Page 61 - Azerbaijan State University of Economics

Shaig E.Kazimov: More Than Just Unemployment: How the Fear Of Not Finding A Job Shapes Life
Satisfaction In Azerbaijan

Model 1: Categorical Job Insecurity
In the first model, where job insecurity was treated as a categorical predictor, the total
effect of job insecurity on life satisfaction was found to be statistically significant and
negative (β = –0.1603, SE = 0.0425, 95% CI [–0.2442, –0.0764]), indicating that
individuals experiencing higher levels of stress or anxiety about employment
prospects reported lower levels of life satisfaction. However, when mediating
variables were accounted for, the direct effect turned positive and remained
statistically significant (β = 0.0864, SE = 0.0375, 95% CI [0.0127, 0.1602]).
The analysis of indirect pathways revealed several significant effects. Notably, significant
negative indirect effects were observed through social impact (β = –0.0322, 95% CI [–
0.0625, –0.0114]), future hope (β = –0.0719, 95% CI [–0.1105, –0.0425]), and the full
sequential mediation pathway involving goal growth, social impact, and future hope (β =
–0.0200, 95% CI [–0.0376, –0.0066]). Additional significant indirect effects emerged
from social impact to future hope (β = –0.0206, 95% CI [–0.0407, –0.0051]). Other two-
step chains (e.g., JS → goal growth → future hope, and JS → goal growth → social
impact) did not reach statistical significance. The total indirect effect across all mediators
was significant (β = –0.2467, SE = 0.0276, 95% CI [–0.2997, –0.1949]).
Model 2: Discrete Job Insecurity and Mediation Variables
When job insecurity and mediators were modeled as discrete (continuous) variables, the
results provided further support for the serial mediation hypothesis. The total effect of job
insecurity on life satisfaction remained strong and statistically significant (B = –0.6469,
SE = 0.0930, p < .001, 95% CI [–0.8296, –0.4643]). Upon introducing the mediators, the
direct effect decreased and became statistically non-significant (B = –0.1965, SE =
0.1148, p = .0875), suggesting a full mediation structure. The total indirect effect was also
significant (Effect = –0.4505, BootSE = 0.0773, 95% CI [–0.6073, –0.3092]).

Table 2. The fisrt model

Path Effect (β) SE 95% CI
Total Effect (JS → LS) -0.1603* 0.0425 [-0.2442, -0.0764]
Direct Effect (JS → LS) 0.0864* 0.0375 [ 0.0127, 0.1602]
Indirect Effects via Mediators
JS → GG → LS -0.0193 0.0123 [-0.0488, 0.0010]
JS → SI → LS -0.0322* 0.0133 [-0.0625, -0.0114]
JS → FH → LS -0.0719* 0.0182 [-0.1105, -0.0425]
JS → GG → SI → LS -0.0061 0.0048 [-0.0176, 0.0008]
JS → GG → FH → LS -0.0117 0.0062 [-0.0266, 0.0005]
JS → SI → FH → LS -0.0206* 0.0086 [-0.0407, -0.0051]
JS → GG → SI → FH → LS -0.0200* 0.0084 [-0.0376, -0.0066]
Total Indirect Effect -0.2467* 0.0276 [-0.2997, -0.1949]

61