Introduction

I do not have intrinsic worth or worthlessness, but merely aliveness. I’d better rate my traits and acts, but not my totality or ‘self.’ I fully accept myself, in the sense that I know I have aliveness and I choose to survive and live as happily as possible, and with minimum needless pain. I require only this knowledge and this choice—and no other kind of self-rating (Ellis 1999, as cited in Bernard 2013)

Self-awareness is one of the most discussed topics in academic and popular psychology. We can find it under different names, in various areas, in different ways elaborated and explained, but always at one of the most prominent places when it comes to importance in the psychological functioning of humans. Self is the fundamental element of every human being, a symbolic construct that reflects our consciousness of our own identity (Crisp and Turner 2010). One of the essential characteristics of human beings—reflective thinking and meta-thinking, the ability to think about the way we think, allows us to think about who we are and how others see us. However, this process does not go on continuously, but it depends on personality traits and specific situations. Empirical psychology uses the self-concept as the subject of research in different domains. Self-concept is more specific than self-awareness and refers to the self-perception of various aspects of one’s personality, which includes body image, attitudes, self-esteem, self-confidence, self-efficacy, etc. (McLeod 2008). Self is, as a multidimensional concept, related to mental health and disorders (McCrae and Costa 1996), and it is often a subject of research in clinical psychology. Based on the representation of self in psychological science, we can conclude that this topic is of great importance because of the time that people are devoting to thinking about who they are and what they are. Also, people often think of certain aspects of the Self as positive or negative, which means that Self implies an evaluative component that is very important.

Several constructs are dealing with some aspects of self-evaluation. The most prominent construct is self-esteem, which represents a subjective positive or negative global assessment of oneself (Sedikides and Gregg 2003). Self-esteem (SE) correlates with positive mental health indicators in many studies (Mann et al. 2004). For example, a higher level of SE correlates with psychological well-being and happiness (Diener and Diener 1995; DeNeve and Cooper 1998; Zimmerman 2000) and with lower levels of emotional distress (Sedikides et al. 2004). Also, lower level of SE correlates with higher levels of psychological distress, depression, and anxiety (Bernet et al. 1993; Tennen and Affleck 1993). That is the reason why is SE seen as a protective factor in a confrontation with different personal and social problems, so programs for boosting SE are implemented in education, health and business sectors in many western countries (Baumeister et al. 2003). However, there is little evidence for the efficacy of these programs (Baumeister et al. 2003). Also, some studies suggest that SE cannot be reduced to high versus low dichotomy (Sava et al. 2011), and it is not a unidimensional construct (Kernis 2005).

Rational Emotive Behavior Therapy (REBT) argues that every global self-assessment, either positive or negative, is non-constructive, even detrimental to mental health. Global self-evaluation is irrational because there is no objective criterion for evaluating the value of human beings (Chamberlain and Haaga 2001a). Human beings are too complex to be evaluated by a single global assessment. Only individual aspects of an individual can be evaluated such as thoughts, feelings, acts, physical aspects from different periods of one’s life, but one unique categorical assessment cannot be given (in the domain “good” or bad”). In addition to the fact that global evaluation is often biased, it is not logical, not empirically sustainable, nor useful because it predisposes a person to feel emotionally disturbed (Popov and Sekulić-Bartoš 2016). Global self-assessment based on the experience of success and failure, acceptance or rejection by other people, comparison with others, is rather unstable, and therefore dysfunctional for mental health (Chamberlain and Haaga 2001b). When the person who holds irrational beliefs which contain global self-evaluation (e.g., self-downing) finds himself in the situation that provokes self-assessment, the result is “ego-disturbance”. The emotional content of “ego-disturbance” usually consists of depression and/or anxiety (Ellis and Dryden 2007). As an alternative, REBT offers unconditional self-acceptance (USA) that excludes a global assessment of human value while maintaining a critical attitude by its characteristics. USA is one of the basic constructs of REBT (Ellis 1994). USA means fully and unconditionally accepting oneself, regardless of the particular outcome of one’s behavior, because the person accepts the idea of human fallibility (Ellis 1977; Sava et al. 2011).

