Consciousness and Cognition 36 (2015) 113130

Contents lists available at ScienceDirect

Consciousness and Cognition

journal homepage: www.elsevier.com/locate/concog

Experimental philosophy of actual and counterfactual free

will intuitions
Adam Feltz
Michigan Technological University, United States

a r t i c l e i n f o a b s t r a c t

Article history: Five experiments suggested that everyday free will and moral responsibility judgments
Received 26 December 2014 about some hypothetical thought examples differed from free will and moral responsibility
Revised 6 June 2015 judgments about the actual world. Experiment 1 (N = 106) showed that free will intuitions
Accepted 8 June 2015
about the actual world measured by the FAD-Plus poorly predicted free will intuitions
Available online 27 June 2015
about a hypothetical person performing a determined action (r = .13). Experiments 25
replicated this result and found the relations between actual free will judgments and free
Keywords:
will judgments about hypothetical determined or fated actions (rs = .22.35) were much
Free will
Moral responsibility
smaller than the differences between them (g2p = .2.55). These results put some pressure
Intuitions on theoretical accounts of everyday intuitions about freedom and moral responsibility.
Fate 2015 Elsevier Inc. All rights reserved.
Determinism
Numeracy

1. Introduction

Many theorists hold that at least a minimal belief in free will is required for us to have healthy relationships with others
and ourselves. Belief in free will may be necessary for autonomy, creativity, desert, reactive attitudes, dignity, love, and friend-
ship (Kane, 1996). Recent advances in psychology and neuroscience may pose some threats to a belief in free will (but see
Mele (2006)). This research suggests that many people appear to have relatively less relevant control over their behavior than
they may think (Bargh & Ferguson, 2000; Libet, 1985; Wegner, 2002; Wegner & Bargh, 1998; Wegner & Wheatley, 1999).
Some worry that if some results from neuroscience were to become widely assimilated, then a belief in free will would be
diminished and the desirable behaviors associated with a belief in free will would also disappear or be dramatically reduced.
For example, in the absence of belief in free will, we may have difculty maintaining meaningful relationships with others and
interpersonal conicts may become more common (Kane, 1996). Empirical research supports these worries to some extent
suggesting that beliefs in free will are linked to judgments about punishment (Carey & Paulhus, 2013; Rakos, Laurene,
Skala, & Slane, 2008). Moreover, belief in free will has been argued to be an important factor for many commonly desirable
behaviors such as refraining from cheating, self-control, and job performance (Baumeister, Masicampo, & DeWall, 2009;
Baumeister, Sparks, Stillman, & Vohs, 2008; Stillman, Baumeister, & Mele, 2011; Vohs & Schooler, 2008). For reasons such
as these, some take it that belief in free will is so important and engrained that if we were to nd out that people really
are not free or morally responsible, we should leave people to their mistaken beliefs (Smilansky, 2002).
But what are peoples attitudes and beliefs about freedom and moral responsibility? Some theorists have extensively
relied on hypothetical thought examples to explore peoples judgments about freedom and moral responsibility (Cova,

Address: 1400 Townsend Drive, Department of Cognitive and Learning Sciences, Michigan Technological University, Houghton, MI 49931, United States.
E-mail address: adfeltz@mtu.edu

http://dx.doi.org/10.1016/j.concog.2015.06.001
1053-8100/ 2015 Elsevier Inc. All rights reserved.
114 A. Feltz / Consciousness and Cognition 36 (2015) 113130

Bertoux, Bourgeois-Gironde, & Dubois, 2012; Nahmias, Morris, Nadelhoffer, & Turner, 2004; Nahmias, Morris, Nadelhoffer, &
Turner, 2005, 2006; Nichols & Knobe, 2007; Sarkissian et al., 2010; Weigel, 2013).1 These thought examples often stipulate a
number of conditions and then probe participants intuitions under those conditions. Other theorists have largely focused on
free will and moral responsibility judgments about the actual world (Carey & Paulhus, 2013; Paulhus & Carey, 2011; Rakos
et al., 2008; Stroessner & Green, 1990; Viney, Waldman, & Barchilon, 1982). One worry is that many people may make similar
judgments about the actual world and a hypothetical world with stipulated conditions. If everyday intuitions about actual cases
do not diverge much from intuitions about hypothetical cases, then the interest in folk intuitions about hypothetical cases may
be diminished. If folk intuitions about actual cases diverge from hypothetical cases, then understanding the extent of that diver-
gence is important to have a complete and accurate understanding of folk theories, concepts, and intuitions about freedom and
moral responsibility.
Five experiments estimated the difference between actual and counterfactual free will judgments. Results suggest that
overall, free will and moral responsibility intuitions about the actual world can be different than free will and moral respon-
sibility intuitions about some hypothetical worlds. First, peoples actual intuitions were weak predictors of their hypothetical
intuitions. Second, the amount of the variance explained by the relation between hypothetical and actual intuitions is sub-
stantially smaller than the variance accounted for by the difference between the two. These results suggest that actual and
hypothetical judgments are not only conceptually different, they are empirically different at least for some people in some
instances. Consequently, only exploring one of actual or hypothetical intuitions is not sufcient to have an accurate picture
of folk intuitions about freedom and moral responsibility.

2. Counterfactual and actual intuitions about free will and moral responsibility

Can one be free and morally responsible if determinism is true?2 There are two prominent answers to this Compatibility
Question. Compatibilists hold that free will and moral responsibility are compatible with determinism. One could be free and
morally responsible if ones conscious mental states (e.g., beliefs, desires, intentions, and plans) are involved in the right way in
the production of an action (along with appropriate background conditions). Those mental states are among the items that
make a person a unique individual (Frankfurt, 1971). If those mental states had been different (i.e., if the person had been dif-
ferent), then a different action could have come about even if determinism is true. Determinism does not rule out this kind of
freedom and moral responsibility because those mental states are just another link in the causal chain that leads to the action.
Incompatibilists do not think that freedom and moral responsibility are compatible with determinism. Some incompatibilists
argue that we cannot be morally responsible for anything that is the complete consequence of things for which we are not
morally responsible. If determinism is true, our actions are the results of events and laws of nature we are not responsible
for (e.g., events and laws that existed before we were born). Therefore, we cannot be morally responsible for our actions
(Strawson, 1994).
Another threat to freedom and moral responsibility is fatalism. Fatalism is the view that whatever happens must hap-
pen (Bernstein, 2002, p. 65). Fatalism rules out some compatibilist styles of freedom and moral responsibility (Nahmias &
Murray, 2010). If an action is fated, then the action must be performed regardless of the past and the laws of nature. There is
nothing that the person could have done not to bring about that action. The persons mental states are in this sense irrelevant
to the production of the action. If the person had different mental states, then the same fated action would have come about.
But the possible differences in ones mental states are precisely the grounds that some compatibilists think can allow for
freedom and moral responsibility. If the persons mental states are irrelevant to the production of the action, then compat-
ibilist style freedom and moral responsibility is ruled out.
Arguments about determinisms relation to free will and moral responsibility are often complicated and nuanced. It is
widely agreed that one important piece of evidence for evaluating those arguments is peoples pre-theoretical intuitions
about freedom and moral responsibility (Sommers, 2010, 2012). Unlike some philosophical pursuits (e.g., logic), philoso-
phers of free will have largely taken themselves to be exploring everyday conceptions of freedom and moral responsibility.
Views that do not take into account everyday intuitions about freedom and moral responsibility run the risk of simply being
irrelevant to the human concepts and practices that have traditionally preoccupied the philosophical study of free will
(Nahmias & Murray, 2010). To put it another way, without being constrained by ordinary thinking about freedom and moral
responsibility, philosophers run the risk of theorizing about philosophical ctions (Mele, 2001).
Understanding everyday intuitions about freedom and moral responsibility is largely an empirical enterprise.3
Unfortunately, there can be some confusion as there are different targets of everyday intuitions about freedom and moral
responsibility. One appropriate target of experimental investigating is peoples intuitions about freedom and moral responsibil-
ity about the actual world. Call these actual intuitions. For example, are people actually free and morally responsible? Is deter-
minism or fatalism actually true? A conceptually distinct set of intuitions that is also an appropriate target for investigation is
peoples intuitions about hypothetical scenarios with stipulated conditions. Call intuitions about these types of hypothetical sit-
uations counterfactual intuitions. For example, can a person be free and morally responsible if determinism or fatalism were

1
Some hold that there is a difference between intuitions and judgments (see Feltz and Bishop (2010) for a review). For the purposes of this paper, there is no
relevant difference.
2
Determinism is the thesis that at any instant exactly one future is compatible with the state of the universe and the laws of nature (Mele, 2006, p. 3).
3
Empirically exploring philosophical issues has been called Experimental Philosophy. See Feltz (2009) for a review.
 A. Feltz / Consciousness and Cognition 36 (2015) 113130 115

true? While it is unknown whether determinism actually is true or false, we can ask people to assume that determinism is true
and then ask them about people in that world. Many philosophers of free will and moral responsibility have favored exploring
counterfactual intuitions whereas many psychologists have favored exploring actual intuitions.4
It is unclear whether and to what degree the conceptual distinction between actual and counterfactual intuitions is
reected in peoples intuitions about freedom and moral responsibility. For this reason, a series of studies was conducted.
The major prediction was that there would be some relation between actual and counterfactual intuitions, but this relation
would be small relative to the difference between those intuitions. There would be a relation between the two sets of intu-
itions because some people would have actual intuitions consistent with the conditions stipulated in the hypothetical situ-
ation or some would view the stipulated conditions as irrelevant to freedom and moral responsibility. Others would have
actual intuitions inconsistent with the stipulated conditions, and this incongruity would change their intuitions about free-
dom and moral responsibility about agents under those conditions. The overall pattern of results from the ve experiments
supported the major hypothesis.

