Humans are creatures of habit, tending to stick to their formed routines. Imagine Paul, who eats the same sandwich every lunchtime. Then, one day, he eats a different sandwich for lunch, an exceptional behavior that deviates from his normal routine. If later that day he experiences severe indigestion, would this deviation from normality impact the way his work colleagues perceive his behavior that day? All other things being equal, we hypothesized and found support for the idea that exceptional behavior resulting in negative outcomes is perceived as involving more agency than routine behavior resulting in the same unfortunate outcomes.
Most people in modern societies tend to believe in free will (Alquist et al., 2015; Monroe & Malle, 2010; Nahmias et al., 2005; Nichols, 2004; Sarkissian et al., 2010; Stillman, Baumeister & Mele, 2011). People generally think of free will as the capacity to choose freely, often associated with perceived responsibility for one’s actions and choices (Monroe, Dillon & Malle, 2014; Nahmias, Coates & Kvaran, 2007; Rigoni, Cleeremans & Brass, 2017). Free will belief also serves an important role in current social systems, since most modern legal systems are based on the premise that individuals are free and therefore accountable for their actions and their misdeeds (Rigoni, Sammicheli & Sartori, 2015). Some researchers view belief in free will as the ability to make choices that are in line with the moral and best long-term interest that does not harm other members of the social group (Alquist et al., 2015; Feldman, Baumeister & Wong, 2014; Monroe & Malle, 2010; Shepherd, 2012; Stillman, Baumeister & Mele, 2011).
Research on free will attributions made the case for ‘motivated free will’, showing that morally bad and good actions are considered as more freely performed than morally neutral actions (Clark et al., 2018). Clark and colleagues (2018) argued that the higher free will attributions for morally bad actions may be a sign of a strong affective punitive motive (e.g., Clark, Baumeister & Ditto, 2017; Martin, Rigoni & Vohs, 2017; Shariff et al., 2014). This may also be driven by pragmatic reasons, based on the expectation that rewards will encourage future good behaviors (Clark et al., 2018).
However, free will attributions may not necessarily relate to moral considerations. For example, regardless of morality, actions and outcomes of negative valence trigger higher attributions of free will comparatively to actions and outcomes of positive valence (Feldman, Wong & Baumeister 2016). A way of interpreting these results is that the perceived negativity of an action or an outcome increases responsibility attributions, regardless of morality. The mere processing of negative circumstances may suffice in triggering stronger free will attributions. If this is indeed the case, then that very processing may serve as an adaptive mechanism to enable accountability, learning, and prospective behavioral change when people are faced with negative events (Feldman, Wong & Baumeister, 2016).
An assessment of normality, namely what is normal, is complex and may refer to perceptions of what is normal for others in society, the normative expectations in a given situation, or past behavior of a person (Feldman, 2020; Feldman, Kutscher & Yay, 2020; Kutscher & Feldman, 2019). In the case of past-behavior exceptionality, we focus on the comparison between the current and past behavior of the same person—the more the present action differs from past behavior, the more it will generate regret. This last effect, sometimes be referred to as ‘exceptionality effect’, is the ‘phenomenon that people associate stronger negative affect with a negative outcome when it is a result of an exception (abnormal behavior) compared to when it is a result of routine (normal behavior)’ (Fillon, Kutscher & Feldman, 2021: 129). This has important consequences for morality as a meta-analysis conducted by Fillon, Kutscher, and Feldman (2021) showed that exceptionality increases victim compensation, offender punishment, and self-blame compared to normality. One possible explanation for an exceptionality effect in moral judgments could be the discrepancy in free will attributions between exceptional and normal circumstances. We seek to test this hypothesis in the present study.
Furthermore, when exceptional behavior results in a negative outcome, it is easier to think of instances of how someone could have acted otherwise and made different choices, since the reference point of what is normal is salient. When considering two people, one that behaves in a usual way and another that behaves unusually, the behavior of the latter may seem as if she or he exercised more free will, behaving inconsistently with her or his habits. When behavior is aligned with what is normal, it is not clear whether the behavior is a result of a deliberate, freely chosen decision to follow what is normal; being coerced into that behavior; or simply not making a decision at all and following the default or in automated mode. Our central hypothesis is that perceptions of exceptionality are associated with perceptions of free will. Accordingly, we predict that the exceptional behavior eliciting a negative outcome will lead to stronger attributions of free will to the enacting agent than routine behavior.
Finally, to enrich our investigation, we also measured moral responsibility and regret. Free will and moral responsibility are traditionally viewed as being intrinsically related (Caruso, 2018), although there are long-lasting and ongoing debates about specific cases in which they can be separated (Vierkant et al., 2019). Beyond moral responsibility, the measurement of regret not only allows us to attempt to replicate the effect of exceptionality on attributions of regret (Kahneman & Miller, 1986) but also works as an indicator of the validity of our recruited sample and methods. It also allows us to examine the links between regret, responsibility, and agency. Overall, our secondary hypotheses were that agents behaving exceptionally and facing negative outcomes are ascribed greater regret and moral responsibility than agents behaving according to routine and facing the same outcomes.
We conducted three experiments to test our predictions. In Experiment 1, we tested whether under negative outcomes people attributed more free will to an agent with exceptional behavior than someone who behaved routinely. We manipulated exceptionality in a between-subject design, and we dissociated self- from other-constrained deviations from routine action. In Experiment 2, we also added a measure of regret and moral responsibility in addition to the measure of attributions of free will. In Experiment 3, we tested a within-subject design with a manipulation of a new between-subject variable contrasting a deterministic and an indeterministic universe, related to classic experimental philosophy paradigms examining folk psychology regarding the abstract notion of free will. In all three experiments, ‘We report how we determined our sample size, all data exclusions (if any), all manipulations, and all measures in the study’ (Simmons, Nelson & Simonsohn, 2012: 4).
We intended to show that exceptionality increases attributions of free will. We subdivided the construct of exceptionality into two types of exceptionality: the self-produced exceptionality (a chosen deviation from routine) and the others-produced exceptionality (an external-constrained deviation from routine). We implemented this subdivision to be able to know whether the exceptional dimension in itself is a sufficient condition to produce the effect on attributions of free will. Because free will is traditionally viewed as something that has its source in people’s own decisions, self-initiated deviation from routine could be the necessary active ingredient required to observe the effect of exceptionality on attributions of free will compared to the routine condition (H1). We also hypothesized that free will attributed to the protagonist of the scenario would be higher in the self-produced exception condition than in the other-produced exception condition (H2). We did not have any specific predictions about the difference between the routine and the other-produced exception conditions, thus, their potentially different effects on attributions of free will were intended to be the object of an exploratory investigation.
