بهزیستی ذهنی در محل کار: اثرات منبع استرس و پشتیبانی در شور و شوق، قناعت و معنادار

Abstract The experience of meaningfulness in one's work is an important predictor of individual and organizational outcomes. We advance a model of nonhedonic (i.e., meaningfulness) subjective well-being (SWB), to assess the potential impact of work role stress (specified by source) in this new model of SWB, and examine the direct and indirect effects of potentially supportive communication (specified by source and type of support). Item Response Theory (IRT) and Classical Test Theory (CTT) frameworks confirmed the proposed factor structure of SWB. Results suggest that positional status of the source of stress determines the magnitude of effect. Specifically, supervisors who are sources of role stressors have the largest negative impact on SWB. Alternatively, social support had the largest positive effect on SWB when the source was a supervisor and communication type was positive or non-work related. In addition, role stressors and communication from supervisors have the strongest direct effect on nonhedonic SWB. Unexpectedly, communication wherein content was negative had the strongest effect when source was a coworker, followed by the subordinate. Finally, positive communication with coworkers attenuated the effect of supervisor role stress on one dimension of SWB (enthusiasm). Overall, supervisors have a strong ability to affect subordinates' SWB, particularly in terms of meaningfulness.

. Introduction The increasing recognition and popularity of positive psychology (Sheldon & King, 2001) has led organizational researchers to investigate employee happiness and other positive emotions (e.g., Luthans, 2002). Applied to health and well-being of workers, positive psychology suggests that well-being is something other than simply the absence of illness. There exists, however, a dichotomy in the definition of well-being. One approach is the hedonic view of well-being and emphasizes such experiences as pleasure, happiness, satisfaction, and the presence of a positive mood ( Diener, 1984 and Kahneman et al., 1999); this approach has dominated the work and well-being literature ( Diener, Suh, Lucas, & Smith, 1999). The other view of well-being emphasizes experiences of greater depth such as meaning ( Jim et al., 2006, Schnell and Becker, 2006, Steger et al., 2006 and Weems et al., 2004), purpose ( Jim et al., 2006, Reker, 1992, Steger et al., 2006 and Weems et al., 2004), importance ( Weems et al., 2004), fulfillment ( Reker, 1992), and eudaimonia ( Waterman, Schwartz, & Conti, 2008). The current study investigates psychological well-being at work and primarily helps to develop the usefulness of the second approach: examining the sense of meaningfulness that employees experience at work. Whereas most previous research on the experience of meaningfulness emphasizes cognitive evaluation of one's life in general, the current study extends prior theory by increasing our understanding of nonhedonic subjective well-being at work ( Rosso, Dekas, & Wrzesniewski, 2010). Designing jobs that enhance feelings of purpose or significance can dramatically enhance individual outcomes and performance (e.g., Grant, 2012). Not only has meaningful work gained popularity among scholars (see Dik, Byrne, & Steger, 2013), but practitioners suggest that meaningful experience at work is a common feature among the most successful and innovative companies ( Bain, 2007). Previous researchers have obfuscated the topics of meaning and meaningfulness by using them interchangeably. Meaning is an outcome of having made sense of something, such as when an individual employee interprets what one's work or organizational life connotes (Gray, Bougon, & Donnellon, 1985). Meaningfulness refers to the magnitude or amount of significance one feels (Pratt & Ashforth, 2003). Two individuals may assign the same meaning to a work activity and yet differ in the valence of meaningfulness. There have been numerous studies of meaning at work (e.g., MOW (Meaning of Work) International Research Team, 1987) and more recently meaning associated with objective virtue of work (Kashdan, Biswas-Diener, & King, 2008). The measurement of meaningfulness, however, as it relates to subjective well-being (SWB) has received relatively little attention (e.g., Ryff, 1989), and, therefore the present research contributes to our understanding of this type of well-being. These two studies advance a new model of SWB, assess the impact of role stress on this model, and test boundary conditions of the model by examining moderating variables. First, measurement qualities of the hypothesized components of SWB were assessed. This first step was intended to create a brief psychometrically appropriate assessment of hedonic and non-hedonic SWB. Next, relationships of work-related stressors with these different types of SWB were examined. Finally, social support as both a direct correlate of SWB and a moderator of the stressor and SWB relationship was tested.

