Respondents
There were 126 participants in Study 1 (88 medical and 38 nursing students), and 186 across both sites in Study 2 (106 medical and 80 nursing students). Demographics are summarised in table 1.
Table 1. Sample demographics
|
Study 1 (n=126)
|
Study 2 (n=186)
|
Newcastle
|
Oxford
|
|
Medicine
(n=88)
|
Nursing
(n=38)
|
Medicine
(n=50)
|
Nursing
(n=23)
|
Medicine
(n=56)
|
Nursing
(n=57)
|
Sex
|
37 female,
49 male,
2 unknown
|
36 female,
2 male
|
25 female,
25 male
|
23 female,
0 male
|
31 female,
25 male
|
50 female,
6 male,
1 unknown
|
Mean age (years)
|
24.7
|
23.7
|
24.0
|
26.5
|
24.4
|
24.6
|
In Study 1, few participants had experience of interprofessional simulation, with just 18% of nursing students (n=7) and 35% of medical students (n=31) reporting any such experience. In Study 2 a majority of medical students at both sites (n=29, 58% at Newcastle, n=29, 52% at Oxford), but still a minority of nursing students (n=8, 35% at Newcastle, n=8, 14% at Oxford), had some experience.
Internal consistency of measures
The internal consistency of measures was generally adequate (see table 2). For the pre-measures of RIPLS-core in Study 1, and Centrality of student group in Study 2, it was marginally below the conventional threshold of alpha>=0.7, but acceptable for the short 4 and 5 item scales used.
Table 2. Cronbach’s Alpha for subscales
|
Study 1
|
Study 2
|
|
Pre
|
Post
|
Pre
|
Post
|
RIPLS-core
|
0.62
|
0.81
|
0.72
|
0.80
|
Identity Strength
|
0.88
|
0.91
|
|
|
Identity Importance
|
0.72
|
0.76
|
|
|
Ingroup Ties: Professional
|
|
|
0.79
|
0.80
|
Ingroup Ties: Student
|
|
|
0.85
|
0.79
|
Ingroup Ties: Team
|
|
|
0.76
|
0.80
|
Centrality: Professional
|
|
|
0.73
|
0.76
|
Centrality: Student
|
|
|
0.68
|
0.74
|
Centrality: Team
|
|
|
0.74
|
0.85
|
Ingroup Affect: Professional
|
|
|
0.86
|
0.87
|
Ingroup Affect: Student
|
|
|
0.85
|
0.86
|
Ingroup Affect: Team
|
|
|
0.80
|
0.84
|
Mixed effects regression: Effects of simulation session and participant group
Our central hypotheses concerned the effects of the simulation session (the Pre-Post factor) and being a medical or nursing student (Participant Group). We expected an increase in all measures, and nursing students to score more highly than medical students. In both studies we observed both of these effects, but significant interactions showed that they were not consistent.
Effects on RIPLS
RIPLS increased for all groups across both studies. In Study 1, the difference was smaller for medical than nursing students, indicated by a significant Pre-Post x Participant Group interaction, but significant for both (estimated marginal means [em-means] are given in table 3). In Study 2, the increase over a session was significant, with no interactions, indicating it was similar for medical and nursing students, and Newcastle and Oxford (overall em-mean pre=4.38, se=0.03; em-mean post=4.60, se=0.04).
Table 3. Estimated marginal means for Pre-Post x Participant Group interaction: RIPLS (Study 1)
|
Estimated marginal means (se)
|
|
Pre
|
Post
|
t-ratio
|
Medicine
|
4.20 (0.05)
|
4.33 (0.05)
|
2.937*
|
Nursing
|
4.52 (0.07)
|
4.88 (0.07)
|
5.501****
|
t-ratio
|
3.708**
|
6.377****
|
|
*p<0.05, **p<0.01, ***p<0.001, ****p<0.0001
p-values adjusted by Tukey method for 4 tests
At Newcastle, medical students had lower RIPLS scores than nursing students, as expected, but this was not true at Oxford (see table 4 and figure 1). Newcastle medical students scored lower than their Oxford peers, while the two nursing groups did not differ.
Table 4. Estimate marginal means for Participant Group x Site interaction: RIPLS (Study 2)
|
Estimated marginal mean (se)
|
|
|
Medicine
|
Nursing
|
t-ratio
|
Newcastle
|
4.19 (0.06)
|
4.73 (0.08)
|
-5.344****
|
Oxford
|
4.43 (0.05)
|
4.60 (0.05)
|
ns
|
t-ratio
|
3.170**
|
ns
|
|
*p<0.05, **p<0.01, ***p<0.001, ****p<0.0001
p-values adjusted by Tukey method for 4 tests
Overall, the hypothesised effect of the simulation session on RIPLS scores was observed. The hypothesised difference between medical and nursing students was partly supported, but the pattern at Oxford suggests this is not just a difference between professions.
Effects on identity scales
Hypothesised effects were observed on the different identity measures in the two studies. Results suggest that measures of identity are affected by participation in simulation, but that effects vary with specific aspects of identity, and sites.
The effect of participant group was consistent across all measures, with nursing students scoring higher than medical students, even within interactions. The Importance scale in Study 1 showed this with no interactions (nursing student overall em-mean=5.50, se=0.15; medical student overall em-mean=4.57, se=0.10).
Pre-post differences were more variable. In Study 1, there was no change for nursing or medical students on the Importance scale, and only nursing students’ scores on the Strength scale increased (see table 5).