We preregistered the hypotheses, desired sample size, exclusion rules, and general analytic strategy on Open Science Framework (OSF: https://osf.io/f2pck/ for the anonymized raw data, analysis codes, and materials of this experiment; see Deviation S.1. in the Supplementary for the preregistration planning and deviation documentation). It is important to specify that our original hypotheses and measures were included as a part of a prior experiment testing another hypothesis (Kutscher & Feldman, 2019). However, in that paper, the key measures of interest were not the same as the one presented here; because the experimental manipulations were identical, the inclusion of our measure of interest (i.e., attributions of free will) was justified. Thus, results presented in this paper are original, going beyond the purpose of Kutscher and Feldman (2019), and transparently reported in the supplementary materials of the original article as extensions meant for a different investigation.
As mentioned in the preregistration, the planned sample size (total N = 336) was determined through a power analysis based on another effect size found in the literature. As this effect size was lower in size than our effect size of interest (i.e., the effect of exceptionality on attributions of free will, η² = 0.16), we had sufficient statistical power to detect this effect.1 In this experiment, in accordance with the preregistration, we planned to run one-tailed t-tests.
We recruited 342 participants (Mage = 39.93, SDage = 12.88; 215 females) from American Amazon Mechanical Turk (MTurk) by using https://www.cloudresearch.com/ (Litman, Robinson & Abberbock, 2017). We disallowed 5% of the top active workers and set a maximum for number of tasks performed on the platform to address MTurker’s possible nonnaïveté (Chandler, Mueller & Paolacci, 2014). As specified in the a priori exclusion criteria, we excluded from the final sample participants who self-reported a low level (self-report < 5, on a 1–7 scale) in English (n = 1) and those who self-reported not being serious (self-report < 4, on a 1–5 scale) when filling in the experiment (n = 5, including one participant already detected on the previous criterion). Our final sample included a total of 337 participants (Mage = 40.04, SDage = 12.94; 212 females, 125 males).
Participants were randomly assigned to one of the three experimental conditions: namely self-produced exception, other-produced exception, and routine. Participants read a scenario describing a robbery at a grocery store with an identified (target) victim named Mr. Paul (Miller & McFarland, 1986). The three scenarios only differed in the exceptionality of the circumstances leading up to the incident. In the routine behavior condition, Mr. Paul visits the store he regularly visits. In the self-produced exception condition, Mr. Paul visits another store because he wants a change of pace. In the other-produced exception condition, Mr. Paul visits another store because the first was closed.
Two convenience stores are located in Mr. Paul’s neighborhood. He frequents Store A more regularly than Store B.
[Routine behavior condition: Last night he visited Store A.
Self-produced exception condition: Last night he visited Store B because he wanted a change of pace.
Other-produced exception condition: Last night he visited Store B because Store A was temporarily closed for renovations.]
He walked in on a robbery occurring in the store. He lost the use of his right arm as a result of a gunshot wound.
As mentioned earlier, in this section, we only reported the measures meant to test our research question. The other measures are detailed in Kutscher and Feldman (2019). An important point is that the regret measure used in Kutscher and Feldman (2019) was presented to the participants before the measures of free will attributions because it aimed to serve as a direct replication of the Miller and McFarland (1986) exceptionality effect study.
The measure of free will attributions (adapted from Clark et al., 2014) was assessed with three items rated on a 7-point scale (1 = strongly disagree, 7 = strongly agree): ‘Mr. Paul could have chosen to do otherwise and visit a different store’; ‘By visiting that specific store, Mr. Paul was exercising his free will’; and ‘Mr. Paul’s choice of which store to visit was his own free choice’. We averaged these three items for the routine condition (α = 0.79), for the self-produced exceptional condition (α = 0.86), and for the other-produced exception (α = 0.80).
We asked participants about their age and gender, their country of birth, their social class (lower class to upper class on a 6-point scale), and how they rate their understanding of the English used in the study on a 7-point scale (1 = very bad, 7 = very good).
A one-way independent ANOVA revealed that experimental conditions (routine, self-produced exception, and other-produced exception) affected attributions of free will, F(2, 334) = 28.01, MSE = 1.08, p < 0.001, ω = 0.14 (see Figure 1 for the general pattern of results).
Then, to test our hypotheses, we ran independent-samples t-tests (planned contrasts, using Fisher’s least significant difference test), expecting higher free will attributions in the self-produced exception condition than in the routine condition (H1) and in the other-produced exception condition (H2). Corroborating our hypotheses, participants in the self-produced exception condition attributed more free will to the target (M = 6.33, SD = 0.92, n = 108) than participants in the routine behavior condition (M = 6.06, SD = 0.97, n = 113), t(334) = 1.92, p = 0.028 (one-tailed), Cohen’s d = 0.29, 95% CI [0.02, 0.55] and participants in other-produced exception condition (M = 5.33, SD = 1.19, n = 116), t(334) = 7.19, p < 0.001 (one-tailed), Cohen’s d = 0.94, 95% CI [0.66, 1.22]. Finally, more exploratory, we observed that participants in other-produced exception condition attributed less free will to the target than participants in the routine condition, t(334) = 5.32, p < 0.001 (one-tailed), Cohen’s d = 0.67 [0.40, 0.94].
In Experiment 1, as predicted by our first hypothesis, participants attributed more free will in the self-produced exception scenario than in the routine scenario. Corroborating our second hypothesis, the free will ascribed to participants was higher when the behavior originated from people’s own internal reasons than from external reasons.
We also discovered that in some cases, people producing a routine behavior are seen as possessing higher free will than people producing an exceptional behavior, which is an interesting exception of our general hypothesis. In line with norm theory (Kahneman & Miller, 1986), this result showed that the effect of exceptionality on attributions of free will only appears when the exception is self-produced. In the case of externally induced exception, participants tended to perceive the target as having no choice to do otherwise and consequently rated lower attributions of free will.
In the next experiment, again using a between-subjects design, we wanted to further generalize aiming to obtain a more reliable estimate of the size of the effect of exceptionality on attributions of free will.
