Microsoft Word ji job Pres Preprint docx
Download 0.9 Mb. Pdf ko'rish
|
JIJobPres Preprint
FOOTNOTES
1 The time spent on surveys was used as an indicator of attention. We used 5 minutes (i.e., 300 seconds) as the cutoff, since it was around the 5th percentile in our sample, and dropped participants who completed each of three surveys in less than 5 minutes. 2 In supplementary analyses, controlling for method factor did not change the results of structural models for either Study 1 or Study 2. 3 They also suggest that research utilizing supervisor-ratings of job performance may want to control for potential self-presentation effects. JOB INSECURITY AND JOB PRESERVATION 42 TABLES Table 1 A Framework of Job Preservation Behaviors Resource Investment Target Task-Oriented Resource Investment Social-Oriented Resource Investment Job Preservation Strategy Promotive Performance Self-presentation Ingratiatory Behaviors Protective Avoiding Counterproductive Work Behavior Evasive Knowledge Hiding JOB INSECURITY AND JOB PRESERVATION 43 Table 2 Means (SDs) and Correlations Among Study Variables of Study 1 Mean (SD) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 1.Job Insecurity1 2.77(1.21) (.88) 2.Job Insecurity2 2.47(1.10) .63** (.84) 3.Job Insecurity3 2.76(1.12) .54** .43** (.84) 4.Counterproductive Work Behavior1 2.83(1.08) .68** .60** .45** (.96) 5.Counterproductive Work Behavior2 2.67(1.00) .70** .62** .41** .80** (.95) 6.Counterproductive Work Behavior3 2.50(1.01) .78** .71** .57** .76** .83** (.95) 7.Knowledge Hiding1 4.42(1.44) .43** .31** .36** .64** .54** .46** (.89) 8.Knowledge Hiding2 3.83(1.4) .50** .45** .20** .66** .76** .57** .61** (.87) 9.Knowledge Hiding3 3.55(1.51) .57** .56** .29** .70** .83** .74** .53** .84** (.90) 10.Self- Presentation1 3.54(.79) .38** .31** .39** .45** .35** .32** .55** .24** .22** (.73) 11.Self- Presentation2 3.37(.72) .24** .22** .35** .22** .22** .24** .32** .18** .12* .50** (.61) 12.Self- Presentation3 3.20(.69) .32** .19** .40** .14** .13* .33** .17** -.01 .09 .40** .36** (.60) 13.Performance1 6.76(1.15) -.43** -.28** -.37** -.43** -.45** -.54** -.37** -.31** -.40** -.19** -.14** -.14* (.87) 14.Performance2 6.61(1.12) -.40** -.21** -.26** -.44** -.49** -.47** -.39** -.41** -.46** -.16** -.20** -.10 .71** (.82) 15.Performance3 6.76(.91) -.32** -.24** -.43** -.42** -.40** -.38** -.51** -.36** -.38** -.20** -.17** -.08 .66** .65** (.77) Note. Diagonal values are Cronbach’s alphas. * p < .05, ** p < .01. JOB INSECURITY AND JOB PRESERVATION 44 Table 3 Fits of CFA Models with Different Number of Factors – Study 1 BIC χ 2 df CFI RMSEA SRMR 3-Factor Model 57119.73 6593.20 1479 0.69 0.097 0.116 6-Factor Model 56284.88 5687.51 1467 0.74 0.089 0.114 12-Factor Model 54833.00 3934.60 1416 0.85 0.070 0.072 15-Factor Model Free factor loadings 53834.52 2705.91 1377 0.92 0.051 0.055 Longitudinal constraints on loadings 53744.14 2780.81 1405 0.92 0.052 0.061 Note. The 15-factor model (JI, performance, counterproductive work behavior, evasive knowledge hiding, and self-presentation ingratiatory behavior at three times) was tested against three alternative models. The first alternative model (i.e., 3-factor model) placed all the indicators of each time into one factor, which ended up with 3 correlated factors (one for each time). The second alternative model (i.e., 6-factor model) placed all the JI indicators into one factor and all the performance-related outcome indicators into another factor, which ended up with 6 correlated factors (two for each time). The third alternative model (i.e., 12-factor model) placed indicators of counterproductive work behavior and knowledge hiding into one factor, which ended up with 12 correlated factors (four for each time). JOB INSECURITY AND JOB PRESERVATION 45 Table 4 Means (SDs) and Correlations Among Study Variables of Study 2 Mean(SD) 1 2 3 4 5 6 7 8 1. Job Insecurity 2.42(1.18) (.87) 2. Job Preservation Motivation 4.05(0.83) .09 (.90) 3.Counterproductive Work Behavior 1.47(0.42) .12 -.37** (.83) 4. Knowledge Hiding 1.56(1.04) .12 -.05 .29** (.92) 5.Self-Presentation 2.42(0.88) .01 .13* .11 .29** (.87) 6.Performance 4.37(0.75) -.14* .32** -.35** -.33** -.01 (.91) 7.Threat Controllability 3.34(1.23) -.28** .08 -.05 .00 .10 .06 (.94) 8.Threat Proximity 1.91(1.04) .64** -.03 .15* .21** .04 -.22** -.38** (.96) Note. Diagonal values are Cronbach’s alphas. * p < .05, ** p < .01. JOB INSECURITY AND JOB PRESERVATION 46 Table 5 Fits of CFA Models with Different Number of Factors – Study 2 BIC χ 2 df CFI RMSEA SRMR 5-Factor Model 18586.13 2117.39 395 0.73 0.121 0.128 6-Factor Model 18185.23 1688.04 390 0.80 0.106 0.083 7-Factor Model I 17966.25 1434.92 384 0.84 0.096 