3. Experiment 1

Experiment 1 was designed to determine the relation between peoples counterfactual and actual intuitions about free-
dom and moral responsibility using a standard description of determinism and a standard measure of actual intuitions about
freedom and determinism.

3.1. Participants

One hundred and six participants were recruited from Amazons Mechanical Turk.5 Three participants were excluded from
analyses for not completing the survey. Sixty-ve were male (63%). Ages ranged from 18 to 67, M = 28.39, SD = 9.61.

3.2. Materials

Participants completed the FAD-Plus (Paulhus & Carey, 2011). The FAD-Plus consists of 27 Likert scale items related to
free will (e.g., People have complete free will, 1 = strongly disagree, 5 = strongly agree). The FAD-Plus has four subscales that
are claimed to measure judgments about Free Will, Scientic Determinism, Fatalistic Determinism, and Unpredictability (see
the Appendix A for a slightly modied version of the FAD-Plus). Participants also received the following standard measure of
counterfactual intuitions about determinisms relation to freedom and moral responsibility:
Determinism concrete
Most respected psychologists are convinced that eventually we will gure out exactly how all of our decisions and actions
are entirely caused. For instance, they think that whenever we are trying to decide what to do, the decision we end up
making is completely caused by the specic thoughts, desires, and plans occurring in our minds. The psychologists are
also convinced that these thoughts, desires, and plans are completely caused by our current situation and the earlier
events in our lives, and that these earlier events were also completely caused by even earlier events, eventually going
all the way back to events that occurred before we were born. So, once specic earlier events have occurred in a persons
life, these events will denitely cause specic later events to occur. For example, one day a person named John decides to
kill his wife so that he can marry his lover, and he does it. Once the specic thoughts, desires, and plans occur in Johns
mind, they will denitely cause his decision to kill his wife. Assume the psychologists are right that the events that
occurred earlier in Johns life denitely caused his decision to kill his wife.
Participants then responded to the following standard free will items on a Likert scale (1 = strongly disagree, 7 = strongly
agree).

1. John decided to kill his wife of his own free will.

2. Johns killing of his wife was up to him.
3. John is morally responsible for killing his wife.

The materials were counterbalanced for order and participants were randomly assigned to one order. Previous research
suggested that responses to the Free Will subscale (Paulhus & Carey, 2011) and Concrete Determinism are associated with

4
Some studies have manipulated psychological distance (Roskies & Nichols, 2008; Weigel, 2011). For example, Roskies and Nichols (2008) told one group
participants that most respected scientists think that determinism is true. However, a different group of participants were told that these scientists live in
Universe A. Participants were then asked to assume that the scientists are right and make judgments about whether a person is free and morally responsible
under those conditions. Those making judgments about Universe A were more incompatibilist than those making judgments in the alternative condition.
Critically, both sets of intuitions are counterfactual intuitions since they ask participants to make judgments under some stipulated conditions (i.e., that the
scientists are right about determinism).
5
Research shows that recruiting participants from Amazons Mechanical Turk is just as good, if not better, than other typical convenience samples (e.g.,
universitys subjects pools) (Buhrmester, Kwang, & Gosling, 2011; Crump, McDonnell, & Gureckis, 2013; Mason & Suri, 2012).
116 A. Feltz / Consciousness and Cognition 36 (2015) 113130

the global personality trait extraversion (Cokely & Feltz, 2009; Feltz, 2013; Feltz & Cokely, 2009; Feltz & Millan, 2015; Feltz,
Perez, & Harris, 2012; Schulz, Cokely, & Feltz, 2011). There has been an increased emphasis on replication in psychology, and
this experiment offered an efcient avenue to replicate some previous work concerning personality. So, after responding to
both the FAD-Plus and Determinism Concrete, participants completed the Ten Item Personality Inventory (TIPI) (Gosling,
Rentfrow, & Swann, 2003). Finally, basic demographic information was collected.

3.3. Results

Responses to Determinism Concrete items 1 (M = 5.3, SD = 2.16), 2 (M = 5.61, SD = 1.94) and 3 (M = 6.08, SD = 1.77)
showed excellent internal reliably (a = .9). A Compatibilist Composite score (mean of response to 13) was constructed to
simplify analyses (M = 5.66, SD = 1.79).6 The four subscales from the FAD-Plus were calculated according to Paulhus and
Carey (2011): Free Will (M = 3.82, SD = 0.65; a = .8), Scientic Determinism (M = 3.03, SD = 0.6; a = .67), Fatalistic
Determinism (M = 1.92, SD = 0.8; a = .86), and Unpredictability (M = 3.39, SD = 0.67; a = .78).
An analysis of variance (ANOVA) with order as the independent variable and Compatibilist Composite score as the depen-
dent variable showed a signicant order effect. Those who received the FAD-Plus rst (M = 6.05, SD = 1.28) had stronger com-
patibilist intuitions than those who received Determinism Concrete rst (M = 5.35, SD = 2.1), F(1, 101) = 4.02, p = .05, g2p = .04.
A multivariate analysis of variance (MANOVA) showed no reliable overall effect of order with the FAD-Plus subscales (F < 1).
The Compatibilist Composite score was not reliably predicted by any of the FAD-Plus subscales: FAD Free Will r = .13, FAD
Scientic r = .01, FAD Fatalistic r = .14, and FAD Unpredictable r. = .19, all ps > .05. Because there was an order effect with
the Compatibilist Composite score, four separate multiple linear regressions were conducted. After controlling for order,
none of the FAD-Plus subscale items were reliably related to the Compatibilist Composite score: Free Will: b = .13,
t = 1.35, p = .18, Scientic Determinism b = .01, t = .05, p = .96, Fatalistic Determinism b = .07, t = .76, p = .45, and
Unpredictability b = .11, t = 1.12, p = .27.
Previous research has suggested that extraversion predicts compatibilist judgments (Andow & Cova, in press; Feltz, 2013;
Feltz & Cokely, 2009, 2013; Feltz & Millan, 2015; Feltz et al., 2012; Schulz et al., 2011). Since order inuenced the
Compatibilist Composite score, a multiple linear regression with order of presentation and extraversion as predictor vari-
ables and Compatibilist Composite score as dependent variable was conducted. The full model was a signicant predictor
of Free Will Composite score, F(2, 100) = 4.23, p = .02, R2 = .08. Order of presentation b = .2, t = 2.14, p = .04 and extraversion
b = .2, t = 2.07, p = .04, R2change = .08 were unique predictors of the Compatibilist Composite score. No other personality vari-
able was reliably related to compatibilist judgments.
In sum, one standard instrument used to measure actual intuitions about freedom and moral responsibility (the
FAD-Plus) did not reliably predict counterfactual free will and moral responsibility intuitions about another standard instru-
ment (Determinism Concrete). The zero-order correlations were not statistically signicant between judgments about the
FAD-Plus and Determinism concrete. Regression analyses indicated that even after controlling for an observed order effect,
the FAD-Plus still did not predict intuitions about Determinism Concrete. Replicating previous work, extraversion predicted
compatibilist judgments.

4. Experiment 2

One worry with Experiment 1 is that the materials used to measure free will intuitions are substantially different. The
wording of the questions was not similar making direct comparisons between actual and counterfactual intuitions difcult.
Experiment 2 attempted to control for these confounds and to allow for more direct comparisons between counterfactual
and actual intuitions.

4.1. Participants

One hundred and six participants were recruited from Amazons Mechanical Turk. Seven participants were excluded for
not completing the survey. One participant was excluded for requesting the answers not be used. Fifty-ve were (56%) were
male. Ages ranged from 18 to 66, M = 33.94, SD = 11.17.

4.2. Materials

Participants completed the FAD-Plus. Participants also read the following scenario (Nichols & Knobe, 2007)7:
Fries concrete
Imagine a universe (Universe A) in which everything that happens is completely caused by whatever happened before it.
This is true from the very beginning of the universe, so what happened in the beginning of the universe caused what hap-
pened next, and so on right up until the present. For example, one day John decided to have French Fries at lunch. Like

6
Using free will composite scores is a standard practice (Feltz & Cokely, 2009; Feltz et al., 2012; Nahmias, Coates, & Kvaran, 2006; Nahmias & Murray, 2010).
7
This scenario has been shown to measure the same construct as Nahmias, Coates, et al.s (2006) scenarios (Feltz & Millan, 2015).
 A. Feltz / Consciousness and Cognition 36 (2015) 113130 117

everything else, this decision was completely caused by what happened before it. So, if everything in this universe was
exactly the same up until John made his decision, then it had to happen that John would decide to have French Fries.
In Universe A, a man named Bill has become attracted to his secretary, and he decides that the only way to be with her is
to kill his wife and 3 children. He knows that it is impossible to escape from his house in the event of a re. Before he
leaves on a business trip, he sets up a device in his basement that burns down the house and kills his family.
Participants were instructed to Please rate your agreement with the following statements about Universe A on a
ve-point scale (1 = strongly disagree, 5 = strongly agree). Participants then responded to 27 modestly modied statements
based on the FAD-plus (see Appendix A). These modications were limited to changes that made the statements consistent
with the scenario.
Numeracy was also measured. Numeracy refers to the general ability to understand and use probabilities and has been
linked to general intelligence, a more reective cognitive style, greater attention and deliberation, and some more normatively
correct answers (Cokely, Galesic, Schulz, Ghazal, & Garcia-Retamero, 2012; Peters, 2012; Peters et al., 2006). Those who are
more numerate may more fully understand key features of scenarios (e.g., to determinism) and may thereby have different
judgments than those who are less numerate. Numeracy was measured with the Berlin Numeracy Test (BNT) (Cokely et al.,
2012). Finally, participants completed the TIPI and basic demographic information was gathered.

