies
are based on peoplesappraisalsofeveryday
situations in which choices are made (e.g.,
electoral voting, consumer choices, interactions
on social media), as generated by other people.
By encouraging a more personalized perspec-
tive of the various genuine scenarios that have
been volunteered by people (Osman, 2020), the
overall relationship between the two (likelihood
of manipulation and level of free choice)
is strengthened. However, the ndings also
show that the relationship is still unstable in spite
of the experimental instructional manipulations
(Osman & Bechlivanidis, 2022). The relation-
ship is stronger in some prototypical contexts
commonly associated with unconscious pro-
cesses being targeted (e.g., sleep research,
subliminal priming, subliminal advertising,
hypnotherapy). The weaker associations were
found in the marketing contexts, which include
a range of scenarios (i.e., communication styles
of political leaders campaigning, jingles in
supermarkets, product placement in supermar-
kets, sales tactics in car dealerships). Overall,
previous work shows that while unconscious
manipulation and free choice are correlated in
the expected (negative) direction, the relation-
ship is much weaker than logically expected and
sensitive to the degree of personalization, and
the type of scenario.
Motivation of Present Study
Although there are multiple experimental
studies in which conscious awareness is manipu-
lated and level of estimated free choice is
investigated (Clark et al., 2017;Schooler et al.,
2014;Shepherd, 2012; for review, see Mudrik
et al., 2022), there is a little data about what
people actually believe about this relationship.
In peoples view, to what extent can there be
freedom of choice in the face of unconscious
manipulation? Is the degree of awareness of
those manipulative tactics associated with the
amount of free choice people believe they are
left with? Understanding folk beliefs on this matter
is important in informing moral, philosophical,
and policy-related questions. For example, while
manipulation is commonly assumed to be morally
reprehensible, especially since it restricts auton-
omy, it is not known whether people see a necessary
link between these two concepts, or whether that
link sufces to determine moral judgments. Which
ever of these intuitions is universally held by people
is critical when deciding between the multitude of
philosophical theories of manipulation that have
appeared recently (Noggle, 2020). Finally, under-
standing how people evaluate manipulative tactics
in different contexts should be a central consider-
ation when the state considers either implement-
ing or regulating interactions in the absence of
clear consent.
There is a logical inconsistency that is con-
fronted when collecting any rating of awareness of
suspected manipulation as well as of the judged
level of manipulation (Osman, 2020;Tallis, 2002).
For both, if manipulative tactics used to target
psychological processes below an individuals
awareness are successful, then how would that
individual be able to detect the presence of the
manipulation, or the amount of manipulation
happening at the time? To get around this, the
proposals made in previous work, and here, are
that people draw from their general knowledge
of the employment of manipulative psychologi-
cal tactics in any context they confront. General
cultural and social knowledge of manipulative
tactics (e.g., subliminal advertising, hypnosis,
microtargeting, subliminal priming) can be
recruited to judge the likelihood of their presence
in any given context. It is also possible that people
form general impressions or feelings of an
impetus to choose to act in a way, as a result
of external mechanisms and techniques, that
is inconsistent, or occasionally consistent, with
their conscious intentions (Holton, 2009;Lefebvre,
2001;Levy, 2011,2014;Osman, 2020;Owens,
2000); evidence of this has also been reported in
studies of young adults (e.g., Hagger et al., 2002).
To this end, the reason for expanding the sample
to include more countries than were in the
original (Australia, Canada, United Kingdom,
2 OSMAN AND BECHLIVANIDIS
United States) was to examine a wider range of
cultural perspectives.
To explore this, the novel contribution the
present study makes by focusing on folk beliefs
is threefold. First, rather than the experimenters
coming up with carefully controlled but possibly
unrealistic examples, participants are instead
asked to generate their own examples where they
suspect psychological tactics are used to manip-
ulate their choices outside of their awareness.
From previous work, the evidence suggests that
people provide a rich range of scenarios. By using
this approach, the present study is also able to
look at the level of consistency with Osmans
(2020) study to assess the contexts that partici-
pants volunteered to determine type and frequency
of examples generated. To that end, we use the
same basic open question as before, and we draw
on the same sample of participants (i.e., Australia,
Canada, United Kingdom, and United States),
while also expanding the sample to include
nationalities from a wide range of countries
and continents (Africa, Asia, Europe, Oceania,
South America), to form a broader cultural
range of perspectives.
Second, as an extension of prior research
(Osman, 2020;Osman & Bechlivanidis, 2021,
2022), the present study asks participants to give
ratings on the examples they provide themselves.
While this means that there is no possibility of
controlling what is generated, only the methods
by which judgments are made, the possibility of
patterns emerging would provide a valuable insight
because of the high number of factors outside
of experimental control. Also, as mentioned, in
previous work participants were presented with
scenarios generated by others, regarding which
they gave ratings such as level of manipulation,
free choice, control, concern, and responsibility.