9. Results As in Study 1, internal consistency estimates for manifest variables were good: enthusiasm (M = 3.94; SD = .67; α = .877), contentment (M = 3.38; SD = .68; α = .873), and meaningfulness (M = 3.83; SD = .72; α = .906). The nonhedonic scale correlated with contentment at r = .566, p < .000 and enthusiasm at r = .758, p < .000. Contentment and enthusiasm correlated r = .674, p < .000. t-Tests to assess equality of means between samples in Study 1 and Study 2 were not significant for enthusiasm (t = .60; p > .05) and meaningfulness (t = .33; p > .05). However, the difference between means for contentment in Study 1 (M = 3.30; SD = 0.70) and Study 2 (M = 3.38; SD = .68) was statistically significant (t = 2.11; p < .05). 9.1. CFA The three-factor CFA model of the SWB indices, with correlated error variances between positive and negative items within factor, was tested to confirm the structure found in Study 1. All loadings were significant at p < .05 on their respective latent factors; the model fit indices were acceptable ( Hu & Bentler, 1999) and were slightly better than the indices in study 1; χ2 (114, N = 1050) = 795.17, p < 0.0, SRMR = 0.06, RMSEA = .08, CFI = .98, NNFI = .97. 9.2. Bivariate correlations Hypothesis 2, that social support will have a stronger positive relationship with well-being when the source of support is higher in the organizational hierarchy (i.e., support from the supervisor will have the strongest relationship, followed by support from coworkers, and their subordinates) was assessed by comparing z transformed correlations and significance test for dependent coefficients ( Steiger, 1980). The positive correlations between total (combined positive, non-work, and negative communication scales) supervisor supportive communication and each type of well-being (contentment, r = .161, p < .000; enthusiasm, r = .299, p < .000; meaningfulness r = .351, p < .000) were significantly greater than those for total coworker communication (contentment, r = − .056, p > .05; enthusiasm, r = − .011, p > .05; meaningfulness r = .025, p > .05). Likewise, most positive correlations between total supervisor supportive communication and well-being were greater than those of total subordinate communication (contentment, r = .022, p > .05; enthusiasm, r = .047, p > .05; meaningfulness r = .134, p < .01) and well-being. Of course aggregate communication obscures the impact of content. The patterns of correlations displayed across different types of contents ( Table 5) suggest the greatest direct positive effect from one's supervisor. Therefore the trend of differences between pairs of correlations provides support, and the significance test of difference in magnitude between correlations provides partial support, for Hypothesis 2 (see Table 5). Table 5. Tests of hypothesized differences in correlations between role stressor source and subjective well-being based on the source and type of communication (Study 2). Subordinate Coworker Supervisor Total Zry,x1–Zry,x2 (p) Zry,x1–Zry,x3 (p) Zry,x2–Zry,x3 (p) Total communication SWBc .022 − .056 .161⁎⁎⁎ .069⁎ .078 (.042) − .140 (.003) − .218 (.000) SWBe .047 − .011 .299⁎⁎⁎ .174⁎⁎⁎ .058 (.130) − .261 (.000) − .319 (.000) SWBm .134⁎⁎ .025 .351⁎⁎⁎ .243⁎⁎⁎ .110 (.004) − .232 (.000) − .342 (.000) Positive work-related communication SWBc .184⁎⁎⁎ .183⁎⁎⁎ .235⁎⁎⁎ .252⁎⁎⁎ .001 (.975) − .053 (.182) − .054 (.086) SWBe .309⁎⁎⁎ .319⁎⁎⁎ .436⁎⁎⁎ .440⁎⁎⁎ − .011 (.744) − .148 (.000) − .137 (.000) SWBm .367⁎⁎⁎ .346⁎⁎⁎ .483⁎⁎⁎ .487⁎⁎⁎ .024 (.484) − .142 (.001) − .166 (.000) Non-work related communication SWBc .028 .002 .155⁎⁎⁎ .083⁎ .026 (.469) − .128 (.005) − .154 (.000) SWBe .031 .026 .237⁎⁎⁎ .141⁎⁎⁎ .005 (.889) − .211 (.000) − .216 (.000) SWBm .090 .058 .255⁎⁎⁎ .178⁎⁎⁎ .032 (.371) − .171 (.000) − .203 (.000) Negative work-related communication SWBc − .166⁎⁎⁎ − .298⁎⁎⁎ − .019 − .204⁎⁎⁎ .140 (.000) − .149 (.002) − .288 (.000) SWBe − .223⁎⁎⁎ − .337⁎⁎⁎ .033 − .216⁎⁎⁎ .124 (.001) − .260 (.000) − .384 (.000) SWBm − .146⁎⁎⁎ − .307⁎⁎⁎ .104⁎⁎ − .137⁎⁎⁎ .170 (.000) − .251 (.000) − .422 (.000) Note: SWBc = contentment, SWBe = enthusiasm, SWBm = meaningfulness, Zry,x1–Zry,x2 = z transformed difference between the first and second (from left to right) correlations in the table; Zry,x1–Zry,x3 = z transformed difference between the first and third correlations in the table; Zry,x2–Zry,x3 = z transformed difference between the second and third correlations in the table; n = 639–1050. ⁎ p < 0.05. ⁎⁎ p < 0.01. ⁎⁎⁎ p < .001. Table options Hypothesis 3 stated that social support will have the strongest positive relationship with well-being when it consists of positive communication (followed by non-work related, and negative communication) and was also assessed by comparing the magnitude of correlation coefficients (Steiger, 1980). This hypothesis was supported, as communication with positive content displayed the strongest positive correlation with each type of well-being, followed by social support where the content is not related to work, and finally a negative correlation between negative work-related communication and each type of well-being. As can be seen in Table 5, there are 27 pairs of correlation coefficients that differ significantly, most of which are those comparing either supervisor versus subordinate as the source of support or supervisor versus coworker, but not subordinate versus coworker. Again, the trend in correlations provides support, whereas the statistical comparison of magnitude of correlations provides partial support, for Hypothesis 3. Hypothesis 4a and Hypothesis 4b were tested with a similar analysis. They stated that the negative relationship of role ambiguity (H4a) and role conflict (H4b) with well-being would be stronger when the source was the supervisor than the coworker or subordinate. For each type of well-being, the relationship with both role ambiguity and role conflict was strongest when the source was one's supervisor (Table 6). Using a z transformation and comparison for dependent correlations ( Steiger, 1980), the supervisor role stressor (i.e., conflict, ambiguity) and well-being correlations were significantly greater than the correlations for both coworker and subordinate, but none of the coworker role stressor and well-being relationships is significantly greater than subordinate relationships. The results suggest a clear trend that role stress has a stronger direct relationship with well-being when its source is one's supervisor, partially supporting Hypothesis 4a and Hypothesis 4b. Table 6. Tests of hypothesized differences in correlations between role stressor by source and subjective well-being (Study 2). Subordinate Coworker Supervisor Total Zry,x1–Zry,x2 (p) Zry,x1–Zry,x3 (p) Zry,x2–Zry,x3 (p) Role ambiguity SWBc − .221⁎⁎⁎ − .253⁎⁎⁎ − .350⁎⁎⁎ − .343⁎⁎⁎ .034 (.373) .141 (.002) .107 (.007) SWBe − .276⁎⁎⁎ − .310⁎⁎⁎ − .466⁎⁎⁎ − .472⁎⁎⁎ .037 (.335) .222 (.000) .184 (.000) SWBm − .337⁎⁎⁎ − .353⁎⁎⁎ − .524⁎⁎⁎ − .515⁎⁎⁎ .018 (.643) .231 (.000) .213 (.000) Role conflict SWBc − .261⁎⁎⁎ − .264⁎⁎⁎ − .393⁎⁎⁎ − .339⁎⁎⁎ .003 (.932) .148 (.001) .145 (.000) SWBe − .254⁎⁎⁎ − .282⁎⁎⁎ − .447⁎⁎⁎ − .390⁎⁎⁎ .030 (.426) .221 (.000) .191 (.000) SWBm − .246⁎⁎⁎ − .280⁎⁎⁎ − .433⁎⁎⁎ − .370⁎⁎⁎ .037 (.335) .212 (.000) .176 (.000) Note: SWBc = contentment, SWBe = enthusiasm, SWBm = meaningfulness, Zry,x1–Zry,x2 = z transformed difference