In parallel with the product development of Empathy VR at Helsa, our research team is focused on developing the most effective and accurate way of measuring its impact. We are focusing on measuring empathy levels, before and after the intervention.
By Elli Fowler • 6th December 2022
Mental Health Research Analyst
In parallel with the product development of Empathy VR at Helsa, our research team is focused on developing the most effective and accurate way of measuring its impact. Accurately measuring impact is a notoriously difficult task for any soft-skills and/or equity, diversity and inclusion (EDI) training company. In order to evaluate the impact of Empathy VR when delivered, we are focusing on a means by which to measure empathy, before the intervention and at several intervals afterwards.
Empathy VR is an immersive soft-skills and EDI training product using virtual reality technology to increase empathy towards minority groups by allowing the user to experience scenarios of racism, sexism, homophobia and transphobia from a first person perspective.
The concept of empathy is a complicated one, with no single definition. It is generally perceived as the ability to understand and experience the feelings of others. Measuring empathy is not a simple task, however, and there are many methods that have been developed in an effort to capture it. This report covers the three main categories of empathy measurement, namely 1) self report measures, 2) behavioural methods and 3) physiological methods. Various examples in each category are highlighted and evaluated with regard to their ability to measure empathy accurately. Ultimately, the conclusion from this body of evidence is that no existing measure alone currently captures empathy in a way that is suitable for Empathy VR, but instead a novel self report tool will be developed that encompasses elements of the Toronto Empathy Questionnaire and an implicit bias scale, that can then be triangulated using physiological methods. This will allow for the impact of Empathy VR to be most accurately and appropriately captured among its users.
Whilst the importance of empathy in human social systems is widely acknowledged and accepted, its exact definition remains vague. The broad perception is that empathy encompasses the ability to understand the emotions of someone else; the ability to walk in someone else's shoes (Gerdes, et al. 2010). Sociologically speaking, empathy has been an important tool in the development of human social systems, facilitating the building of relationships and large group living. Empathy is consistently linked to the occurrence of prosocial behaviour and overall wellbeing, and generally is perceived as a thoroughly positive human trait.
Empathy has two capacities, cognitive and emotional, both of which are required for effective social interactions. Whilst cognitive empathy can be acquired and learnt, emotional empathy is integrated within human development, but remains heavily influenced by one’s own perspectives and life experiences (Reiss, 2017). Lack of emotional empathy can occur when there are racial, ethnic, religious or physical differences and is heavily impacted by conscious or unconscious bias. We are most likely to empathise with those who look like us and have the most similar experiences to our own.
For individuals with a lower level of emotional empathy, there is the ability to compensate with cognitive empathy. Perspective taking through imagining the world through another person's eyes is considered a powerful approach in invoking much needed cognitive empathy.
As the corporate sphere becomes increasingly aware of its responsibility to provide diverse and inclusive workspaces, both for employee wellbeing and economic success, many companies have chosen to employ EDI Training. This type of training is designed to improve tolerance and acceptance in the workplace, with the aim to make workplaces healthy environments for people from all groups and walks of life. However, as a field, EDI training is often criticised for being ineffective. Though short term outcomes tend to be positive, by several weeks later, the effects of the training have too often all but disappeared. As a result, many workplaces remain ineffective at providing inclusive environments. Fundamentally, EDI is the key to creating representative and successful companies in the modern corporate sphere, but it cannot be maintained without empathy, and so, a novel approach is needed.
Virtual reality as a technology is slowly creeping its way into the public sphere, primarily through the gaming industry, allowing players to experience an immersive version of the world in which they are playing. Virtual realities can be effective means of eliciting an emotional response from its users. With regard to its use for eliciting empathy in particular, a powerful element of the VR experience is embodiment. Embodiment is, quite simply, a person’s sense of possessing a body. The human brain creates a simulation of a body within the world and thus experiences that environment as if physically present (Wiederhold, 2020).
