A Structural Model on the Role of Perceived Multi-Dimensional Social Support in Attitudinal Variables

This research employs survey-based correlational model. This study has been carried out to reveal the relationships among perceived multi-dimensional social support and satisfaction with life, job satisfaction and job stress. The research data were gathered from 355 teachers working at schools in Hatay city in 2018-2019 academic year. Data were achieved by “Perceived MultiDimensional Social Support Scale”, “Job Satisfaction”, “Job Stress” and “Satisfaction with Life”. The relationships between multidimensional social support and job satisfaction, job stress and satisfaction with life have been determined in the measurement model. Structural equation model shows that perceived social support has a positive impact on teachers’ job satisfaction and satisfaction with life, in reverse, a negative impact on teachers’ job stress. Besides this, perceived social support positively affects teachers’ satisfaction with life through the partial mediation effects of job satisfaction and job stress. It has been concluded in this study that increasing perceived social support is an effective strategy for increasing job and life satisfaction and decreasing stress of teachers.


Introduction
One of the mostly researched subjects in recent years is social subject.It is a remarkable aspect of perceived social support that it has positive effects on work and private lives of individuals.When we take into consideration that most of the time of individuals passes at work, positive attitudes at work and their positive reflections on private life are seen crucial.Basaran (1982) has emphasized that individuals can generalize their important feelings by reflecting them on other fields, too.Dikmen (1995) has emphasized that individuals can transmit their good and bad memories at work to their private life, in reverse, they transmit their good and bad memories with their families and friends to their work.
Sincerity at work, family meetings out of work, union memberships and neighbourhood have been seen as social support components (Ozkalp & Kirel, 2010).Dikmen (1995) has expressed that people in this century ascribe blessing meanings to work, and they show this importance by their works.In order to stay away from harmful effects of job stress and to achieve happiness in all roles of their work, social support elements are seen important (Guney, 2012).Yildirim (2007) has stated that supportive social environment presenting whatever physical and psychological assistance has positive effects on individuals.Turkmenoglu (2015) has explained that humankinds have a wish to be happy and succeed all roles they take on.It comes to mind that people can achieve satisfaction in profession life by means of social support.
Previous researchers indicate that teachers may experience problems due to some reasons such as teacher-student incompatibility, teacher-family incompatibility, indisciplines of students, not rise in vocation, insufficient support from society and inadequate attendance to decisions (Russell, Altmaier & Van Velzen, 1987;Sorenson, 2007).It is required to give physical and psychological supports to teachers who are struggling such problems and support them to improve quality of education.
Social support is known one of the most important factors to enhance physical and psychological power.This paper focused on teachers' work outcomes when they perceive social support.Researcher hasn't met a conceptual framework for determining the relationship between social support, job satisfaction, job stress and satisfaction with life in educational organizations.Thus, it can be stated that social support is a new concept in this relationship.In the literature, most of studies in social support focus on the relationships between social support and attitudes in the other organizations except educational organizations (Korte & Simonsen, 2018).Social support perception can be utilized as an effective performance variable to enhance desired attitudes and to reduce undesired attitudes of teachers at school.There is a necessity to examine the relationship between attitudinal outcomes and social support to get more conceptual knowledge.Therefore, the relationships between social support and attitudinal variables were examined in a single structural model.The aim of this research is to find the roles of perceived social support on teachers' job satisfaction, job stress and satisfaction with life.Also, in this research, the mediation effects of job stress and job satisfaction have been examined in the relationships between perceived social support and satisfaction with life.The effects of teachers' multi-dimensional social support taken from their friends (in and out of work), their family or special people on attitudinal variables have been identified with only one structural model.

