Empathy from Davis’s Multidimensional Model

According to ^{Fernández-Pinto et al. (2008}), two main approaches have been identified in the history of empathy research: the cognitive versus affective perspective, and the integrative approach viewing empathy as both cognition and emotion. The first approach is characterized by various conceptualizations that emphasize either the cognitive or the affective dimension of empathy; in contrast, the second approach encompasses conceptualizations that integrate both dimensions. From this integrative standpoint, ^{Davis (1980}) proposed that empathy consists of a set of cognitive and emotional reactions in an observer who perceives another’s experiences. He further described empathy as a complex, multidimensional construct comprising four components: perspective‑taking, defined as the tendency to adopt others’ viewpoints; fantasy, reflecting the inclination to identify with fictional characters in books, films, or theatrical works; empathic concern, which involves feeling compassion, care, and warmth in response to others’ negative experiences; and personal distress, referring to the propensity to feel discomfort or anxiety at others’ suffering. The first two components are considered cognitive, while the latter two belong to the affective domain. Although subsequent integrative conceptualizations share Davis’s vision, they sometimes include additional dimensions or frame empathy as a dynamic process. For example, ^{Decety and Jackson (2004}) incorporate mental flexibility and emotion regulation into their model.

To measure these four facets, ^{Davis (1983}) developed the Interpersonal Reactivity Index (IRI), which has become one of the most widely used scales in empathy research (^{Fernández-Pinto et al., 2008}; ^{Israelashvili & Karniol, 2018}). One methodological trend has been to compute an overall empathy score by averaging two or more IRI subscales. For instance, some studies examining prejudice and empathy have averaged all four components-perspective‑taking, empathic concern, fantasy, and personal distress-to yield a single empathy measure (^{Álvarez-Castillo et al., 2018}; ^{Bäckström & Björklund, 2007}), whereas others have averaged only the cognitive subscales (^{Díaz-Lázaro et al., 2014}; ^{Díaz-Lázaro & Toro-Alfonso, 2013}; ^{McFarland, 2010}; ^{Nicol & Rounding, 2013}). However, this approach poses three main problems: first, by combining cognitive and affective components, it obscures the independent contributions of each dimension to other variables; second, by selectively omitting fantasy and/or personal distress from certain composites, it nullifies the specific relevance of those factors; and third, by collapsing all subscales into one score, it prevents any analysis of the interrelations among the distinct components of empathy (^{Israelashvili & Karniol, 2018}).

Empathy from the Affect‑to‑Cognition Model

^{Israelashvili and Karniol (2018}) noted that it remains unclear whether affective factors precede cognitive ones in empathy, as empirical findings have been inconsistent. To address this, they tested two statistical models based on ^{Davis (1980}) multidimensional framework. Their results provided stronger support for a model in which affective components influence cognitive ones, leading them to propose the Affect‑to‑Cognition Model of Empathy. In this model, empathic concern and personal distress independently predict perspective‑taking and fantasy. They argue that empathic concern and personal distress arise when one perceives another’s difficulty: concern fosters perspective‑taking, whereas distress inhibits it (^{Israelashvili & Karniol, 2018}).

A key theoretical contribution of the Affect‑to‑Cognition Model is that it establishes both conceptual and empirical links among the empathy components, thereby enhancing the precision of related research and interventions. Methodologically, the model restores the complete and independent use of all four IRI subscales, preserving empathy’s multidimensional conceptualization instead of collapsing them into a single aggregate score.

However, despite using multiple samples, the Affect‑to‑Cognition Model has only been empirically tested by ^{Israelashvili and Karniol (2018}). Nonetheless, the model aligns with neuroscientific research showing that affective components of empathy develop earlier than cognitive ones and constitute the initial subjective experience of empathy (^{Decety & Holvoet, 2021}; ^{Lamm et al., 2007}). Another limitation is that the predictive relationship between affective components and fantasy remains conceptually ambiguous. Although not consistently supported across Israelashvili and Karniol’s three samples, this link merits further theorizing to bolster the model’s coherence.

It is possible that personal distress triggers the activation of fantasy as a means of regulating stress or anxiety in response to another’s difficult situation (^{Shiota & Nomura, 2022}). This interpretation is consistent with research showing that imagination or fantasy can be employed as a coping mechanism under conditions of stress or uncertainty (^{Rubinstein et al., 2021}; ^{Rubinstein et al., 2023}).

Likewise, empathic concern may enhance fantasy or identification with others to better understand their emotions and thus reduce one’s own discomfort. This is consistent with research linking fantasy to emotion recognition (^{Bukach et al., 2018}; ^{Namba et al., 2021}) and correlating empathic concern with emotional recognition (^{Israelashvili et al., 2020}).