between the first and second (from left to right) correlations in the table; Zry,x1–Zry,x3 = z transformed difference between the first and third correlations in the table; Zry,x2–Zry,x3 = z transformed difference between the second and third correlations in the table; n = 610–1050. ⁎⁎⁎ p < .001. Table options 9.3. MSEM Finally, positive communication was hypothesized (H5a and H5b) to exhibit a moderating (buffering) effect on the relationship between role stressors and subordinates' well-being when the source of communication was either one's supervisor or one's coworker. This was tested using an unconstrained (Marsh et al., 2004) MSEM approach. Item parcels were created for all scales in order to reduce the overall number of indicators. Parceling is conventional in SEM when there are a high number of indicators and has been shown to have less impact on covariance estimates when items have five ordinal response options (Yang, Nay, & Hoyle, 2010), which is similar to items in the current study. An odd-even item number parceling approach was taken. Unstandardized path coefficients are presented in Table 7. The interactions between supervisor role stressors and positive supervisor communication were not significantly related to well-being, which does not support Hypothesis 5a. There was a significant interaction, however, between supervisor role stressors and coworker social support when the type of subjective well-being was enthusiasm. Probes of both interactions were in the hypothesized direction for H5b. Specifically, positive coworker support attenuates the negative relationship between supervisor role stress and enthusiasm at work. Table 7. Moderated structural equation models of hypothesized role stressor, communication source and content, and subjective well-being interactions (Study 2). φ λx DV Fit 1 2 3 X1 X2 Z1 Z2 X1Z1 X2Z2 SWBc SWBe SWBm RMSEA RMR CFI 1. RaSup 0.897⁎⁎⁎ 1.00 0.930⁎⁎⁎ − 0.295⁎⁎⁎ − 0.095⁎⁎⁎ − 0.114⁎⁎⁎ 2. PosSup − 0.476⁎⁎⁎ 0.789⁎⁎⁎ 1.00 1.039⁎⁎⁎ 0.055 0.213⁎⁎⁎ 0.158⁎⁎⁎ 3.1 × 2 3.464 1.00 0.208 0.014 0.002 − 0.007 0.112 0.115 0.950 1. RcSup 0.711⁎⁎⁎ 1.00 1.072⁎⁎⁎ − 0.370⁎⁎⁎ − 0.096⁎⁎⁎ 0.037 2. PosSup − 0.285⁎⁎⁎ 0.760⁎⁎⁎ 1.00 1.076⁎⁎⁎ 0.126⁎⁎⁎ 0.236⁎⁎⁎ 0.238⁎⁎ 3. 1 × 2 1.885⁎⁎⁎ 1.00 0.365 0.004 0.014 − 0.025 0.098 0.081 0.955 1. RaSup 0.921⁎⁎⁎ 1.00 .902⁎⁎⁎ − 0.313⁎⁎⁎ − 0.163⁎⁎⁎ − 0.153⁎⁎⁎ 2. PosCo − 0.221⁎⁎⁎ 0.641⁎⁎⁎ 1.00 1.294⁎⁎⁎ − 0.096⁎⁎ − 0.153⁎⁎⁎ 0.141⁎⁎⁎ 3. 1 × 2 0.471⁎⁎ 1.00 1.525⁎ 0.010 0.055⁎⁎ 0.041 0.098 0.078 0.953 1. RcSup 0.709⁎⁎⁎ 1.00 1.077⁎⁎⁎ − 0.394⁎⁎⁎ − 0.163⁎⁎⁎ − 0.018 2. PosCo − 0.096⁎⁎⁎ 0.634⁎⁎⁎ 1.00 1.307⁎⁎⁎ 0.152⁎⁎⁎ 0.187⁎⁎⁎ 0.180⁎⁎⁎ 3. 1 × 2 0.050 1.00 13.367 0.040 0.056⁎⁎ 0.000 0.097 0.062 0.951 1. RaSup 0.921⁎⁎⁎ 1.00 0.904⁎⁎⁎ − 0.287⁎⁎⁎ − 0.186⁎⁎⁎ 0.177⁎⁎⁎ 2. NegCo − 0.175⁎⁎⁎ 0.566⁎⁎⁎ 1.00 1.338⁎⁎⁎ − 0.259⁎⁎⁎ − 0.131⁎⁎⁎ 0.035 3. 1 × 2 1.403⁎⁎⁎ 1.00 0.471 − 0.026 − 0.002 − 0.003 0.098 0.065 0.953 1. RcSup 0.714⁎⁎⁎ 1.00 1.065⁎⁎⁎ − 0.329⁎⁎⁎ − 0.137⁎⁎⁎ − 0.021 2. NegCo − 0.265⁎⁎⁎ 0.601⁎⁎⁎ 1.00 1.257⁎⁎⁎ 0.213⁎⁎⁎ 0.128⁎⁎⁎ 0.045 3. 1 × 2 0.423⁎⁎ 1.00 1.419⁎⁎⁎ − 0.068 − 0.074⁎ − 0.005 0.098 0.065 0.953 Note: RaSup = role ambiguity from supervisor. RcSup = role conflict from supervisor. PosSup = positive communication with supervisor. PosCo = positive communication with coworker. NegCo = negative communication with coworker. SWB variables were allowed to covary. φ = unstandardized phi coefficient. λx = unstandardized lambda-x coefficient (1.00 = fixed). DV = dependent variables. SWBc = contentment, SWBe = enthusiasm, SWBm = meaningfulness. RMSEA = root mean squared error of approximation. RMR = root mean square residual. CFI = comparative fit index. n = 610–1050. ⁎ p < 0.05. ⁎⁎ p < 0.01. ⁎⁎⁎ p < 0.001.