Experiment 2 was designed to test the effect of exceptionality on free will attributions. We also included a measure of moral responsibility and regret. We argued that measuring these three dependent variables in the same study could help to compare the effect size of exceptionality on regret (a known reliable and easily replicable effect) via previously used scenarios that we can borrow to replicate the effect of exceptionality on regret (Kahneman & Miller, 1986; Miller & McFarland, 1986) and test the effect of exceptionality on moral responsibility. Additionally, we aimed to explore the interrelations between these three variables.
The two following hypotheses were central to this experiment. We predicted that participants would attribute more regret to the character in the exceptional situation than in the routine situation (H1), a replication of the effect of exceptionality on regret (Kahneman & Miller, 1986). Concerning the central claim defended in this paper, we predicted that participants would attribute more free will to the character in the exceptional situation than in the routine situation (H2).
The last two hypotheses were considered secondary hypotheses. We hypothesized that participants would attribute more moral responsibility to the character in the exceptional situation than in the routine situation (H3). Finally, we hypothesized that measures of attributions of regret, free will, and responsibility would be positively correlated with each other (H4).
We preregistered the hypotheses, desired sample size, exclusion rules, and general analytic strategy on Open Science Framework (https://osf.io/f2pck/ for the anonymized raw data, analysis codes, and materials of this experiment; see Deviation S.2. in the Appendices for the preregistration planning and deviation documentation). We planned to recruit 554 participants, with a statistical power of 0.95, an α set to 0.10, and an effect size of Cohen’s d = 0.28. This value was based on the effect size of exceptionality on attributions of free will found in Experiment 1.
We recruited 589 MTurk participants online using https://www.cloudresearch.com/ (Litman, Robinson & Abberbock, 2017). As specified in the a priori exclusion criteria, after keeping only participants who complete the study until the end (N = 578), we excluded from the final sample participants who self-reported a low level (self-report < 5, on a 1–7 scale) in English (n = 10) and participants who self-report not being serious (self-report < 4, on a 1–5 scale) when filling in the experiment (n = 10, including three participants already detected on the previous criterion). Our final sample included a total of 561 participants (Mage = 39.44, SDage = 12.43; 306 females, 253 males, and 2 ‘others’).
Participants were instructed to read three different scenarios presented in a random order (i.e., the car accident; the hitchhiker scenarios from Kahneman & Miller, 1986; and the convenience store scenario we used in Experiment 1 from Miller & McFarland, 1986). The type of event (routine vs. exceptional) was a between-subjects variable manipulated for each scenario: participants could only see one of the two experimental conditions (i.e., routine or exceptional condition) for each scenario. The condition in which the participant is randomly allocated in one scenario is independent of the conditions in which the participant was randomly allocated in the other scenario.
The scenario for the routine condition was ‘Mr. Jones frequently takes hitchhikers in his car. Yesterday he gave a man a ride and was robbed’. The scenario for the exceptional condition was ‘Mr. Jones almost never takes hitchhikers in his car. Yesterday he gave a man a ride and was robbed’. Participants then had to answer comprehension questions as follows: ‘Does Mr. Jones typically tend to take hitchhikers in his car?’ (Yes/No) and ‘What happened yesterday?’ with four possible answers, one correct, ‘Mr. Jones gave a man a ride and was robbed’, and three wrong, ‘Mr. Jones did not give a ride and was robbed’, ‘Mr. Jones gave a man a ride but was not robbed’, and ‘Mr. Jones did not give a ride and was not robbed’. Participants had to answer the comprehension questions correctly before being allowed to proceed to the next page.
The scenario for the routine condition was ‘Mr. Adams was involved in an accident when driving home after work on his regular route’, and the scenario for the exceptional condition was ‘Mr. Adams was involved in an accident when driving home on a route that he only takes when he wants a change of scenery’. Participants then had to answer the following comprehension questions: ‘Was Mr. Adams driving home on his regular or irregular route?’ with two possibilities: ‘Regular route’ and ‘Irregular route only taken when Mr. Adams wants a change of scenery’, and the question ‘What happened?’ with a correct answer, ‘Mr. Adams was involved in an accident’, and a wrong answer, ‘Mr. Adams made it home without any incidents’. Participants were not allowed to continue to the next page until they had correctly answered the previous comprehension questions.
The scenario for the routine condition was ‘Two convenience stores are located in Mr. Paul’s neighborhood. He frequents Store A more regularly than Store B. Last night he visited Store A. He walked in on a robbery taking place at the store and was shot. He lost the use of his right arm as a result of the gunshot wound’. The scenario for the exceptional condition was ‘Two convenience stores are located in Mr. Paul’s neighborhood. He frequents Store A more regularly than Store B. Last night he visited Store B because he wanted a change of pace. He walked in on a robbery taking place at the store and was shot. He lost the use of his right arm as a result of the gunshot wound’. Participants then had to answer these comprehension questions: ‘Which convenience store does Mr. Paul visit frequently?’ (Store A/Store B); ‘Which convenience store did Mr. Paul visit last night?’ (Store A/Store B); and ‘What happened?’ with a correct answer, ‘Mr. Paul walked in on a robbery and was shot’ and a wrong answer, ‘Mr. Paul completed his shopping without any incidents’. Correct answers to these questions had to be given before being allowed to proceed to the next page.
Measures were the same for all conditions and scenarios. Only the name of the target involved changed, depending on the target’s name of the scenario. We randomized the presentation of the measures.
We measured free will attributions with two items from Clark et al. (2014) on a 7-point scale (0 = strongly disagree, 6 = strongly agree): ‘Mr. Jones/Adams/Paul could have chosen to act otherwise’ and ‘Mr. Jones/Adams/Paul’s choice was his own free choice’. For the hitchhiker scenario, the car accident scenario, and the convenience store scenario, the reliability of two items about free will attributions were high enough to be averaged (routine event: Spearman-Brown coefficient = 0.82; exceptional event: Spearman-Brown coefficient = 0.78; routine event: Spearman-Brown coefficient = 0.80; exceptional event: Spearman-Brown coefficient = 0.87; routine event: Spearman-Brown coefficient = 0.85; exceptional event: Spearman-Brown coefficient = 0.86, respectively; see Eisinga, Te Grotenhuis, and Pelzer, 2013, for the reason of privileging Spearman-Brown reliability estimate rather than Cronbach’s alpha for two-item scales).
We measured regret with one item on a 7-point scale (0 = no regret, 6 = very strong regret). For example, for the hitchhiker scenario, the item was ‘How much regret does Mr. Jones feel over giving a ride?’
We measured moral responsibility with one item on a 7-point scale (0 = no responsible at all, 6 = completely responsible) with the question ‘How responsible does Mr. Jones/Adams/Paul feel for the unfortunate outcome?’