Those ndings suggested that when instructional
manipulations are introduced to increase the
personalization of the scenarios, then this in turn
increases the strength of the negative relationship
between judged level of manipulation and judged
level of free choice. Therefore, it may be the case
that, by rating the examples participants have
themselves generated, the association between
the ratings would become even stronger.
Third, in an attempt to explain the weak
relationship between unconscious manipulation
and free choice, we also consider a new judgment
that investigates the relationship between the level
of awareness of the manipulative technique and
the level offree choice. To the authorsknowledge,
this relationship has yet to be explored in the
arguably more valid context of examples of
participants volunteer themselves.
Method
Participants
A total of 1,019 participants were recruited
but due to some respondents not completing all
judgments, and some errors generated by the
crowdsourcing platform where some respon-
dents took part in the study twice, a total of 961
participants remained that generated unique
scenarios. Participants were from 46 different
nationalitiesand lived in 27 different countries(for
details of sampling, see Osman, 2023).
In the nal analysis, 874 participants were
included (for details, see https://osf.io/3xvwq/)
for reasons of comparability. A proportion of
participants (n=87) generated multiple scenar-
ios in their examples, so to be able to compare
across judgments analyses were based only on
those that generated a single example. Split by
continent, 167 reside in Europe, 131 in North
America, 140 in South America, 170 in Africa,
132 in Oceania, and 134 in Asia (https://osf.io/
3xvwq/). The average age of the sample was M=
31.81 (SD =10.53), and ranged from 19 to 69,
with 429 males, 432 females, and 13 preferring
not to say. For full details of age, gender,
ethnicity, political afliation, education level,
and religiosity, see Table 1.
The study was presented via Qualtrics (https://
www.qualtrics.com/), an online platform for hosting
experiments, while participants were recruited
through Prolic(https://www.prolic.co/), a
crowdsourcing platform. To take part in the
study, theinitial set of criteria were thatparticipants
were born and currently reside in Australia,
Canada, United Kingdom, and United States
that their age ranged between 18 and 80, and
their rst language was English. The initial
inclusion criteria were designed to be identical
to Osman (2020, Study 1) in order to replicate
the original sample for comparison. In addition, to
ensure that the sample was also global, partici-
pants were also recruited from Asia, Africa, and
South America (full details regarding the settings
used for recruitment can be found at https://osf.io/
3xvwq/).
FOLK BELIEFS ON MANIPULATION AND AWARENESS 3
All participants were nancially compensated
for their time (1.40 USD). When taking part in
the study, participants were asked to give a free
text response to the open question requesting
an example of a context where they suspect
manipulation outside of awareness is taking
place, along with ratings for six questions and ve
demographic questions (age, gender, education
level, political afliation, religiosity), which are
summarized in Table 1; note that for analytic
purposes details of ethnicity were included based
on the participant prole details that Prolic
automatically collects. The present study received
ethical approval from Judge Business School,
University of Cambridge ethics board, JBS/23-
02/11.01.2023.
Design and Materials
In the present study, there were three sets of
variables that each participant generated: responses
to ve demographic questions (i.e., age, gender,
education level, political afliation), ratings to six
judgment probes (see Table 2), free text generated
in response to the open question (see Table 2)
concerning an example where manipulation
outside of awareness was suspected to have
occurred. For each demographic question parti-
cipants were provided with the option prefer not
to say.The six ratings were as follows: ratings
of awareness of unconscious manipulation,
ratings of free will, ratings of othersexperience,
ratings of concern, ratings of ultimate control,
and ratings of success of the manipulation, each
of which was on a scale ranging from 0 to 10
(see Table 2). The presentation of the open
question was xed, and always presented rst
to every participant and, thereafter, the order
of presentation of the six judgment probes was
randomized for each participant.
Procedure
Participants were rst asked to provide their
consent in order to take part in the study. They
were told that they would be presented with an
open question and six follow-up questions. They
were informed that after they had completed all
ratings, and lled ve demographic questions, the
study would be complete.
Ratings
To establish reliability, approximately one
third of the total sample of volunteered free
text scenarios were coded independently by two
researchers (n=376). Specically, eachresearcher
assigned a context to each example, by choosing
from one of the following categories, taken from
Osman (2020):marketing,”“research,”“media,
political,”“therapy,and other.There was an
84% agreement between the coders. Most dis-
agreements were due to participants providing more
than one example (which the rst researcher coded
as multiple,while the second coded based on
the rst example), excluding multipleexam-
ples, the agreement rose to 87%. The remaining
disagreements weredue to ambiguity, for example,
an example of a marketing manipulation taking
place in social media. To resolve these disagree-
ments both coders reviewed where disagreements
had occurred and nalized the classication of the
examples according to a consistent coding frame
that focused on the primary source of manipulation
by context (e.g., marketing), rather than where the
manipulation had occurred (social media). The
remaining two thirds of scenarios were coded by
a single researcher. Since some of the participants
(n=74) volunteered more than one example in
their free text responses, to ensure comparisons
with the original study (Osman, 2020, study 1),
and for later analysis, only single examples that
Table 1
Participants Prole
Sample World sample
Total participants N=961
Females 480 (49.9%)
Males 467 (48.6%)
Prefer not to say 14 (1.5%)
Age M31.90 (SD =10.57) ranging from
19 to 69
Educational
background
15.9 completed education at 16 or
18 years, 56.8% qualied with a
degree (bachelor degree), 17.5%
qualied with a postgraduate
degree (masters or PhD),
9.9% other/prefer not to say.