Perceived Social Support and Job Satisfaction
Ozkalp and Kirel (2010) have identified that the concept of social support as satisfying the basic social needs of the individual as a result of their interaction with other individuals.Guney (2012) has explained social support as communication with groups and other people.That's to say, in this research, perceived social support has been expressed as all the help and support received from social support elements, such as family, friends or a special person in an individual's life (Cohen & Hoberman, 1983).Social support is an important resource to cope with negative manners, such as stress and burnout among classroom teachers (Paine, 1981;Russell et al., 1987).
Job satisfaction is a concept that states how happy an individual is with his or her job (Saruhan & Yildiz, 2014;Schermerhorn, Hunt, Osborn & Uhl-Bien, 2011).The people who attain what they expect from their jobs reach saturation as much as they obtain (Simsek, Akgemci & Celik, 2011).The higher the level of satisfaction that the individual reaches as a result of his/her work life evaluation, the higher the satisfaction he/she reaches from his/her work (Basaran, 1982).Kreitner and Kinichi (2009) have described job satisfaction as an emotional and soulful reaction to different aspects of work.Kreitner and Kinichi (2009) have expressed that an individual may relatively be satisfied with some aspects of the job and may not be satisfied with some aspects of the job.Job satisfaction can provide corporate performance increase by increasing performances and successes of employees (Saruhan & Yildiz, 2014).Job satisfaction deprivation can cause some negative results such as, absence at work, leave of employment, being adversely affected by coworkers and environment (Saruhan & Yildiz, 2014;Schermerhorn et al., 2011).The individual's good friendship, solidarity and collaborations with other employees at work affect job satisfaction (Basaran, 1982).In addition, support from their family, special people in their life and friends out of work can also contribute to teachers' work life quality.Yilmaz and Aslan (2018) have already determined positive relationship between teachers' job satisfaction and social support perceptions.
Perceived Social Support and Job Stress Schermerhorn et al. (2011) have depicted stress as a throttle that causing by chances, restrictions and extraordinary demands.Restrictions are relevant to retention from doing the desired thing, and demands are related to the loss of the desired thing (Robbins, Decenzo & Coulter, 2013).Stress is one of the elements causing negative effects on structural behaviours and health problems.Stress has positive relationship with non-attendance, dismissal, job turnover, heart related disorders and viral infections (Kreitner, 1982).People who have a feeling of loathing show hostile attitudes and behaviors to their organizations and to them (Simsek, Celik & Akgemci, 2014).
People are mostly stressed at work.Stress arises due to too much or little responsibility and too much or too little control (Simsek et al., 2014).Some employees at work may coerce to others.Insufficient support from coworkers, economical problems, weak interpersonal relations and personality characteristics are factors for stress (Robbins et al., 2013).Kreitner and Kinichi (2009) have interpreted that domineering family worries may cause stress and individuals can ease this problem with sufficient manager support about regulating hours of labor.Robbins et al. (2013) have said that employees would reflect their problems to their work.That's to say, teachers may be exposed to job stress as a result of some situations such as, insufficient support from their family, their manager and their friends.Cohen and Hoberman (1983) have expressed that perceived social support is a crucial factor to handle distressing situations.Guney (2012) has told that people living alone and people not accepted to any groups are more sensitive and under risk than other people.Russell et al. (1987) have found a significant relationship between school administrator support and teachers' work stress.

Perceived Social Support and Satisfaction with Life
The life of an individual is nothing more than time passing in and out of business life.Time spent out of work is also free time and other non-work time.Job satisfaction is individual's sensual reaction to life consisting of the time in-and outside of work (Dikmen, 1995).Individuals has some different roles at work, at home and in free times and defining individual's life satisfaction in these roles has been preferred in this research.Topkaya and Buyukgoze Kavas (2015) have concluded that the higher the perception a person at work, the higher the satisfaction with life would be.

Job Satisfaction and Job Stress
Managers can positively affect some organisational outputs such as performance, production and customer delight etc. by increasing employees' job satisfaction (Kreitner & Kinichi, 2009).In meta-analysis consisting of 11063 people and 32 surveys, it has been detected that there is a negative relationship between stress and job satisfaction (Kreitner, 1982).Wang, Hall and Rahimi (2015) have found that controlling job stress factors of teachers may decrease burnout and intent to leave.On the other hand, it may increase job satisfaction.Also, Demir (2018) revealed that teachers' job satisfaction is negatively correlated to their level of stress.Basaran (1982) has expressed that leave of employment can cause many psychosomatic disorders.
Job Satisfaction and Satisfaction with Life Basaran (1982) has emphasized that for individuals, it is normal to reflect their emotions at work to their family life.In previous investigations, some researchers have found that there is a positive relationship between job satisfaction and satisfaction with life at medium levels (Dikmen, 1995;Reizer, 2015;Rice, McFarlin, Hunt & Near, 1985).When job satisfaction and satisfaction with life have been investigated, it has been seen that satisfaction with life has been included to analysis as a dependent variable (Reizer, 2015;Rice et al., 1985).
Perceived social support increases employees' job satisfaction (Basaran, 1982).Besides, if job employees' job satisfaction increases, their satisfaction with life will also arise (Dikmen, 1995;Reizer, 2015;Rice et al., 1985).Positive emotional situations of individuals raise their environment interest and their wishes to be in enjoyable activities (Judge & Ilies, 2004).Shortly, it has been examined in this survey that job satisfaction has had a mediation effect in the relationship between social support and satisfaction with life.