The means2 and standard deviations of all dependent variables by experimental condition and by scenario are detailed in Table 1 and plotted in Figures 2, 3, 4. The means and standard deviations of all dependent variables and their bivariate correlations are reported in Table 2. Excluding outliers did not affect the results; therefore, all the subsequent reporting is using the entire sample. Results excluding outliers are available in Table S.1. in the Appendices.
|SCENARIO||HITCHHIKER||CAR ACCIDENT||CONVENIENCE STORE|
|TYPE OF EVENT||ROUTINE(N = 282)||EXCEPTIONAL(N = 279)||ROUTINE(N = 280)||EXCEPTIONAL(N = 281)||ROUTINE(N = 280)||EXCEPTIONAL(N = 281)|
|1. Free will||6.41||0.83||6.46||0.80||5.56||1.45||6.23||0.98||5.93||1.18||6.25||1.05|
|2. Moral responsibility||5.23||1.56||5.5||1.51||3.80||1.55||4.65||1.66||2.79||1.81||3.49||1.99|
We ran a set of one-sided independent samples Welch t-tests for each scenario, summarized in Table 3. As predicted by our hypotheses, participants attributed more regret (H1) and moral responsibility (H3) to the person involved in an exceptional event than to the person involved in a routine event for all the scenarios (ps < 0.001, from Cohen’s d = 0.21 to Cohen’s d = 1.32). Free will (H2) was also attributed more to a person involved in an exceptional event than to a person involved in a routine event for two out of the three scenarios (Cohen’s d = 0.29 in the convenience store scenario and Cohen’s d = 0.54 in the car accident scenario). However, we did not find support for an exceptionality effect in free will attributions in the hitchhiker scenario (Cohen’s d = 0.06). Effect sizes of the impact of exceptionality on attributions of regret were higher than the effect of exceptionality on attributions of free will and moral responsibility (Table 3).
Finally, in line with H4, for all scenarios, we found support for a positive bivariate correlation between all measures (Table 2). The only exception is the weaker correlation between attributions of moral responsibility and attributions of free will for the convenience store scenario, r(287) = 0.05, 95% CI [–0.04, 0.13], p = 0.274.
|1. Free will||6.44||0.81||–|
|2. Moral responsibility||5.39||1.54||0.24*** [0.16, 0.31]||–|
|3. Regret||6.17||1.19||0.42*** [0.35, 0.49]||0.36*** [0.28, 0.43]||–|
|1. Free will||5.90||1.28||–|
|2. Moral responsibility||4.23||1.66||0.24*** [0.16, 0.31]||–|
|3. Regret||4.40||2.06||0.20*** [0.12, 0.28]||0.48*** [0.41, 0.54]||–|
|1. Free will||6.09||1.13||–|
|2. Moral responsibility||3.14||1.93||0.05 [–0.04, 0.13]||–|
|3. Regret||5.80||1.58||0.34*** [0.27, 0.41]||0.21*** [0.12, 0.28]||–|
|Routine versus exceptional event|
|Free will attributions||0.69||558.44||0.245||0.06 [–0.10, 0.23]|
|Moral responsibility attributions||2.51||558.83||0.006||0.21 [0.04, 0.37]|
|Regret attributions||7.66||498.66||<0.001||0.65 [0.48, 0.82]|
|CAR ACCIDENT SCENARIO|
|Routine versus exceptional event|
|Free will attributions||6.41||488.38||<0.001||0.54 [0.37, 0.71]|
|Moral responsibility attributions||6.25||557.02||<0.001||0.53 [0.36, 0.70]|
|Regret attributions||15.68||523.73||<0.001||1.32 [1.14, 1.51]|
|CONVENIENCE STORE SCENARIO|
|Routine versus exceptional event|
|Free will attributions||3.32||550.30||<0.001||0.29 [0.12, 0.45]|
|Moral responsibility attributions||4.42||553.98||<0.001||0.37 [0.20, 0.54]|
|Regret attributions||6.57||506.06||<0.001||0.55 [0.38, 0.72]|
In Experiment 2, we reached the same conclusion as in Experiment 1, that participants attributed more free will to a person who acted exceptionally than to a person who acted normally. We also replicated classic effects for regret and moral responsibility, showing that people attributed stronger regret and accountability over negative outcomes resulting from exceptions compared to routine (Kahneman & Miller, 1986; Kutscher & Feldman, 2019; Miller & McFarland, 1986). Attributions of free will were associated with attributions of regret and moral responsibility.
Having provided a clear demonstration of the effects of exceptionality on attributions of free will, regret, and moral responsibility via Experiment 2, Experiment 3 tested whether these effects would show in different contexts with varying degrees of the possibility for free will (for more details on the background for this experiment, see the Note S.1. in the Appendices).
We adjusted a classic method used in experimental philosophy (Knobe et al., 2012). This method consists of presenting hypothetical universes and asking people about these scenarios to test their philosophical intuitions about free will (e.g., Nahmias et al., 2005; Nahmias & Murray, 2011; Nichols & Knobe, 2007). In these scenarios, participants are presented with hypothetical universes, with either a deterministic or an indeterministic universe.3
We predicted (H1) that people will ascribe more free will (H1a), moral responsibility (H1b), and regret (H1c) to the target who changed his habits (through an exceptional action leading to a negative outcome) than to the target who did not change (through a routine action leading to the same negative outcome) and, more importantly (H2), that within a deterministic universe, the above-mentioned differences will be weakened compared to an indeterministic universe (H2a, H2b, H2c).4
We preregistered hypotheses, desired sample size, exclusion rules, and general analytic strategy on the Aspredicted website (https://aspredicted.org/mg4ia.pdf). We aimed for a sample size of 128 based on 80% power and alpha of 0.05 to detect predicted within–between interaction (H2a,b,c) corresponding to a 2 × 2 mixed-design ANOVA (predicted correlation among the repeated measures, r = 0.00; with a small-to-medium effect size of the predicted interaction, η2 = 0.03). Anonymized raw data, analysis codes, and materials of this experiment can be found here: https://osf.io/f2pck/. See Deviation S.3. in the Appendices for the preregistration planning and deviation documentation.
One hundred and twenty-eight French undergraduate students in psychology (Mage = 19.46, SDage = 3.40; 110 females) participated in an experiment with a duration of approximately 15 minutes in exchange of course credits. Participants were run in groups of 20 to 30 and randomly assigned to conditions of the 2 (type of universe: deterministic universe vs. indeterministic universe) × 2 (type of event: exceptional vs. normal) mixed design with the first factor between subjects and the second factor within subjects.