Political afliation 43.2% identifying as liberal, 21.2%
as center, 17.3% as conservative,
and 18.3% as unsure/other
Religiosity 31% religious, 55.6% not religious,
13.4% unsure/other
Ethnicity 7.1% Asian, 16.3% Black, 13.3%
mixed race, 55.3% White, 8%
other/prefer not to say.
4 OSMAN AND BECHLIVANIDIS
were reported were used to conduct analyses
presented in the results section (n=874).
Results
The analyses were focused on exploring two
factors: (a) the extent to which the contexts
volunteered in the present study are consistent
with those reported by Osman (2020),which
were initially generated in 2018, and what, if
any, new contexts emerged, and (b) the strength of
the relationship between judgments of awareness
and free choice overall, as well as by context
and demographics.
Contexts Volunteered
The noteworthy patterns highlighted here
concern the location of similarities between the
present study and the original, and then where
there are key differences (see Table 3). First, by
far the most common context volunteered in the
present sample overall (n=874), as well as by
the four countries (n=184) that were directly
comparable to the replicated study (Osman, 2020),
was marketing and advertising (present study
overall [n=874] 44.9%, present study subsample
[n=184, Australia, Canada, United Kingdom,
United States] 51.6%, Osman, 2020 [n=399]
Table 2
Judgment Probes Alongside the Main Open-Ended Question
Types of questions Instructions and response options
Open-ended question Instruction to open-ended question
In psychology the unconscious is taken to mean many things. The simplest
description is that the unconscious is a type of process that inuences what
we do (thoughts, feelings, behaviours, attitudes, beliefs, judgments) in some
way without us being consciously aware of HOW it inuences what we do.
That is, there is something that is guiding what we are doing at the back of
our minds, but we cant easily explain what it is, and how it might be doing
that. In the space provided below, all you need to do is describe a
TYPICAL context, it could be any context, in which you tend to think that
psychological research on the unconscious has been used in some way to
manipulate behaviour. This question is left deliberately open so that you
can answer in whichever way you think captures a typical experience in
which you think the unconscious was inuenced in some way that would in
turn have changed your behaviour. There is no right or wrong answer, and
the answers that you will provide will be extremely informative.
Judgment probes All judgment probes begat with the sentence: Based on the example you just
described, respond to the following question. Imagine you are in the
situation you described
Awareness of unconscious
manipulation at the time
To what extent are you aware at the time that you are being unconsciously
manipulated through the processes you described?
(0 =I have NO IDEA that something could be inuencing me to 10 =I have
A STRONG FEELING that something could be inuencing me)
Level of free choice To what extent do you think your critical choices in this situation are FREE,
given that they are being unconsciously manipulated?
(0 =NOT AT ALL Free to 10 =COMPLETELY Free)
Extent to which other would
experience the same situation
To what extent do you think the experience you just described is an
experience that is also experienced by others?
(0 =I DO NOT think anyone else experiences what I described to
10 =I DO think that everyone experiences what I described)
Level of concern To what extent do you care given that your critical choice in the experience
you described has been unconsciously manipulated?
(0 =DO NOT Care at all to 10 =I Care HUGELY)
Level of control To what extent do you have ultimate control over your critical choice in the
experience you just described, given that the choice is being unconsciously
manipulated?
(0 =NO CONTROL AT ALL to 10 =COMPLETE CONTROL)
Extent to which the unconscious
manipulation was successful
To what extent do you think your critical choices are being SUCCESSFULLY
unconsciously manipulated through the processes you described?
(0 =NOT AT ALL SUCCESSFUL to 10 =ENITRELY SUCCESSFUL)
FOLK BELIEFS ON MANIPULATION AND AWARENESS 5
45.6%). This suggests that marketing and adver-
tising are likely the most salient and perhaps the
most prototypical context that people call to mind
when thinking about where they are manipulated
without their conscious awareness.
Unlike Osmans (2020) study (18%) for which
researchwas the next most common context
volunteered, in the present study, it was media
(including, entertainment media, news media,
social media; present study overall 10.7%, present
study subsample 14.2%). It is hard to draw any
rm conclusions as to why this is the case. The
only difference between studies is when they were
conducted (i.e., original study was run in 2018,
and the present study was run in 2023), given that
the sampling method and the question presented
to both samples itself were identical.
Given that the context media generated more
examples from the present sample compared to
the original study, the responses falling under
the category of media were examined in more
detail. From this, the examples clustered into
two subcategories of media coded according to
whether they explicitly referred to social media
(SM), separate from conventional media (CM;
i.e., news media, entertainment media; CM).
