Our study has found that there was no significant difference in emotional empathy between first year and final year medical students. But there was significantly higher cognitive empathy among final year students as compared to first year students. Sex was a predictor variable for both cognitive and emotional empathy. In addition, future specialization interest area and being active user of online social media like facebook were also predictor variables for emotional empathy among medical students.
Compared with the stated norms of empathy used by an instruments used for this study [32, 34], there was lower mean emotional and cognitive empathy scores among Ethiopian first year and final year medical students. Preserving the low emotional empathy at final year similar to first year, may be a positive result of the students training and experience through clinical years of the students. It has been stated that ‘emotional relationships that elicit emotional response are conceptually more relevant to sympathy than to empathy’ .
Both cross-sectional and longitudinal studies have indicated that females have generally more emotional empathy than males [16, 19, 21–26]. According to psychoanalytic and evolutionary theory of parental investment, women are believed to develop greater care-giving attitudes toward their offspring than men  and these caring characteristics can be associated with high emotional empathy. Even the norm for empathy measure has set much higher empathy score of female than male . As it has been indicated with all the above findings, male students had much lower emotional empathy scores in this particular study. But gender based comparative study is required with representative proportions to make inference in this regard.
Even though we found that final year students have scored statistically higher cognitive empathy score than first year students, practically the difference was below the standard; according to RME-R test, a mean score under 22 indicates low score . This may be related with the validity of the instruments used to the culture of the study participants (the instrument used was with a Caucasian faces in the photos). In any case, final year students had higher cognitive empathy score than first year students. Similar to emotional empathy score, this may be also associated with the training in medical education or experience during the clinical years. Similar studies have found contradicting findings in this regard; studies in Japan and Korea found the highest values for measures of empathy, by year of medical school, among senior medical students [7, 36] while another study in Iran did not find variations in empathy . This difference might be attributed to the instruments used; since, in the Korean study they measured clinical empathy than general empathy.
Unlike that of emotional empathy, females have scored higher cognitive empathy than the males which may suggest that female may provide a better type of medical care [38–40] based on a better understanding of the patient’s experiences and feelings (cognitive empathy). In a number of studies, a higher empathy level in females was found as compared to males [4, 5, 7, 16, 19, 21–26].
Even though many findings of studies recorded a decline in empathy during medical school proceedings [3–5, 16–22], in a normal circumstances we expect emotional empathy to decrease and cognitive empathy to increase as students progress through the years of medical school training. Our finding is also compatible with such a normal situation. The weak positive correlation between the two measures may be indicative of this explanation.
Previous studies found an association between the choice of medical students’ future specialization and their empathy level scores [4, 22, 26, 27]. In our study, we found that students who did not decide about future interest specialization area had low emotional empathy than who have decided which may be attributed to first year students may not be familiar about some of the medical specialization areas and were still in undecided situation.
Students who were using online social media like facebook had significantly higher both emotional and cognitive empathy score as non-users. There was no documented previous study on the effect of using social media like facebook on empathy. Hence, further study is required to give more explanations for such differences. Another issue that needs further study is why there was inverse relationship between cognitive empathy and the number of brothers and sisters students had.
There was no statistically significant difference with other socio-demographic variables in this study. This may be due to similarity of the different cultures with regard to empathy in Ethiopia. One of the strength of this study is we have tried to measure two dimensions of empathy and the method of data collection was self administered study so that there may be less social desirability bias. Since the study is of explanatory nature, it is not worth adjusting for multiple testing. The validity of the study may be limited by a cross-sectional rather than longitudinal design of the study. The relatively small sample size and the fact that both instruments of the emotional and cognitive empathy scales were not validated in Ethiopia are the main limitations of this study. Nonetheless, this is the first study in the area and we believe it will add valuable information to the existing knowledge gap.