Core self-evaluation is a broad, latent, and higher-order personality trait that has wide and important effects on all people.1 For example, prior studies have indicated core self-evaluations can affect a person's life satisfaction and emotional health.1,2 Furthermore, prior studies have indicated core self-evaluation is closely related to mental health symptoms,3 and it also has a negative predictive effect on envy.4 In addition, envy also has a positive relationship with a person's mental health symptoms.5 However, at the present time, no study has explored how envy mediates the relationship between core self-evaluation and mental health symptoms. In addition, many researchers also have pointed out that envy can be divided into benign and malicious envy.6 Therefore, we explored the mediating roles of malicious envy and benign envy between core self-evaluation and mental health symptoms on the basis of previous studies.
Previous studies have shown that core self-evaluations can reflect a person's self-esteem, self-efficacy, self-control, and emotional stability.1 What's more, many researchers have proved that self-esteem, self-efficacy, self-control, and emotional stability all can inhibit a person's mental health symptoms.7–10 Zhang and Zhao3 have shown that core self-evaluation is significantly and negatively related to mental health symptoms, as people who have higher core self-evaluation tend to have more beliefs in their abilities and values, their emotions tend to be more stable and optimistic, and they are less susceptible to external environmental factors, giving them better psychological quality and allowing them to overcome or alleviate mental health symptoms. However, no further study has been done to explore the specific mechanisms between core self-evaluation and mental health symptoms.
Prior studies have found that core self-evaluation affects a person's perception of emotions.1,2 For example, people with high core self-evaluations are less likely to experience envy.4 Envy is a typical and widespread negative emotion that is caused by the success (eg, achievement, wealth) of others.11 There are two types of envy: malicious envy and benign envy.6 Lange and Crusius12 have found that envy reported in previous studies was mostly malicious envy, not benign envy. Therefore, we could initially assume that core self-evaluation is negatively correlated with malicious envy. Benign envy has a positive motivation;13 a large number of studies have shown that there is a significant and positive correlation between benign envy and high self-esteem or self-control.14,15 Also, self-esteem and locus of control are the two important personality traits that make up core self-evaluations.1 Thus, we could further assume that core self-evaluations and benign envy are positively correlated.
Many studies have found that general envy, which is almost always malicious envy,12 has a significant predictive impact on mental health symptoms.5 General envy not only negatively predicts a person's subjective well-being and life satisfaction, but also positively predicts a person's depression and anxiety,16 which may lead to other mental health symptoms.17 However, benign envy and malicious envy are different in feelings, thoughts, and behaviors.6 Benign envy motivates people to improve themselves and “catch up” with the people they envy.12 Moreover, the positive motivation has a significant and negative predictive effect on mental health symptoms.18 Therefore, we assumed that there is a positive correlation between malicious envy and mental health symptoms, and a negative correlation between benign envy and mental health symptoms. Combined with the above, we proposed a further assumption—that core self-evaluation inhibits mental health symptoms by inhibiting malicious envy and enhancing benign envy. That is, malicious envy and benign envy both play mediating roles in the relationship between core self-evaluations and mental health symptoms.
Our study explored the mediating roles of malicious envy and benign envy between core self-evaluations and mental health symptoms. Based on previous studies, we proposed three hypotheses: (1) core self-evaluation has a negative correlation with malicious envy and a positive correlation with benign envy; (2) there is a positive correlation between malicious envy and mental health symptoms, and a negative correlation between benign envy and mental health symptoms; and (3) malicious envy and benign envy both play mediating roles in the relationship between core self-evaluations and mental health symptoms.
We randomly recruited 811 participants from South China Normal University, Jinan University, South China University of Technology, and Hunan Normal University, of whom 217 were men and 594 were women. The ages ranged from 17 to 26 years, with an average age of 19.54 ± 1.86 years. The present study was approved by the Academic Committee of the School of Psychology of Hunan Normal University. All participants provided informed written consent before completing the questionnaire, and they were paid after fully completing the questionnaire.
The questionnaire consisted of four parts: the investigation of basic demographic variables, the Core Self-Evaluations Scale (CSES), Benign and Malicious Envy Scale (BeMaS), and Symptom Checklist 90 scale (SCL-90).
The Core Self-Evaluations Scale. The CSES was developed by Judge et al.1 The scale consists of 10 items and is scored on a 5-point Likert-type scale, with higher scores meaning higher core self-evaluation. In this study, we used the adapted version by Song et al.19 to evaluate the core self-evaluations of our participants, which has been proved to have high reliability and validity in Chinese populations. In the present study, the Cronbach's alpha coefficient for this scale was 0.89.