4.3. Results

To determine if a similar factor structure existed for Fries Concrete and the FAD-Plus, two factor analyses were conducted.
These analyses restricted the number of factors to four because there were a priori grounds for this factor structure (see
Tables 1 and 2).
The factor analyses suggested that the same general factors emerged for actual and counterfactual judgments. Fries
Concrete items 21, 12, and 4 that are normally part of the Free Will subscale had the strongest loadings on the Fatalistic
Determinism subscale. However, there were strong cross loadings for those items on the Free Will subscale (>.31) and
expected negative loadings on the Fatalism subscale. Despite these minor differences, the internal reliability of the scales
remained acceptable to excellent and consistent with previous reports of the scales internal reliability (Paulhus & Carey,
2011).
Means and standard errors are reported in Fig. 1. Correlations among dependent variables are reported in Table 3. To
determine whether judgments about the FAD-Plus and Fries Concrete subscale items differed, a mixed-model ANOVA with
subscale items as within-participants factors and order of presentation as between-subjects factors was conducted. There
was an overall difference between judgments about FAD-Plus and Fries Concrete F(4, 93) = 8.31, p < .001, g2p = .26. Order
did not reliably interact with judgments F < 1. There were signicant differences between Fries Concrete Free Will
(a = .92) and Fad-Plus Free Will (a = .82) F(1, 96) = 23.64, p < .001, g2p = .2, Fries Concrete Fatalism (a = .91) and Fad-Plus
Fatalism (a = .81) F(1, 96) = 23.66, p < .001, g2p = .2, Fries Concrete Unpredictable (a = .82) and Fad-Plus Unpredictable
(a = .79) F(1, 96) = 15.61, p < .001, g2p = .14, and a trend in the right direction for a difference between Fries Concrete
Determinism (a = .65) and Fad-Plus Determinism (a = .66) F(1, 96) = 2.78, p = .1, g2p = .03.8
To summarize, Experiment 2 suggested that the factor structures of judgments about the FAD-Plus and Fries concrete
were similar so comparisons were legitimate. While there was some relation between actual and counterfactual free will
intuitions about Fries Concrete, the relation is smaller than the difference between the two judgments. Zero order correla-
tions suggested that the FAD-Plus factors were modest predictors of counterfactual free will judgments (rs < .26).
Importantly, the relation between free will judgments was much smaller than the difference (g2p = .2). Relations between
judgments about the nature of the worlds were stronger. For example, FAD-Plus Scientic Determinism and
Unpredictability were good predictors of their counterfactual counterparts (rs = .44) and there were no reliable differences
between those judgments. However, there was no reliable relation between the Fatalistic Determinism judgments.

4.4. Experiment 3

Experiment 3 was the next in the planned series of experiments. Experiments 1 and 2 measured counterfactual intuitions
about concretely described persons John and Bill. One could omit descriptions of a person acting in a deterministic universe
and simply ask the abstract question if some unidentied individual could be free and morally responsible in a deterministic
universe. There is substantial evidence that counterfactual intuitions about freedom and moral responsibility vary as a func-
tion of concretely or abstractly described individuals (Cova et al., 2012; Nichols & Knobe, 2007). People tend to have more
compatibilist friendly intuitions in the concrete frame compared to the abstract frame. The FAD-Plus may measure most
directly abstract actual intuitions about freedom and moral responsibility. If the FAD-Plus measures abstract intuitions,
the FAD-Plus should be a good predictor of intuitions about abstract counterfactual intuitions.

8
Even though order did not reliably inuence judgments, one may think that if these items were read in isolation, there would be no appreciable differences
in responses. Analyses of rst responses were conducted revealing largely the same pattern of responses. Unless otherwise noted, order did not reliably interact
with the overall patterns of response. Additionally, this experiment failed to replicate the relation of extraversion with counterfactual free will judgments.
However, agreeableness was related to counterfactual free will judgments r = .24, p < .05. Consciousness was related to FAD Free Will judgments r = .3, p > .01.
No other personality traits were related to counterfactual or actual judgments.
118 A. Feltz / Consciousness and Cognition 36 (2015) 113130

Table 1
Pattern matrix for fries concrete.

Item # Free will Scientic determinism Fatalistic determinism Unpredictability

8 .95 .13 .02 .05
23 .87 .06 .07 .08
16 .87 0 .03 .03
26 .53 .16 .33 .11
10 .18 .62 .16 .03
22 .01 .5 .21 .06
2 .15 .44 .04 .01
6 .07 .43 .07 .01
14 .01 .41 .19 .19
24 .13 .37 .07 .12
18 .1 .35 .35 .11
5 .04 0 .92 .06
1 .03 .03 .9 .05
13 .02 .05 .82 .02
9 .07 .17 .79 .15
21 (FW) .36 .27 .54 .02
12 (FW) .31 .25 .53 .03
17 .34 .15 .52 .05
4 (FW) .32 .1 .43 .02
11 .07 .02 .04 .84
19 .04 .09 .04 .83
7 .14 .08 .15 .57
3 .1 .05 .19 .51
25 .04 .01 .19 .51
20 .17 .15 .09 .5
15 .31 .04 .13 .45
27 .01 .2 .31 .43

Maximum likelihood extraction with Oblimin rotation. Accounts for 51.21% of the total variance. Bold indicates items loading on that factor.

Table 2
Pattern Matrix for FAD-Plus in Experiment 2.

Item # Free will Scientic determinism Fatalistic determinism Unpredictability

4 .72 .11 .14 .21
8 .71 .04 .13 .14
21 .69 .1 .03 .03
23 .68 .18 .01 .03
16 .57 .22 .06 .04
12 .2 .11 .08 0
26 .5 .02 .04 .21
10 .1 .66 .25 .08
14 .11 .55 .05 .17
2 .12 .47 .04 .29
24 .09 .38 .04 .19
22 .12 .35 .06 .08
6 .01 .32 .08 .07
18 .02 .27 .02 .14
9 .12 .11 .9 .07
5 .07 .12 .78 .06
1 .03 .12 .74 .05
13 .28 .17 .49 .37
17 .13 .02 .4 .07
25 .2 .18 .05 .68
3 .07 .28 .03 .67
11 .04 .15 .19 .63
19 .12 .14 .11 .6
20 .07 .02 .01 .58
15 .25 .13 .02 .25
7 .36 .01 .07 .53
27 .16 .06 .12 .28

Maximum likelihood extraction with Oblimin rotation. Accounts for 40.52% of the total variance. Bold indicates items loading on that factor.

4.5. Participants

One hundred and nineteen participants were recruited from Amazons Mechanical Turk. Seventeen participants were
excluded for not completing the survey. One participants responses were excluded because the data were requested not
to be used. Forty-seven (47%) were male. Ages ranged from 18 to 68, M = 35.27, SD = 13.
 A. Feltz / Consciousness and Cognition 36 (2015) 113130 119

Table 3
Correlations for concrete determinism and FAD-Plus in experiment 2.

1 2 3 4 5 6 7 8
1. Determinism concrete free will 1 .02 .7** .3** .22* .2* .25* .15
2. Determinism concrete science 1 .16 .13 .02 .44** .13 .24*
3. Determinism Concrete Fatalism 1 .19 .02 .01 .06 .04
4. Determinism concrete unpredict 1 .17 .24* .25* .44**
5. FAD free will 1 .06 .17 .14
6. FAD science 1 .29** .23*
7. FAD fatalism 1 .28**
8. FAD unpredict 1
*
p < .05.
**
p < .01.

Table 4
Pattern matrix for fries abstract.

Item # Free will Scientic determinism Fatalistic determinism Unpredictability

16 .89 .03 .08 .04
8 .87 .03 .03 .03
23 .71 .03 .15 .05
21 .54 .07 .07 .32
4 .52 .02 .21 .21
12 .48 .07 .19 .19
26 .48 .21 .17 .16
2 .14 .66 .09 .14
10 .04 .62 .16 .01
22 .05 .6 .14 .07
24 .27 .52 .07 0
18 .12 .47 .05 .04
14 .21 .42 .31 .11
6 .17 .32 .32 .29
9 .03 .01 .92 0
1 .03 .04 .8 .12
17 .02 .03 .62 .23
5 .23 .04 .62 .02
13 .33 .06 .51 .1
11 .08 .03 .02 .83
19 .14 .01 .04 .73
27 .08 .1 .21 .65
3 .01 .17 .14 .61
7 .29 .05 .03 .59
15 .33 .19 .04 .53
20 .24 .06 .01 .42
25 .09 0 .07 .41

Maximum likelihood extraction with Oblimin rotation. Accounts for 51.7% of the total variance. Bold indicates items loading on that factor.

4.6. Materials

Participants received the FAD-Plus and Fries Abstract in one of two counterbalanced orders. Fries Abstract is identical to
Fries Concrete except the second paragraph was deleted. After reading Fries Abstract, participants were instructed to Please
rate your agreement with the following statements about Universe A and then were given the exact items on the FAD-Plus.
Participants then completed the BNT, the TIPI, and basic demographic information was gathered.

4.7. Results

To determine if a similar factor structure existed for Fries Abstract and the FAD-Plus, two factor analyses proceeded in the
same fashion as in Experiment 2 (see Tables 4 and 5).
The factor analyses suggested that similar factors emerged for actual and counterfactual intuitions. Item 4 of the Scientic
Determinism subscale of the FAD-Plus had the strongest loading on the Free Will scale. The loading on the Free Will subscale
was negative and there was substantial loading on the Scientic Determinism subscale (>.35).
Some support for the concrete/abstract difference in counterfactual intuitions was found. Compared to responses to Fries
Concrete from Experiment 2, a MANOVA indicated an overall difference F(4, 192) = 3.55, p = .008, g2p = .07. Order of presen-
tation did not reliably interact with judgments F(4, 192) = 1.04, p = .39, g2p = .02. ANOVAs found trends for differences
120 A. Feltz / Consciousness and Cognition 36 (2015) 113130

Fig. 1. Experiment 2 means and standard errors.

