nursing students 4.1 Objectives
4.3.13. Additive effects
Significant differences in the change between ‘increased’ and ‘decreased’ groups on the full dimensional scores was explored. In order to correct for alpha inflation due to conducting multiple tests, Bonferroni correction was used. Based on 26 tests, a p value of < .002 was considered to be statistically significant. The results are described in table 4.12. It can be seen from the table that all variables were highly significant at p<.001. Time point 1 Mean SD Time point 2 Mean SD P value
Work Resources Increase Group 31.70 5.97 35.97 5.62 <.001 Work Resources Decrease Group 36.14 5.71 31.06 6.75 <.001 Job Demands Increased Group 16.71 4.55 20.74 4.27 <.001 Job Demands Decreased Group 21.16 4.09 17.94 4.73 <.001 Personal Stressors Increased Group 11.85 6.41 17.49 7.52 <.001 Personal Stressors Decreased Group 15.39 7.77 11.12 6.41 <.001 Academic Stressors Increased Group 24.04 5.42 28.78 5.23 <.001 Academic Stressors Decreased Group 28.55 5.09 24.91 5.51 <.001 Clinical Stressors Increased Group 16.67 7.16 22.87 7.76 <.001 Clinical Stressors Decreased Group 22.26 7.83 17.73 7.13 <.001 Lack of Support on Placement Increased 13.58 5.32 18.95 5.82 <.001 Lack of Support on Placement Decreased 17.58 5.28 13.41 4.60 <.001 Perceived Job Stress Increased Group 18.91 4.42 22.14 4.05 <.001 Perceived Job Stress Decreased Group 22.17 4.24 19.18 4.45 <.001 Perceived Life Stress Increased Group 8.81 4.03 12.57 3.95 <.001 Perceived Life Stress Decreased Group 12.51 4.30 9.95 4.33 <.001 Mental Health Problems Increased Group 25.45 6.69 29.53 5.91 <.001 Mental Health Problems Decreased Group 29.01 6.13 24.44 6.77 <.001 Physical Health Increased Group 9.19 3.48 11.13 3.52 <.001 Physical Health Decreased Group 11.75 3.15 8.77 3.35 <.001 Emotion-Focused Coping Increased Group 19.23 6.55 23.66 6.34 <.001 Emotion-Focused Coping Decreased 24.16 6.41 19.72 6.36 <.001 Social Support Coping Increased Group 14.22 3.58 16.19 3.00 <.001 Social Support Coping Decreased Group 16.25 2.79 13.33 2.94 <.001 Positive Personality Traits Increased 30.24 7.43 33.79 6.94 <.001 Positive Personality Traits Decreased 35.42 6.77 30.73 7.37 <.001
Table 4.12 Change between increased and decreased groups on the full dimensional
The dimensions that showed a significant difference on the full dimensional scores were then summed together to compute the global NOF score and this score was split into tertiles. As all dimensions were significant, all variables met the criteria for inclusion in the NOF analyses.
The NOF was computed by recoding the increase/decrease scores so that “bad” factors were recoded as 1 and “good” factors were recoded as 0. For example, a participant that increased in job demands between T1 and T2 was coded as 1, whereas a participant that decreased in job demands between T1 and T2 was coded as 0. Decisions about the direction of the recoded variables were made based on relationships with the outcome variables and theoretical understanding. These scores were then added together to produce the global NOF score and this score was split into tertiles. Logistic regression was used to assess the association between the NOF score with stress and mental health outcomes. Table 4.13 shows dose-response relationships between increases in the NOF score and increases in stress and mental health problems. Specifically, participants in the highest tertile of the NOF were over four times more likely to report increases in clinical stress, and over seven times more likely to report increases in academic stress and mental health problems than participants in the lowest tertile.
Outcome: Academic Stress OR 95% CI p
Lowest tertile 1.000
Middle tertile 3.407 2.008-5.780 <.001 Highest tertile 7.624 4.313-13.476 <.001
Outcome: Clinical Stress
Lowest tertile 1.000
Middle tertile 1.774 1.067-2.950 .027 Highest tertile 4.255 2.482-7.294 <.001
Outcome: Mental Health Problems
Lowest tertile 1.000
Middle tertile 3.493 2.008-6.077 <.001 Highest tertile 7.904 4.449-14.043 <.001
Table 4.13. Logistic regression (enter method) for the relationship between NOF and stress and mental health outcomes.
Next, each individual nursing group (mental health, adult and child) were investigated further using the NOF score. It can be seen by figure 4.6 that the mental health branch had the highest NOF mean score (6.33), followed by the adult branch (6.31), and then the child branch (5.93).
Figure 4.6. Means for the global NOF score for each branch of nursing students 5.7 5.8 5.9 6 6.1 6.2 6.3 6.4
Adult Mental health Child
M e an o f N OF Nursing Branch
Summary of key findings Multi-factor level findings
A dose-response relationship was found between the global NOF score and the outcomes academic stress, clinical stress and mental health problems.
When taking into account combinations of variables, the mental health nursing students had the highest mean NOF score, followed closely by the students in the adult nursing branch, and then finally child nursing.
Factor level findings
Direct effects on mental health problems included academic stressors, clinical stressors, work resources, social support coping, personality, perceived life stress, weekend alcohol consumption and physical health.
Interactions were found between work resources and perceived job stress, academic stressors and personal stressors, and clinical stressors and weekend alcohol consumption.
Single-item level findings
Direct effects included academic dissatisfaction, friendship problems, time pressures, financial problems, heavy clinical workload and discrimination.
Interaction effects were found between being inadequately prepared and death and dying on placements. A further interaction between heavy clinical workload and academic dissatisfaction was also found.
4.4. Discussion
Occupational stress researchers adopt a variety of methods when carrying out their research, but the problem this creates is that studies can often find contradictory evidence due to these methodological differences. The method described here could, if developed further, provide a more systematic approach to stress measurement that can be flexible enough to be applied to any occupational group. Indeed, approaches to managing stress must be flexible enough to represent individual circumstances (Bliese & Jex, 1999), as well as being able to consider the multiple dimensions of the stress process (Smith, 2015).
Measuring multiple constructs using suitably designed single items can be helpful when including potential confounders in the model that would otherwise not be controlled for. Indeed, the effects reported here are present even when a multi- dimensional approach is adopted and a variety of other constructs are taken into account.