What REBT has recognized when the concept of SE is concerned, has also been recognized by contemporary research practice in the field of personality and social psychology. Similarly to USA, they also speak of a self-respecting attitude that is more constructive regarding mental health than famous SE and introduce concepts such as contingent and non-contingent self-esteem (Crocker and Wolfe 2001). REBT considers unconditional self-acceptance as a continuum or as an intellectual and emotional habit which is expressed to a greater or lesser extent in different individuals. People with the high USA are more resistant to ego-provoking situations such as failure or rejection (Davies 2006, 2007, 2008). In the experimental study which investigated whether USA was a better protective factor than SE in an ego-threatening situation (“trigger” of possible upset to the self-image), USA and SE made a similar, though not quite identical, impact on the mental health indicators: anxiety, depression, positive and negative affect. USA had the moderating role between positive feedback and positive affect, and it was a stronger predictor of negative affect (Popov et al. 2015). In a study that follows the previous one, results supported predictions of REBT theory about the protective role of USA in depressive and anxious emotional reactions (Popov et al. 2016). In earlier studies, increase in USA lead to the reduction in depression when SE was statistically controlled (Chamberlain and Haaga 2001b). USA also represented a mediator of the relationship between maladaptive perfectionism and depression (Flett et al. 2003; Scott 2007). Macinnes (2006) examined the relationship of USA and SE with psychological health. He found the concepts of USA and SE behaved similarly in relation to the dependent variables in a study, but not quite the same. USA was lower among participants who had higher levels of anxiety and depression and participants with higher USA had higher scores on psychological well-being. Although SE showed similar relations with mental health variables, it was more closely associated with affect, while USA appeared to be more closely associated with general psychological well-being (Macinnes 2006).

According to REBT, USA and SE are conceptually different phenomena, but in research practice, they tend to be closely associated both with each other and with mental health outcomes (Sava et al. 2011; Popov et al. 2015). Researchers generally emphasize the association of these constructs, but there are not many studies comparing the contribution of these constructs in the prediction of mental health variables. This study aimed to explore the relation between USA and SE, and mental health indicators, such as subjective well-being, a state of positive and negative affect, as well as state of anxiety and depression, in cases when positive and negative trait affectivity is also concerned. Mental health indicators in this study were chosen following REBT theoretical assumptions (Chamberlain and Haaga 2001b) that the higher level of USA should be associated with a lower level of anxiety, depression and negative mood, as well as a higher level of happiness and general psychological well-being. The novelty in this study is the inclusion of trait affectivity that could affect the relationship between measures of self-assessment and emotional state, which has not been taken into account in previous studies.

Methods

Participants

The sample consisted of 275 university students (80% female, mean age 19.85, SD = 2.67) who voluntarily participated in the research. We selected study groups that are less familiar with the psychotherapeutic concepts: Science Teacher Education, Serbian and French Language, Tourismology, Sociology, Food Technology, and Journalism. Our assumption was that selected study groups are less likely to give answers loaded with previous knowledge. As motivation for participation in the research, additional credits were offered to students for one of the subjects, in agreement with the authorities at the University. After the elimination of multivariate outliers (according to the Mahalanobis distances; Tabachnick and Fidell 2001) 268 participants remained in the final sample.

Instruments

  • The Big Five plus Two (VP + 2: Smederevac et al. 2010) is a standardized personality inventory. In this study, we used only items related to Negative Affectivity (12 items) and Positive Affectivity (8 items) as a personality traits. Positive affectivity is a part of the Extraversion subscale, while negative affectivity is a part of Neuroticism. Affectivity items encompass the temperamental aspects of positive and negative mood, the indicators of optimism and pessimism, the lifestyle and behavior that are not situatively determined, but come from personality. Internal consistency of the PA trait scale was α = . 79, and α = . 82 of the NA trait scale.

  • The Short Subjective Well-Being Scale (SWB: Jovanović 2010) was used for assessment of participants’ subjective well-being. It consists of 8 items related to the experience of pleasant emotions and positive evaluation of life. Accordingly, the instrument consists of two dimensions: Positive affectivity and Positive attitude toward life. The response format is 5-point Likert scale. In this study, we used total scale score as a subjective well-being variable. Internal consistency of the whole scale was α = .91.