Benign and Malicious Envy Scale. BeMaS was developed by Lange and Crusius.12 The scale consists of 10 items and is divided into two subscales: malicious envy scale and benign envy scale, each of which has five items. The scale is scored on a 6-point Likert-type scale, with higher scores meaning stronger tendency of malicious/benign envy. In this study, we used the adapted version of Xiang et al.20 to evaluate the malicious/benign envy of our participants, which has been proved to have high reliability and validity in Chinese populations. In the present study, the Cronbach's alpha coefficient for the malicious envy scale and benign envy scale was 0.85 and 0.80, respectively.
Symptom Checklist 90 Scale. The SCL-90 scale was developed by Derogatis.21 The scale consists of 90 items and is divided into 10 categories: somatization, obsessive-compulsive symptoms, interpersonal sensitivity, depression, anxiety, hostility, phobic anxiety, paranoid ideation, psychoticism, and others. The scale is scored on 5-point Likert-type scale, with higher scores meaning severer mental health symptoms. In this study, we used the adapted version of Tang and Cheng,22 which has been proved to have high reliability and validity in Chinese populations, to evaluate the mental health symptoms of our participants. In the present study, the Cronbach's alpha coefficient for this scale was 0.96, and the Cronbach's alpha coefficient for each dimension was as follows: somatization 0.86, obsessive-compulsive symptoms 0.83, interpersonal sensitivity 0.83, depression 0.89, anxiety 0.86, hostility 0.79, phobic anxeity 0.75, paranoid ideation 0.74, psychoticism 0.81, and others 0.71.
First, we used AMOS 22.0 to test the measurement model we built, to determine if our indicators were good predictors of latent variables.22 We divided the items for evaluating core self-evaluations into three parts, divided the items for evaluating malicious envy and benign envy into two parts, respectively, and separated the 10 dimensions of SCL-90 to serve as indicators of the factors using an item-to-construct balance approach.23 Once we determined that the measurement model fit well, we created the structural model, choosing to use the chi-square statistic, standardized root-mean-square residual (SRMR), root-mean-square error of approximation (RMSEA), and comparative fit index (CFI) as the indicators to test the model's goodness of fit.24 At the same time, we used the Akaike information criterion (AIC) as an indicator to compare which model fit better (smaller value means better goodness of fit).25 Finally, we used the expected cross-validation index (ECVI) as an indicator to evaluate the potential for replication of the model (smaller value means greater potential for replication).26
The latent variables in the measurement model include core self-evaluations, malicious envy, benign envy, and SCL-90. The results showed that the data fit well with the measurement model (chi-squared [113, N = 811] = 627.262, P < .001; RMSEA = 0.075; SRMR = 0.038; and CFI = 0.952). In addition, the latent variables were significantly correlated with the factors that they were loading (P < 0.001). This demonstrated that latent variables can represent observed variables well. In addition, there were significant bivariate correlations among all latent variables (Table 1). The means and the standard deviations of the core self-evaluations, malicious envy, benign envy, and the scores of SCL-90 are shown in Table 1.
Descriptive Statistics and Zero-Order Correlations for All Measures
The Evaluation of Rationality in Structural Model
When there is no mediating variable (malicious envy and benign envy), the scores of SCL-90 can be significantly predicted by core self-evaluations (beta = −0.509, P < 0.001) (criterion). This allowed us to establish Model 1, in which the scores of SCL-90 can both directly and indirectly predicted through malicious envy and benign envy by core self-evaluations. The results showed the goodness of fit in Model 1 (chi-squared [114, N = 811] = 653.938, P < .001; RMSEA = 0.076; SRMR = 0.044; and CFI = 0.950) (Table 2). In addition, we found that there was an acceptable correlation between the error item of SCL somatization and the error item of SCL others in SCL-90. Therefore, based on Model 1, we constructed Model 2. The results showed that Model 2 had better goodness of fit on observed variables (chi-squared [113, N = 811] = 580.56, P < .001; RMSEA = 0.071; SRMR = 0.043; and CFI = 0.956). Compared with Model 1, we found that Model 2 had a smaller delta chi-squared value ([1, N = 811] = 73.378, P < .001) and smaller AIC. This indicated that Model 2 had a better goodness of fit and was more suitable than Model 1; therefore, we used Model 2 as the final structural model (Figure 1).