Table 5
Factor matrix for FAD-Plus in experiment 3.

Item # Free will Scientic determinism Fatalistic determinism Unpredictability

8 .75 .16 .02 .05
12 .7 .01 .03 .11
16 .67 .02 .09 .08
23 .6 .18 .08 .17
26 .56 .1 .1 .02
4 .51 .13 .07 .06
21 .48 .1 .03 .2
10 (SD) .4 .35 .29 .12
24 .09 .81 .13 .09
14 .07 .56 .02 .03
18 .11 .56 .15 .02
22 .13 .48 .08 .02
2 .19 .37 .17 .29
6 .09 .37 .06 .06
9 .02 .03 1 .02
1 .01 .02 .7 .06
17 .23 .02 .57 .03
5 .13 .01 .43 .18
19 .05 .04 .02 .7
11 .1 .01 .15 .64
15 .15 .19 .18 .6
25 .05 .09 .21 .58
27 .17 .13 .09 .55
7 .2 .09 .09 .53
3 .17 .18 .1 .47
13 (FD) .32 .24 .07 .41
20 0 .31 .09 .4

Maximum likelihood extraction with Oblimin rotation. Accounts for 41.24% of the total variance. Bold indicates items loading on that factor.

between Free Will F(1, 195) = 2.67, p = .1, g2p = .01, Scientic Determinism F(1, 195) = 3.42, p = .07, g2p = .02, but not for
Fatalistic Determinism F < 1 or Unpredictability F(1, 195) = 1.75, p = .19, g2p = .01.
Means and standard errors are reported in Fig. 2, and correlations among dependent variables are reported in Table 6 A
MANOVA found no reliable overall order effect for the Fad-Plus items F(1, 192) = 1.38, p = .24, g2p = .03. To determine whether
judgments about the FAD-Plus subscale items signicantly differed from the Fries Abstract subscale items, a mixed-model
ANOVA with subscale items as within-participants factors and order of presentation as between-participants factors was con-
ducted. There was an overall difference between judgments about FAD-Plus and Fries Abstract F(4, 96) = 13.68, p < .001,
g2p = .36. Order did not reliably interact with judgments F(4, 96) = 2.08, p = .09, g2p = .08. There were signicant differences
between sub-scale items Fries Abstract Free Will (a = .92) and Fad-Plus Free Will (a = .81) F(1, 99) = 42.64, p < .001, g2p = .3,
Fries Abstract Scientic Determinism (a = .74), Fad-Plus Scientic Determinism (a = .72) F(1, 99) = 7.55, p = .007, g2p = .07,
 A. Feltz / Consciousness and Cognition 36 (2015) 113130 121

Table 6
Correlations in Experiment 3 for determinism abstract and FAD-Plus subscale items.

1 2 3 4 5 6 7 8
1. Determinism abstract free will 1 .03 .59** .62** .35** .11 .15 .24*
2. Determinism abstract science 1 .31** .07 .05 .41** .14 .04
3. Determinism abstract fatalism 1 .3** .05 .06 .19 .02
4. Determinism abstract unpredict 1 .16 .1 .14 .57**
5. FAD free will 1 .01 .08 .21*
6. FAD science 1 .34** .18
7. FAD fatalism 1 .36**
8. FAD unpredict 1
*
p < .05.
**
p < .01.

Table 7
Pattern matrix for fatalism concrete.

Item # Free will Scientic determinism Fatalistic determinism Unpredictability

12 .85 .1 .09 .08
26 .53 .22 .11 .24
21 .37 .33 .21 .17
23 (FW) 0 .92 .03 .01
16 (FW) .1 .79 .07 .1
8 (FW) 0 .73 .11 0
18 .04 .58 .14 .1
14 .06 .5 .02 .18
24 .17 .48 .03 .29
22 .18 .43 .07 .24
2 .15 .38 .12 .12
10 .1 .29 .18 .07
6 .19 .27 .18 .14
9 .06 .08 .88 .15
5 .16 .14 .8 .03
13 .22 .03 .82 .09
1 .15 .02 .76 .03
17 0 .04 .7 .05
4 (FW) .33 .21 .39 .07
19 .12 .01 .04 .82
20 .03 .09 .14 .75
7 .04 .02 .13 .7
11 .13 .08 .01 .68
25 .24 .24 0 .65
27 .02 .26 .09 .59
15 .15 .28 .08 .38
3 .09 .24 .06 .34

Maximum likelihood extraction with Oblimin rotation. Accounts for 57.15% of the total variance. Bold indicates items loading on that factor.

Fries Abstract Fatalistic Determinism (a = .88) and Fad-Plus Fatalistic Determinism (a = .75) F(1, 99) = 49.45, p < .001, g2p = .33,
and Fries Abstract Unpredictability (a = .86) and Fad-Plus Unpredictability (a = .79) F(1, 99) = 7.93, p = .006, g2p = .07.9
Consequently, Experiment 3 largely reproduced previous results. Again, there were similar factors structures of judg-
ments about the FAD-Plus and Fries Abstract. Analyses indicated additional evidence for the concrete/abstract difference
in counterfactual intuitions about determinism. The relation between free will judgments was moderate (r = .35) but again
the difference between the judgments was larger (g2p = .3). While there was no relation between actual and counterfactual
Fatalism Determinism judgments, the relations between actual and counterfactual Scientic Determinism (r = .41) and

9
Extraversion predicted Fries Abstract Free Will judgments r = .26, p = .009. Conscientiousness (r = .33, p = .001) and openness to experience (r = .23, p = .02)
predicted FAD Free Will judgments. Linear regressions indicated that BNT was signicantly related to the Fries Abstract Unpredictability R = .23, F(1, 99) = 5.28,
p = .02 and to Fries Abstract Free Will R = .27, F(1, 99) = 7.54, p = .007. This suggested a possible mediation relation between the variables (Baron & Kenny, 1986).
A multiple linear regression with BNT and Fries Abstract Unpredictability as predictor variables and Fries Abstract Free Will as the outcome variable was
conducted. The full model was a signicant predictor of the Fries Abstract Free Will R = .64, F(2, 98) = 33.16, p < .001. Fries Abstract Unpredictabilitys relation
with Fries Abstract Free Will remained signicant, b = .59, t = 7.4, p < .001, but the relation between BNT and Fries Abstract Free Will was not b = .13, t = 1.66,
p = .10. The difference between the direct effect of BNT and the Fries Abstract Free Will and the mediated effect was signicant (Sobel test = 2.18, p = .03).
Unstandardized indirect effects were computed for 10,000 bootstrapped samples. At the 95% condence interval, the indirect effect was .12, SE = .06,
LLCI = .26, ULCI = .03. Two different models were tested in path models. These models were roughly statistically equivalent. BNT ? Unpredictability ? Free
Will: v2 = 2.8, p = .1, RMSEA = .13 (90% CI 0.33), pclose = .14, AIC = 834, BIC = 850, CFI = .97, TLI = .9. The alternative model was BNT ? Free Will ? Unpre-
dictability: v2 = 0.63, p = .42, RMSEA = 0 (90% CI 0.24), pclose = .48, AIC = 832, BIC = 848, CFI = 1, TLI = 1.
122 A. Feltz / Consciousness and Cognition 36 (2015) 113130

Table 8
Pattern matrix for FAD-Plus in Experiment 4.

Item # Free will Scientic determinism Fatalistic determinism Unpredictability

12 .76 .24 .04 .08
21 .66 .02 .09 .06
8 .6 .23 .06 .08
4 .56 .01 .09 .1
26 .54 .09 .25 .13
16 .52 .2 .27 .16
23 .46 .21 .09 .17
18 (SD) .21 .18 .01 .01
6 .08 .43 .04 .19
24 .14 .43 .01 .05
14 .27 .42 .08 .29
22 .28 .41 .05 .13
10 .06 .36 .24 .05
9 .01 .11 .94 .02
1 .02 .03 .85 .09
17 .22 .04 .69 0
5 .14 .35 .59 0
13 .23 .14 .58 .3
2 (SD) .11 .15 .21 .09
19 .03 .13 .09 .72
27 .01 .16 .1 .72
11 .12 .03 .08 .67
25 .1 .25 .1 .67
7 .12 .14 .18 .64
3 .02 .12 .12 .61
15 .21 .28 .06 .56
20 .13 .21 .04 .51

Maximum likelihood extraction with Oblimin rotation. Accounts for 42.57% of the total variance. Bold indicates items loading on that factor.

Fig. 2. Experiment 3 means and standard errors.

Unpredictability (r = .57) judgments were strong. Unlike Experiment 2, there were reliable differences between all counter-
factual and actual intuitions.

5. Experiment 4

Determinism constitutes one threat to freedom and moral responsibility. Fatalism is another threat. Overall, people tend
to judge a person who performs a fated action as less free and morally responsible than when a person performs a similar
determined action (Feltz & Millan, 2015). Fatalism scenarios were used because one of the subscales of the FAD-plus is said
to measure fatalistic determinism. Since this subscale is supposed to measure actual fatalistic intuitions, it could predict
counterfactual intuitions about freedom and moral responsibility in fated worlds. Experiments 4 and 5 were designed to test
whether actual intuitions predicted counterfactual intuitions about fated actions.
 A. Feltz / Consciousness and Cognition 36 (2015) 113130 123

5.1. Participants

One hundred and eighteen participants were recruited from Amazons Mechanical Turk. Eleven were excluded for not
completing the survey. One persons data were excluded because the participant requested the data not be used.
Forty-ve (43%) were male. Ages ranged from 18 to 70, M = 34.15, SD = 12.69.

