Empathy is an essential skill in our social lives, allowing us to understand and share others' feelings. In the last decade, empathy has been a hot topic in cognitive neuroscience for its important role in social interaction. To date, there are a variety of definitions of the term "empathy", such as "a reaction to the observed experiences of another" and "the ability to experience and understand what others feel without confusion between oneself and others". Based on different definitions of empathy, different methodologies have been developed and adopted in the field of empathy research. Valid empathy questionnaires are developed to provide ones' subjective evaluation about their empathy ability (i.e., trait empathy); behavioral paradigms are designed to measure individuals' empathy levels under a specific context (i.e., state empathy); physiological techniques and brain imaging techniques are useful in detecting objective physiological signals related to empathy. However, some findings on empathy are mixed and inconsistent, which might be due to the different definitions and methodologies adopted by each study. To deepen our understanding of empathy and its psychological mechanisms, it is necessary to integrate different methodologies and interpret results from different aspects. To this end, this review first summarized how empathy is typically measured and quantified at different measurement levels with an emphasis on emotional empathy and cognitive empathy, given that existing studies on empathy mainly focus on the two categories and their combination. Specifically, methodologies in empathy research were introduced as followed: (1) Widely-used empathy questionnaires, including the detailed definitions and measurement dimensions; (2) classic experimental materials and behavioral paradigms for measuring emotional empathy and/or cognitive empathy; (3) empathy-related physiological signals detected by facial electromyography, electroencephalography (EEG), magnetoencephalography (MEG), and magnetic resonance imaging (MRI). Next, we discussed the advantages and disadvantages of each methodology and pointed out the difficulties in compiling findings from studies using different methodologies. For example, different empathy questionnaires applied to the same group may yield contradictory results and the quality of experimental stimulus in the same kind of empathy paradigms potentially influences participants' brain activities. More importantly, the physiological signals (e.g. event-related potential components) recorded in empathy experiments are not "specific" but only "related" to empathy. Then, we suggested that researchers should carefully select appropriate methodologies according to their research needs. For instance, EEG is suitable for exploring empathic processing in the time domain, while functional magnetic resonance imaging (fMRI) is conducive to pinpoint the specific functions (e.g. self-other discrimination) of empathy-related brain regions. Finally, we proposed that the conceptual framework of empathy should be unified and more naturalistic paradigms should be developed to reveal the interactions among different empathic subprocesses in the future. In conclusion, this review highlights the importance of comparing and combining different methodologies to reveal the underlying mechanisms of empathy, which yields new insights into social interaction.