Splitting the responses by countries comparable
to the original study (Osman, 2020), there does
appear to be some further differences between
countries: Australia (n=48, SM =10.4%,
CM =6.3%), Canada (n=45, SM =4.4%,
CM =8.9%), United Kingdom (n=48, SM =
6.3%, CM =4.2%), and United States (n=43,
SM =7%, CM =9.3%). If the general split
between social media and CM is applied to the
overall sample of the present study, then the
distribution of examples is approximately equal
(n=874, SM =5.3%, CM =5.4%). Overall,
what this may reveal is that as more people treat
social media and CM as different entities, their
Table 3
Proportion of Respondents Volunteering Examples Correspondent With the Main Contexts by Study (Present
Study Overall Sample n =887, Present Study Subsample n =184, Osman, 2020, Study 1)
(%)
Marketing/
advertising Research Therapy Media Political Other
None/do
not know
Overall (Osman, 2020,N=399)
Australia, Canada, United
Kingdom, United States
45.1 18 4.3 4.8 6.3 18.3 3.3
Present (N=184)
Australia, Canada, United
Kingdom, United States
51.6 5.4 1.6 14.2 2.2 25 0
Overall (present, N=887) 44.9 10.6 1.6 10.7 2.7 28.1
Social =45
Nonsocial =55
.7
Breakdown of individual countries for comparison between the present study and Osmans (2020) study
Australia (Osman, 2020,N=96) 51 20.8 3.1 4.2 6.3 11.5 3.1
Australia (present, N=48) 50 2.1 016.7 2.1 29.2
Social =25
Nonsocial =75
0
Canada (Osman, 2020,N=104) 43.3 21.2 1.9 3.8 5.8 22.1 1.9
Canada (present, N=45) 51.1 8.9 6.7 13.3 2.2 17.8
Social =37.5
Nonsocial =62.5
0
United Kingdom (Osman, 2020,
N=100)
44 10 6.0 9.0 6.0 19 6.0
United Kingdom (present,
N=48)
47.9 2.1 010.5 2.1 37.5
Social =50
Nonsocial =50
0
United States (Osman, 2020,
N=99)
42.4 20.2 6.1 2.0 7.1 20.2 2.0
United States (present, N=43) 58.1 9.3 016.3 2.3 14
Social =16.7
Nonsocial =83.3
0
6 OSMAN AND BECHLIVANIDIS
experiences of what they see as manipulative
tactics employed in each becomes more nuanced.
Nonetheless, other than being a pattern to
highlight, without exploring this in depth in future
research, wehold forth on speculating any further,
or drawing conclusions.
The second main difference between the
present study and the original is the proportion
of examp les that were volunteer ed that fell under
the category other,in the original sample
(18.3%) and the replicated sample (present study
overall 28.1%, present study subsample 25%).
Given that close to a third of the overall sample
generated contexts that do not fall under the ve
categories original identied by Osman (2020),
this required further exploration. A detailed
analysis of the examples generated under the
othercategory revealed two overall clusters
basedonthosethatexplicitlyreferredtoother
agents in social contexts, often around inter-
personal interactions with family, friends,
dating partners, spouses, and children. This
corresponds to situations referred to by Deci
and Ryan (2012) in their self-determination
theory. Social pressures were identied in infor-
mal settings where conforming behaviors were
expressed, or reciprocal behaviors were per-
formed but were not ones that were intended by
the participant themselves (Deci & Ryan, 2012).
Social pressure of this kind often identied social
agents that were highly familiar to the participant
describing the scenario, and emotional manipu-
lation (e.g., guilt) was the tactic often employed
that could be construed as subtly coercive.
Additionally, social pressures were described
in reference to hierarchically structured relation-
shipwhereconformitytosocialnormswas
expected (e.g., in religious, educational, military,
management settings where authority gures
were present). The second type described
scenarios where the mechanism of inuence or
process that was being inuenced did not
explicitly included reference to people (e.g.,
sleeping, driving, dreaming). For the purposes
of analysis, the examples initially coded as
otherwere recoded as othernonsocial
(55%) and othersocial(45%) to distinguish
between examples that explicitly referred to another
agent from those that did not (see Table 3). This
suggests that a nontrivial proportion of examples
(12%) volunteered by participants reected
beliefs that others employed tactics attempting to
inuence them, which at least at the time, they were
deemed to be outside of their awareness.
Relationship Between Ratings and Context
The pattern indicated in Figure 1 suggests that
participants still rated that they had some degree
of free choice, M=5.28, SD =2.49, t(874) =
3.353, p<.001, d=0.11, 95% CI [5.12, 5.45],
1
and control, M=5.86, SD =2.53, t(874) =
10.019, p<.0001, d=0.34, 95% CI [5.69, 6.02],
despite the fact that the manipulations they
described were deemed to be relatively success-
ful, M=6.16, SD =2.21, t(874) =15.531,
p<.0001, d=0.53, 95% CI [6.01, 6.31].