Fit Indices Between Model 1 and Model 2
The mediating model factor loadings are standardized. CS1, CS2, and CS3 are three parcels of core self-evaluation (CS). BE1 and BE2 are two parcels of benign envy (BEnvy). ME1 and ME2 are two parcels of malicious envy (MEnvy). SCL somatization (SCL_S), SCL obsessive-compulsive symptoms (SCL_OC), SCL interpersonal sensitivity (SCL_IS), SCL depression (SCL_D), SCL anxiety (SCL_A), SCL hostility (SCL_H), SCL phobic anxiety (SCL_PA), SCL paranoid ideation (SCL_PA), SCL psychoticism (SCL_P) and SCL others (SCL_O) are dimensions of the multi-dimensional scale of the Symptom Checklist 90 Scale (SCL-90).
The Significance Test of Mediating Variables
Based on the above, we used bootstrap estimation procedure to explore the stability of mediating effects. We generated 2,000 bootstrap samples (N = 811) from the original data set by random sampling. The results showed that mediating variables play significant roles in 95% confidence intervals (CI). As we can see in Table 3, using malicious envy (95% CI, −0.2 to −0.050) and benign envy (95% CI, 0.019–0.096) as mediating variables, core self-evaluations had a significant indirect effect on the scores of SCL-90 (ie, mental symptoms).
Standardized Indirect Effects and 95% Confidence Intervals
The Comparison Test of Mediating Variables
Finally, we compared the mediating effects between the two mediating variables: malicious envy and benign envy. The results showed that malicious envy plays a significantly stronger mediating role than benign envy (change in beta = −0.178; 95% CI, −0.278 to −0.092).
We used the four latent variables to test gender differences. The results showed that malicious envy (t  = 0.672, P = 0.502) and mental health symptoms (t  = −0.805, P = 0.421) had no significant gender difference, but the gender differences in core self-evaluation (t  = 2.732, P = 0.006) and benign envy (t  = −3.096, P = 0.002) were significant. Additionally, the score of men in core self-evaluation was significantly higher than that of women, and the score of women in benign envy was significantly higher than that of men. Based on these, we further investigated the stability of the gender differences we found in the structural model.
We used a multigroup analysis to determine whether there were significant differences in the path coefficients of the differences between gender models. To do so, we established two models on the basis of keeping the basic parameters (factor loadings, error variances, and structural covariances) unchanged. One model allowed free estimations of transgender paths (unconstrained structural paths), whereas the other limited the equality of path coefficients between two genders (constrained structural paths). The results showed that there were significant differences between the two models (delta chi-squared [20, N = 811] = 52.81, P < 0.001]. At the same time, both models meet the criteria of fitness (Table 4); therefore, the parameter-limited models in multiple groups were acceptable. What's more, because delta chi-squared values can be easily influenced by the large sample size and reach a significant level, we further calculated the critical ratio of standard deviation (CRD), which was used to investigate the difference between male and female participants. According to the decision rule, when the absolute value of CRD is greater than 1.96, there is a significant difference between the two parameters at a significance level (P < 0.05). Our results showed that there was no significant difference in the structural paths of all variables (CRD [core self-evaluaton --> benign envy] = 1.609, CRD [core self-evaluaton --> malicious envy] = −0.409, CRD [benign envy --> SCL-90] = 1.438, CRD [malicious envy --> SCL-90] = 0.967, CRD [core self-evaluation --> SCL-90] = 0.357). Therefore, we believe that there is no significant difference in the gender comparison between the two models, and there is no significant gender difference in the specific coefficients according to the value of CRD.
Unconstrained and Constrained Structural Paths Across Genders
The purpose of this study was to investigate the mediating effects of malicious envy and benign envy between core self-evaluation and mental health symptoms. The results showed that (1) core self-evaluation has a negative correlation with malicious envy and a positive correlation with benign envy; (2) there is a positive correlation between malicious/benign envy and mental health symptoms; and (3) malicious envy and benign envy both play mediating roles in the relationship between core self-evaluations and mental health symptoms, and malicious envy plays a significantly stronger mediating role than benign envy.
Firstly, according to the regression coefficients in the structure model, there is a negative correlation between core self-evaluation and malicious envy, and a positive correlation between it and benign envy, which was consistent with our first hypothesis. We thought that core self-evaluation may affect different kinds of envy by improving a person's self-esteem and self-control. Previous studies have shown that people with higher core self-evaluations tend to have both higher self-esteem and higher self-control.1 Duarte14 showed that self-esteem is negatively correlated with malicious envy and positively correlated with benign envy, and Van de Ven et al.15 have shown that self-control was negatively correlated with malicious envy and positively correlated with benign envy. Therefore, we believed that core self-evaluation is negatively correlated with malicious envy and positively correlated with benign envy.
