Affective temperament , social support and stressors at work as the predictors of life and job satisfaction among doctors and psychologists Temperament , socijalna podrška i faktori stresa na poslu u predvi đ anju zadovoljstva životom i poslom kod lekara i psihologa

Background/Aim. Affective temperament, social support and work-related stresors belong to the group of life and job satisfaction indicators. The aim of this research was to examine predictive roles of the basic affective temperament traits, social support and work-related stressors in the feeling of job and life satisfaction among doctors and psychologists. Methods. The sample consisted of 203 individuals out of whom there were 28% male and 72% female doctors (61%) and psychologists (39%), 25–65 years old (39.08 ± 9.29), from the two university towns in Serbia. The set of questionnaires included Serbian version of the Temperament Evaluation of Memphis, Pisa, Paris and San Diego – autoquestionnaire version (TEMPS-A), Satisfaction with Life scale, Job Satisfaction Survey, short Interpersonal Support Evaluation List, and Source of Stress at Work Scale (IRSa) for estimating the frequency of stressors at work. Results. According to the existing norms our examinees are satisfied with their life, but considerably less satisfied with their work, specially with pay and benefits, while they are most satisfied with nature of work itself and social relations with co-workers and supervisors. Our results show that depressive and hyperthymic, and to some extent cyclothymic temperament traits of the affective temperament significantly predict 21% of life satisfaction variance. Situational factors, such as stressors at work and social support, are important in predicting job satisfaction (58% of variance) with no significant contribution of temperament traits. The analysis did not point out any significant relation of sex, occupation, and age with life and job satisfaction. Conclusions. Affective temperaments can be regarded as predictors of life satisfaction, but in order to better predict satisfaction the aspects of wider social surrounding and sources of stressors at work must be taken in consideration. Future studies should consider other indicators of life satisfaction such as family or health satisfaction, stressors outside of work situations, and negative aspects of caregiving, for example hyperactivation.

There are three perspectives in literature which try to explain the relationship between job and life satisfaction 3 : according to the 'bottom-up' perspective, job satisfaction is an integral part of life satisfaction and together with other life domains it influences total life satisfaction; the 'bottomdown' perspective suggests that life satisfaction influences job satisfaction; regarding the third perspective, much of the relationship between job and life satisfaction is spurious, resulting from common influences, such as environmental variables (eg, job income) and personality traits.
In accordance with the last perspective, differences in basic personality traits and 'core self-evaluations' are pointed out as relevant predictors of both life and job satisfaction 4,5 .

Affective temperaments
Contemporary theory of affective temperaments based on Aristotle, Krecmer and Kraepelin's ideas about basic temperamental types 6 , refers to temperament as a stable behavior trait with strong affective coloring.These different temperaments (depressive, cyclothymic, hyperthymic, irritable and anxious) are described as sub-syndromic conditions of the affective pathology, and they are likely to represent liability factors in the subsequent development of affective disorders 7 .
The majority of earlier studies on temperaments is used for predicting ill-being, or psychological disfunctioning and only a few studies explore where the traits of the affective temperament predict positive characteristics of mental health and concepts of life and job satisfaction.According to Diener et al. 8 , the affective component of the subjective well-being construct is comprised of the frequency and intensity of positive and negative affect experience, which is essentially the description of temperament.Fogle et al. 9 established that the traits of extraversion and neuroticism, as major temperament dimensions, are important predictors of life satisfaction among adolescents.Moreover, neuroticism trait has been a significant negative predictor of life satisfaction among medical professionals 10 .
It was previously established that some aspects of job satisfaction are also moderately correlated with emotionality 11 .For instance, the correlation of stress at work with affective temperament traits is confirmed 5 .Moreover, the studies showed that choosing occupational orientation can also correlate with temperament traits 12 .These data point out that the affective temperament characteristics should be considered for predicting job satisfaction as well as life satisfaction, with obligatory observation of situational factors in that correlation 9 .Job satisfaction results from the characteristics of the job and experience a person has with certain aspects of job such as: pay, chances of promotion, managers, the system of rewarding, work procedures, co-workers and the nature of work 13 .