  • Serbian Inventory of Affect based on the Positive and Negative Affect Schedule-X (SIAB-PANAS: Novovic and Mihic 2008) was used to assess states of positive and negative affect. It is a Serbian translation and adaptation of the Positive and Negative Affect Schedule-X (PANAS-X: Watson and Clark 1994). We used a short version of the questionnaire, consisted of 20 items. Internal consistency of the scale was α = .90 for positive affect and α = .87 for negative affect.

  • Unconditional Self-Acceptance Questionnaire (USAQ: Chamberlain and Haaga 2001a) is a Serbian translation of the original instrument (Chamberlain and Haaga 2001a). USAQ consists of 20 statements with a 7-point Likert scale, that reflect various aspects of USA philosophy, as used in REBT. The scale was translated from English into Serbian and then independently backtranslated by a second translator (α = .70).

  • The Rosenberg Self-Esteem Scale (RSES: Rosenberg 1965) is the most widely-used instrument for measuring explicit self-esteem. It is a 10-item instrument with a 5-point Likert-type response format (α = .81).

  • Anxiety Scale from Depression, Anxiety, Stress Scale (DASS-42:Lovibond and Lovibond 1995), assesses autonomic arousal, skeletal muscle effects, situational anxiety, and subjective experience of anxious affect. The scale contains 14 items with a 4-point Likert-type response format (α = .87).

  • The State Depression Scale (SD: Novovic et al. 2009) contains 20 items regarding the presence of depressive affect and its cognitive content, as well as other behavioral and motivational interferences relating to a depressive state. The respondents’ task was to indicate to what extent they felt as described in the scale items, on a 5-point Likert scale (α = .92).

Design and Procedure

We used a correlational, cross-sectional design, to examine relations among variables in this study. Participants filled out printed questionnaires during regular classes at University after they initially consented to participate in the research. Considering the suggestions that the order of the questionnaires assignment may influence the results of the study (Brown and Brown 2011), the instruments were sorted in different, randomly sequential order. The estimated time for the examination was 30 min, including the previous instructions by the examiner.

Results

Table 1 shows the descriptive characteristics of variables in the study. All scales show satisfactory reliability, and values of skewness and kurtosis suggested the deviation of data from normality was not severe and fulfill the basic conditions for the implementation of the proposed data analysis.

Table 1 Descriptive statistics for variables in the study
Full size table

In Table 2 can be seen that intercorrelations among the variables go in expected theoretical directions. Positive and Negative Affectivity traits are negatively correlated, as well as Positive and Negative Affectivity state. Also, the correlation between USA and RSE is moderate and positive. Both variables have similar connections with mental health variables.

Table 2 Intercorelations among the variables
Full size table

Secondly, a series of linear regression analyses were conducted where affectivity traits, SE and USA were entered as predictors. Criterion variables were States of Anxiety and Depression, States of Positive and Negative Affect and Subjective Well-being. We used the linear regression analysis to understand the unique contribution of every independent variable to the prediction of mental health. Summary of linear regression analyses is presented in Table 3.

Table 3 Summary of linear regression analysis
Full size table

As can be seen, all regression models are statistically significant (p < .000). The results showed that besides affectivity traits, USA has the significant contribution to the prediction of mental health indicators. On the other hand, SE is not the significant predictor of mental health variables. In Table 3, we can see that affectivity traits have the strongest relationship with mental health variables, while USA also becomes the significant predictor of all criterion variables except PA state. PA trait in the negative direction predicts the state of depression (β = −.12, p < .05) and NA state (β = −.16, p < .01), and in the positive direction predicts PA state (β = .49, p < .001) and SWB (β = .65, p < .001). NA trait in the positive direction predicts state of anxiety (β = .45, p < .001), depression (β = .38, p < .001) and NA state (β = .37, p < .001) and in the negative direction SWB (β = −.10, p < .05). USA in the negative direction predicts the state of anxiety (β = −.15, p < .05) and depression (β = −.13, p < .05), as well as NA state (β = −.16, p < .05), and in the positive direction predicts SWB (β = .11, p < .05). That means a higher level of USA is associated with a lower level of states of anxiety, depression, and negative affect and with a higher level of subjective well-being. Self-esteem is not a predictor of either emotional outcome in these analyses.