Secondly, according to the regression coefficients in the structure model, we also found that both malicious envy and benign envy are positively correlated with mental health symptoms, which was inconsistent with our second hypothesis. In view of this, we thought that although previous studies have shown that envy can be divided into two kinds of envy with different properties (ie, malicious envy and benign envy6,12), and envy has a positive effect of motivating people to improve themselves.15 In truth no matter what kind of envy is elicited, people will feel that their self-esteem is threatened by others who have higher social standing and then have a feeling of social pain.27 That is to say that both malicious envy and benign envy are essentially negative emotions, and this has been demonstrated by a previous study.13 In addition, negative emotions are considered to be one of the main factors leading to the occurrence of mental health symptoms.28 Thus, it is easy to understand why both malicious envy and benign envy can positively predict mental health symptoms.
In addition, we also found that both malicious envy and benign envy have mediating effects between core self-evaluations and mental health symptoms. There are two kinds of partial mediating relations: (1) core self-evaluations --> malicious envy --> mental health symptom and (2) core self-evaluations --> benign envy --> mental health symptoms, which is consistent with our third hypothesis. We further found that the mediating effect of malicious envy is significantly stronger than that of benign envy. Combined with the direction of the regression coefficient, we found a significant and negative correlation between core self-evaluations and mental health symptoms, which is consistent with the results by Zhang and Zhao.3 Furthermore, people with higher core self-evaluation had lower malicious envy and higher benign envy, and both malicious envy and benign envy can positively predict mental health symptoms—ie, malicious envy and benign envy both played active mediating roles in the relationship between core self-evaluation and mental health symptoms. The core of these mediating mechanisms is that malicious envy and benign envy are both compound negative emotions,13 and negative emotions can promote the occurrence of mental health symptoms.29 Moreover, previous studies have also shown that people with malicious envy have more destructiveness and aggression than those who have benign envy,11 and that these two traits are more likely to cause the occurence of mental health symptoms.30 Therefore, we believe that malicious envy plays a more significant role than benign envy in mediating the relationship between core self-evaluations and mental health symptoms.
The present study has some limitations. Firstly, our participants are all Chinese and college students, which may limit the promotion of results because of cultural influence. Secondly, this study used the method of questionnaires to collect data, but self-reporting measurements are essentially subjective and susceptible to bias.
The present study proves the correlation between core self-evaluations and malicious/benign envy, and the correlation between malicious/benign envy and mental health symptoms. It also demonstrates the mediating effects of malicious envy and benign envy between core self-evaluations and mental health symptoms. This study provides a theoretical basis for preventing mental health symptoms by intervening when a person experiences envy, especially malicious envy. It is also expands the theory of the relationships among core self-evaluation, envy, and mental health symptoms.
- Judge TA, Erez A, Bono J E, et al. . The core self-evaluations scale: development of a measure. Pers Psychol.2003;56(2):303–331. doi:. doi:10.1111/j.1744-6570.2003.tb00152.x [CrossRef]
- Tsaousis I, Nikolaou I, Serdaris N, et al. Do the core self-evaluations moderate the relationship between subjective well-being and physical and psychological health?Pers Individ Dif.2007;42(8):1441–1452. doi:. doi:10.1016/j.paid.2006.10.025 [CrossRef]
- Zhang X, Zhao Y. Relationship between core-self evaluation and mental health of normal school students. Chin J School Health. 2009;30(2):129–130.
- Hiroaki N. Examining a self-evaluation maintenance model of jealousy/envy [in Japanese]. Kobe College Studies. 1992;39:79–90.
- Mujcic R, Oswald AJ. Is envy harmful to a society's psychological health and well-being? A longitudinal study of 18,000 adults. Soc Sci Med.2018;198(3):103–111. doi:. doi:10.1016/j.socscimed.2017.12.030 [CrossRef]
- Van de Ven N, Zeelenberg M, Pieters R. Leveling up and down: the experiences of benign and malicious envy. Emotion. 2009;9(3):419–429. doi:. doi:10.1037/a0015669 [CrossRef]
- Hiçdurmaz D, Inci F, Karahan S. Predictors of mental health symptoms, automatic thoughts, and self-esteem among university students. Psychol Rep. 2017;120(4):650–669. doi:. doi:10.1177/0033294117707945 [CrossRef]
- Jiang Q, Dongling QU, Wang E, et al. The relationship among physical exercise, self-efficacy, and mental health of college students. Chin J Health Psychol. 2017;25(5):763–766.
- Zhu Q, Pan G, Management SO. The influence of self-control ability on mental health of medical college students. Chin J Health Psychol.2014;22(1):111–113.
- Zhu WM, Wang BQ, Li-Ping HE, et al. Research of emotional stability and mental health of higher vocational students. J Zhejiang Inst Commun. 2009;10(4):69–73.