5.2. Materials

Participants received the FAD-Plus and the following scenario in one of two counterbalanced orders:
Fatalism concrete
Imagine a universe, (Universe A) where there is a special book that has all of our decisions and actions truly written in its
content. For instance, whenever we are trying to decide what to do, the decision we end up making is completely and
truly written in this book and the decision will happen regardless of our thoughts, beliefs, desires, and plans. The special
book has these events truly written in it lifetimes before the events took place.
So, if the book has an event written in it, the event will denitely occur regardless of the past events and the laws of nat-
ure. For example, one day a person named Bill decides to kill his wife so that he can marry his lover, and he does it. Once
the specic event is truly written in the book, it is impossible for Bill not to kill his wife regardless of his thoughts, beliefs,
desires, and plans.
Assume the books contents made it impossible for Bill not to kill his wife.
Participants were then instructed to Please rate your agreement with the following statements about Universe A.
Participants were given the same modied FAD-Plus items used in Experiment 2 (see Appendix A). Participants then com-
pleted the BNT, the TIPI, and basic demographic information was collected.

5.3. Results

To determine if a similar factor structure existed for Fatalism Concrete and the FAD-Plus, two factor analyses were con-
ducted in the same way as in Experiments 2 and 3 (see Tables 7 and 8).
For Fatalism Concrete, items 23, 16, and 8 of the Free Will scale had the strongest loading on the Scientic Determinism
subscale. These items had minimal loadings on the Free Will subscale. One reason for the clustering of these three items is
that all three dealt with attributions of moral responsibility that could be importantly different from attributions of freedom
for fated actions. Item 4 of the Free Will subscale had the strongest loading on the Fatalistic Determinism subscale. That load-
ing was negative and the item had substantial loading on the Free Will subscale (>.33). Still, each of these subscales had
excellent internal reliability (see below).
Previous research found that there is a difference in judgments between fated and determined actions (Feltz & Millan,
2015). A MANOVA with Determinism Concrete (from Experiment 2) and Fatalism Concrete as independent variables and
the four subscales as dependent variables indicated an overall difference in judgments F(4, 197) = 10.79, p < .001, g2p = .18.
There was no main effect of order and order did not interact with judgments (Fs < 1). Subsequent ANOVAs revealed differ-
ences with each of the dependent variables: Free Will F(1, 197) = 33.45, p < .001, g2p = .14, Scientic Determinism
F(1, 197) = 18.85, p < .001, g2p = .09, Fatalistic Determinism F(1, 197) = 23.77, p < .001, g2p = .11, and Unpredictability
F(1, 197) = 20.01, p < .001, g2p = .09.
Means and standard errors are reported in Fig. 3, and correlations among the dependent variables are reported in Table 9.
10
A mixed-model ANOVA indicated an overall difference between judgments about FAD-Plus and Fatalism Concrete
F(4, 101) = 37.63, p < .001, g2p = .6. Order did not reliably interact with judgments F < 1. There were signicant differences
between sub-scale items for Fatalism Concrete Free Will (a = .93) and FAD-Plus Free Will (a = .8) F(1, 104) = 127.74, p < .001,
g2p = .55, Fatalism Concrete Scientic Determinism (a = .81), FAD-Plus Scientic Determinism (a = .6) F(1, 104) = 38.88,
p = .007, g2p = .28, Fatalism Concrete Fatalistic Determinism (a = .91) and FAD-Plus Fatalistic Determinism (a = .86)
F(1, 104) = 482.51, p < .001, g2p = .44, and Fatalism Concrete Unpredictability (a = .88) and FAD-Plus Unpredictability (a = .84)
F(1, 104) = 93.36, p < .001, g2p = .47.11

10
No personality traits were reliably related to Fatalism Concrete Free will judgments. However, agreeableness (r = .25, p = .01), conscientiousness (r = .41,
p < .001), and openness to experience (r = .24, p = .02) were related to FAD Free Will judgments.
11
Linear regressions indicated that BNT was signicantly related to Fatalism Concrete Unpredictability R = .25, F(1, 104) = 7.14, p = .009 and to the Fatalism
Concrete Free Will R = .25, F(1, 104) = 6.67, p = .01. A multiple linear regression with BNT and Fatalism Concrete Unpredictability scores as predictor variables
and the Fatalism Concrete Free Will subscale score as the outcome variable was conducted. The full model was a signicant predictor of the Fatalism Concrete
Free Will R = .65, F(2, 103) = 38.67, p < .001. The Fatalism Concrete Unpredictability scores relation with the Fatalism Concrete Free Will subscale score
remained signicant, b = .63, t = 8.15, p < .001 but the relation between BNT and Fatalism Concrete Free Will subscale score was not b = .09, t = 1.21, p = .27.
The mediated effect was signicant (Sobel test = 2.52, p = .01). Unstandardized indirect effects were computed for 10,000 bootstrapped samples. At the 95%
condence interval, the indirect effect was .18, SE = .07, LLCI = .34, ULCI = .06. suggesting a fully mediated relation. Two different models were tested in
path models. These models were roughly statistically equivalent. BNT ? Unpredictability ? Free Will: v2 = 1.29, p = .26, RMSEA = .05 (90% CI 0.27),
pclose = .32, AIC = 847, BIC = 863, CFI = .995, TLI = .986. The alternative model was BNT ? Free Will ? Unpredictability: v2 = 1.74, p = .18, RMSEA = .08 (90% CI
0.28), pclose = .24, AIC = 847, BIC = 863, CFI = .988, TLI = .965.
124 A. Feltz / Consciousness and Cognition 36 (2015) 113130

Fig. 3. Experiment 4 means and standard errors.

Table 9
Correlations in Experiment 4 for fatalism concrete and FAD-Plus subscale items.

1 2 3 4 5 6 7 8
1. Fatalism concrete free will 1 .74** .62** .65** .16 .07 .31* .02
2. Fatalism concrete science 1 .38** .61** .02 .23* .37** .08
3. Fatalism concrete fatalism 1 .41** .05 .15 .01 .11
4. Fatalism concrete unpredict 1 .05 .09 .18 .13
5. FAD free will 1 .12 0 .3**
6. FAD science 1 .19 .3**
7. FAD fatalism 1 .09
8. FAD unpredict 1
*
p < .05.
**
p < .01.

To summarize, the patterns of intuitions in Experiment 4 largely mirrored the pattern of intuitions in the previous 2
experiments. The factor structure of judgments about the FAD-Plus and intuitions about Fatalism concrete were similar.
Comparisons between Fries Concrete and Fatalism Concrete again found difference in counterfactual intuitions between
fated and determined actions. There was no reliable relation between free will judgments concerning the FAD-Plus and
Fries concrete, and the difference between the two sets of judgments was large, g2p = .55. The only reliable relation concern-
ing the nature of the world was between Scientic Determinism judgments (r = .23), but again the difference between actual
and counterfactual Scientic Determinism intuitions was large g2p = .28.

6. Experiment 5

The nal experiment in the planned series gathered judgments about abstractly described fated actions and compared
them to judgments about the FAD-Plus. The main prediction for this experiment remained. While there would be some rela-
tion between actual and counterfactual intuitions, that relation would be modest compared to the difference between the
intuitions. Experiment 5 also had an additional benet of being able to determine whether the pervasive abstract/concrete
difference concerning counterfactual intuitions would be found not only in deterministic scenarios but also in fatalistic
scenarios.

6.1. Participants

One hundred and twelve participants were recruited from Amazons Mechanical Turk. Sixteen participants were excluded
for not completing the survey. Three participants were excluded for requesting their answers not be used. Forty (43%) were
male. Ages ranged from 19 to 73, M = 34.54, SD = 11.76.

6.2. Materials

Participants received the FAD-Plus and Fatalism Abstract. Fatalism Abstract was identical to Fatalism Concrete except the
second paragraph was deleted. The order of presentation was counterbalanced and participants were randomly assigned to
 A. Feltz / Consciousness and Cognition 36 (2015) 113130 125

Table 10
Pattern matrix for fatalism abstract.

Item # Free will Scientic determinism Fatalistic determinism Unpredictability

26 .89 .02 .03 .03
23 .72 .04 .11 .04
22 (SD) .68 .18 .05 .08
16 .63 .08 .13 .12
12 .63 .07 .02 .26
24 (SD) .6 .24 .11 .08
21 .57 .01 .32 .12
8 .55 .06 .28 .12
4 .45 0 .26 .18
10 .01 .93 .03 .06
2 .4 .53 .09 .06
14 .19 .41 .01 .2
6 .08 .4 .09 .04
18 .09 .26 .14 .24
1 .13 .01 .99 .05
9 .02 .01 .85 .02
17 .1 .02 .73 .02
5 .21 0 .71 .03
13 .4 .01 .58 .1
19 .11 .14 .1 .81
11 .03 .09 .12 .74
20 .21 .39 .01 .65
15 .36 .03 .02 .58
25 .04 .27 0 .55
27 .27 .12 .14 .5
3 .15 .22 .01 .49
7 .18 .08 .31 .44

Maximum likelihood extraction with Oblimin rotation. Accounts for 58.75% of the total variance. Bold indicates items loading on that factor.

Table 11
Pattern matrix for FAD-Plus in Experiment 5.