Discussion

Unconditional self-acceptance and explicit self-esteem, although theoretically different, in empirical studies show more similarities than differences (Sava et al. 2011). REBT stands for the concept of USA and emphasizes its advantage over the concept of SE which is more popular in psychological theory and practice outside of REBT. SE represents a global evaluation of self-worth (Baumeister 1998, as cited in De Cremer and Sedikides 2005), but according to REBT awarding oneself a general assessment (positive or negative) is considered to be irrational and leads to the psychological problems (Chamberlain and Haaga 2001a). However, none of the previous studies has unequivocally supported this idea, since both constructs are significant positive predictors of mental health (Popov et al. 2015). Previous studies supported hypotheses related to the beneficial influence of USA on mental health indicators (Chamberlain and Haaga 2001a, b; Popov et al. 2016). However, SE has also proved to be a significant predictor of mental health variables, although it does not behave quite as same as USA (Jibeen 2016; Popov et al. 2015). Previous experimental research in the field of SE has also shown that people who scored higher for this variable show less emotional distress in situations of negative feedback (Brown 2010). All previous studies that tried to compare USA and SE used the same instruments for measuring these constructs. It is important to note that the reason why it seems difficult to empirically demonstrate the advantage of USA in comparison to SE is the method by which these variables are measured. The instruments that have been used for that purpose overlap somewhat (David et al. 2013). The instruments used to measure USA were contaminated with SE items. On the other hand, items from the Rosenberg self-esteem scale could be understood as derivatives of high and low USA.

Instead of using different instruments for measuring USA and SE, we tried to compare them in the prediction of mental health variables in the situation when positive and negative affectivity traits are also included in the model. Why is that important? Positive and negative affectivity traits represent dispositional measures of affect. These personal characteristics are expressed by the tendency to emotionally respond to situations in life in a stable and predictive way. People with higher positive affectivity tend to perceive reality much more optimistic than people with high negative affectivity (Barsade and Gibson 2007). If we do not consider dispositional affectivity in studies about emotional reactivity as a measure of mental health, we risk to lowering the importance of specific therapeutic mechanism that may affect the regulation of the current emotional state (which in the long term has beneficial effects on mental health in general). We believe that this is one of the reasons why we have not been able to demonstrate the greater therapeutic utility of USA in comparison to the more popular construct of SE, in previous studies. This study aimed to explore the relation between affectivity traits, USA, SE, and mental health indicators, such as SWB, states of PA and NA, as well as states of anxiety and depression. The results of this study suggest that our assumption that USA will be a better predictor of mental health then SE if we include the affectivity traits in the model has proven to be correct. The most influential predictors of state anxiety, state depression and positive and negative affect are dispositional affectivity traits. However, USA has also shown a protective role in mental health indicators. USA in the negative direction predicts the states of anxiety, depression, and NA. On the other hand, USA in the positive direction predicts SWB. The only criterion variable which USA does not predict is PA state, which is theoretically expected, because USA aims to generate functional feelings, rather than positive (David et al. 2013).

Strengths, Limitations and Future Directions

The main advantage of this study has been the idea that USA as a constructive attitude about Self could be the better predictor of positive mental health outcomes than SE when we consider dispositional measures of affect. This study is maybe the first which supports the hypothesis about the superiority of USA in comparison to SE in predicting mental health indicators.

The limitations of the study are mostly related to the correlational design and sample selection from a non-clinical population. Moreover, the sample consisted of young and educated population (students), and it was mostly female. It seems that this research problem could be better examined in a psychotherapeutic setting with continuous monitoring of the effects of treatment based on USA in comparison to treatments based on a philosophy of SE. Also, an experimental design that manipulates ego-provoking factors that can affect self-assessment (e.g. positive and negative feedback for achievement) would significantly improve the research of this topic. We also suggest the use of other instruments for measuring SE and USA. When it comes to SE, researchers have proposed different tools that include various aspects of self-evaluation, instead of estimation of general self-worth: sensitivity, awareness, and dependence of varying life standards (Kernis 2005). When it comes to USA, researchers proposed making a difference between philosophical and psychological self-acceptance (David et al. 2013), which should be taken into account in future studies.