Job Stress and Satisfaction with Life
Having a vivacious or stable life affects the stress level of individuals.During life change, system may be enforced because of excessive loading, and it may show weakness (Simsek et al., 2014).There is a positive relationship between a stressed life and disruption of work-life balance (Schermerhorn et al., 2011).It has been seen that if people expose to excessive job stress, they cannot get satiety from their work, too.In short, the lower the job stress is, the more the pleasure in their life will be.

Research Frame
In the research frame below (figure 1), it's been seen that while social support perception gets strong, teachers' job and life satisfaction increase, and job stress reduces.The researcher has asserted that perceived social support can reduce teachers' negative attitudinal outcomes and positive attitudinal outcomes.H5b: Job satisfaction has a mediation role in the relationship between perceived social support and satisfaction with life.
H6a: Job stress has a negative effect on satisfaction with life.
H6b: Job stress has a mediation role in the relationship between social support and satisfaction with life.

Research Design
This research has been designed according to correlational model that is one of the quantitative investigation approach.Correlational model is a model aiming to determine the existence or/and the level of interchange between variables (Buyukozturk, Cakmak, Akgun, Karadeniz, & Demirel, 2012;Christensen, Johnson & Turner, 2015;Karasar, 2012).The relationships between multi-dimensional social support and job satisfaction, job stress and satisfaction with life have been determined.Then, a structural model has been created in line with the conceptual framework.

Population and Sample
The study population of the present research consisted of teachers working at public schools in Hatay in 2018-2019 academic year.Disproportionate cluster sampling model defining as choosing groups for samples was used in this survey.Schools in Hatay were taken as clusters, and 39 schools were randomly selected for study population of this study.Based on the principle of voluntariness, 355 teachers working in these schools have accepted to participate in the study.A form consisting of personal information and scales in this survey were given to teachers.Teachers were informed shortly about the variables and the aim of study.Researcher have told teachers to tick up the most suitable options by their own ideas.Forms were taken by the researcher a few days later from the schools.The sampling number was enough to represent study population at 95 percent for confidence level (Field, 2009).A participant group with substantial numbers (200 and above) is required for complex analyses in structural model (Cokluk, Sekercioglu & Buyukozturk, 2014).Kline (2011) has suggested that the sample size should be 10 times of the total number of items in the scales.The number of items of the scales in this research is 29.The multiplication of 10 and 29 equals to 290.Due to the fact that the research participant group consists of 355 teachers, this number was available for the research objective and statistical analysis.42.8% of teachers (n=152) work at primary schools, 35.2% of them (n=125) work at middle schools, and 22% (n=78) of them work at high schools.

Research Instruments
Survey data were acquired with "multi-dimensional perceived social support scale", "job satisfaction", "job stress" and "satisfaction with life" scales.Scale data acquired in consequences of reliability analysis and exploratory and confirmatory factor analysis (CFA) were explained below.