Helping professions
Helping professions are professions that provide support to people with physical, psychological, and social problems.Doctors and psychologists were the subjects of this research because the area of work in which they are engaged, covers all areas of helping professions such as medicine, nursing, social work, psychotherapy, counseling, education and coaching.These professions are high risk professions for ill-being, low life and job satisfaction 14 .Both medical doctors and psychologists are exposed to numerous factors that can have negative influences on mental health such as high intensity of work, constant exposure to pain, psychic and physical suffering of clients, disbalance between made effort and the reward for that effort, personal and family stresses that are reflected through job, etc.
The impact of organizational factors on subjective wellbeing and life satisfaction among health workers and other helping professions have been confirmed in numerous studies all over the world [15][16][17] .This research examined if and to what extent the affective temperament traits were inserted in the correlation between the situational factors (eg workrelated stress) and life/job satisfaction among psychologists and doctors.According to the model of Argyle and Lu 18 social competence plays the role of the mediator in the correlation between temperament and happiness, due to which perception of social support, as one of the objective measures of the social competence, was introduced in our model.
The aim of this research was to investigate predictive roles of the basic affective temperament traits, social support and work-related stressors in the feeling of job and life satisfaction among doctors and psychologists.

Sample
The sample consisted of 203 male and female medical doctors and psychologists, 25-65 years old, from the two university towns in Serbia (Table 1).The doctors were employees in governmental health institutions, while psychologists were employees at schools, kindergartens, centres for social work and health institutions.The total number of administered questionnaires was 320 (reply rate of 63%).
The research was conducted from January till March, 2015 after the approval of the Ethics Committee at the Faculty of Medical Sciences in Kragujevac.All the included examinees signed written consent for participating in this research.

Instruments
The short form of Temperament Evaluation of Memphis, Pisa, Paris and San Diego -autoquestionnaire version (TEMPS-A) was used in the research, namely the standardized Serbian version of the scale 19 .So far, this self-evaluation questionnaire has been validated in more than twenty languages, representing the most frequently used instrument for assessing basic temperament traits 19 .The scale measures emotional reactivity (eg, depressiveness, irritability), cognitive components of the emotional reactions (pessimism or optimism), psychomotor reactions and circadian rhythm (the energy of reactions, sleep schedule, etc.), and social aspects of emotional reactions 20 .In the Serbian version, 41 yes-or-no items are grouped into six instead of five subscales (the anxious temperament is divided into two subdimensions): depressive, cyclothymic, hyperthymic, irritable, anxious-somatic and anxious-cognitive temperament.The scale showed good internal consistency and the construct validity (α = 0.83), as well as moderately high alpha coefficient of subscales (0.73-0.80) among adult non-clinical population.The average test-retest coefficient (rho = 0.82) suggests stable reliability 19 .
The scale Satisfaction with Life (SWLS) measures the cognitive component of life satisfaction, namely it is referred to the personal assessment of well-being and happiness as the most frequent indicators of the quality of life 1 .Participants state the level of agreement with five items (eg, "In most ways my life is close to my ideal") on the seven-point Likert scale ranging from 1 -strongly disagree to 7 -strongly agree.The scores range from 5 to 35, while higher scores indicate greater life satisfaction.Test-retest correlation of the scale is 0.82, and Cronbach's α ranges from 0.83 to 0.87 21 .
The Job Satisfaction Survey (JSS) is a nine facet scale, comprised of 36 items describing employees' attitudes about their job and aspects of the job: pay, chances of promotion, supervision, fringe benefits, contingent rewards, operating conditions, co-workers, nature of work, and communication 22 .The summative score, which points out the general degree of job satisfaction, can also be calculated.Although the JSS was originally developed for use in human service organizations, it is applicable to all types of organizations which are the subject of our research, including medical and mental health centres, schools, social services 2 .It is translated into more than 20 languages, while the official Serbian version was used in this research 23 .The realiability of the whole instrument is α = 0.91, while the reliability of certain subscales ranges from 0.62 to 0.82 2 .
The Serbian version 24 of the short Interpersonal Support Evaluation List (ISEL) 25,26 was used in this research.The scale comprises 12 items and measures the perceived availability of Appraisal (advice or guidance), Belonging (empathy, acceptance, concern), and Tangible (material or financial aid) social support 26 .Items are valued on a fourpoint scale ranging from 1 (definitely false) to 4 (definitely true).Considering general population, the scale reliability ranges from 0.80 to 0.90 25 .Moreover, there is a positive correlation of this scale with other scales of social support (eg, Inventory of Socially Supportive Behaviors), with the number of close friends, as well as with the level of the quality of marital relationships 26 .
The IRSa is a multidimensional scale for estimating the frequency of stressors at work 27 .The scale consists of 38 items, out of which 34 have the form of the five-point Likert scale (from "hardly ever" to "almost always") and assesses  the presence of adverse job conditions and events, eg, "How often you do not have time to complete all your tasks?" or "Can you decide when to take a break?".The other four Yes/No items assess job insecurity, eg, "Are you worried that you'll get fired?".The IRSa comprises seven subscales that assess different sources of stress at work (demands at work, supervision, the support of the manager, the support of colleagues, relations and working environment, position at work, distribution of work assignments).The reliability of subscales ranges from 0.72 to 0.82, with the average reliability of the whole scale α = 0.74.

