**3.3 Associations between mobbing and potential correlates: results from multivariate regression analysis**

Results of multiple regression model fit with dependent variable "Score of negative issues at work" and independent variables the demographic features. Null hypothesis that model does not fit and its data is rejected (p-value = 0.002 < 0.05). Level of adjustment is moderate as R2 = 0.212 < 0.4.

**Table 4** indicates coefficients of multiple regression model fit with dependent variable "Score of negative issues at work" and independent variables the demographic features. Null hypothesis is that coefficients are zero and the alternative is that they are statistically significant. Statistical significant are considered the coefficients of variables "Age" (p-value = 0.035 < 0.05), "Sector-Specialty" (p-value = 0.001 < 0.05). Linear model is described by the following mathematical formula:


#### **Table 4.**

*Results of the multivariate regression analysis examining the associations between NAQ (dependent variable), sociodemographic and work-related parameters.*

"Score of negative issues at work" ¼ 0*:*001 � 0*:*365 � Age þ 0*:*222 � Sector � Specialty*:* (1)

Results of independent variables affection in dependent variable "Score of negative issues at work" using BETA coefficients were as follows: Variable "Age" affects negatively (BETA = �0.238) dependent variable of model, while "Sector-Specialty" positively (BETA = 0.349). BETA coefficients have values in the interval [�1.1]. Values close to 1 indicate maximum positive affection while values close to �1 maximum negative. Values close to 0 indicate no affection.

The sample of medical residents in this study shows an average score of 56.32 � 17.080 for NAQ-22, which means on average that there were very high levels of mobbing- frequent and extreme aggressive behaviors in the medical residents of UGHP. No statistically significant difference occurred in NAQ-22 relative to sex (p = 0.14), training (p = 0.735), the specialty years (p = 0.478), or years of work as a trainee in UGHP (p = 0.052). On the other hand, statistically significant difference (p = 0.0174) occurs with respect to: (a) age, where at the age group 25–35 the highest value is observed (mean = 58.5) in relation to ages groups 35–35 (mean = 51.0) and 45–55 (mean = 29.50), and (b) Sector, (p = 0.0043) with higher score in the Division of Surgery (mean = 62.55), and lowest in the Laboratory sector (mean = 45.4).

#### **3.4 Multivariate regression analysis for subscales of NAQ**

**Table 5** indicates the results of Cronbach Alpha coefficient for subscales of NAQ. Reliability of "labor related intimidation" was a = 0.864 while "labur related intimidation regarding the worker's personality" 0.902. Reliability of the total scale was 0.932.

Using as dependent variable "Labor related intimidation", statistically significant are considered the coefficients of variables "**Age**" (p-value = 0.003 < 0.05), "**Sector-Specialty**" (p-value = 0.001 < 0.05) and the coefficient of constant (p-value < 0.001). Linear model is described by the following mathematical formula:

Labor related intimidation ¼ 2*:*987 þ 0*:*087 � Gender–0*:*544 � Age þ 0*:*237

� Sector � Specialty–0*:*118 � Level of education Að Þ

þ 0*:*112 � Years of work as a medical resident Að Þ–0*:*098

�Years of workas a medical resident at UGHP Að Þ*:* (2)

*Correlates of Mobbing among Medical Residents in a University General Hospital… DOI: http://dx.doi.org/10.5772/intechopen.104041*


**Table 5.**

*Reliability analysis for subscales of NAQ.*

Using as dependent variable "Labor related intimidation regarding the worker's personality", statistically significant is considered the coefficient of variable "Sector-Specialty" (p-value = 0.006 < 0.05). No appropriate model was formulated.

Using as dependent variable "Victim of mobbing in work", statistically significant is considered only the constant (p-value = 0.034 < 0.05). No appropriate model was formulated.

Using as dependent variable "Score of negative issues at work" statistically significant are considered the coefficients of variables "Age" (p-value = 0.035 < 0.05), "Sector-Specialty" (p-value = 0.001 < 0.05) and the coefficient of constant (p-value = 0.0001 < 0.05). Linear model is described by the following mathematical formula:

Score of negative issues at work ¼ 2*:*122 þ 0*:*092 � Gender–0*:*365

�Age þ 0*:*222 � Sector � Specialty–0*:*044 � Level of education Dð Þ

(3)

þ 0*:*132 � Years of work as a medical resident Dð Þ–0*:*084

�Years of work as a medical resident at UGHP Dð Þ*:*