Turning specically to the correlations
between judgments (see Figure 2)overall
and by context. Across all contexts, there was
a consistently strong correlation between free
choice and control, r(872) =0.503, p<.001. In
most contexts, but for social media, there was
a moderate correlation found between aware-
ness and control, r(872) =0.245, p<.001,
indicating that the more aware someone is of
a manipulation the more control they retain.
Ratings of free choice are weakly negatively
correlated with concern, r(872) =0.113,
p<.001, indicating that the less free choice one
has, the more concerned they are (not found in
research and politics). Free choice is also weakly
positively correlated with awareness, r(872) =
0.167, p<.001, meaning that the more aware
someone is of the manipulation the more free their
choice is thought to be.
Clearly, the relationship between awareness
and free choice is unreliable and varies by the
context in which the manipulation has occurred,
such that it is stronger in therapeutic contexts and
CM but not so for social media and political
contexts. Given that the relationship between
awareness and free choice is a novel avenue, we
examined this further. In particular, the interest
was to examine whether at extreme ends of the
scale, the relationship was more distinct or was a
consistently weak relationship. To address this,
responses 0or 1were classied as low
awareness(n=111) and those responses 9or
10were classied as high awareness(n=80).
Applying the same rationale, responses 0or 1
1
The one-sample ttests are against the midpoint of the
scale (5), on the assumption that participants may use it to
represent uncertainty.
FOLK BELIEFS ON MANIPULATION AND AWARENESS 7
(n=68) were classied as restricted free choice
and responses 9or 10were classied as
unrestricted free choice(n=103). Comparing
mean ratings of free choice by low (M=4.21,
SD =3.04) and high aware respondents (M=6.09,
SD =3.23), t(189) =4.10, p<.001, d=.61, 95%
CI [2.78, .98], reveals a distinct pattern in line with
a view that high awareness is associated with
higher free choice compared with low awareness.
Complementing this, comparing mean ratings of
awareness by low restricted free choice (M=3.91,
SD =3.48) and unrestricted free choice (M=5.75,
SD =3.11), t(169) =3.60, p<.001, d=.56, 95%
CI [2.84, .83], also indicates a reasonably distinct
difference in the expected direction.
Finally, several demographic details were
collected (age, ethnicity, gender, continent of
residence, education, political afliation, religi-
osity) and these were used as predictor variables
in a single regression analysis for every rating.
The only signicant predictors (a=0.01) were
age for awareness, age for concern, and age
for control, F(1, 850) <7.459, p<.006, partial
η<.033. Also, the continent that participants
currentlyresided in predicted the judgedgenerality
of their offered example. Putsimply, the continent
predicted the extent to which participants judged
that their examples were experienced by others,
F(5, 850) =5.770, p<.001.
Conclusion
Previous work has examined folk beliefs
regarding situations where psychological tactics
are suspected to subvert our ability to make
choices freely because they target our uncon-
scious. The ndings from that research suggest
that the relationship between judged level of
manipulation and free choice is highly labile and
particularly sensitive to the context in which
judgments are made (e.g., political, marketing,
media, therapy). The ndings from the present
study reveal that the lability also extends to the
relationship between the amount of awareness of
the manipulative tactics used at the time and the
level of free choice that can still be maintained.
The current sample of participants was drawn
from a wider range of countries compared to the
original (Osman, 2020) so as to expand the range
of cultural perspectives that participants would
likely be drawing from. Consistent with previous
research, and now extended to a different judgment
(namely awareness of manipulation at the time of
choice), overall the relationship between judged
Figure 1
Mean Ratings for the Six Judgments Collapsed Across the Contexts Volunteered
Note. Error bars represent 95% condence intervals. See the online article for the color version of this gure.
8 OSMAN AND BECHLIVANIDIS
awareness and free choice is weak but does
strengthen when looking at specic contexts
(e.g., marketing and advertising). In addition,
consistent with previous research, the examples
that participants volunteered were similar to those
generated in a previous study (Osman, 2020),
for which the most common illustration where
manipulative tactics directed at the unconscious
was marketing and advertising contexts.
A new insight revealed in the present study is
that approximately 15% (n=107) of the sample
provided examples of social situations where they
described manipulative tactics (e.g., emotional
manipulation) that they interpreted as an illustration
of a psychological technique that targeted their
unconscious. This is consistent with Deci and
Ryans (2012) self-determination theory which
asserts that unless people believe that they are no
external coercive inuences from others, then their
actions are not free, and they cannot attribute to
themselves personal responsibility. In addition,
work by philosophers (e.g., Kligman & Culver,
1992;Susser et al., 2019) that treat manipulation as
something that is covert or a hidden inuence on
the recipient includes examples such as gaslight-
ing, guilt tripping, deception, and misdirection
consider the interpersonal domain of manipula-
tion. Recent developments in the philosophical
domain (e.g., Rachar, 2023) also consider this
from the view of boundary condit ions in contexts
where reciprocal actions occur. Where it is
expected that both agents have aligned intentions
because there are shared interpersonal obliga-
tions, manipulation is a means of subverting
social conventions to serve ones own ends at
the expense of others. While this was not an area
of focus in the present study, the ndings help
reveal how sensitive people are when it comes
to actions in interpersonal settings. Participants
volunteered examples of an interpersonal nature
as an illustration of where psychological tactics
target the unconscious. This in turn suggests that
even when reciprocity in social interactions is
expected to occur, the intentions behind socially
oriented actions are not always aligned, and in
turn, free choice is impacted.