Statistical analyses
The data were analyzed using SРЅЅ Statistics v.20.Pearson's correlation coefficients were used to examine associations between continuous variables, while t-test was used to determine the significance of differences among arithmetic means.Kolmogorov-Smirnov test was used for analyzing normality of the distribution of scores on the scales of life and job satisfaction.Furthermore, the hierarchical and logistic regression analyses were used for establishing the degree where predicting variables indicate criterion variables.

Results
All the scales showed relatively high reliability, for the subscales TEMPS-A ranging from 0.70 to 0.80, ISEL 0.78, IRSa 0.91, SWLS 0.84 and JSS 0.92.Kolmogorov-Smirnov test showed normal distribution of scores on the scales of life and job satisfaction.Considering life satisfaction, the values are to some extent higher than theoretical frameworks, therefore the curve is slightly curved toward higher values, but it is still within the bounds of significance.Concerning job satisfaction, the values are in the range of normal distribution (Table 2).
The analysis of sociodemographic characteristics did not point out any significant difference considering sex and occupation (doctors and psychologists) in regard to life and job satisfaction.The relationship status of the examinees (in a relationship/single) showed that those who are in a relationship are more satisfied with life than those who are single (Table 3).
Life satisfaction is positively correlated with the number of children and the years of education, while job satisfaction is negatively correlated with the years of working experience (Table 4).
According to the already existing norms in the world 28 , our examinees are satisfied with their life, but considerably less satisfied with their work [t(2) = 11.20;p < 0.01)] 29 .Namely, our workers are the least satisfied with pay (9.03 ± 4.61) and benefits (9.37 ± 4.45), while they are most satisfied with nature of work itself (18.72 ± 3.52), social relations with co-workers (15.56 ± 4.15) and supervisors (15.57± 5.33).
The correlation among affective temperaments, life and job satisfaction, social support, and stress at work is shown in Table 5.The effects of sex and age variables were controlled in this analysis owing to the fact that there are certain differences of dominant temperament traits in regard to these variables 19,30 .
Life and job satisfaction are in low to moderate negative correlation with almost all temperament traits, moderate positive correlation with social support, and moderate to high negative correlation with sources of stress at work (Table 5).Life satisfaction has moderate correlation with all the three types of social support (Appraisal r = 0.254**, Belonging r = 0.259**, and Tangible r = 0.239**), and the situation is similar regarding job satisfaction (Appraisal r = 0.351**, Belonging r = 0.298**, and Tangible r = 0.346**).In regard to the sources of stressors at work, the lack of the manager's support (r = 0.712**), distribution of work assignments (r = 0.650**) and bad work environment (r = 0.640**) are highly correlated with job satisfaction, as well as the support of colleagues (r = 0.528**).
In order to examine the relation between explanatory and criterion variables, analyses of hierarchical and linear regression were conducted.According to the theoretical model and the fact that temperament is a basic personality trait, in the first part of these analyses, life satisfaction was the dependent variable, while the independent ones were those temperament traits that moderately correlated with life satisfaction.In the second step, social support and sources of stress at work were introduced in the model (Table 6).
When life satisfaction is predicted by temperament traits, the model is significant and presents 21% of the variance in life satisfaction.Introducing additional variables, work-related stress and social support, the additional 6% of the variance in life satisfaction is explained and this change is also significant.
Linear regression analysis of job satisfaction considered only variables 'social support' and 'stress at work', because the other ones did not show moderate correlation with the dependent variable.Approximately 58% of the life satisfaction variance was explained by this model (Table 7).