Thus, although both personal distress and empathic concern may positively predict fantasy, the underlying motivations differ-emotion regulation versus emotional comprehension-with empathic concern typically exerting a stronger effect (β = .38 to .50) than personal distress (β = .15 to .22) (^{Israelashvili & Karniol, 2018}). Personal distress may exert a weaker influence on fantasy because individuals can also mitigate distress by avoiding the stressor (^{Israelashvili & Karniol, 2018}). Additionally, personal distress and empathic concern often correlate positively, as both arise from emotional reactions to similar situations (^{Davis, 1983}; ^{Israelashvili & Karniol, 2018}; ^{Israelashvili et al., 2020}).

Cognitive Rigidity: A Direct Predictor of Empathy

In order to develop interventions to promote empathy, some studies have identified its predictors (^{Guilera et al., 2019}; ^{Song & Shi, 2017}). However, few research studies have examined predictors of empathy considering the affect-to-cognition model of empathy due to its relative novelty (^{Israelashvili & Karniol, 2018}; ^{Israelashvili et al., 2020}). While personality research has pinpointed predictors for each empathy facet, these predictors have not been integrated into a unified statistical framework. Openness to experience is particularly relevant, as it influences all empathy facets-especially perspective‑taking and personal distress-showing greater stability and effect size in this regard (^{Guilera et al., 2019}; ^{Melchers et al., 2016}; ^{Song & Shi, 2017}).

Cognitive rigidity, which integrates openness to experience and other individual differences, is defined as a cognitive style and capacity marked by a heightened need to resolve uncertainty, complexity, and ambiguity, and a preference for certainty, structure, order, and closure; with openness to experience being a negative expression of rigidity (^{Jost et al., 2009}; ^{Jost et al., 2003}; ^{Rottenbacher, 2012a}). Cognitive rigidity may negatively impact perspective taking by reducing or inhibiting openness to others’ thoughts and emotions in complex or uncertain contexts, thereby undermining cognitive needs for certainty and order. This view is corroborated by evidence that openness to experience predicts perspective taking (^{Di Fabio & Kenny, 2021}; ^{Levin et al., 2016}; ^{Song & Shi, 2017}) and aligns with ^{Decety and Jackson (2004}), theory that empathic understanding requires mental flexibility.

Cognitive rigidity’s contribution to personal distress may be explained by viewing personal distress as a manifestation of rigidity in interpersonal contexts. The distress that arises in response to others’ crises likely reflects the perception of those situations as uncertain and unpredictable stimuli -that rigid individuals seek to avoid or minimize (^{Davis, 1980}; ^{Rottenbacher, 2012a}). This interpretation is consistent with prior evidence showing that openness to experience -a low cognitive rigidity indicator- negatively predicts personal distress (^{Song & Shi, 2017}).

Theatrical Improvisation Practice: Direct and Indirect Predictor of Empathy

Certain interventions impact not only cognitive rigidity but also empathy facets. Meta‑analytic reviews indicate that mindfulness and theater training enhance empathy (^{Hu et al., 2022}; ^{Lewandowska & Węziak-Białowolska, 2022}). Theater practice is especially pertinent, as it affects both cognitive and affective empathy facets as well as related social and communication skills (^{Schmidt et al., 2021}). Theatrical Improvisation, in particular, is defined as a theater technique-practiced individually or in groups-entailing the spontaneous creation of scenarios, narratives, or performances based on ongoing agreements among improvisers, collaborators, and/or audience (^{Argentino, 2013}).

Regarding the influence of improvisation practice on cognitive rigidity, in two experiments with samples of university students it was found that this practice decreased the levels of two expressions of cognitive rigidity, which is explained by the comfortable encounters with uncertainty or ambiguity, in addition to the demand for cognitive flexibility that continuous improvisation practice implies (^{Baldeon & Espinosa, 2024}; ^{Felsman et al., 2020}; ^{Felsman et al., 2023}; ^{Hainselin et al., 2018}; ^{Lewis, 2012}; ^{Lewis & Lovatt, 2013}). If personal distress reflects rigidity, improvisation should likewise reduce distress. Indeed, quantitative studies report that improvisation lowers anxiety levels (^{Felsman et al., 2019}; ^{Felsman et al., 2023}). Complementarily, from a qualitative approach, perceptions that improvisation reduces stress and even personal distress have been reported (^{Stewart, 2016}; ^{Zelenski et al., 2020}).