To manipulate the type of universe, participants were invited to read a scenario that describes a deterministic or an indeterministic universe (Nichols & Knobe, 2007; French translation from Cova, 2011), as follows:
Deterministic universe: Imagine a universe 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 happened next and so on right up until the present. For example, one day, John decided to have French fries at lunch. Like 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.
Indeterministic universe: Imagine a universe in which almost everything that happens is completely caused by whatever happened before it. The one exception is human decision-making. For example, one day, John decided to have French fries at lunch. Since a person’s decision in this universe is not completely caused by what happened before it, even if everything in the universe was exactly the same up until John made his decision, it did not have to happen that John would decide to have French fries. He could have decided to have something different.
Then, participants read two scenarios (the order was randomly assigned) about a car accident (occurring on an exceptional route vs. a routine route). These scenarios were based on a modified version of the car accident scenario from Kahneman and Miller (1986). We insisted on the fact that scenarios were taking place in the previously described universe:
In this universe, Mr. Smith left the office at the usual hour. Two routes only exist to come home: one route that he takes regularly and one along the shore that he rarely takes.
Exceptional event: It’s closing time. Mr. Smith is driving home. He drives on the route that he rarely takes, along the shore, to come home. Then, Mr. Smith car is crashed by a truck. He escaped alive but seriously wounded.
Routine event: It’s closing time. Mr. Smith is driving home. He drives on the regular route to come home. Then, Mr. Smith car is crashed by a truck. He escaped alive but seriously wounded.
All measures were on a 7-point Likert scale (1 = not at all, 7 = completely). Regret attributions were measured first with one item adapted from Towers et al. (2016): ‘Indicate to what extent Mr. Smith feels regret for what happened to him’. Second, moral responsibility was measured with one item: ‘Indicate to what extent Mr. Smith could be held accountable for what happened to him’. Third, free will attributions were measured with three items from Study 1b of Clark et al. (2018): ‘Indicate how much Mr. Smith decision to take this route was freely chosen’, ‘Indicate how much Mr. Smith could have made other choices’, and ‘Indicate how much the fact that Mr. Smith took this route was due to his own free choice’. We averaged the three items measuring free will attributions (α = 0.73 and 0.83, for the routine and exceptional condition, respectively). Then, as a comprehension check, we again presented the corresponding scenario used as the manipulation of the type of universe followed with a comprehension check item, ‘In the universe previously described, individuals have a total free will’, measured with a 7-point Likert scale (1 = completely disagree, 7 = completely agree). Finally, participants informed their gender and age and had the opportunity to leave a comment in a free space before being debriefed.
All relevant descriptive statistics are detailed in Table 4 and plotted in Figure 5. The comprehension check showed that participants agreed more that individuals have a total free will in the indeterministic universe (M = 5.17, SD = 1.50, n = 63) than in the deterministic universe (M = 3.29, SD = 2.08, n = 65), t(126) = 5.86, p < 0.001, Cohen’s d = 1.03, 95% CI [0.66, 1.41].
|TYPE OF UNIVERSE||DETERMINISTIC (n = 65)||INDETERMINISTIC (n = 63*)||TOTAL|
|TYPE OF EVENT||ROUTINE||EXCEPTIONAL||r||ROUTINE||EXCEPTIONAL||r|
|Free will||4.64||1.74||5.04||1.80||0.78 [0.66, 0.86]||5.33||1.20||5.90||1.11||0.56 [0.36, 0.71]||5.22||1.44|
|Moral responsibility||2.28||1.65||2.98||1.96||0.37 [0.14, 0.56]||2.30||1.46||2.54||1.55||.48 [0.27, 0.65]||2.53||1.40|
|Regret||3.98||2.06||5.37||1.84||0.28 [0.04, 0.49]||4.43||1.95||5.79||1.47||0.61 [0.43, 0.75]||4.89||1.56|
To test our hypotheses (H1a,b,c and H2a,b,c), we ran three mixed ANOVA 2 (type of universe: deterministic vs. indeterministic) × 2 (type of event: exceptional vs. routine) on the three dependent variables (i.e., attributions of free will, moral responsibility, and regret). In accordance with H1, participants attributed more free will (H1a), moral responsibility (H1b), and regret (H1c) to the target in the exceptional situation than in the routine situation. Moreover, participants attributed more free will to the target in the indeterministic scenario than in the deterministic one. Finally, we failed to find support for H2, as we did not find support for any interactions between the type of universe and the type of event. Descriptively, only the results for attributions of free will were in the predicted direction. Results are detailed in Table 5.5
|ATTRIBUTIONS FACTOR||FREE WILL||MORAL RESPONSIBILITY||REGRET|
|Type of event||22.95||(1,125)||0.65||<0.001***||0.16 [0.07, 0.25]||8.69||(1,126)||1.65||0.004**||0.06 [0.01, 0.14]||60.23||(1,126)||2.01||<0.001***||0.32 [0.21, 0.42]|
|Type of universe||9.80||(1,125)||3.86||0.002**||0.07 [0.02, 0.15]||0.72||(1,126)||3.93||0.398||0.01 [0.00, 0.05]||2.51||(1,126)||4.80||0.115||0.02 [0.00, 0.08]|
|Type of event × Type of universe||0.70||(1,125)||0.65||0.403||0.01 [0.00, 0.05]||2.14||(1,126)||1.65||0.146||0.02 [0.00, 0.07]||<0.01||(1.126)||2.01||0.956||<0.01 [0.00, 1]|
In this experiment, in line with a part of our predictions, participants attributed more free will, moral responsibility, and regret to the target who took the exceptional route than the routine route. We found that participants attributed more free will in the indeterministic universe than in the deterministic universe, regardless of the exceptionality of the event. However, contrary to our predictions, we did not find evidence to suggest that the type of universe moderated the effect of the type of event on attributions of free will, moral responsibility, and regret. The effect of exceptionality on the three dependent variables could be relatively too strong in comparison with the effect of the type of universe on these same dependent variables, leading to a too great difficulty to detect their interaction effects without a significantly larger sample. This may also indicate that the effect of exceptionality on the attribution of free will is robust, regardless of whether the action takes place in a deterministic universe or not.