This sense of self can be manipulated either by the individual themselves or externally for therapeutic reasons. VR has the potential to enhance this process and create virtual versions of the body, thus allowing someone to temporarily become someone else. If empathy is as conceptualised above, namely the feeling of seeing the world from someone else's perspective, VR provides an ideal medium to allow virtual embodied experiences to be created. This will allow people to quite literally feel the experiences of others, thus facilitating greater empathetic response and understanding, without having to rely on their own imagination or experiences (Wiederhold, 2020).
VR has shown significant promise in increasing altruistic intention and compassion, with examples of reducing short term implicit racial bias or altering the behaviour of aggressive populations (Schoeller, et al. 2019). One example is The Machine To Be Another (TMBA), which allows people to experience the lives of migrants, religious prejudice and victims of military violence. This technology has been successfully used to foster compassion towards humans often considered to be outgroups. This shows that virtual reality has the potential to be hugely influential in fostering more empathetic attitudes towards social minority groups (Bertrand, et al. 2018).
Empathy VR is built on the principle that virtual reality has significant potential in promoting more empathetic attitudes towards minority groups. Empathy VR serves to simulate the lives of a range of different minority groups, with the aim of highlighting the impacts of sexism, racism, homophobia and transphobia on their day to day lives. Being aware of lived experiences of minority groups will aid in being able to empathise with them and increase people’s understanding of the stigma and discrimination minority groups experience. Additionally, it will improve the wellbeing of these groups through increased empathy and decrease of prejudices and implicit bias. Empathy VR will be transformative for any company seeking to create a more inclusive and empathetic workplace environment.
To assess the impact of Empathy VR’s immersive training workshops it is imperative to measure empathy in the participants before and after the training. However, measuring empathy is a notoriously difficult task due to both the lack of consensus on its definition and its highly subjective, nuanced nature. There are many different development frameworks designed to measure empathy, which broadly can be split into three categories: 1) self report, 2) behavioural/observational methods, and 3) physiological methods (Gerdes, et al. 2010). For each category, key examples will be highlighted to demonstrate their use and validity in measuring empathy. We therefore developed our own scale based on this body of evidence.
Self report methods of measuring empathy encompass the very simplest, pen-to-paper types of psychometric measurement. They are the easiest and thus the most common method utilised, both clinically and vocationally. There are a vast number of self report questionnaires that exist designed to capture human empathy, however, those highlighted are those which show the greatest potential for accurately measuring empathetic affect.
The first two descriptions of empathy scales are to provide insight into the origins of this convoluted science, and to demonstrate that whilst we sit over 50 years on from these early examples, we are yet to fully establish an accepted, all encompassing measure of empathy. The following examples will be of scales developed in more recent decades that reflect our growing understanding of what human cognitive and affective (emotional) empathy actually entail.
The Carkhuff & Truax scales of empathy were developed as tools to evaluate psychologists and therapists in their ability to empathise with their patients. They are an example of a number of such therapist empathy rating scales that offer the advantage of an objective report of empathy, less likely to be impacted by personality bias on the behalf of the subject (see section Comparing Measures of Empathy for comments on the advantage of informant based empathy measures). However, they are generally considered to only reflect certain cognitive components of empathy and overly focused on responses from the subject, without taking wider behaviour into account (Decker, et al. 2014)
The Empathy Scale is one of the earliest adopted measures of empathy, and continued to be used long after its conception, an impressive feat given the discrepancies about what empathy actually entailed in the literature of the time. This scale encompassed four distinct components of empathy; social self confidence, even temperedness, sensitivity and non-conformity (Froman & Peloquin, 2001). Despite its adoption as a scale of empathy, more recent analyses have demonstrated poor levels of validity (Froman & Peloquin, 2001). It has instead been suggested to be a better measure of social and interpersonal skills as opposed to “true emotional empathy” per se. Whilst social skills and empathy are considered to be intrinsically linked, the former cannot be used as an absolute proxy for the latter.
One of the earliest “replacements” for the scales described above is the Interpersonal Reactivity Index (IRI). The IRI contains four subscales: (1) perspective taking, (2) fantasy, (3) empathetic concern and (4) personal distress. These subscales are then paired up, with the idea being that these pairs target the dual nature of empathy respectively. So the pair of (1) perspective taking and (2) fantasy are geared towards the measurement of cognitive empathy, whereas (3) empathetic concern and (4) personal distress are more aimed at capturing affective/emotional empathy (Stocks & Lishner, 2012).