Regarding the contribution of improvisation practice on perspective taking, although there is no quantitative evidence of this correlation or prediction, from a qualitative approach, there have been reports of perceptions of perspective taking learning and higher levels of empathy in Peruvian adolescents (^{Bachani, 2020}) and U.S. undergraduates (^{Stewart, 2016}) who practiced improvisational theatre, which is explained by perceptions of greater listening, consideration and understanding of other people after practicing improvisation. Furthermore, two core improvisation principles-active listening and unconditional acceptance-are proposed to foster perspective‑taking skills by promoting emotion recognition and acceptance (^{Cai et al., 2019}; ^{Davis, 1980}; ^{Shivarajan & Andrews, 2021}).

Finally, although no studies have directly examined improvisation’s effect on the fantasy subscale, it is possible that regular identification with fictional characters enhances fantasy levels. This hypothesis is supported by findings of significant fantasy differences between general populations and those practicing role‑play techniques (^{Rivers et al., 2016}; ^{Schmidt et al., 2021}). Notably, existing improvisation research has focused on practice effects rather than cumulative practice, underscoring the need to examine practice time.

The importance of researching empathy and its associated factors

Historically, conflicts have persisted worldwide. As of February 2024, over 240,000 deaths have been attributed to armed conflicts in Gaza and Ukraine (^{Khatib et al., 2024}; ^{The Lancet, 2023}). Promoting a culture of peace to prevent violent conflict and resolve disputes through dialogue is therefore imperative (^{United Nations Educational, Scientific and Cultural Organization, 2002}). Empathy plays a crucial role in both personal and social transformation toward peace (^{de Rivera et al., 2011}). However, global challenges-such as the COVID‑19 pandemic and financial crises-have heightened anxiety and stress, potentially diminishing empathy (^{Arceneaux, 2017}; ^{Nair et al., 2024}; ^{Nitschke & Bartz, 2023}). Investigating strategies to counteract these effects and foster empathy -and, by extension, peace- is thus paramount.

In Peru, discrimination remains a persistent conflict. National surveys report that one in three Peruvians has experienced discrimination, 53 % perceive their compatriots as racist, and 63 % of LGBTI individuals report discrimination (^{Instituto Nacional de Estadística e Informática (2017}, ²⁰¹⁹; ^{Ministerio de Cultura, 2018}). Prejudice is a key explanatory factor (^{Agadullina et al., 2022}; ^{Pauwels & Williamson, 2022}).

^{McFarland (2010}) identified empathy as one of the three major predictors of generalized prejudice, although subsequent findings have been mixed (^{Álvarez-Castillo et al., 2018}; ^{Bäckström & Björklund, 2007}; ^{Díaz-Lázaro & Toro-Alfonso, 2013}; ^{Nicol & Rounding, 2013}). Interventions aimed at promoting empathy have also been reported to reduce prejudice (^{Matera et al., 2021}; ^{Miklikowska, 2018}; ^{Olivier et al., 2019}; ^{Suarez et al., 2024}; ^{Vezzali et al., 2015}) , but prejudice is also shaped by political conservatism, socioeconomic context, historical factors and other variables (Esponisa et al., 2007; ^{Jost et al., 2003}; ^{Sidanius et al., 2004}). Given empathy’s role in prejudice reduction, further research into enhancing empathy is warranted.

Beyond conflict prevention, empathy contributes to health promotion. It directly predicts mental health in young samples (^{Chung et al., 2021}; ^{Li et al., 2024}; ^{Niu et al., 2023}; ^{Wang et al., 2023}), and indirectly predicts subjective well‑being, positive relationships, and social adjustment through its influence on prosocial behavior (^{Chen, 2023}; ^{Fu et al., 2022}; ^{Pang et al., 2022}). Thus, investigating empathy and its associated factors is crucial due to its individual and societal impact.

Objective of the study

The objective of this study was to analyze the statistical contribution of months of theatre improvisation practice and cognitive rigidity on empathy from the affect -to-cognition model of empathy using an integrated statistical model. The hypotheses stated are shown in Figure 1.

Figure 1: Hypothesised model 

Design

The present study employed a non‑experimental, cross‑sectional, correlational‑predictive design. Although two groups of participants were included, they were not randomly assigned as would be required in an experimental design. Instead, this approach allowed examination of the relationships outlined in the theoretical framework at a single point in time (^{Hernández et al., 2010}).

Participants

Non-probabilistic convenience sampling was used. The sample consisted of 204 Peruvian young people and adults. The participants were selected according to their belonging to two groups: one group of practitioners of theatre improvisation in the last 6 months (Group does practice improvisation) and another group of non-practitioners (Group does not practice improvisation). Improvisers were recruited through instructors at ten improvisation schools, whereas non‑practitioners were contacted at universities and workplaces. The total sample size provided the minimum statistical power required for path analysis (^{Kline, 2011}). Sociodemographic characteristics of the sample are presented in Table 1.