The main contribution of this paper was to establish the effect of exceptionality on free will attributions: the more exceptional a behavior is, the more free will would be attributed to an actor producing this behavior. The three experiments we conducted supported this hypothesis. We also replicated the effect of exceptionality on the attribution of regret and found that exceptionality also affected responsibility attributions. Overall, exceptionality not only increases regret attributions but also increases free will and moral responsibility attributions.
In Experiment 1, we found that the exceptional behavior should be seen as originating from the actor to affect free will attributions, as other-produced exceptionality did not increase free will attributions. In the latter case, people could infer that the target could not have had a choice to do otherwise. In Experiment 2, we strengthened the confidence we have regarding the existence of the effect of exceptionality on free will attributions with more participants and the use of more scenarios, although we found that the effect did not seem to generalize well to one of the added scenarios. We also found that attributions of free will, regret, and moral responsibility were positively related. In Experiment 3, participants ascribed more free will, moral responsibility, and regret in the exceptional condition than in the routine condition, whether within a deterministic world or not, strengthening the results of the two other experiments. These effects were robust and persisted whether the design is between or within subjects.
Our findings could be linked with those by Bear and Knobe (2016), who found that people relied on two cues to determine the active or passive feature of a behavior: mental effort and spontaneity. They found that a behavior that was performed ‘actively’ rather than ‘passively’ modifies people’s judgment about the compatibility of this behavior with causal determinism thesis. More concretely, people perceived a behavior that was performed actively (such as composing a highly technical legal document) as less possible (i.e., less compatible) in a causally deterministic universe than a behavior performed passively (such as impulsively shoplifting from a convenience store; Bear & Knobe, 2016). By adopting this framework, we may link an exceptional behavior to an active behavior (because its ‘breaking off from the flow of things’ and requires mental effort and spontaneity) and a routine behavior to a passive effort (because it is ‘going with the flow’ and does not require a mental effort or spontaneity). In the same vein, an agent acting spontaneously is considered freer than an agent acting deliberately (Vierkant et al., 2019). Although Vierkant et al. (2019) manipulated the agent’s choice (spontaneous vs. deliberate) in a within-subject design in their study, it may suggest that when deliberation (or mental effort) and spontaneity are experimentally contrasted, it is spontaneity that seems to be the driving force behind the increase of perceived free will of the agent.
However, unlike Vierkant et al. (2019), we did not observe dissociation between attributions of free will and responsibility. They reported that when an action was based on deliberation (versus spontaneity), it was judged to be less free, but the person who initiated it was considered to be more responsible for it. In our case, the person at the source of the routine behavior (versus exceptional behavior) was judged both less free and less morally responsible. In addition, our results on moral responsibility also contradict what could have been predicted by Miller and McFarland (1986). Indeed, if we consider that victim compensation is partly based on the perceived moral responsibility of the victim herself, then the to-be-compensated victim should be considered less morally responsible in the exceptional condition than in the routine condition. This prediction could be in line with the evidence (Miller & McFarland, 1986) that exceptionality increased victim compensation compared to normality (for instance, operationalized in terms of money allocated to compensate for the lost arm of a victim following a gunshot). This result failed to find support in a replication by Kutscher and Feldman (2019), and we found to the contrary, that exceptionality increased rather than decreased perceived moral responsibility. As a result, our findings suggest new directions and contribute to the understanding of the determinant of attributions of free will and moral responsibility.
In one specific instance, we failed to find support for an exceptionality effect: the hitchhiker scenario in Experiment 2. Although this could be the result of a false negative, and despite the result being in the expected direction of the hypothesis, there is one difference between this scenario and the two others that may potentially explain the differences in effect. In the car accident and convenient store scenarios, the actor does the exceptional action ‘because of …’ something (i.e., ‘because he wanted a change of scenery’; ‘because he wanted a change of pace’), which is not the case in the hitchhiker scenario. The two scenarios could have been seen as active deliberation scenarios and the hitchhiker scenario as an autopilot scenario, leading to a more pronounced effect in the former because of this confound.
In all of our experiments, the means of attributions of free will were between 4 and 7, upper to the midpoint. This is not the case for the other dependent variables. Although exceptionality increases attributions of free will, it looks like people generally attribute free will to others. This result is in accordance with the compatibilist view of the world, in which people are seen as free, even in a deterministic universe. Furthermore, it may also mean that exceptionality effect for attributions of free will is weaker than for moral responsibility and regret because the range of response is restricted. This lack of sensitivity may be a limit to the use of Likert scale in the study of free will attributions.
Several limitations of the present research should be noted.
We used the original materials from Kahneman and Miller (1986) and Miller and McFarland (1986). Future work may extend to test the generalizability of exceptionality by using new (or other) scenarios and further extend this effect by manipulating other characteristics of normality (e.g., action effect, the effect of availability, the effect of mutability). Furthermore, because the functional theory of counterfactual thinking (Roese & Epstude, 2017) suggests that regret is mitigated by the justifiability of the action (Inman & Zeelenberg, 2002), one might investigate the extent to which justifiability can change free will attributions in the same way it changes regret attributions. Our design suggests it would also be useful to incorporate measures of moral responsibility because of its conceptual links to the notions of blame and justification.
In our three experiments, we used scenarios in which participants had to attribute free will to a target. In doing so, we did not know if the effect of exceptionality on attributions of free will would be found if the target of attributions is oneself (i.e., self-attribution). This question arises because some research suggests that people tend to attribute more free will to themselves than to others (Earp, 2011; Pronin & Kugler, 2010; see also Feldman, Wong & Baumeister, 2016, for exceptions). Hence, the potential presence of a ceiling effect (i.e., a high score on the scale of self-attribution of free will) could potentially weaken the difference of self-attribution of free will between exceptionality and routine conditions. Therefore, future experiments are needed to ensure the generalizability of this finding and explore self-other differences regarding the exceptionality effect in attributions of free will.
Although we looked at the relationships between different types of attributions made by participants, the specific design of our experiments prevents any causal interpretations of the results of these analyses. As it was our predictor of interest, we only manipulated exceptionality in our experiments. In the future, researchers may further examine whether attributions of free will play a mediating role between exceptionality and attributions of regret. Further, researchers may also incorporate measures of counterfactuals to assess the way they vary with the attributions of free will and regret. Additionally, future work may investigate how belief in free will moderates the effect of exceptionality on attributions of free will (Monroe, Brady & Malle, 2016).