Breaking down the scale into subdimensions falls in line with scientific best practices, accounting for the complex nature of empathy by acknowledging both cognitive and emotional empathy, due to the former being more of a learned skill than the latter..
Whilst the principles are widely commended, there are criticisms of the IRI itself. One such observation is that the fantasy component of the scale better measures imagination than empathy. The second is that the personal distress component does not measure actual personal distress caused by emotional empathy but rather emotional self control. Though emotional self control is a component of social behaviour, it is generally not considered to be a central component of empathetic affect (Neumann, et al 2015)
Despite these issues, the empathetic concern and perspective taking subscales of the IRI are considered to be good measures of emotional and cognitive empathy respectively, and are commonly used to validate or critically evaluate novel empathy questionnaires.
The Empathy Quotient (EQ) can be considered an antagonist to Baron-Cohen and colleagues’ other infamous scale, the Autism Quotient (AQ). This is based on the principle that Autism Spectrum Disorder (ASD) can be evaluated by a deficit in empathy, and thus those with ASD will score highly on the AQ and low on the EQ, whereas those with higher empathy will have opposite results. This is a somewhat controversial perception of ASD, however, it is possible to evaluate the EQ in isolation to establish whether it is a valid test of empathy.
These authors chose to define empathy as “the drive to identify another person’s emotions and thoughts, and to respond to these with an appropriate emotion” (Baron-Cohen & Wheelwright, 2004). The definition they have chosen to embody however could well be considered more a definition of cognitive empathy alone, namely understanding other people's emotions and knowing the correct response, something that can be a learned response, as opposed to an innate emotional response to another's emotional state. That said, the survey itself was designed to be a short, easily interpretable survey targeted at evaluating both the cognitive and emotional aspects of empathy. It consists of 60 items, with 40 to test empathy and 20 control questions designed to prevent emotional overwhelm for the participants. These control questions have the additional advantage in allowing for researchers to check for response bias.
In terms of reliability the EQ scores well, with reported Cronbach alpha values ranging from 0.78 to 0.92 and test-retest reliability coefficients ranging from 0.84 to 0.97 (Neumann, et al 2015). In terms of validity, the EQ converges well with a range of other scales, including the perspective taking subscale of the IRI, a commonly used validity tool as mentioned above. It does diverge with the IRI personal distress subscale however, which as mentioned may be considered more a measure of emotional self control, so does not necessarily indicate a deficit of the EQ’s ability to measure empathy (Neumann, et al 2015).
A major observation with the EQ however is that it displays significant sex differences, with female participants on average scoring much higher scores than male participants. Whether sex differences in scores indicates a positive or negative capture of empathy is a matter of personal opinion. Some believe that there is no inherent difference in empathy between men and women, and so discrepancies in scores show an inherent female bias in the scale. However, on the other hand, some researchers, such as Baron-Cohen himself, consider women to be significantly more “naturally empathetic” than men, something that is then compounded by social gender stereotypes of women being kind and caring. If this is accepted, then a discrepancy in scores between men and women on a scale of empathy would be considered a positive indication of empathy measurement.
This is a controversial topic for multiple reasons. Firstly, as a society, perceptions of sex, gender and stereotyped perceptions of behavioural expression are significantly developing, and so such binary comparisons are being moved on from. Secondly, Baron-Cohen himself is the propagator of the controversial “extreme male brain” theory of autism, which ties into the concept that ASD represents a deficit in empathy, and that there is a higher level of ASD diagnosis in the male population than the female (Baron-Cohen, 2002). As a theory, this understandably has a range of intense criticisms.
Therefore, due to these factors, sex differences in empathy scales will not be further discussed, as their relevance is highly debated, and not useful in developing a novel, more inclusive and representative scale of measuring empathy.
The Questionnaire of Cognitive and Affective Empathy (QCAE) aims to build on previous measures of empathy that were considered to be too narrow in the definition of empathy that they used. Designed to measure both the cognitive and emotional aspects of empathy, this 31 item questionnaire measures with a four-point forced response scale, akin to a Likert Scale.. Based on a mixture of the EQ, The Empathy Scale and the IRI, this questionnaire aims to incorporate the best components of each for maximum accuracy, validity and reliability for measuring empathy (Reniers et al, 2011).