Item # Free will Scientic determinism Fatalistic determinism Unpredictability

23 .75 .1 .04 .06
16 .7 .01 .02 .11
4 .65 .1 .03 .01
8 .62 .21 .1 .08
21 .46 .19 .12 0
26 .45 .05 .19 .17
12 .42 .1 .17 .15
14 .26 .33 .17 .22
10 .13 .8 .01 .12
2 .15 .55 .17 .06
22 .17 .47 .03 .05
24 .14 .47 .01 .05
3 (U) .13 .43 .1 .34
6 .05 .42 .08 .07
9 .01 .07 .92 .04
1 .04 .11 .91 .03
5 .05 .09 .63 .08
17 .15 .04 .43 .07
18 (SD) .21 .17 .24 .1
19 .03 .01 .01 .68
15 .01 .18 .01 .63
11 .07 .14 0 .6
20 .17 .23 .17 .55
25 .2 .21 .04 .5
7 .02 .03 .08 .49
13 (FD) .05 .06 .37 .46
27 .17 .21 .18 .4

Maximum likelihood extraction with Oblimin rotation. Accounts for 39.8% of the total variance. Bold indicates items loading on that factor.

one of the orders. After reading Fatalism Concrete, participants were instructed to Please rate your agreement with the fol-
lowing statements about Universe A and then were given the exact items on the FAD-Plus. Then participants completed the
BNT, the TIPI, and basic demographic information was collected.
126 A. Feltz / Consciousness and Cognition 36 (2015) 113130

6.3. Results

Factors analyses revealed similar factor structures for Fatalism Abstract and the FAD-Plus (see Tables 10 and 11).
For the FAD-Plus, Unpredictability item 3 had the strongest loading on Scientic Determinism, Scientic Determinism
item 18 had the strongest loading on Fatalistic determinism, and Fatalistic Determinism item 13 had the strongest loading
on Unpredictability. However, items 3 and 13 still had substantial loadings on the appropriate factors.
The internal validity of each subscale item was good (see below). A MANOVA indicated an overall difference between
Fatalism Abstract and Determinism Abstract from Experiment 3 F(4, 187) = 4.62, p = .001, g2p = .09. Order did not interact with
judgments F < 1. Subsequent ANOVAs revealed that the major factor in this differences was the Scientic Determinism sub-
scale F(1, 190) = 16.45, p < .001, g2p = .08. Judgments about the other sub-scales were not reliably different: Free Will
F(1, 190) = 1.47, p = .23, g2p = .01, Fatalistic Determinism F(1, 190) = 1.2, p = .28, g2p = .01, Unpredictability F(1, 190) = 1.81,
p = .18, g2p = .01.
There was an abstract-concrete difference with respect to responses to the fatalism scenario: MANOVA F(4, 192) = 5.93,
p < .001, g2p = .11. Subsequent ANOVAs found signicant differences between all dependent variables: Free Will
F(1, 195) = 11.6, p < .001, g2p = .06, Scientic Determinism F(1, 195) = 4.67, p = .03, g2p = .02, Fatalistic Determinism
F(1, 195) = 14.7, p < .001, g2p = .07, and Unpredictability F(1, 195) = 20.02, p < .001, g2p = .09. This overall effect was qualied
by an interaction of order F(4, 192) = 2.94, p = .02, g2p = .06. ANOVAs revealed that the only reliable interaction occurred with
the Fatalistic Determinism subscale F(1, 195) = 9.0, p = .003, g2p = .04. In this case, when the FAD-Plus was presented rst,
Judgments about Fatalism Abstract Fatalistic Determinism (M = 3.63, SD = 1.12) and Fatalism Concrete Fatalistic
Determinism (M = 3.76, SD = 1.24) were much closer than when FAD-Plus was presented second (Fatalism Abstract
Fatalistic Determinism M = 2.91, SD = 1.33; Fatalism Concrete Fatalistic Determinism M = 4.03, SD = 0.93).
Means and standard errors are reported in Fig. 4, and correlations among the dependent variables are reported in
Table 12.12 A mixed-model ANOVA with subscale items as within-participants factors and order of presentation as
between-participants factors was conducted. There was an overall difference between the FAD-Plus and Fatalism Abstract
F(4, 88) = 11.04, p < .001, g2p = .33. Order did not reliably interact with judgments F(4, 88) = 1.84, p = .13, g2p = .08. There were sig-
nicant differences between Fatalism Abstract Free Will (a = .92) and FAD-Plus Free Will (a = .76) F(1, 91) = 42.21, p < .001,
g2p = .32 Fatalism Abstract Scientic Determinism (a = .73), FAD-Plus Scientic Determinism (a = .69) F(1, 91) = 11.9, p = .001,
g2p = .12, Fatalistic Abstract Fatalistic Determinism (a = .92) and FAD-Plus Fatalistic Determinism (a = .8) F(1, 91) = 39.17,
p < .001, g2p = .3, and Fatalism Abstract Unpredictability (a = .88) and FAD-Plus Unpredictability (M = 3.3, SD = 0.62; a = .75)
F(1, 91) = 19.61, p < .001, g2p = .18.13
To summarize, the factor structure of judgments was similar between the FAD-Plus and Fatalism abstract. There was a
reliable difference in judgments between Determinism Abstract (from Experiment 3) and Fatalism Abstract suggesting
the fatalism/determinism difference applies also to abstract scenarios. Comparing responses to Fatalistic Concrete
(Experiment 4) to responses to Fatalism Abstract suggested the concrete/abstract difference also exists for fatalistic scenar-
ios. FAD-Plus reliably predicted only Fries Abstract Free Will (r = .23) and Fries Abstract Scientic Determinism (r = .51). But
again, the differences between judgments were much larger than the similarities (g2p = .12.32).

7. General discussion

The results from ve experiments supported the prediction that some actual and counterfactual intuitions about freedom
and moral responsibility are different. Experiment 1 did not nd a relation between counterfactual and actual free will intu-
itions using standard and pervasively used instruments. Using modied materials to facilitate more direct comparisons,
Experiments 25 found some moderate relations between counterfactual and actual intuitions. The relation between actual
and counterfactual intuitions persisted for both fated and determined actions framed concretely and abstractly. However,
these relations were modest compared to the differences in intuitions.
To put the relations into perspective, there were sizable differences in actual and counterfactual intuitions about free will
and moral responsibility in Experiments 25 (where they could be directly compared). Experiment 3 had one of the smaller
differences between actual and counterfactual Free Will judgments and one of the stronger relations between to two. The

12
No personality factors were associated with Fatalism Abstract Free Will or with FAD-Plus Free Will.
13
Linear regressions indicated that there was a trend for a relation between BNT was and Fatalism Abstract Unpredictability R = .17, F(1, 92) = 2.77, p = .1. BNT
was signicantly related to the Fatalism Abstract Free Will subscale R = .22, F(1, 92) = 4.74, p = .03. A multiple linear regression with BNT and Fatalism Abstract
Unpredictability scores as predictor variables and the Fatalism Abstract Free Will subscale score as the outcome variable was conducted. The full model was a
signicant predictor of the Fatalism Abstract Free Will subscale Score R = .66, F(2, 90) = 35.1, p < .001. The Fatalism Abstract Unpredictability scores relation
with the Fatalism Abstract Free Will subscale score remained signicant, b = .63, t = 7.89, p < .001 but the relation between BNT and Fatalism Abstract Free Will
subscale score was reduced b = .11, t = 1.42, p = .16. The difference between the direct effect of BNT and the Fatalism Abstract Free Will score and the
mediated effect trended toward signicance (Sobel test = 1.62, p = .1). Unstandardized indirect effects were computed for 10,000 bootstrapped samples. At the
95% condence interval, the indirect effect was .11, SE = .07, LLCI = .26, ULCI = .01 suggesting a fully mediated relation. Two different models were tested in
path models. These models were roughly statistically equivalent. BNT ? Unpredictability ? Free Will: v2 = 1.59, p = .21, RMSEA = .08 (90% CI 0.3), pclose = .26,
AIC = 784, BIC = 800, CFI = .989, TLI = .968. The alternative model was BNT ? Free Will ? Unpredictability: v2 = 0.19, p = .66, RMSEA = 0 (90% CI 0.21),
pclose = .7, AIC = 783, BIC = 798, CFI = 1, TLI = 1. Given the persistent relation between these judgments but a lack of a unique meditational model, more work
should be done exploring these relations.
 A. Feltz / Consciousness and Cognition 36 (2015) 113130 127

Fig. 4. Experiment 5 means and standard errors.

Table 12
Correlations for Experiment 5.

1 2 3 4 5 6 7 8
** ** ** *
1. Fatalism abstract free will 1 .48 .72 .65 .23 .04 .17 .02
2. Fatalism abstract science 1 .3** .5** .14 .51** .19 .04
3. Fatalism abstract fatalism 1 .53** 0 .16 .16 .15
4. Fatalism abstract unpredict 1 .1 .01 .31** .16
5. FAD free will 1 .1 .19 .09
6. FAD science 1 .27** .15
7. FAD fatalism 1 .16
8. FAD unpredict 1
*
p < .05.
**
p < .01.

difference between the actual and counterfactual intuitions accounted for 30% of the total variance in peoples free will judg-
ments. The relation between these two measures accounted for about 12% of the total variance. This estimated difference is
undoubtedly too low. All instruments are limited, and one limitation on instruments used in these experiments is that there
is random variation in measurement. That means some of the total variance will be the result of randomness and is therefore
not explainable. One way to gauge the amount of random variation in an instrument is testretest reliability. Some instru-
ments measuring counterfactual free will intuitions have good testretest reliability (r = .85) (Feltz, Cokely, & Nadelhoffer,
2009). Using this testretest reliability as an estimate of the reliable variance for both the counterfactual and actual intu-
itions, the maximum amount of variance that could be explained is .72 (1 (.85 .85) = .72). In this light, the difference
between counterfactual and actual intuitions accounted for 42% of the explainable variance whereas the relation between
these two judgments only accounts for 17%.
These differences provide some support for the use of hypothetical thought examples in the empirical exploration of free-
dom and moral responsibility (and for their use more generally). Some have argued that the use of hypothetical thought
examples in empirical research is problematic because that research has been handicapped by the lack of an instrument
that does not confound free will belief with alternative worldviews (Carey & Paulhus, 2013). Clearly, there are conceptual
differences between actual and counterfactual intuitions. Many people appear to appreciate this conceptual difference and
their intuitions are not confounded. Of course, some people make similar judgments in actual and counterfactual conditions,
but that does not mean that their judgments are confounded. One plausible explanation for many of these responses is that
those individuals simply think the conditions posited in the counterfactual situation are the same as the actual world. Or,
those stipulated conditions are not relevant to freedom and moral responsibility. In either case, the judgments are not con-
founded even if there is no measurable difference between actual and counterfactual intuitions. Consequently, exploring
both actual and counterfactual intuitions is important in order to have an adequate and complete account of folk intuitions,
concepts, or theories surrounding free will and moral responsibility.14