People attribute stronger regret to someone who experienced a bad outcome following an exceptional behavior than a routine behavior. We demonstrated that exceptionality also increases attributions of free will. People attributed more free will to others in exceptional situations, yet only when the exceptional behavior was produced by the target and not produced by others.
Our pre-registrations, materials, codes and data are available at https://osf.io/f2pck/.
The additional file for this article can be found as follows:Supplementary file of the exceptionality effect in agency
The supplementary file has tables of deviation from preregistration, a note about the studies order, the results of experiment 2 without outliers, and a summary table of effect sizes across all experiments. DOI: https://doi.org/10.5334/irsp.591.s1
3These hypothetical scenarios set a Laplacian conception of determinism (Clark, Winegard & Baumeister, 2019; Nahmias et al., 2006). More concretely, it means that theoretically, by knowing all the laws of nature and the location of each particle of the universe at one moment, we would be able to know with certainty the state of the universe at another moment.
5In line with criteria set in the preregistration, we detected an observation with a too large studentized deleted residual (i.e., a studentized deleted residual strictly greater than 4, see McClelland, 2014) on statistical analyses involving attributions of free will as output. Following our preregistering document, we removed this participant from the sample. Keeping this participant did not change the conclusions.
Part of this work was based on the first author’s Master’s thesis under the supervision of Anthony Lantian and Ahogni N’gbala at Univ. Paris Nanterre.
The authors have no competing interests to declare.
Alquist, J. L., Ainsworth, S. E., Baumeister, R. F., Daly, M., & Stillman, T. F. (2015). The making of might-have-beens: Effects of free will belief on counterfactual thinking. Personality and Social Psychology Bulletin, 41(2), 268–283. DOI: https://doi.org/10.1177/0146167214563673
Bear, A., & Knobe, J. (2016). What do people find incompatible with causal determinism? Cognitive Science, 40(8), 2025–2049. DOI: https://doi.org/10.1111/cogs.12314
Caruso, G. (2018). Skepticism about moral responsibility. In E. N. Zalta (Ed.), The Stanford Encyclopedia of Philosophy, Spring 2018 Edition. https://plato.stanford.edu/archives/spr2018/entries/skepticism-moral-responsibility/
Chandler, J., Mueller, P., & Paolacci, G. (2014). Nonnaïveté among Amazon Mechanical Turk workers: Consequences and solutions for behavioral researchers. Behavior Research Methods, 46, 112–130. DOI: https://doi.org/10.3758/s13428-013-0365-7
Clark, C. J., Baumeister, R. F., & Ditto, P. H. (2017). Making punishment palatable: Belief in free will alleviates punitive distress. Consciousness and Cognition, 51, 193–211. DOI: https://doi.org/10.1016/j.concog.2017.03.010
Clark, C. J., Luguri, J. B., Ditto, P. H., Knobe, J., Shariff, A. F., & Baumeister, R. F. (2014). Free to punish: A motivated account of free will belief. Journal of Personality and Social Psychology, 106(4), 501–513. DOI: https://doi.org/10.1037/a0035880
Clark, C. J., Shniderman, A., Luguri, J. B., Baumeister, R. F., & Ditto, P. H. (2018). Are morally good actions ever free? Consciousness and Cognition, 63, 161–182. DOI: https://doi.org/10.1016/j.concog.2018.05.006
Clark, C. J., Winegard, B. M., & Baumeister, R. F. (2019). Forget the folk: Moral responsibility preservation motives and other conditions for compatibilism. Frontiers in psychology, 10, 215. DOI: https://doi.org/10.3389/fpsyg.2019.00215
Cova, F. (2011). Neurosciences et droit pénal : le déterminisme peut-il sauver la conception utilitariste de la peine ? Klesus—Revue Philosophique, 21, 32–77. https://www.revue-klesis.org/pdf/Klesis-Philosophie-analytique-du-droit-3-Cova.pdf
Earp, B. D. (2011). Do I have more free will than you do? An unexpected asymmetry in intuitions about personal freedom. New School Psychology Bulletin, 9(1), 21–27. DOI: https://doi.org/10.1037/e505232012-003
Eisinga, R., Te Grotenhuis, M., & Pelzer, B. (2013). The reliability of a two-item scale: Pearson, Cronbach, or Spearman-Brown? International Journal of Public Health, 58, 637–642. DOI: https://doi.org/10.1007/s00038-012-0416-3
Feldman, G. (2020). What is normal? Dimensions of action-inaction normality and their impact on regret in the action-effect. Cognition and Emotion, 34(4), 728–742. DOI: https://doi.org/10.1080/02699931.2019.1675598
Feldman, G., Baumeister, R. F., & Wong, K. F. E. (2014). Free will is about choosing: The link between choice and the belief in free will. Journal of Experimental Social Psychology, 55, 239–245. DOI: https://doi.org/10.1016/j.jesp.2014.07.012
Feldman, G., Kutscher, L., & Yay, T. (2020). Omission and commission in judgment and decision making: Linking action-inaction effects using the concept of normality. Social and Personality Psychology Compass, 14(8), e12557. DOI: https://doi.org/10.1111/spc3.12557
Feldman, G., Wong, K. F. E., & Baumeister, R. F. (2016). Bad is freer than good: Positive–negative asymmetry in attributions of free will. Consciousness and Cognition, 42, 26–40. DOI: https://doi.org/10.1016/j.concog.2016.03.005
Fillon, A., Kutscher, L., & Feldman, G. (2021). Impact of past behaviour normality: Meta-analysis of exceptionality effect. Cognition and Emotion, 35(1), 129–149. DOI: https://doi.org/10.1080/02699931.2020.1816910
Inman, J. J., & Zeelenberg, M. (2002). Regret in repeat purchase versus switching decisions: The attenuating role of decision justifiability. Journal of Consumer Research, 29(1), 116–128. DOI: https://doi.org/10.1086/339925
Kahneman, D., & Miller, D. T. (1986). Norm theory: Comparing reality to its alternatives. Psychological Review, 93(2), 136–153. DOI: https://doi.org/10.1037/0033-295X.93.2.136