However, for this goal, the QCAE falls somewhat short. Its internal consistency is fair, with reported Cronbach’s alpha ranging from 0.65 to 0.85, however its test-retest values are generally poor, with only the original authors being able to demonstrate reliability and validity. Additionally, despite being based on multiple previous measures of empathy, the only one with which the QCAE shows significant convergence is the Basic Empathy Scale, which is the oldest and least accurate of the questionnaires (Nuemann, 2015).
Though the principle of combining the successful aspects of previous questionnaires is a sound one, for this particular questionnaire, the desired outcome has not quite been achieved.
One scale that was built using a similar principle of combining previous methods is the Toronto Empathy Questionnaire (TEQ). The researchers behind the TEQ factor analysed responses from 11 self report measures of empathy and built a questionnaire based on the outcome. The result is a 16 item scale scored using a 5 point Likert type scale, generally considered to lean more heavily towards the emotional aspects of empathy. Generally the test shows moderate to good reliability (Cronbach's alpha ~ 0.85) and validity (stability coefficient after time interval ~ 0.81) (Spreng, 2009).
In terms of convergence with other tests, the TEQ correlates positively with IRI Empathetic Concern, IRI perspective taking and EQ scores, as well as showing a negative correlation with the Autism Quotient (which would be expected given the principle on which the scale was built) (Neumann, 2015). The positive correlation with IRI Empathic concern is no surprise given several items on the TEQ are reworded versions of questions from the former, however the positive correlation with Perspective Taking is more notable, as zero questions were adapted from this section. The only notable divergence the TEQ has been recorded as having is with IRI Personal Distress (Neumann, 2015), however, as mentioned, many do not consider this subscale to be a central component of empathy. Subsequently, this observation does not necessarily negate the effectiveness of the TEQ in measuring emotional empathy.
The authors of the TEQ consider it a representation of the very best of what came before it in terms of empathy measurement, however they emphasise that in the vast majority of situations in which empathy is being measured, it is best to engage a multi faceted approach. It is concluded that unidimensional concepts of empathy are neither useful nor accurate, therefore the frameworks used to measure it should reflect this (Spreng, 2009).
Though self report methods of measuring empathy are the most common due to their ease of delivery, there are a number of other methods that have been created and tested for measuring empathy. They broadly fall into two categories; behavioural or observational methods that involve observing participants' reactions to emotional scenarios and physiological methods that utilise more neuroscientific principles using measures of the nervous system.
N.B. The distinctions between these categories of measures are not always clear. There is sometimes overlap in what quantifies a behavioural vs a self report measure. For example the “Picture Reviewing Paradigm” is classified as a behavioural measure but relies on a self report of empathy from a subject who is exposed to empathetic scenarios.
The Comic Strip Task is a classic example of a behavioural paradigm used to measure empathy. In this test, an individual’s empathy is judged based on how well they can interpret the mental states of other individuals. This is an adaptation of the “attribution of intention” task in which participants must complete a story based on the assumed intentions of the subject. Vollm and colleagues designed comic strips that were designed to test cognitive empathy alongside theory of mind and physical attribution of intention.
The sample used in the development of the paradigm was, however, very small (n=13) and further analyses have been unable to replicate the original study’s positive results. Therefore, though the general concept is well supported and logical, this particular test is poorly validated (Neumann, 2015).
MRI’s are commonly used in many diagnostic processes, but can also be used to image the anatomical features of the brain to gain insight into the various functions different areas might perform. Functional MRI’s have the additional feature of showing live brain activation through blood flow and metabolic factors. For exploring empathy, focus tends to be on the areas of the brain thought to subserve this neurological process, namely the Prefrontal cortex and the amygdalae (Hillis, 2014). Participants can be shown emotional stimuli and the brain scans then used to quantify empathetic response. Standard MRIs cannot show the process of empathy in action but can be used to inform brain regions involved in empathy. fMRIs can in theory show empathy in action, but the general consensus is that fMRI results reflect the neural/emotional responses to empathy, as opposed to empathy itself (Singer, 2006).