0.077 7-Factor Model II 17678.10 1146.77 384 0.88 0.082 0.092 8-Factor Model 17277.34 706.17 377 0.95 0.054 0.048 Note. The 8-factor (i.e., JI, job preservation motivation, counterproductive work behavior, self- presentation ingratiatory behavior, knowledge hiding, job performance, perceived threat controllability, and threat proximity) model was tested against four alternative models. The first alternative model (i.e., 5-factor model) combined the four outcomes into one factor. The second alternative model (i.e., 6-factor model) combined threat controllability, threat proximity, and JI indicators into one factor. The third alternative model (i.e., 7-factor model I) combined threat controllability and threat proximity into one factor. The fourth alternative model (i.e., 7-factor model II) combined counterproductive work behavior and knowledge hiding into one factor. JOB INSECURITY AND JOB PRESERVATION 47 Table 6 Unstandardized Path Coefficients (SEs) of the Latent Moderated Mediation Model of Study 2 Job Preservation Motivation Job Performance CWB Self- Presentation Knowledge Hiding Job Preservation Motivation 0.41 (0.08)** -0.49 (0.08)** 0.16 (0.07)* -0.10 (0.07) Job Insecurity (JI) 0.17 (0.10) -0.22 (0.08)** 0.22 (0.08)** -0.01 (0.07) 0.15 (0.07)* Threat Proximity -0.29 (0.12)* JI*Proximity 0.25 (0.08)** Control 0.01 (0.07) JI*Control 0.03 (0.07) Indirect effect at low proximity -0.03 [-0.14, 0.08] 0.04 [-0.08, 0.17] -0.01 [-0.07, 0.03] 0.01 [-0.02, 0.05] Indirect effect at high proximity 0.18* [0.06, 0.31] -0.21* [-0.37, -0.07] 0.07* [0.01, 0.15] -0.04 [-0.12, 0.02] Note. The 95% Monte Carlo bootstrapped confidence intervals (CIs) for conditional indirect effects are in square brackets. * p < .05, ** p < .01. Significant indirect effect based on 95% CIs excluding zero is marked with *. JOB INSECURITY AND JOB PRESERVATION 48 Table 7 Example Open-Response Descriptions of Job Preservation Behaviors as Matched to the Framework of Job Preservation Strategies and Targets. Resource Investment Target Job Preservation Strategy Task-Oriented Resource Investment Social-Oriented Resource Investment Promotive “I started working extra hard to prove my job was important to keep. I worked overtime and at nights to show that I was hardworking and dedicated.” “I have taken on new tasks to help management and the front desk, so I am trying to [...] increase my worth.” “Make an extra effort so that [my] work product is as good as possible, with no errors. Accept extra work if asked.” “Lately, I've been communicating weekly with my boss to keep him informed on what I've been accomplishing and make sure he knows about all the areas where I'm contributing.” “I just try to make sure that my manager likes me and go above and beyond to try to show my worth.” “Promoting oneself modestly is a key part of protecting my job.” Protective “I focus on the things that I can control, like showing up to work on time. Getting my work done. Not creating drama in the workplace. etc.” “Always ensuring I was on time, never took long lunches and tried my best to clock out on time to avoid overtime if possible.” “I’ve tried to put my head down and work. I tried to become a ghost and just not draw attention to myself.” “There are definitely some processes that only I know how to do. [...] If I were to leave these processes would not get done.” “They would have trouble replacing me in the short term because not many others have learned how to perform my role yet.” “I have made sure I have a set of skills that is hard to replace at my work. I am one of the few that knows all our systems. I learned early on pay and job security is not based on how hard your [sic] work but on how hard it is to replace you.” JOB INSECURITY AND JOB PRESERVATION 49 FIGURES Figure 1 Unstandardized Path Coefficients of the Latent Cross-Lagged Panel Model of Study 1 Note. Autoregressive correlations (dash arrows) and cross-lagged coefficients (solid arrows) were constrained to be equal across time to test a more parsimonious model. Measurement models and synchronous correlations for latent variables were not shown for a cleaner view. * p < .05, ** p < .01. JOB INSECURITY AND JOB PRESERVATION 50 Figure 2 Moderating Effect of Perceived Threat Proximity on the Relationship between JI and Subsequent Job Preservation Motivation – Study 2 -1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8 1 Low JI High JI Job P re se rva ti on M ot iva ti on High Threat Proximity Low Threat Proximity View publication stats Download 0.9 Mb. Do'stlaringiz bilan baham: |
Ma'lumotlar bazasi mualliflik huquqi bilan himoyalangan ©fayllar.org 2024
ma'muriyatiga murojaat qiling
ma'muriyatiga murojaat qiling