Knobe, J., Buckwalter, W., Nichols, S., Robbins, P., Sarkissian, H., & Sommers, T. (2012). Experimental philosophy. Annual Review of Psychology, 63, 81–99. DOI: https://doi.org/10.1146/annurev-psych-120710-100350
Kutscher, L., & Feldman, G. (2019). The impact of past behaviour normality on regret: Replication and extension of three experiments of the exceptionality effect. Cognition and Emotion, 33, 901–914. DOI: https://doi.org/10.1080/02699931.2018.1504747
Litman, L., Robinson, J., & Abberbock, T. (2017). TurkPrime.com: A versatile crowdsourcing data acquisition platform for the behavioral sciences. Behavior Research Methods, 49, 433–442. DOI: https://doi.org/10.3758/s13428-016-0727-z
Martin, N. D., Rigoni, D., & Vohs, K. D. (2017). Free will beliefs predict attitudes toward unethical behavior and criminal punishment. Proceedings of the National Academy of Sciences, 114(28), 7325–7330. DOI: https://doi.org/10.1073/pnas.1702119114
McClelland, G. H. (2014). Nasty data: Unruly, ill-mannered observations can ruin your analysis. In H. T. Reis, & C. M. Judd (Eds.), Handbook of research methods in social and personality psychology (2nd ed., pp. 608–626). New York: Cambridge University Press. DOI: https://doi.org/10.1017/CBO9780511996481.028
Miller, D. T., & McFarland, C. (1986). Counterfactual thinking and victim compensation. Personality and Social Psychology Bulletin, 12(4), 513–519. DOI: https://doi.org/10.1177/0146167286124014
Monroe, A. E., Brady, G. L., & Malle, B. F. (2016). This isn’t the free will worth looking for. Social Psychological and Personality Science, 8(2), 191–199. DOI: https://doi.org/10.1177/1948550616667616
Monroe, A. E., Dillon, K. D., & Malle, B. F. (2014). Bringing free will down to Earth: People’s psychological concept of free will and its role in moral judgment. Consciousness and Cognition, 27, 100–108. DOI: https://doi.org/10.1016/j.concog.2014.04.011
Monroe, A. E., & Malle, B. F. (2010). From uncaused will to conscious choice: The need to study, not speculate about people’s folk concept of free will. Review of Philosophy and Psychology, 1, 211–224. DOI: https://doi.org/10.1007/s13164-009-0010-7
Nahmias, E., Coates, D. J., & Kvaran, T. (2007). Free will, moral responsibility, and mechanism: Experiments on folk intuitions. Midwest Studies in Philosophy, 31(1), 214–242. DOI: https://doi.org/10.1111/j.1475-4975.2007.00158.x
Nahmias, E., Morris, S., Nadelhoffer, T., & Turner, J. (2005). Surveying freedom: Folk Intuitions about free will and moral responsibility. Philosophical Psychology, 18(5), 561–584. DOI: https://doi.org/10.1080/09515080500264180
Nahmias, E., Morris, S. G., Nadelhoffer, T., & Turner, J. (2006). Is incompatibilism intuitive?. Philosophy and Phenomenological Research, 73(1), 28–53. DOI: https://doi.org/10.1111/j.1933-1592.2006.tb00603.x
Nahmias, E., & Murray, D. (2011). Experimental philosophy on free will: An error theory for incompatibilist intuitions. In J. H. Aguilar, A. A. Buckareff, & K. Frankish (Eds.), New waves in philosophy of action (pp. 189–216). London: Palgrave Macmillan. DOI: https://doi.org/10.1057/9780230304253_10
Nichols, S. (2004). The folk psychology of free will: Fits and starts. Mind & Language, 19(5), 473–502. DOI: https://doi.org/10.1111/j.0268-1064.2004.00269.x
Nichols, S., & Knobe, J. (2007). Moral responsibility and determinism: The cognitive science of folk intuitions. Nous, 41(4), 663–685. DOI: https://doi.org/10.1111/j.1468-0068.2007.00666.x
Pronin, E., & Kugler, M. B. (2010). People believe they have more free will than others. Proceedings of the National Academy of Sciences of the United States of America, 107(52), 22469–22474. DOI: https://doi.org/10.1073/pnas.1012046108
Rigoni, D., Cleeremans, A., & Brass, M. (2017). Causes and consequences of the belief in free will. In C. Zedelius, B. Müller, & J. Schooler (Eds.), The science of lay theories (pp. 229–242). Cham: Springer. DOI: https://doi.org/10.1007/978-3-319-57306-9_10
Rigoni, D., Sammicheli, L., & Sartori, G. (2015). Looking for the right intention: Can neuroscience benefit from the law? Frontiers in Human Neuroscience, 9, 432. DOI: https://doi.org/10.3389/fnhum.2015.00432
Roese, N. J., & Epstude, K. (2017). The functional theory of counterfactual thinking: New evidence, new challenges, new insights. Advances in Experimental Social Psychology, 56, 1–79. DOI: https://doi.org/10.1016/bs.aesp.2017.02.001
Sarkissian, H., Chatterjee, A., De Brigard, F., Knobe, J., Nichols, S., & Sirker, S. (2010). Is belief in free will a cultural universal? Mind & Language, 25(3), 346–358. DOI: https://doi.org/10.1111/j.1468-0017.2010.01393.x
Shariff, A. F., Greene, J. D., Karremans, J. C., Luguri, J. B., Clark, C. J., Schooler, J. W., … & Vohs, K. D. (2014). Free will and punishment: A mechanistic view of human nature reduces retribution. Psychological Science, 25(8), 1563–1570. DOI: https://doi.org/10.1177/0956797614534693
Shepherd, J. (2012). Free will and consciousness: Experimental studies. Consciousness and Cognition, 21(2), 915–927. DOI: https://doi.org/10.1016/j.concog.2012.03.004
Simmons, J. P., Nelson, L. D., & Simonsohn, U. (2012). A 21 word solution. Dialogue: The Official Newsletter of the Society for Personality and Social Psychology, 26, 4–7. DOI: https://doi.org/10.2139/ssrn.2160588
Stillman, T. F., Baumeister, R. F., & Mele, A. R. (2011). Free will in everyday life: Autobiographical accounts of free and unfree actions. Philosophical Psychology, 24(3), 381–394. DOI: https://doi.org/10.1080/09515089.2011.556607
Towers, A., Williams, M. N., Hill, S. R., Philipp, M. C., & Flett, R. (2016). What makes for the most intense regrets? Comparing the effects of several theoretical predictors of regret intensity. Frontiers in Psychology, 7, 1941. DOI: https://doi.org/10.3389/fpsyg.2016.01941
Vierkant, T., Deutschländer, R., Sinnott-Armstrong, W., & Haynes, J.-D. (2019). Responsibility without freedom? Folk judgments about deliberate actions. Frontiers in Psychology, 10, 1133. DOI: https://doi.org/10.3389/fpsyg.2019.01133