Electromyography is the measurement of electric potentials produced by muscles when they contract. When used on facial muscles, this technique can allow for the capture of facial expressions and reactions not visible to the human eye. Using this technology to measure facial reactions in response to empathetic situations could be highly insightful for two main reasons. The first is that mimicry is a key component of emotional empathy, and is one of the earliest skills that human infants develop. As a result, the second benefit is that this is considered an innate empathetic response which cannot be consciously prevented. This is therefore the closest we have achieved to observing genuine, non-biased empathetic reactions (Westbury & Neumann, 2008).
Measurements of facial electromyography correlate well with other methods of measuring empathy, however there is some concern in ensuring any muscle contraction observed is due to the reaction to the empathetic scenario rather than any other external stimuli (Neumann, 2015).
It is clear that there are many methods, both past and present, that have been conceptualised as capable of measuring empathy. The question therefore becomes which form of measurement is the best to use, particularly with regard to measuring the impact of Empathy VR effectively.
As mentioned, self report measures are the most efficient models of measuring empathy, due to their ease of delivery, and are unsurprisingly the most commonly used. However, as demonstrated, there is no single self report test for empathy that is widely accepted or considered a “best-fit” measure. Many consider that these measures act as proxies for empathy rather than direct indications of personal empathetic affect (Grainger, 2022). In particular, it has been reported by a number of studies that self report measures do not converge with results from standardised social cognition tasks. This suggests that these measures may not be as effective at measuring cognitive empathy as they may be at emotional empathy.
It has also been suggested that self-report measures of empathy are vulnerable to personal bias and personality “blindspots” (Grainger, 2022). This means that a participant, understanding the fact that empathy is looked upon favourably in society, may overestimate their own empathetic ability and compassion. Whilst a level of bias may be considered somewhat inevitable, it is considered a major downside to self report measures of empathy, if not otherwise accounted for. One potential solution is using informant report measures of empathy, which entails the empathy questionnaire being filled out on behalf of the participant by someone who knows them very well. This approach is less vulnerable to bias and has been demonstrated to be an effective manner of measuring empathy, converging most successfully with measures of broader social functioning such as social engagement (Grainger, 2022). This suggests that an independent, but known, observer may provide the most accurate reflection of an individual empathetic capacity .At the present time, however, this approach is not feasible for Empathy VR’s distribution model.
A meta analysis comparing self report and behavioural methods found that these measures generally do not converge well (Melchers, et al. 2015), adding to a large body of literature that finds these two approaches vaguely incompatible. Other literature reviews have demonstrated however that self report measures tend to converge well with each other and can be validated using physiological methods such as facial electromyography(Neumann, 2015). The behavioural methods however barely correlated either with the self report measures or each other. It was therefore concluded that these behavioural tests should only be used for measuring very specific endotypes of empathy in specific research questions (Melchers, et al. 2015).
The overall conclusion held by authors of these large meta analyses of different measures of empathy tends to warn clinicians or similar away from relying on single approaches, and in a situation where the outcome of these tests are poignant, a multi method assessment approach is likely warranted (Grainger, 2022).
Due to the nature of Empathy VR, and its purpose to effectively promote empathy towards minority groups, it is worth considering incorporation of elements of implicit bias testing. According to social identity theory, implicit bias can be defined as the prejudicial attribution of particular qualities to a member of an outgroup (Greenwald & Banaji, 1995). Therefore, when considering empathy with regards to social outgroups, it may be an effective measure of impact to explore whether Empathy VR can alter implicit biases in participants, which in turn will promote future empathy.
The most popular method for measuring implicit bias is the Implicit Association Test (IAT). The IAT measures the strength of associations between concepts (eg. members of a minority group), evaluations (good vs bad) and stereotypes. It bases itself on the categorisation of words and how quick people are to associate a concept with an evaluation or stereotype. The faster someone associates a concept with a positive evaluation or stereotype, the more likely they are to have an implicit bias towards that concept.
