- Results
3.1 Data Screening
Three participants scores were removed from the data set because they had missing data. Maximum and minimum scores were computed for each questionnaire item to check for out of range values that indicated an input error. Following on from this questionnaire items that needed to be reversed were recoded. Combined totals for each scale were computed by adding the relevant questionnaire item scores for each participant.
In order to check for univariate outliers scale totals were converted into z-scores for each participants. Participants with z-scored of more than +/-3.0 on any single scale were considered outliers and removed from the analysis. This process resulted in the exclusion of three participants from further analysis. A Mahalanobis distance analysis was used to check for multivariate outliers but none were identified in this sample.
3.2 Descriptive Statistics
Table 1 presents the means and standard deviations for measures of attachment styles, emotional competencies and marital satisfaction.
Table 1: Descriptive Statistics for attachment styles, emotional competencies & marital satisfaction measures.
| Marital Satisfaction | Mean | Std. Deviation. |
| Emotional Competencies | 16.53 | 3.82 |
| Self Insight | 75.54 | 8.73 |
| Expressivity | 73.57 | 6.82 |
| Sensitivity | 77.03 | 8.68 |
| Drive | 70.53 | 7.12 |
| Foundations | 72.58 | 9.98 |
| Secure attachment style | 20.57 | 3.66 |
| Anxious attachment style | 13.97 | 4.06 |
| Avoidant attachment Style | 14.46 | 4.21 |
3.3 Correlations
Table 2: Correlation Matrix for attachment styles, emotional intelligence and marital satisfaction measures (*p<0.05, **p>0.01).
A significant positive correlation was found between overall emotional competencies and Secure attachment (r(91)=.43; P<0.01). Significant positive correlations were also found for Secure attachment and all the EI subscales; Sensitivity (r(91)=.4;p<0.01), Self insight, (r(91)=.39;p<0.01), Drive, (r(91)=.39;p<0.01), Expressivity, (r(91)=.38;p,0.01) and Foundation, (r(91)=.32;p<0.01).
A significant negative correlation was found between Romantic satisfaction and both anxious attachment (r(91=-.32;p<0.01) and Avoidant attachment(r(91)=-.33;p<0.01).
Finally, a significant negative correlation was found between anxious attachment and self-insight, (r(91)=-.30;p<0.01). A significant negative correlation was also found between avoidant attachment and foundation, (r(91)=-.41;p<0.01), Self-insight, (r(91)=-.39;p<0.01) and drive, (r(91)=-.36;p<0.01), respectively.
3.4 Multiple Regression Analysis
A multiple regression was used to assess the extent to which adult attachment styles predict overall emotional intelligence. Scores for the secure, anxious and avoidant subscales of the Adult Attachment Scale were used as predictors of overall MSEQ scores. This analysis produced a significant predictive model accounting for 18% of the variability in the dataset (R2adj=0.18; F(1,89)=20.45; p<0.05). Examination of Beta coefficients revealed that only the secure attachment scores contributed significantly to the model (β=0.43; t=4.52; p<0.05). Anxious (β=-0.13; t=-1.32; p=0.19) and avoidant (β=-0.21; t=-1.84; p=0.07) scores did not contribute significantly to the model. Thus, this model reflects a predictive relationship whereby as secure attachment scores rise so do MSEQ scores.
A multiple regression was used to assess the extent to which Adult Attachment Styles and Self Insight predicts KMS. The analysis produced a significant predictive model accounting for 10% of the variability in the data set (R2adj=0.10; F(1,89)=11.01; p<0.05). Only the avoidant attachment scale contributed significantly to the model (β=-.30; t=-3.32; p<0.05). Secure (β .03; t=.21; p>0.05), Anxious (β=-..20; t=1.8; p>0.05) insight (β=.10; t=.96; p>0.05) scores did not contribute significantly to the model. Thus, this model reflects a predictive relationship whereby anxious attachment style predicts less marital satisfaction.
