Descriptive statistics
In today’s study, analyses regarding your very first purpose of the study have been carried out in 2 stages. 1st, the correlations ranging from narcissism and you can motives with the unfaithfulness was basically examined (Figs step 1 and you can 2), and https://datingranking.net/es/citas-pansexual/ you will afterwards, the fresh new moderating effect data try carried out. Prior to the design evaluating, preliminary analyses was carried out to disclose new no-buy correlations, function, and standard deviations one of several 10 seen variables, which are claimed inside the Dining table step 1. The latest skewness beliefs towards the observed variables ranged out of .eleven to just one.08, plus the kurtosis parameters had been ranging from .08 and you may step one.38. Such philosophy show that the fresh new analytical appropriateness of your parameters is ensured regarding normal shipment presumptions.
Take to of one’s dimensions model
In the aspect model, new observed variables was appointed into latent variables about architectural design. Since same latent variables (narcissism, objectives on cheating, and you can dating satisfaction) was in fact utilized in each other models, one dimension model sample are performed. Like with the latest noticed details, the full many brand new NPI subscales offered because the hidden varying for narcissism. Toward intentions with the unfaithfulness and relationships pleasure hidden variables, parcels was indeed tasked proportionally on quantity of facts due to the fact scales one assess the variables had been couched inside a-one-dimensional framework. That it parcelling means is actually did by assigning bills to the parcels according to the computed number of parcels, according to the items-complete relationship values.
According to the tests of the measurement model utilized as the model for the study, the goodness of fit values were optimal: ? 2 (32, N = 184) = , p = .001; GFI = .94; AGFI = .89; CFI = .92; IFI = .92; RMSEA = .073 (90% confidence interval for RMSEA = .046–.099). It was determined that the entire factor loads related to the observed variables of all latent variables were high and statistically significant (standardized values ranged from .44 to .91, p < .05; see Table 2).
Sample of architectural models
In this phase of the analysis, two structural models were tested. First, the partial mediation model given in Fig 1 was tested, and the goodness of fit values were found to be close to perfect: ? 2 (32, N = 184) = , p = .001; GFI = .94; AGFI = .89; CFI = .92; IFI = .92; RMSEA = .073 (90% confidence interval for RMSEA = .046–.099). 13, p>.05). However, when the effect of intentions towards infidelity was fixed to zero, the standardized means coefficient between narcissism and relationship satisfaction was found to be statistically significant (? = -.31, p < .01). The path between narcissism and relationship satisfaction was subsequently excluded from the model in order to test the fully mediated model, and the goodness of fit values were calculated as ? 2 (33, N = 184) = , p = .001; GFI = .93; AGFI = .89; CFI = .92; IFI = .92; RMSEA = .072 (90% confidence interval for RMSEA = .046–.098). The chi-square difference test results (1.66, 1: p>.05) indicated that the exclusion of this path from the model did not cause a significant deterioration in the model. The final model after excluding the path is shown in Fig 4. Ultimately, as the model demonstrates, intentions towards infidelity has a full mediating effect on the relationship between narcissism and relationship satisfaction.
Second, the partial mediation model given in Fig 2 was tested and the goodness of fit values were found to be close to perfect: ? 2 (32, N = 184) = , p = .001; GFI = .94; AGFI = .89; CFI = .92; IFI = .92; RMSEA = .073 (90% confidence interval for RMSEA = .046–.099). Additionally, the standardized means coefficient between narcissism and intentions towards infidelity was not found to be statistically significant (? = .14, p>.05). However, when the effect of relationship satisfaction on intentions towards infidelity was nullified, the standardized means coefficient between narcissism and intentions towards infidelity was found to be statistically significant (? = .35, p < .01). Later, the path between narcissism and intentions towards infidelity was subsequently excluded from the model in order to test the fully mediated model, and the goodness of fit values were calculated as ? 2 (33, N = 184) = , p = .001; GFI = .94; AGFI = .89; CFI = .92; IFI = .92; RMSEA = .073 (90% confidence interval for RMSEA = .046–.098) (chi-square difference test: 1.95, 1: p>.05). This result indicates that the exclusion of this path from the model did not cause a significant deterioration in the model. The final model, shown in Fig 5, indicates that relationship satisfaction has a fully mediating effect between narcissism and intentions towards infidelity.