One concern that has been raised with the IAT, however, is that it shows low test-retest reliability, and therefore cannot be used to imply individual differences in implicit biases based on single scores, due to the level of variability. One analysis found that through methods of aggregation, the reliability correlations improved (Carpenter, 2022). This involves taking multiple scores and aggregating them to create a much more reliable interval average. The inclusion of an implicit bias scale is therefore possible, but study should be done to ensure that individual scores do not show too much variability due to factors other than individual differences.
A range of approaches could be taken from the evidence presented above, however there are some key elements that are presented repeatedly within the literature that should be considered.
Use a self report measure: the ease of delivery is a major advantage of this method, particularly given the setting in which Empathy VR will be used in the long term. There is also a large body of questionnaires that already exist, with several (such as the TEQ) based on prior renditions, that show high levels of effectiveness in measuring emotional empathy in particular. Adaptation of one of these existing models is likely to render the most effective outcome.
Cross validate and triangulate: Though the recommendation is to use a self report measure, it is clear that they are not without fault, and thus they require validation by additional methods of empathy measurement. In terms of developing an empathy scale, it is possible to enlist much more complex techniques for the purpose of triangulating the final outcome. The two validation techniques that stand out from the literature would be informant-report measures and physiological methods such as facial electromyography. Whilst using these tools actually in practice is not currently feasible, there is potential for them to become technologically viable in the future, and using them in development of the scale will ensure maximum convergence with measures of social skills.
Incorporate an implicit bias scale: Due to the nature of Empathy VR as a tool for promoting more empathy and acceptance towards minority groups, exploring its impact on implicit bias is an important component. Measuring implicit bias is as difficult in many respects as measuring empathy itself. However, incorporation of an implicit bias scale, based on the Implicit Association Test (IAT) is a viable option that will strengthen the evaluation of Empathy VR’s impact.
Empathy’s nature as a highly subjective, complex, nuanced element of human nature makes it very difficult to measure in a way that is valid and repeatable. However, for the purpose of measuring the impact of Empathy VR, a self report measurement system, developed using cross framework triangulation, with the incorporation of an implicit bias scale is likely to be the most powerful, feasible option.
As mentioned above, a major criticism that circles the diversity and inclusion training available currently to corporate clients is that it only “works” in a very short term sense. The lessons taught in these sessions, whilst valuable ones, simply are not having true impact in terms of long term presence in company behaviour and culture. Empathy VR presents a novel, cutting edge option to companies that will, through its powerful embodiment of the lived experience of minority groups, promote long-lasting empathetic attitudes within the workplace. Measuring this impact is therefore crucial to demonstrate and highlight how Empathy VR stands out from the existing landscape of diversity and inclusion corporate services.
Above, there is significant discussion on the methods of effectively measuring empathy and incorporating those measures into a framework with a scale of implicit bias. In terms of measuring the impact of Empathy VR, this novel tool aims to combine the very best of what came before it, allowing for accurate, reliable measurement of emotional and cognitive empathy alongside implicit biases towards social outgroups.
The principle would be to deliver this self report measure before the empathy VR session to give a baseline for each individual. The same measure would then be delivered immediately after the intervention to observe short term impact. This would be combined with questions regarding the programme itself to collect data on participants' experience of Empathy VR itself to inform future development. Overall this would allow for the impact of Empathy VR to be effectively quantified, and positioned for optimal future development
Long term impact is an area in which previous diversity and inclusion interventions fail to perform. However, Empathy VR, with its powerful capacity for immersive embodiment, aims to have a much more lasting effect. In order to measure this, the same empathy based questionnaire will be delivered at an interval of a number of weeks after the original session. This will allow for insight into whether the impact is lasting and succeeding in changing personal behaviour and actions.
A more ambitious goal in the future landscape of Empathy VR will also be to establish strategic partnerships with corporate clients, which would facilitate exploration of company culture and policy. Policy both leads and is influenced by company culture. Therefore, introduction of inclusive corporate policies such as gender neutral parental leave may be an indicator that there has been an impactful shift in company culture as a result of the experiences of Empathy VR.