Multi-dimensional Perceived Social Support Scale:
Multi-dimensional social support scale which was developed by Zimet, Dahlem, Zimet and Farley (1988) has been used to determine the level of multi-dimensional social support.Also, it was adapted to Turkish by Eker and Arkar (1995), and the psychometric features were reviewed by Eker, Arkar and Yaldiz (2001).There are not any reverse items in this scale that is three dimensional with 12 items.Each dimension contains four items.The examples of the scale items are: "My family really try to help me", "My friends really try to help me" and "I can talk my problems with my friends".In current study, the Kaiser-Meyer-Olkin measure of sampling adequacy was .911,and Bartlett's test of sphericity was significant (p<.05).This result has revealed the adequacy of the sample for explanatory factor analysis.In the result of exploratory factor analysis, the scale was obtained as three dimensional and with 12 items like in its original form.Every dimension consisted of 4 items.These three dimensions have explained 78.19% of total variance.According to rotation explanation values, first dimension consisted of family dimension items, and second dimension consisted of items of special person dimension.Third dimension contained friend dimension items.Cronbach's Alpha's reliability coefficient was determined .904 in family dimension, .916 in special person dimension, .880 in friend dimension and .935throughout this scale.That's to say, it has been seen that multi-dimensional social support scale was extremely reliable for this research.
Job Satisfaction Scale: Global job satisfaction scale which was developed by Griffin, Hogan, Lambert, Tucker-Gail and Baker (2010) has been used in this present investigation.This scale consists of five items.One of these items is a reverse item.The examples of the scale items are: "I don't like my job at all", "I am doing my job with pleasure" and "I am too pleased to do this job".In current research, the Kaiser-Meyer-Olkin measure of sampling adequacy was .840,and Bartlett's test of sphericity was significant (p<.05).This result has illustrated the adequacy of the sample for exploratory factor analysis.The scale was unidimensional, and this dimension explained 68.35% of total variance.Cronbach's Alpha's reliability coefficient was determined as .882.This result has proved that used job satisfaction scale was extremely reliable.
Job Stress Scale: Job stress scale developed by House and Rizzo (1972) and adapted to Turkish by Efeoglu (2006) was used in this study.There are not any reverse items in this scale that consists of seven items.The examples of the scale items are: "My job tends to influence my health directly", "I work under a big tension" and "I feel irritable because of my job".In this survey, the Kaiser-Meyer-Olkin measure of sampling adequacy was .877and Bartlett's test of sphericity was significant (p<.05).This result has revealed the adequacy of the sample for exploratory factor analysis.In this investigation, after exploratory factor analysis, one dimension was obtained like its original scale form.It has been determined that this dimension could explain 59.96% of total variance.Cronbach's Alpha Reliability Coefficient was calculated as .884.In short, scale has been extremely reliable.
Satisfaction with Life Scale: "Satisfaction with Life Scale" was developed by Diener, Emmons, Larsen and Griffin (1985) and adapted to Turkish by Imamoglu (2004) has been used to determine the level of satisfaction with life of teachers in this research.There are not any reverse items in this scale that consists of five items.The examples of the scale items are: "My living conditions are great", "I am content with my life" and "If I lived my life again, I changed nothing".In current study, the Kaiser-Meyer-Olkin measure of sampling adequacy was .840and Bartlett's test of sphericity is significant (p<.05).This result has illustrated the adequacy of the sample for factor analysis.Original form of scale consisting of unidimensional items was confirmed in current research.This dimension explained 64.14% of total variance.Cronbach's Alpha reliability coefficient was calculated as .856.This reliability result has showed that satisfaction with life scale is extremely reliable.
Confirmatory factor analysis has been done for every scale benefited in current study and for some criterions, fit values are seen on Table 1 below.(Celik & Yilmaz, 2016).It has been expressed that if TLI, CFI, and IFI values are between .90 and .95,there is an acceptable goodness of fit index value and if values are over .90, it is evidence for a good fit index (Arbuckle, 2009;Byrne, 2010;Kline, 2011).After CFA, path model was designed in accordance with the theoretical framework.Path analysis was performed with latent variables.Because, it has produced more reliable results than path analysis with observed variables (Meydan & Sesen, 2015).