14
One may worry that the items in the FAD-Plus do not measure intuitions about freedom and moral responsibility. For example, Whatever will be, will be
theres not much Bill can do about it does not directly bear on freedom or moral responsibility (it is consistent with many views about freedom, moral
responsibility, and the causal nature of the universe). This worry can be deected somewhat because each item is part of a factor, or helps indicate one
underlying structure to responses that could be responsible for a cluster of judgments. Additionally, responses to the counterfactual version for the FAD-Plus
were similar to counterfactual judgments in Experiment 1.
128 A. Feltz / Consciousness and Cognition 36 (2015) 113130

To illustrate, one prominent account holds that the folk conception of freedom and moral responsibility is compatibilist
(Nahmias et al., 2004, 2005; Nahmias, Morris, Nadelhoffer, & Turner, 2006). Of course, not all people have compatibilist intu-
itions and in some instances people can be inuenced to have incompatibilist intuitions. To account for these incompatibilist
intuitions, some have argued that folk incompatibilist intuitions rest on a mistaken view of determinism. Those who express
incompatibilist intuitions mistakenly interpret determinism as bypassing an agents conscious agency (e.g., the agents
beliefs, desires, plans, and intentions) (Nahmias, Coates, et al., 2006; Nahmias & Murray, 2010). According to the compati-
bilists, those mental states allow for compatibilist style freedom and moral responsibility. The mental states are just one
more link in the causal chain. If one interprets determinism as bypassing those mental states, then one understands deter-
minism in a way that is inconsistent with how compatibilists understand determinism. As such, one would appear to have
incompatibilist intuitions but for the wrong reasons.
There are two competing causal models concerning the relation of bypassing intuitions and incompatibilist intuitions
(Rose & Nichols, 2013). The bypassing model posits the following causal pathway:

Determinism ? Bypassing judgments ? Moral responsibility and freedom judgments

The Incompatibilist Model posits the following causal pathway.
Determinism ? Moral responsibility and freedom judgments ? Bypassing judgments.

According to Rose and Nichols (2013), the Incompatibilist model is the statistically better model. One explanation for the
Incompatibilist model is that some people do not think that a person makes a decision in a deterministic universe. The folk
concept of a decision requires indeterminism. If a person does not make a decision, then of course the compatibilist favored
mental states play no role in the production of an action. So rather than an understanding of determinism causing bypassing
intuitions that then cause judgments of freedom and moral responsibility, determinism causes freedom and moral respon-
sibility judgments that then cause bypassing judgments (because no decision is made in the rst place).
The data from the current series of studies support Rose and Nichols. There were consistent and strong relations between
the Unpredictability subscale and the Free Will subscale for both counterfactual and actual judgments. Unpredictability
appears to be an important element in free will judgments. If determinism is true, then theoretically there should be no
unpredictable eventseven human choices. While there may be limitations on human cognition that may not allow for pre-
dictability in practice, that epistemic limitation is not a metaphysical limitation. Given that some people persisted in think-
ing that there are unpredictable elements in determined and fated worlds, they may not have internalized the crucial
elements of determinism or fatalism. Rather, they may have a more robust notion of freedom and moral responsibility that
is not compatibilist friendly. They may have free will no matter what intuitions (Feltz & Millan, 2015; Feltz et al., 2009).
Numeracy was also negatively related to counterfactual Free Will and Unpredictability intuitions. The meditational anal-
yses suggested that BNTs relation to counterfactual Free Will judgments went through counterfactual Unpredictability judg-
ments (see Footnotes 9, 11, and 13). If the correct interpretation of determinism and fatalism entails that no events are
unpredictable, then those who are higher in numeracy understand the deterministic and fatalistic elements of the scenarios
more correctly than those who are less numerate. This result is consistent with numerate individuals being more attentive
and providing more normatively correct decisions across a number of different decisions (Cokely & Kelley, 2009; Cokely
et al., 2012). If those who are numerate pay more attention or understand determinism more correctly, then we should pref-
erence those intuitions than intuitions that plausibly rest on relatively less attention or some mistaken interpretations.
For these reasons, those who think folk intuitions do not support compatibilism are in a rhetorically strong position. First,
statistically preferable models suggest that folk intuitions are not necessarily compatibilist friendly. Second, those who have
compatibilist intuitions may have a wider, more permissive non-compatibilist friendly set of intuitionsfree will no matter
what intuitions. Third, those who were more numerate, and therefore more likely to be more attentive and interpret critical
elements correctly, had weaker counterfactual free will and moral responsibility judgments for fated and determined actions
than the less numerate. As such, there is growing evidence that folk intuitions do not support compatibilism.
Finally, it is almost trivially true that actual intuitions about freedom and moral responsibility have some real-world con-
sequences. What is less clear is whether counterfactual intuitions (or the tendency to have such intuitions) also have impor-
tant real-world consequences. There is some reason to think that they do. Some counterfactual intuitions about freedom and
moral responsibility were related to some personality traits (e.g., extraversion) and cognitive abilities (e.g., numeracy).
Extraversion and numeracy are related to a number of real-world outcomes including health, professional satisfaction,
and happiness (Cokely, Ghazal, Garcia-Retamero, & J. Shulkin (Eds.), 2014; Revelle, Wilt, & Condon, 2011). It stands to reason,
then, that ones counterfactual intuitions about free will and moral responsibility (or tendency to have those intuitions) also
have real world implications. To the extent that these intuitions are consistent and stable across time, they may constitute
part of ones philosophical character, or stable tendencies to make some philosophical judgments (Feltz & Cokely, 2013).

Appendix A

1. I believe that Bills future has already been determined by fate.

2. Bills biological makeup determines his talents and personality.
3. Chance events seem to be the major cause of Bills history.
 A. Feltz / Consciousness and Cognition 36 (2015) 113130 129

4. Bill has complete control over the decisions he makes.

5. No matter how hard Bill tries, Bill cant change his destiny.
6. Psychologists and psychiatrists will eventually gure out of Bills behavior.
7. No one can predict what will happen in Universe A.
8. Bill must take full responsibility for any bad choices he makes.
9. Fate already has a plan for Bill.
10. Bills genes determine his future.
11. Universe A seems unpredictablejust like throwing dice or ipping a coin.
12. Bill can overcome any obstacles if he truly wants to.
13. Whatever will be, will betheres not much Bill can do about it.
14. Science has shown how Bills past environment created his current intelligence and personality.
15. Bill is unpredictable.
16. Bill is totally responsible for the bad things he does.
17. Whether Bill likes it or not, mysterious forces seem to move his life.
18. As with other animals, Bills behavior always follows the laws of nature.
19. Universe A is hard to predict because it is almost totally random.
20. Luck plays a big role in Bills lives.
21. Bill has complete free will.
22. Bills character will determine the character of his children.
23. Bill is always at fault for his bad behavior.
24. Childhood environment will determine Bills success as an adult.
25. What happens to Bill is a matter of chance.
26. Strength of Bills mind can always overcome his bodys desires.
27. Bills future cannot be predicted.

Free Will: 4, 8, 12, 16, 21, 23, 26.

Scientic Determinism: 2, 6, 10, 14, 18, 22, 24.
Fatalistic Determinism: 1, 5, 9, 13, 17.
Unpredictability: 3, 7, 11, 15, 19, 20, 25, 27.