Baron-Cohen (2002) The extreme male brain theory of autism, Trends in Cognitive Sciences, Volume 6, Issue 6, Pages 248-254
Baron-Cohen, S., & Wheelwright, S. (2004). The empathy quotient: An investigation of adults with Asperger syndrome or high functioning autism, and normal sex differences. Journal of Autism and Developmental Disorders, 34, 163175.
Bertrand et al (2018) Learning Empathy Through Virtual Reality: Multiple Strategies for Training Empathy-Related Abilities Using Body Ownership Illusions in Embodied Virtual Reality. Front. Robot. AI 5:26. doi: 10.3389/frobt.2018.00026
Carpenter, et al. (2022). Individual Differences in Implicit Bias Can Be Measured Reliably by Administering the Same Implicit Association Test Multiple Times. Personality and Social Psychology Bulletin, 0(0).
Davis, (1983). Measuring individual differences in empathy: Evidence for a multidimensional approach. Journal of personality and social psychology, 44(1), 113.
Decker, S. E., Nich, C., Carroll, K. M., & Martino, S. (2014). Development of the Therapist Empathy Scale. Behavioural and cognitive psychotherapy, 42(3), 339–354.
Froman, R. D., & Peloquin, S. M. (2001). Rethinking the use of the Hogan Empathy Scale: a critical psychometric analysis. The American journal of occupational therapy : official publication of the American Occupational Therapy Association, 55(5), 566–572.
Gerdes, et al (2010) Conceptualising and Measuring Empathy. British Journal of Social Work 40, 2326–2343 doi:10.1093/bjsw/bcq048
Grainger, et al (2022). Measuring Empathy Across the Adult Lifespan: A Comparison of Three Assessment Types. Assessment, 0(0)
Greenwald, Anthony G.; Banaji, Mahzarin R. (1995). "Implicit social cognition: Attitudes, self-esteem, and stereotypes". Psychological Review. 102 (1): 4–27.
Hillis A. E. (2014). Inability to empathize: brain lesions that disrupt sharing and understanding another's emotions. Brain : a journal of neurology, 137(Pt 4), 981–997.
Hogan, R. (1969). Development of an empathy scale. Journal of consulting and clinical psychology, 33(3), 307.
Melchers, et al (2015) Assessment of empathy via self-report and behavioural paradigms: data on convergent and discriminant validity, Cognitive Neuropsychiatry, 20:2, 157-171
Neumann, et al (2015). Measures of Empathy. 10.1016/B978-0-12-386915-9.00010-3.
Reniers, et al (2011). The QCAE: A questionnaire of cognitive and affective empathy. Journal of personality assessment, 93(1), 84-95.
Reiss (2017) The Science of Empathy. Journal of Patient Experience 2017, Vol. 4(2) 74-77
Schoeller et al (2019) Combining Virtual Reality and Biofeedback to Foster Empathic Abilities in Humans. Front. Psychol. 9:2741. doi: 10.3389/fpsyg.2018.02741
Singer (2006) The neuronal basis and ontogeny of empathy and mind reading: Review of literature and implications for future research, Neuroscience & Biobehavioral Reviews, Volume 30, Issue 6, Pages 855-863
Spreng, R. N., McKinnon, M. C., Mar, R. A., & Levine, B. (2009). The Toronto Empathy Questionnaire: scale development and initial validation of a factor-analytic solution to multiple empathy measures. Journal of personality assessment, 91(1), 62–71.
Stocks & Lishner (2012) Empathy, Editor(s): V.S. Ramachandran, Encyclopedia of Human Behavior (Second Edition), Academic Press, 2012, Pages 32-37, ISBN 9780080961804
Truax & Carkhuff (1967) Toward effective counseling and psychotherapy. Chicago: Aldine.
Völlm, et al (2006). Neuronal correlates of theory of mind and empathy: a functional magnetic resonance imaging study in a nonverbal task. Neuroimage, 29(1), 90-98.
Westbury & Neumann (2008). Empathy-related responses to moving film stimuli depicting human and non-human animal targets in negative circumstances. Biological psychology, 78(1), 66-74.
Wiederhold (2020) Embodiment Empowers Empathy in Virtual Reality. Cyberpsychology, behaviour and social networking. Volume 23, Number 11.