Finally, regardless of whether psychological
manipulative tactics are successful in changing
Figure 2
Full Matrix of Correlations Between the Six Judgments Probes (Awareness, Free Choice, Others Experiencing
the Same Thing, Concern, Control, Success)
Note. The correlation color coding is from dark blue (1) to dark brown (+1), with lighter shades indicating weaker or no
correlations. See the online article for the color version of this gure.
FOLK BELIEFS ON MANIPULATION AND AWARENESS 9
behavior, and regardless of whether they are
judged as substantively impacting free choice,
from a folk beliefs point of view the most
common examples given are negative in nature.
In philosophical (e.g., Barnhill, 2022)and
psychological (e.g., Sunstein, 2016), research
manipulation has also been conceptually ana-
lyzed from different moral positions that include
neutral and positive illustrations. Therefore, it is
important to recognize that the term manipula-
tionis loaded, and this may be a limitation of
the present line of work. For instance, if the main
study question posed was framed as inuence
rather than manipulation,it may well be that
more neutral or even positive examples would
be volunteered. Thus, a further area of investi-
gation is to examine how sensitive folk beliefs
are to different framings of the key question
posed. Where denitions are provided, it may
be the case that, this in turn impacts the judged
moral status of psychological manipulative
(otherwise referred to as inuential) tactics
designed to operate outside of awareness.
References
Barnhill, A. (2022). How philosophy might contribute
to the practical ethics of online manipulation. In F.
Jongepier & M. Klenk (Eds.), The philosophy of
online manipulation (pp. 4971). Routledge.
https://doi.org/10.4324/9781003205425-4
Clark, C. J., Baumeister, R. F., & Ditto, P. H. (2017).
Making punishment palatable: Belief in free will
alleviates punitive distress. Consciousness and
Cognition,51,193211. https://doi.org/10.1016/j
.concog.2017.03.010
Deci, E. L., & Ryan, R. M. (2012). Self-determination
theory. In P. A. M. Van Lange, A. W. Kruglanski, &
E. T. Higgins (Eds.), Handbook of theories of social
psychology (pp. 416436). Sage Publications.
https://doi.org/10.4135/9781446249215.n21
Hagger, M. S., Chatzisarantis, N. L., & Biddle, S. J.
(2002). The inuence of autonomous and control-
ling motives on physical activity intentions within
the Theory of Planned Behaviour. British Journal of
Health Psychology,7(3), 283297. https://doi.org/
10.1348/135910702760213689
Holton, R. (2009). Willing, wanting, waiting. Oxford
University Press. https://doi.org/10.1093/acprof:oso/
9780199214570.001.0001
Kligman, M., & Culver, C. M. (1992). An analysis of
interpersonal manipulation. The Journal of Medicine
and Philosophy,17(2), 173197. https://doi.org/10
.1093/jmp/17.2.173
Lefebvre, V. (2001). Algebra of conscience. Kluwer
Academic Publishers. https://doi.org/10.1007/978-
94-017-0691-9
Levy, N. (2011). Resisting weakness of the will.
Philosophy and Phenomenological Research,82(1),
134155. https://doi.org/10.1111/j.1933-1592.2010
.00424.x
Levy, N. (2014). Consciousness, implicit attitudes and
moral responsibility. Noûs,48(1), 2140. https://
doi.org/10.1111/j.1468-0068.2011.00853.x
Mudrik, L., Arie, I. G., Amir, Y., Shir, Y., Hieronymi,
P., Maoz, U., OConnor, T., Schurger, A., Vargas,
M., Vierkant, T., Sinnott-Armstrong, W., & Roskies,
A. (2022). Free will without consciousness? Trends
in Cognitive Sciences,26(7), 555566. https://
doi.org/10.1016/j.tics.2022.03.005
Noggle, R. (2020). Pressure, trickery, and a unied
account of manipulation. American Philosophical
Quarterly,57(3), 241252. https://doi.org/10.2307/
48574436
Osman, M. (2020). Overstepping the boundaries of
free choice: Folk beliefs on free will and determin-
ism in real world contexts. Consciousness and
Cognition,77, Article 102860. https://doi.org/10
.1016/j.concog.2019.102860
Osman, M. (2023). World folk beliefs unconscious
manipulation.https://osf.io/3xvwq/
Osman, M., & Bechlivanidis, C. (2021). Public
perceptions of manipulations on behavior outside
of awareness. Psychology of Consciousness: Theory,
Research, and Practice. Advance online publica-
tion. https://doi.org/10.1037/cns0000308
Osman, M., & Bechlivanidis, C. (2022). Impact of
personalizing experiences of manipulation outside of
awareness on autonomy. Psychology of Consciousness:
Theory, Research, and Practice. Advance online
publication. https://doi.org/10.1037/cns0000343
Owens, D. J. (2000). Reason without freedom: The
problem of epistemic normativity. Psychology Press.