Discussion
The aim of this paper was to examine the predictive role of affective temperaments in the feeling for life satisfacti-on among doctors and psychologists, as to well as examine their intermediary roles in predicting job satisfaction.
Regarding the life satisfaction variance, depressive and hyperthymic and to some extent cyclothymic temperament traits, make the greatest contribution.These results are expected since hyperthymic temperament refers to cheerful, optimistic, energetic (active) and self-confident persons 31 , and optimism and self-esteem are actually the traits that are positively correlated with life satisfaction [32][33][34] .Moreover, it is also expected that depressive temperament has negative influence on life satisfaction, because it can be described as the contrast to hyperthymic temperament.Cyclothymic temperament showed the weakest predictive power, particularly when situational variables are included in the model.Nevertheless, it is obvious that persons with rapid mood and energy swings, which are characteristics of the cyclothymic temperament 31 , are less satisfied with their life.This is proved by numerous studies that present great power of this type of temperament for predicting early depressivity and suicidal tendencies 35 , higher rates of relapse of depression, higher rates of irritability and mixed feelings, as well as addictions 36 .Introducing stressors at work and social support in the model, the significance of predicting is to some extent increased which points out their potential intermediary role in this model.Other authors also came to similar conclusions 37,38 , emphasizing that approximately 50% of individual differences referred to experiencing happiness can be explained by the inborn predisposition for a certain level of positive affect, 10% can be explained by life experience, while the other 40% can be under our own control.
In respect to demographic variables, the relationship status, the number of children and years of education showed a correlation with life satisfaction.It is expected that people in relationships have the possibility to satisfy psychological, reproductive and social needs, but first of all, the need for belonging as one of basic psychological needs.Being married/cohabiting and social support explained a half of the total variance of life satisfaction among physicians 10 .However, until now most papers either did not confirm the correlation, or showed a negative correlation between the number of children and life satisfaction.According to Angeles 39 , previous researches did not manage to identify these effects, because the key role of marital status was not taken into consideration.The experience of individuals who are married and share duties and problems, differs from the experience of the ones who are separated, or have never been married and who have to take care of children themselves.Furthermore, financial status has not been considered in the earlier researches.Regarding the fact that our research includes highly educated people who are employed and have stable incomes, the number of children does not imply worrying for family existence.
National studies in Sweden and Spain confirms our results considering the correlation between years of education and life satisfaction 40,41 .Incomes and level of education have a direct and indirect impact on most aspects of life satisfaction in these studies.More educated people can better or-ganize their free time, take various activities, make numerous social contacts in their life, and all these issues indirectly influence life satisfaction.
The second model confirms that situational factors such as stressors at work and social support are significant predictors of job satisfaction, while it does not confirm our hypothesis that temperament traits also play a role in such a model.Stressors at work as the most significant predictors for job satisfaction together with social support predict 58% of the variance in job satisfaction.As we have already mentioned, researches in different countries show a high prevalence of work stress among helping professions [15][16][17] .Exposing oneself to numerous kinds of stressors, among healthcare professionals, can cause increased anxiety, depressivity, irritability, sleep disorder, lack of concentration and aggressive behavior, and it can also contribute to the development of burnout syndrome, mental problems, absence from work, earlier retirement 42 .The sources of stressors at work have a tendency to make its effect complex and multiplied because they are combined together with the sources of stressors from other domains of life (family life, relationship with friends), which all together contribute to not only lower job satisfaction but to low life satisfaction in general.