Table 1: Socio-demographic data of the simple 

Measures

Sociodemographic Questionnaire. Information was collected on age, sex, professional career, theatre improvisation practice in the last 6 months and improvisation practice time over the life span.

Interpersonal Reactivity Index (IRI; ^{Davis, 1983}). Empathy was assessed using the Peruvian adaptation of IRI (^{Acasiete, 2015}). The 28‑item scale uses a 5‑point response format (1 = Does not describe me well; 5 = Describes me very well). Adequate reliability coefficients were found for the dimensions studied: perspective taking α = .75, empathic concern α = .73, personal distress α = .73 and fantasy α = .70.

Integrated and reduced cognitive rigidity questionnaire. A survey was developed that integrated several items from instruments used in previous studies to assess cognitive rigidity in a more reduced form (^{Baldeon, 2020}; ^{Rottenbacher, 2012a}, ^2012b). It is composed of 20 items whose response option is a 6-point Likert scale (1= Strongly disagree and 6= Strongly agree). The overall score has a good level of reliability (α = .88).

Data collection procedure

Data were collected individually and in person between March and June 2019. Participants completed the printed questionnaires in approximately 15 minutes on average.

Half of the participants were contacted with the collaboration of teachers from 10 theatre improvisation schools and the other half were contacted at universities and workplaces in Lima. The ethical standards of the Pontifical Catholic University of Peru were followed, based on the Helsinki ethical guidelines for research involving human subjects. In addition, the ethical considerations of the Psychology Ethics Committee 2019 were followed and a declaration of commitment to the ethical principles of the Research Ethics Committee of the Pontificia Universidad Católica del Perú was signed. All participants received informed consent.

Data Analysis

The data were processed and analyzed with the statistical software R Studio and the Lavaan package (version 0.5-23.1097). Reliability, Pearson's correlation and path analyses were performed, using the Maximum Likelihood estimator because the variables studied showed a normal distribution (^{Kline, 2011}; ^{Li, 2016}). Model fit was evaluated against the following criteria: χ²/df < 2; Bentler-Bonett Comparative Fit Index (CFI) > .95; Incremental Fit Index (IFI) > .95; Steiger-Lind Root Mean Square Error of Approximation (RMSEA) < .08 and Standardized root mean square residual (SRMR) < .08 (^{Kline, 2011}).

Affect‑to‑Cognition Model of Empathy in the Latin American Context

First, hypotheses 1 through 5-concerning the interrelations among perspective‑taking, fantasy, empathic concern, and personal distress posited by the Affect‑to‑Cognition Model-were confirmed, indicating that this framework can be generalized to the Latin American context while preserving its core structure (^{Israelashvili & Karniol, 2018}).

According to the results, personal distress contributes negatively to perspective taking variability, but positively to fantasy. While this is consistent with the study by ^{Israelashvili and Karniol (2018}), it is not fully clarified in that study. Therefore, the present research suggests that it is possible that personal distress increases fantasy as a form of escape from the situation that produces distress, regulating one's own emotions, while empathic concern drives fantasy to facilitate the emotional understanding of other people (^{Bukach et al., 2018}; ^{Israelashvili et al., 2020}; ^{Namba et al., 2021}; ^{Shiota & Nomura, 2022}). This conceptual refinement advances the theoretical precision of the Affect‑to‑Cognition Model by delineating distinct functional roles for its affective predictors.

Second, these findings carry practical implications for empathy training. Interventions aiming to enhance empathy should incorporate activities that (a) mitigate the detrimental effects of personal distress through targeted emotion‑regulation strategies and (b) cultivate empathic concern, given its stronger statistical contribution to cognitive facets of empathy. This dual focus aligns with evidence linking emotion regulation and empathic concern to reduced prejudice (^{Bobba & Crocetti, 2022}; ^{Burns et al., 2016}; ^{Levin et al., 2016}; ^{Makwana et al., 2021}).

Cognitive Rigidity from a Multidimensional Perspective

Third, hypotheses 6 and 7 concerning cognitive rigidity were supported: cognitive rigidity negatively predicted perspective taking and positively predicted personal distress, consistent with prior research (^{Di Fabio & Kenny, 2021}; ^{Levin et al., 2016}; ^{Rottenbacher, 2012a}; ^{Song & Shi, 2017}). Particularly, the finding of the contribution of cognitive rigidity on personal distress contributes to the scarce line of research on cognitive rigidity (^{Rottenbacher, 2012a}). Moreover, this result indicates that cognitive rigidity is not a purely cognitive concept as it also influences affective reactions such as personal distress. Thus, it is possible that emotional uncertainty, an affective dimension of cognitive rigidity, has a stronger association with personal distress than other dimensions; similarly, perspective taking may have a stronger correlation with more cognitive dimensions of cognitive rigidity, such as openness to experience (^{Jost et al., 2003}; ^{Rottenbacher, 2012a}; ^{Song & Shi, 2017}).