Rachar, M. (2023). Conditional intentions and shared
agency. Noûs. Advance online publication. https://
doi.org/10.1111/nous.12452
Schooler, J. W., Nadelhoffer, T., Nahmias, E., & Vohs,
K. D. (2014). Measuring and manipulating beliefs
and behaviors associated with free will. In A. R.
Mele (Ed.), Surrounding free will: Philosophy,
psychology, neuroscience (pp. 7294). Oxford
University Press. https://doi.org/10.1093/acprof:
Oso/9780199333950.003.0005
Shepherd, J. (2012). Free will and consciousness:
Experimental studies. Consciousness and Cognition,
21(2), 915927. https://doi.org/10.1016/j.concog.2012
.03.004
10 OSMAN AND BECHLIVANIDIS
Sunstein, C. R. (2016). The ethics of inuence:
Government in the age of behavioral science.
Cambridge University Press. https://doi.org/10
.1017/CBO9781316493021
Susser, D., Roessler, B., & Nissenbaum, H. (2019).
Technology, autonomy, and manipulation. Internet
Policy Review,8(2). https://doi.org/10.14763/2019
.2.1410
Tallis, F. (2002). Hidden minds: A history of the
unconscious. Arcade Publishing.
Received May 30, 2023
Revision received August 16, 2023
Accepted August 28, 2023
FOLK BELIEFS ON MANIPULATION AND AWARENESS 11
Article
Full-text available
When we are victims of manipulation, then theory expects that our ability to freely choose is threatened by default. But, when looking at folk beliefs, we see nuance in the judgments made about choice and manipulation, which appear to be strongly informed by perspective and context. Regarding perspective, the negative relationship between free choice and manipulation is stronger when we explicitly see ourselves rather than others in the same autonomy-threatening contexts. When it comes to context, free choice is impacted more in some manipulative situations (e.g., microtargeting, subliminal priming, subliminal advertising, hypnosis) than others (e.g., jingles, product placement, political campaigning). The labile nature of the relationship between manipulation and free choice, which is in contrast to theoretical expectations, requires further investigation to expose whether these two factors (perspective, context) are robust. The present study stress tests these two factors by examining order effects (Experiments 1 and 2: manipulation vs. autonomy, autonomy vs. manipulation), as well as varying verbal descriptions (Experiments 1 and 2: manipulation vs. influence) and the way judgments are expressed (Experiment 1 ratings on scales 0–10, Experiment 2 forced choice Y/N). Consistent with previous research, the strength of the relationship between manipulation and free choice is amplified in specific contexts and when considered from a personalized perspective. The implications of this for theories on free choice and free will are discussed.
Article
Full-text available
Generative AI enables automated, effective manipulation at scale. Despite the growing general ethical discussion around generative AI, the specific manipulation risks remain inadequately investigated. This article outlines essential inquiries encompassing conceptual, empirical, and design dimensions of manipulation, pivotal for comprehending and curbing manipulation risks. By highlighting these questions, the article underscores the necessity of an appropriate conceptualisation of manipulation to ensure the responsible development of Generative AI technologies.
Article
Full-text available
Shared agency is a distinctive kind of sociality that involves interdependent planning, practical reasoning, and action between participants. Philosophical reflection suggests that agents engage in this form of sociality when a special structure of interrelated psychological attitudes exists between them, a set of attitudes that constitutes a collective intention. I defend a new way to understand collective intention as a combination of individual conditional intentions. Revising an initial statement of the conditional intention account in response to several challenges leads to a specification of the properties these intentions need to have in order to be genuine commitments. I then show how a structure of conditional intentions with these properties settles a collection of agents on engaging in social interactions that display all the features typically associated with shared agency.
Article
Full-text available
Since 2016, when the Facebook/Cambridge Analytica scandal began to emerge, public concern has grown around the threat of “online manipulation”. While these worries are familiar to privacy researchers, this paper aims to make them more salient to policymakers—first, by defining “online manipulation”, thus enabling identification of manipulative practices; and second, by drawing attention to the specific harms online manipulation threatens. We argue that online manipulation is the use of information technology to covertly influence another person’s decision-making, by targeting and exploiting their decision-making vulnerabilities. Engaging in such practices can harm individuals by diminishing their economic interests, but its deeper, more insidious harm is its challenge to individual autonomy. We explore this autonomy harm, emphasising its implications for both individuals and society, and we briefly outline some strategies for combating online manipulation and strengthening autonomy in an increasingly digital world.