Descriptive Analysis Related to Variables and Correlation Matrix
Arithmetic average, standard deviation and standard error values related to variables investigated in survey are given in Table 1 below.2, teachers' perceptions related to their social support and job satisfaction are at partially high levels.Job stress perceptions of teachers are at partially low levels.On the other hand, their satisfaction with life perceptions are at medium levels.When correlation matrix have been examined, perceived social support is positively associated with job satisfaction (r=.245, p<.001) at a low level and satisfaction with life (r=.314,p<.001) at a medium level, but it is negatively associated with job stress (r=-.184,p<.001) at a low level.Satisfaction with life is negatively associated with job stress (r=-.235,p<.001) at a low level.Job satisfaction is negatively associated with job stress (r=-.331,p<.001) at a medium level.There is a positive relationship between satisfaction with life and job satisfaction (r=.334, p<.001) at a medium level.It has been seen that there is a statistically significant relationship at .01 level between all variables investigated in this research.
After CFA has been performed to every scale, measurement model has been constituted.Error covariance have been added between D5 and D10, D3 and D4, S1 and S2 because the errors of these items are related to each other.It has been seen that all measured variables in this measurement model have a statistically significant relationship with each other.Measurement model has showed that all of the scales used in the study are in good agreement with the research data.
Significant relationships between latent variables in measurement model has been shown in figure 2. The fit indexes related to measurement model were given in Table 3.The fit index of the measurement model built was obtained as follows: Chi squared (x 2 ) = 897.591;df=366; CMIN/DF (x 2 /df) = 2.452; GFI=.854;IFI=.925,TLI=.916,CFI=.924;RMSEA =.064.This result has illustrated that the model fit index is at an acceptable level.After the measurement model has showed a good fit, covariances between latent variances have been cleared and one-way paths have been added to the latent variances in accordance with the hypotheses to be tested.
Standardized path coefficients and structural equation model that produces the best fit values have been presented in Figure 3.The fit index of the model built has been achieved as follows: Chi squared (x 2 ) = 897.067;df=365; CMIN/DF (x 2 /df) = 2.457; GFI=.854;IFI=.925,TLI=.916,CFI=.924;RMSEA =.064.This result has depicted that the model fit index is at an acceptable level.According to the results of structural equation model in Figure 3, perceived social support has positive effects on job satisfaction and satisfaction with life of teachers.On the other hand, it has a negative effect on teachers' job stress.In addition, perceived social support has a positive impact on life satisfaction through the partial mediation effects of job satisfaction and job stress.

Discussion and Recommendations
Perceived social support increases teachers' job and life satisfaction, but it decreases teachers' job stress.Perceived social support increases teachers' life satisfaction by increasing job satisfaction and decreasing teachers' job stress.Teachers' job satisfaction, directly or through the partial mediation of job stress indirectly increases their life satisfaction.Job stress negatively influences satisfaction with life.Getting support or help from family, friends (at-out of work), administrators and special people positively affects teachers' attitudes.Yilmaz and Aslan (2008) have found that there is a positive relationship between social support and teachers' job satisfaction.Also, in this study, similarly, it has come up that perceived social support is positively associated with teachers' job satisfaction.Backing up by their family, a special person in their lives or other people out of work contribute to work life quality of teachers.If teachers get good feedback from their environment, they will do their jobs better.Thus, their job satisfaction will also be higher.Cohen and Hoberman (1983) have determined that perceived social support is an important factor to cope with stressful situations.Guney (2012) has found that support from people in their family, society or colleagues, the sharing levels, being happy together can enhance the success level of teachers in coping with stressful situations.Russell et al. (1987) have detected a significant relationship between school administrator support and job stress.Similarly in this research, support and help can lower teachers' job stress.Topkaya and Buyukgoze-Kavas (2015) have found that if the social support perception of teachers is high, satisfaction with life will also be high.It has also been seen in this research that perceived social support can increase life satisfaction of teachers.It has been determined that perceived social support affected teachers' satisfaction with life indirectly by the partial mediation effects of job satisfaction and job stress.In another word, job satisfaction and job stress are mediators in relationship between social support and job satisfaction.
In this research it has been detected that job satisfaction has a negative impact on job stress.Similarly, previous researchers (Demir, 2018;Wang et al., 2015) found out that teachers' job satisfaction is negatively correlated to their job stress.Also, in previous researches, it was concluded that there is a moderate and positive relationship between job satisfaction and life satisfaction (Dikmen, 1995;Reizer, 2015;Rice et al., 1985).It has been encountered with a similar results also in this research.If job satisfaction of individuals increases, their satisfaction with life will increase, too.
Differently in this research, job satisfaction increases life satisfaction with the partial effect of job stress.Satisfaction with life lets up the stress at work.As the natural result of this situation, life satisfaction is affected positively.Schermerhorn et al. (2011) have alleged that if stress at work increases too much, their lives will negatively be affected.
Teachers obtaining useful and beneficial information with the aid of social support can do their jobs easier and overcome uncertainties in their lives.These situations can positively influence all their lives.
Based on the literature review, this research has constructed an integrated framework to illustrate the effects of the perceived social support on teachers' attitudinal outputs.The perceived social support is a strong variable that has impacts on teachers psychologically in private and organizational outcomes in general.The results of this research have pointed out that perceived social support can improve job satisfaction and satisfaction with life, and reduce job stress.
The findings also have revealed that job satisfaction and job stress act as the mediators embedded in the relationship between perceived social support and satisfaction with life.Job satisfaction and job stress are important variables between perceived social support and satisfaction with life.
It is required to realize that social support perception can be used as an effective performance variable to enhance positive attitudes and to reduce negative attitudes of teachers at work.Examining multi-dimensional social support concept with important outputs would contribute to the developing well-designed training programs towards school leaders and teachers.Therefore, these programs may help the increase of awareness related to the importance of social support.
It has been concluded in this study enhancing perceived social support is an effective strategy for increasing job and life satisfaction and decreasing stress of teachers.Based on the findings of this study, it hasn't been concluded that there is a causal connection with perceived social support and job satisfaction, job stress and satisfaction with life.Data of this study have been collected in a short time with the help of scales.In further studies, the causal effects of perceived social support on attitudinal variables can be carefully evaluated.Furthermore, the effects of the basic points of social support (family, friends and special person) on job stress, job satisfaction and life satisfaction can also be investigated.It can be revealed which social support element is effective on these attitudinal variables.Some suggestions can be presented for the development of this social support element.