- Smith RH, Kim SH. Comprehending envy. Psychol Bull.2007;133(1):46–64. doi:. doi:10.1037/0033-2909.133.1.46 [CrossRef]
- Lange J, Crusius J. Dispositional envy revisited: unraveling the motivational dynamics of benign and malicious envy. Pers Soc Psychol Bull.2015;41(2):284–294. doi:. doi:10.1177/0146167214564959 [CrossRef]
- Van de Ven D, Zeelenberg M, Pieters R. The envy premium in product evaluation. J Consum Res. 2011;37(6):984–998. doi:. doi:10.1086/657239 [CrossRef]
- Duarte JL. The Effects of Scarcity and Self-Esteem on the Experience of Envy [master's thesis]. Tempe, AZ: Arizona State University; 2011.
- Van de Ven D, Zeelenberg M, Pieters R. Appraisal patterns of envy and related emotions. Motiv Emotion. 2012;36(2):195–204. doi:. doi:10.1007/s11031-011-9235-8 [CrossRef]
- Smith RH, Parrott WG, Diener EF, et al. Dispositional envy. Pers Soc Psychol Bull.1999;25(8):1007–1020. doi:10.1177/01461672992511008 [CrossRef]
- Taylor SE, Brown JD. Illusion and well-being: a social psychological perspective on mental health. Psychol Bull.1988;103(2):193–210. doi:10.1037/0033-2909.103.2.193 [CrossRef]
- Nes LS, Segerstrom SC. Optimism, motivation, and mental Health. In: Friedman H, ed. Encyclopedia of Mental Health. 2nd ed. Amsterdam, Netherlands: Elsevier; 2016:232–236.
- Song G, Kong F, Jin W. Mediating effects of core self-evaluations on the relationship between social support and life satisfaction. Soc Indic Res. 2013;114(3):1161–1169. doi:10.1007/s11205-012-0195-5 [CrossRef]
- Xiang Y, Chao X, Ye Y. Effect of gratitude on benign and malicious envy: the mediating role of social support. Front Psychiatry. 2018;9:139. doi:. doi:10.3389/fpsyt.2018.00139 [CrossRef]
- Derogatis LR. The Symptom Checklist-90-R (SCL-90-R). Towson, MD: Clinical Psychometrics Research; 1975.
- Tang Q, Cheng Z. The use and reanalysis of SCL-90 in China. Chin J Clin Psychol.1999;7(1):16–20.
- Little TD, Cunningham WA, Shahar G, et al. To parcel or not to parcel: exploring the question, weighing the merits. Struct Equ Modeling. 2002;9(2):151–173. doi:. doi:10.1207/S15328007SEM0902_1 [CrossRef]
- Byrne BM. Structural Equation Modeling with AMOS: Basic Concepts, Applications, and Programming. Abingdon, UK: Routledge; 2009:343–344.
- Akaike H. Factor analysis and AIC. Psychometrika. 1987;52(3):317–332. doi:10.1007/BF02294359 [CrossRef]
- Browne MW, Cudeck R. Alternative ways of assessing model fit - Institute for Social and Economic Research (ISER). Clin Diagn Lab Immun. 1992;7(2):161–167. doi:10.1177/0049124192021002005 [CrossRef].
- Johnson C. Behavioral responses to threatening social comparisons: from dastardly deeds to rising above. Soc Personal Psychol Compass. 2012;6(7):515–524. doi:. doi:10.1111/j.1751-9004.2012.00445.x [CrossRef]
- Piqueras AJ, Ramos. Negative emotions impact on mental physical health [in Spanish]. Suma Psicologica. 2009;16(2):85–112.
- Berking M, Poppe C, Luhmann M, Wupperman P, Jaggi V, Seifritz E. Is the association between various emotion-regulation skills and mental health mediated by the ability to modify emotions? Results from two cross-sectional studies. J Behav Ther Exp Psychiatry.2012;43(3):931–937. doi:. doi:10.1016/j.jbtep.2011.09.009 [CrossRef]
- Suls J, Bunde J. Anger, anxiety, and depression as risk factors for cardiovascular disease: the problems and implications of overlapping affective dispositions. Psychol Bull. 2005;131(2):260–300. doi:. doi:10.1037/0033-2909.131.2.260 [CrossRef]
Descriptive Statistics and Zero-Order Correlations for All Measures
|Core of self-evaluation
Fit Indices Between Model 1 and Model 2
Standardized Indirect Effects and 95% Confidence Intervals
||Confidence Interval (95%)
|Core self-evaluation --> malicious envy --> mental health symptoms
|Core self-evaluation --> benign envy --> mental health symptoms
Unconstrained and Constrained Structural Paths Across Genders