Article
Full-text available
Previous studies examining the relationship between beliefs about manipulation of behavior outside of awareness and free will reveal a weak association between both constructs. That is, people reconcile the two in such a way that there is room for free will despite mechanisms determining their behavior. The present study further explores the association between judgments of manipulation outside of awareness and free will, along with other judgments (ultimate control, personal experience, concern) in real-world scenarios where manipulation is expected to occur. The aim was to examine whether the relationship between manipulation outside of awareness and free will could be amplified if personalizing the experiences of the real-world examples was emphasized. To this end, Studies 1a (N = 111) and 1b (N = 106; replication) emphasized that people could have been manipulated, and in Studies 2a (N = 106) and 2b (N = 104; replication) emphasized that they had actually been manipulated. The findings revealed that when the degree of personalization increased, so did the strength of the negative correlation between judgments of manipulation without awareness and free will. A meta-analysis of all studies (8 studies, N = 1,230) that examined this relationship was conducted to locate where (by context, by individual scenario) the relationship between manipulation outside of awareness and free will was strongest. The findings reveal that the strongest associations were found in scenarios that are most prototypically associated with the unconscious (e.g., sleep research, subliminal priming, subliminal advertising, hypnotherapy).
Article
Full-text available
Findings demonstrating decision-related neural activity preceding volitional actions have dominated the discussion about how science can inform the free will debate. These discussions have largely ignored studies suggesting that decisions might be influenced or biased by various unconscious processes. If these effects are indeed real, do they render subjects’ decisions less free or even unfree? Here, we argue that, while unconscious influences on decision-making do not threaten the existence of free will in general, they provide important information about limitations on freedom in specific circumstances. We demonstrate that aspects of this long-lasting controversy are empirically testable and provide insight into their bearing on degrees of freedom, laying the groundwork for future scientific-philosophical approaches.
Article
Full-text available
The present study examined the role of individual differences (e.g., age, gender, education level, political affiliation, religiosity) and stance (general vs. personal) on contexts associated with manipulations without awareness. In all three studies, people were presented with several real-world contexts. They first rated the extent to which there was manipulation of behavior without awareness, and then provided additional ratings of agentic (e.g., free choice, Conscious Intentions, Conscious Control, responsibility) and affective (Certainty, satisfaction, concern) experiences. Study 1 (N = 222) replicated prior findings: When taking a general stance, the relationship between ratings of manipulation without awareness and ratings of agentic experiences was determined by context. These findings extended to Study 2 (N = 377) and Study 3 (N = 283) where people were asked to take a personal stance, that is, to consider situations of possible manipulation that they themselves have experienced, and provided ratings of their experiences. Across all three studies, people showed remarkable agreement, indicating that individual differences played no substantive role in the patterns of ratings, but stance and context did. People taking a general stance rated Research and Therapy as the most common contexts where they suspected manipulation without awareness, but for those taking a personal stance, Media and Marketing were the most common. The findings are discussed in reference to key theories (e.g., Dynamic monitoring and control theory, Reactance theory, Self-determination theory, Social learning theory) that explain why people place high such a premium on agentic experiences.
Article
Although manipulation is neither rational persuasion nor coercion, a more precise definition remains elusive. Two main accounts have been offered. One characterizes manipulation as a form of trickery. The other characterizes manipulation as a form of non-coercive pressure. Each account properly identifies only a subset of intuitively clear cases of manipulation. That is, some instances of manipulation apparently involve pressure, while others apparently involve trickery. Yet trickery and pressure seem distinct, so it is puzzling how they could be instances of the same phenomenon. This puzzle can be solved once we recognize that manipulative pressure only works if the target acts akratically. This fact provides the basis for a unified account of manipulation as the attempt to get someone to make a mistake.
Article
We know little about the commonality of folk beliefs around applications of psychological research on the unconscious control of behaviours. To address this, in Experiment 1 (N = 399) participants volunteered examples of where research on the unconscious has been applied to influence their behaviours. A subset of these were presented in Experiment 2 (N = 198) and Experiment 3 (N = 100). Participants rated the extent to which the behaviour being influenced in these contexts was: (1) via the unconscious, (2) free, (3) the result of prior conscious intentions, (4) under conscious control. Relative to judgements about the extent to which behaviour was influenced via the unconscious, the remaining judgements regarding conscious control of behaviours were either higher (e.g., political contexts) or lower (e.g., therapy). This study is the first to show, using ecologically valid examples, the folk beliefs people share on psychological constructs concerning free will and determinism.
Book
In recent years, ‘Nudge Units’ or ‘Behavioral Insights Teams’ have been created in the United States, the United Kingdom, Germany, and other nations. All over the world, public officials are using the behavioral sciences to protect the environment, promote employment and economic growth, reduce poverty, and increase national security. In this book, Cass R. Sunstein, the eminent legal scholar and best-selling co-author of Nudge, breaks new ground with a deep yet highly readable investigation into the ethical issues surrounding nudges, choice architecture, and mandates, addressing such issues as welfare, autonomy, self-government, dignity, manipulation, and the constraints and responsibilities of an ethical state. Complementing the ethical discussion, The Ethics of Influence: Government in the Age of Behavioral Science contains a wealth of new data on people’s attitudes towards a broad range of nudges, choice architecture, and mandates.