Figure 1 :
Figure 1: Research Framework Notes: Social support: Perceived social support, Family: Support from family, Special person: Support from special person, Friend: Support from friend, JSat: Job satisfaction, JStress: Job stress, SWL: Satisfaction with life The researcher has tested hypotheses below in the light of the related literature.H1: Social support perception has a positive effect on job satisfaction.H2: Social support perception has a negative effect on work stress.

Figure 2 .
Figure 2. Measurement model with standardized coefficients Notes: Family: Support from family, Special person: Support from special person, Friend: Support from friend, JSat: Job satisfaction, SWL: Satisfaction with life

Figure 3 .
Figure 3. Standardized path coefficients and the last structural equation model Notes: Family: Support from family, Special person: Support from special person, Friend: Support from friend, JSat: Job satisfaction, SWL: Satisfaction with life

Table 1 .
The fit indexes of the scales as a result of CFA The fit indexes of scales according to confirmatory factor analysis has been given in Table1.Confirmatory factor analysis has been done for every scale used in current study.The best fit indexes calculated in consequence of confirmatory factor analysis in Table1has showed that used scales are extremely coherent for this research data.When CMIN/DF (x 2 /df) rate is below three as it is shown in Table1, it means an acceptable fit index.RMSEA values are below than .08,so it is acceptable.GFI, TLI, CFI, and IFI values are over than .95.Therefore, it is evidence for a good fit index.This result has revealed that the model fit indexes of the scales are at an acceptable and desired level.Data have been examined in terms of reliability.Factor analysis results and Cronbach's Alpha Reliability Coefficient have showed that scales are reliable and valid.In this research, descriptive analyses, correlation and structural equation modeling analyses have been used.During CFA, factor structures of every scale were examined and looked whether acceptable adaptive values are produced or not in terms of some criteria.These criteria are as below.If CMIN/DF (x 2 /df) rate is below two, it means there is a good concordance.If CMIN/DF (x 2 /df) rate is between two and five, it indicates an acceptable fit index.If RMSEA values are between .08 and .05, it is acceptable.If values are below .05, it means there is a good fit

Table 2 .
Descriptive statistics and correlation values of variables in this survey.**p<.01Support: Perceived social support, JSat: Job satisfaction, JStress: Job Stress, SWL: Satisfaction with life According to Table

Table 3 .
Fit indexes of measurement model.

Table 4 .
The fit indexes related to structural equation modeling