References

Andow, J., & Cova, F. (in press). Why compatibilist intuitions are not mistaken: A reply to Feltz and Millan. Philosophical Psychology.
Bargh, J. A., & Ferguson, M. J. (2000). Beyond behaviorism: On the automaticity of higher mental processes. Psychological Bulletin, 126(6), 925945. http://
dx.doi.org/10.1037//0033-2909.126.6.925.
Baron, R. M., & Kenny, D. A. (1986). The moderator mediator variable distinction in social psychological-research Conceptual, strategic, and statistical
considerations. Journal of Personality and Social Psychology, 51(6), 11731182. http://dx.doi.org/10.1037/0022-3514.51.6.1173.
Baumeister, R. F., Masicampo, E. J., & DeWall, C. N. (2009). Prosocial benets of feeling free: Disbelief in free will increases aggression and reduces
helpfulness. Personality and Social Psychology Bulletin, 35(2), 260268. http://dx.doi.org/10.1177/0146167208327217.
Baumeister, R. F., Sparks, E. A., Stillman, T. F., & Vohs, K. D. (2008). Free will in consumer behavior: Self-control, ego depletion, and choice. Journal of
Consumer Psychology, 18(1), 413. http://dx.doi.org/10.1016/J.Jcps.10.002. 2007.
Bernstein, M. (2002). Fatalism. In R. Kane (Ed.), The Oxford handbook of free will (pp. 6581). Oxford: Oxford University Press.
Buhrmester, M., Kwang, T., & Gosling, S. D. (2011). Amazons mechanical turk: A new source of inexpensive, yet high-quality, data? Perspectives on
Psychological Science, 6(1), 35. http://dx.doi.org/10.1177/1745691610393980.
Carey, J. M., & Paulhus, D. L. (2013). Worldview implications of believing in free will and/or determinism: Politics, morality, and punitiveness. Journal of
Personality, 81(2), 130141. http://dx.doi.org/10.1111/J.1467-6494.2012.00799.X.
Cokely, E. T., & Feltz, A. (2009). Adaptive variation in judgment and philosophical intuition Reply. Consciousness and Cognition, 18(1), 356358. http://
dx.doi.org/10.1016/J.Concog.01.001. 2009.
Cokely, E. T., Galesic, M., Schulz, E., Ghazal, S., & Garcia-Retamero, R. (2012). Measuring risk literacy: The Berlin numeracy test. Judgment and Decision
Making, 7(1), 2547.
Cokely, E. T., Ghazal, S., & Garcia-Retamero, R. (2014). Measuring numeracy. In B. L. Anderson & J. Shulkin (Eds.), Numerical reasoning in medical decision
making (pp. 1138). Cambridge: Cambridge University Press.
Cokely, E. T., & Kelley, C. M. (2009). Cognitive abilities and superior decision making under risk: A protocol analysis and process model evaluation. Judgment
and Decision Making, 4(1), 2033.
Cova, F., Bertoux, M., Bourgeois-Gironde, S., & Dubois, B. (2012). Judgments about moral responsibility and determinism in patients with behavioural variant
of frontotemporal dementia: Still compatibilists. Consciousness and Cognition, 21(2), 851864. http://dx.doi.org/10.1016/J.Concog.02.004. 2012.
Crump, M. J. C., McDonnell, J. V., & Gureckis, T. M. (2013). Evaluating Amazons mechanical turk as a tool for experimental behavioral research. PloS One, 8(3).
http://dx.doi.org/10.1371/journal.pone.0057410 (ARTN e57410 DOI).
Feltz, A. (2009). Experimental philosophy. Analyse & Kritik, 31, 201219.
Feltz, A. (2013). Pereboom and premises: Asking the right questions in the experimental philosophy of free will. Consciousness and Cognition, 22(1), 5363.
http://dx.doi.org/10.1016/j.concog.2012.11.007.
Feltz, A., & Bishop, M. (2010). The proper role of intuitions in epistemology. In M. Milkowski & K. Talmont-Kaminski (Eds.), Beyond description: Normativity in
naturalised philosophy (pp. 101122). London: College Publications.
Feltz, A., & Cokely, E. T. (2009). Do judgments about freedom and responsibility depend on who you are? Personality differences in intuitions about
compatibilism and incompatibilism. Consciousness and Cognition, 18(1), 342350. http://dx.doi.org/10.1016/J.Concog.08.001. 2008.
Feltz, A., & Cokely (2013). Predicting philosophical disagreement. Philosophy Compass, 8, 978989.
Feltz, A., Cokely, E. T., & Nadelhoffer, T. (2009). Natural compatibilism versus natural incompatibilism: Back to the drawing board. Mind & Language, 24(1),
123. http://dx.doi.org/10.1111/J.1468-0017.2008.01351.X.
Feltz, A., & Millan, M. (2015). An error theory for compatibilist intuitions. Philosophical Psychology, 28(4), 529555. http://dx.doi.org/10.1080/
09515089.2013.865513.
130 A. Feltz / Consciousness and Cognition 36 (2015) 113130

Feltz, A., Perez, A., & Harris, M. (2012). Free will, causes, and decisions individual differences in written reports. Journal of Consciousness Studies, 19(910),
166189.
Frankfurt, H. (1971). Freedom of the will and the concept of a person. The Journal of Philosophy, 68, 520.
Gosling, S. D., Rentfrow, P. J., & Swann, W. B. (2003). A very brief measure of the Big-Five personality domains. Journal of Research in Personality, 37(6),
504528. http://dx.doi.org/10.1016/S0092-6566(03)00046-1.
Kane, R. (1996). The signicance of free will. New York: Oxford University Press.
Libet, B. (1985). Unconscious cerebral initiative and the role of conscious will in voluntary action. Behavioral and Brain Sciences, 8(4), 529539.
Mason, W., & Suri, S. (2012). Conducting behavioral research on Amazons Mechanical Turk. Behavior Research Methods, 44(1), 123. http://dx.doi.org/
10.3758/S13428-011-0124-6.
Mele, A. (2006). Free will and luck. Oxford; New York: Oxford University Press.
Mele, A. (2001). Acting intentionally: Probing folk intuitions. In B. F. Malle, L. J. Moses, & D. A. Baldwin (Eds.), Intentions and intentionality (pp. 2743).
Cambridge: MIT Press.
Nahmias, E., Coates, D. J., & Kvaran, T. (2006). Free will, moral responsibility, and mechanism: Experiments on folk intuitions. Philosophy and the Empirical,
31, 214242.
Nahmias, E., Morris, S., Nadelhoffer, T., & Turner, J. (2004). The phenomenology of free will. Journal of Consciousness Studies, 11(78), 162179.
Nahmias, E., Morris, S., Nadelhoffer, T., & Turner, J. (2005). Surveying freedom: Folk intuitions about free will and moral responsibility. Philosophical
Psychology, 18(5), 561584. http://dx.doi.org/10.1080/09515080500264180.
Nahmias, E., Morris, S., Nadelhoffer, T., & Turner, J. (2006). Is incompatibilism intuitive? Philosophy and Phenomenological Research, 73(1), 2853. http://
dx.doi.org/10.1111/J.1933-1592.2006.Tb00603.X.
Nahmias, E., & Murray, D. (2010). Experimental philosophy on free will: An error theory for incompatibilst intuitions. In J. A. Aguilar, A. Buckareff, & K.
Frankish (Eds.), New waves in philosophy of action (pp. 189215). New York: Palgrave Macmillan.
Nichols, S., & Knobe, J. (2007). Moral responsibility and determinism: The cognitive science of folk intuitions. Nous, 41(4), 663685. http://dx.doi.org/
10.1111/J.1468-0068.2007.00666.X.
Paulhus, D. L., & Carey, J. M. (2011). The FAD-Plus: Measuring lay beliefs regarding free will and related constructs. Journal of Personality Assessment, 93(1),
96104 (doi: Pii 931472849 Doi 10.1080/00223891.2, 0000).
Peters, E. (2012). Beyond comprehension: The role of numeracy in judgments and decisions. Current Directions in Psychological Science, 21(1), 3135. http://
dx.doi.org/10.1177/0963721411429960.
Peters, E., Vastfjall, D., Slovic, P., Mertz, C. K., Mazzocco, K., & Dickert, S. (2006). Numeracy and decision making. Psychological Science, 17(5), 407413. http://
dx.doi.org/10.1111/J.1467-9280.2006.01720.X.
Rakos, R., Laurene, K., Skala, S., & Slane, S. (2008). Belief in free will: Measurement and conceptualization innovations. Behavior and Social Issues, 17, 2039.
Revelle, W., Wilt, J., & Condon, D. (2011). Individual differences and differential psychology: A brief history and prospect. In T. Chamorro-Premuzic, S. Von
Stumm, & A. Furnham (Eds.), The Wiley-Blackwell handbook of individual differences (pp. 338). Malden: Blackwell.
Rose, D., & Nichols, S. (2013). The lesson of bypassing. Review of Philosophy and Psychology, 4, 599619.
Roskies, A., & Nichols, S. (2008). Bringing moral responsibility down to earth. Journal of Philosophy, 105, 228247.
Sarkissian, H., Chatterjee, A., De Brigard, F., Knobe, J., Nichols, S., & Sirker, S. (2010). Is belief in free will a cultural universal? Mind & Language, 25(3),
346358.
Schulz, E., Cokely, E. T., & Feltz, A. (2011). Persistent bias in expert judgments about free will and moral responsibility: A test of the expertise defense.
Consciousness and Cognition, 20(4), 17221731. http://dx.doi.org/10.1016/J.Concog.04.007. 2011.
Smilansky, S. (2002). Free will, fundamental dualism, and the centrality of illusion. In R. Kane (Ed.), The Oxford handbook of free will (pp. 489505). Oxford:
Oxford University Press.
Sommers, T. (2010). Experimental philosophy and free will. Philosophy Compass, 5, 192212.
Sommers, T. (2012). Relative justice: Cultural diversity, free will, and moral responsibility. Princeton: Princeton University Press.
Stillman, T. F., Baumeister, R. F., & Mele, A. R. (2011). Free will in everyday life: Autobiographical accounts of free and unfree actions. Philosophical
Psychology, 24(3), 381394. http://dx.doi.org/10.1080/09515089.2011.556607.
Strawson, G. (1994). The impossibility of moral responsibility. Philosophical Studies, 75(12), 524. http://dx.doi.org/10.1007/Bf00989879.
Stroessner, S. J., & Green, C. W. (1990). Effects of belief in free will or determinism on attitudes toward punishment and locus of control. Journal of Social
Psychology, 130(6), 789799.
Viney, W., Waldman, D. A., & Barchilon, J. (1982). Attitudes toward punishment in relation to beliefs in free will and determinism. Human Relations, 35(11),
939949. http://dx.doi.org/10.1177/001872678203501101.
Vohs, K. D., & Schooler, J. W. (2008). The value of believing in free will Encouraging a belief in determinism increases cheating. Psychological Science, 19(1),
4954. http://dx.doi.org/10.1111/J.1467-9280.2008.02045.X.
Wegner, D. M. (2002). The illusion of conscious will. Cambridge: MIT Press.
Wegner, D. M., & Bargh, J. A. (1998). Control and automaticity in social life. In D. Gilbert, S. Fiske, & G. Lindzey (Eds.), Handbook of social psychology
(pp. 446496). Oxford: Oxford University Press.
Wegner, D. M., & Wheatley, T. (1999). Apparent mental causation Sources of the experience of will. American Psychologist, 54(7), 480492. http://
dx.doi.org/10.1037/0003-066x.54.7.480.
Weigel, C. (2011). Distance, anger, freedom: An account of the role of abstraction in compatibilist and incompatibilist intuitions. Philosophical Psychology,
24(6), 803823. http://dx.doi.org/10.1080/09515089.2011.562640.
Weigel, C. (2013). Experimental evidence for free will revisionism. Philosophical Explorations, 16(1), 3143. http://dx.doi.org/10.1080/
13869795.2012.708774.