Family and social relationships, as segments of social and emotional support a person perceives, are also important for maintaining the quality of life 43 , though this variable made only a small contribution in our models.Researches emphasize the relevance of social relationships for mental health, therefore the individuals with weak social relationships are liable to numerous health and social problems 44 .Social support is correlated with many positive indices of mental health, such as optimism, happiness, and life satisfaction, but also negatively correlated with perceived stress 45 .
Regarding sociodemographic variables our study points out that only working experience is (negatively) correlated with job satisfaction.Some researches show that age is positively correlated with job satisfaction, but more and more authors suggest that the relationship between age (or working experience) and job satisfaction is U-shaped, declining until around the mid-30s and gradually increasing until the late 60s 46,47 .On the other hand, our research considers specific professions, so we have to regard only the studies with similar population.The correlation between working experience and depression, anxiety and chronic diseases among doctors can be seen in Caplan's 48 research, where he found that older doctors are more exposed to the significant amount of stress.Working experience results in chronic diseases, general exhaustion, impossibility to adapt to new technologies, which all cause harder fulfilling work obligations and less job satisfaction.
Finally, the aspects of job satisfaction that we examined in this paper should also be mentioned.Our examinees are least satisfied with those aspects of job referred to extrinsic motivation, namely external rewarding, such as pay and benefits.These results are expected considering the fact that Serbia belongs to developing countries with budget activities of limited financial possibilities, and organizations mostly relied on other motivation factors for achieving employees' efficiency at work 49 .On the other hand, they are most satis-fied with aspects of intrinsic motivation, characteristics of job, but with social relationships at work as well, which probably contributed to better total job satisfaction, because support from superiors and co-workers are significant components of high job satisfaction [49][50][51] .
The studies show that caring for others is associated with improvement in health indicators of caregivers, which supports the hypothesis that providing support to others can sometimes be more advantageous than receiving it 50 .The effect of caregiving on the person who gives care quite depends on some psychological characteristics, therefore it would be good to examine the influence of negative aspects of caregiving among helping professions in some future researches.
This considers particularly caregiving hyperactivation, which represents self-sacrificing approach to caregiving with an intense desire for increased closeness 51 .Hyperactivated caregivers put demands on themselves so they are constantly stressed which results in lower quality of life and job satisfaction.

Conclusion
The results of this study are mainly theoretical and point out that affective temperaments can be regarded as predictors of life satisfaction, but in order to predict job satisfaction we have to take in consideration the aspects of wider social surrounding and sources of stressors at work.Future researches should also consider other indicators of life satisfaction such as family satisfaction, health satisfaction, free time etc.It is entirely possible that these aspects are differently influenced not only by personality traits, but also by situational factors that we examined.Furthermore, there are evidences that domestic stresses as well as the aspects of work environment lead to psychology disorders in healthcare professionals, therefore the lack of this study is that we did not examine stressors outside of work situations.This fact is important because it is known that job interference with family life and constant interruptions correspond with high levels of job dissatisfaction and negative well-being in healthcare professionals.

.05; AS -average score; SD -standard deviation. Table 4 Correlation of life and job satisfaction with demographic characteristics
* **p < 0.