Fourth, conceptualizing cognitive rigidity as a multidimensional construct suggests potential interactions with other affective predictors of empathy. Empirical evidence indicates that specific facets of cognitive rigidity influence emotional self‑regulation (^{Alghamdi et al., 2017}; ^{Kumar & Tankha, 2023}), which in turn impact in empathy (^{Makwana et al., 2021}; ^{Zirenko & Krasavtseva, 2021}). These findings support the inclusion of cognitive rigidity within broader Affect‑to‑Cognition models of empathy prediction.

Fifth, the contribution of cognitive rigidity in the variability of empathy has practical implications as interventions designed to increase empathy and reduce prejudice could include techniques to decrease cognitive rigidity (^{Hillen et al., 2017}), as it has been found to have a direct influence on prejudice (^{Sassenberg et al., 2022}; ^{Sekerdej et al., 2018}).

Direct and Indirect Contributions of Improvisation Practice

Sixth, although people who practice theater improvisation may be characterized by lower cognitive rigidity before practicing it, the results of the study and the experiments of ^{Felsman et al. (2020}; ²⁰²³) indicate that the time spent practicing theater improvisation contributes negatively to cognitive rigidity, decreasing its levels. Thus, hypothesis 8 is corroborated. Also, consistent with reported differences between people in general and those who practice theater and role-playing (^{Rivers et al., 2016}; ^{Schmidt et al., 2021}), months of improvisation practice directly and positively predict fantasy; that is, identification with fictional characters by interacting with them through improvised and fictional stories (^{Argentino, 2013}). Thus, hypothesis 11 is confirmed. These findings suggest that improvisation practice time contributes significantly in individual cognitive variables.

Seventh, contrary to what was expected in hypothesis 9, the time spent practicing improvisation has no predictive relationship with personal distress, which can be attributed to the greater influence of this practice on cognitive reactions compared to affective ones. Although the experiments of ^{Felsman et al. (2020}; ²⁰²³) report that improvisation practice reduces tolerance to uncertainty, an expression of cognitive rigidity, it is possible that a more detailed analysis of the cognitive, emotional, and behavioral dimensions of rigidity will confirm this conjecture. Furthermore, although cognitive rigidity has a predictive relationship with personal distress with a large effect size (.58), personal distress also receives statistical contribution from other variables not contemplated in the study, on which improvisation practice may not have any contribution, even if the practice is performed over a period of months.

Similarly, Hypothesis 10 was not supported: practice duration did not predict perspective‑taking, a result at odds with qualitative reports of enhanced perspective‑taking following improvisation training (^{Bachani, 2020}; ^{Stewart, 2016}). This can be attributed to the insufficiency of improvisation practice to train a complex, multi-causal reaction such as perspective taking. Although theatrical improvisation practice contributes to the variability of cognitive rigidity (^{Felsman et al., 2020}; ^{Felsman et al., 2023}) and consequently contributes to greater openness to the thoughts and feelings of others in complex situations, skills such as recognition, clarification, regulation and understanding of emotions, as well as compassion, are required for perspective taking. (^{Davis, 1980}, ¹⁹⁸³; ^{Israelashvili et al., 2020}; ^{Makwana et al., 2021}; ^{Shiota & Nomura, 2022}). Although theatre improvisation training can cover emotion management, the approach tends to be reduced, intuitive and ephemeral because the aim of theatre improvisation is to create improvised stories without necessarily delving into the stories, their fictional characters or emotions (^{Argentino, 2013}; ^{Baldeon, 2020}).

Eighth, nevertheless, the indirect contribution of improvisation practice on empathy through cognitive rigidity has practical implications as exercises that are used to train improvisation can be used and targeted to develop empathy, complemented by training in other relevant skills such as emotion management. In addition, such exercises can be combined with techniques like exposure to counter-stereotypical exemplars, one of the most effective interventions for reducing prejudice according to a review of 30 studies (^{FitzGerald et al., 2019}). For instance, requiring participants to embody or interact with counter‑stereotypical characters within a playful, accepting framework may bolster empathic responses while preserving the spontaneity that defines theatrical improvisation (^{Argentino, 2013}; ^{Felsman et al., 2023}).