In Study 2, some significant changes were found for all subscales. Ingroup Affect – the positive feelings associated with group membership – improved consistently, with no interaction with Participant Group, Site or Target (Pre overall em-mean=4.30, se=0.04; Post overall em-mean=4.38, se=0.04).
By contrast, Ingroup Ties and Centrality for Professional and Student Group did not change. These measures did change for Team however – Centrality across both sites (table 6), Ingroup Ties just at Oxford (Table 7, Figure 2). The Centrality result is therefore comparable with that for Importance in Study 1, where only Professional Group was addressed.
Table 5. Estimated marginal means for Pre-Post x Participant Group interaction: Identity Strength (Study 1)
|
Estimated marginal means (se)
|
|
|
Pre
|
Post
|
t-ratio
|
Medicine
|
5.50 (0.08)
|
5.60 (0.08)
|
ns
|
Nursing
|
6.11 (0.12)
|
6.39 (0.12)
|
4.138***
|
t-ratio
|
4.446****
|
5.693****
|
|
*p<0.05, **p<0.01, ***p<0.001, ****p<0.0001
p-values adjusted by Tukey method for 4 tests
Table 6. Estimated marginal means for Pre-Post x Target interaction: Centrality (Study 2)
|
Estimated marginal means (se)
|
t-ratio
|
|
Pre
|
Post
|
|
Professional
|
4.01 (0.05)
|
4.01 (0.05)
|
ns
|
Student
|
3.87 (0.06)
|
3.75 (0.06)
|
ns
|
Team
|
3.36 (0.05)
|
3.55 (0.06)
|
-3.373*
|
t-ratio
|
Prof-Student: ns
Prof-Team: 11.645****
Student-Team: 8.902****
|
Prof-Student: 4.419***
Prof-Team: 8.117****
Student-team: 3.479*
|
|
*p<0.05, **p<0.01, ***p<0.001, ****p<0.0001
p-values adjusted by Tukey method for 9 tests
Table 7. Estimated marginal means for Pre-Post x Target x Site interaction: Ingroup Ties (Study 2)
|
Estimated marginal means (se)
|
t-ratio
|
|
Pre
|
Post
|
Newcastle, Professional
|
3.84 (0.08)
|
3.95 (0.08)
|
ns
|
Newcastle, Student
|
4.11 (0.08)
|
3.96 (0.08)
|
ns
|
Newcastle, Team
|
3.63 (0.08)
|
3.86 (0.08)
|
ns
|
t-ratio
|
Prof-Student: ns
Prof-Team: ns
Student-team: 5.663****
|
Prof-Student: ns
Prof-Team: ns
Student-team: ns
|
|
Oxford, Professional
|
4.12 (0.06)
|
4.18 (0.06)
|
ns
|
Oxford, Student
|
4.29 (0.06)
|
4.27 (0.06)
|
ns
|
Oxford, Team
|
3.56 (0.06)
|
4.13 (0.06)
|
8.671****
|
t-ratio
|
Prof-Student: ns
Prof-Team: 8.683****
Student-Team: 11.276****
|
Prof-Student: ns
Prof-Team: ns
Student-team: ns
|
|
*p<0.05, **p<0.01, ***p<0.001, ****p<0.0001
p-values adjusted by Tukey method for 18 tests
All three scales showed an interaction between Participant Group, Target and Site (ie collapsing Pre and Post measures). This can be hard to interpret, but it appears that Newcastle medical students exhibit a different pattern of results in some areas (Table 8, Figure 3). In particular, Ingroup Affect for Team is relatively higher at Newcastle than for Oxford medical students or the nursing students, suggesting Newcastle medical students feel more positively about team membership.
Table 8. Estimated marginal means for Participant Group x Target x Site interaction (Study 2) Ingroup Affect
|
Estimated marginal means (se)
|
t-ratio
|
|
Medicine
|
Nursing
|
|
Newcastle, Professional
|
3.80 (0.08)
|
4.69 (0.11)
|
6.564****
|
Newcastle, Student
|
3.70 (0.08)
|
4.61 (0.12)
|
6.572****
|
Newcastle, Team
|
4.08 (0.08)
|
4.52 (0.11)
|
ns
|
t-ratio
|
Prof-Student: ns
Prof-Team: 4.441**
Student-Team: 5.995****
|
Prof-Student: ns
Prof-Team: ns
Student-team: ns
|
|
Oxford, Professional
|
4.28 (0.07)
|
4.70 (0.07)
|
4.125**
|
Oxford, Student
|
4.28 (0.07)
|
4.63 (0.07)
|
ns
|
Oxford, Team
|
4.24 (0.07)
|
4.55 (0.07)
|
ns
|
t-ratio
|
Prof-Student: ns
Prof-Team: ns
Student-team: ns
|
Prof-Student: ns
Prof-Team: ns
Student-team: ns
|
|
Centrality
|
Estimated marginal means (se)
|
t-ratio
|
|
Medicine
|
Nursing
|
|
Newcastle, Professional
|
3.54 (0.08)
|
4.28 (0.12)
|
4.977***
|
Newcastle, Student
|
3.47 (0.08)
|
3.97 (0.13)
|
ns
|
Newcastle, Team
|
3.04 (0.08)
|
3.66 (0.13)
|
4.056**
|
t-ratio
|
Prof-Student: ns
Prof-Team: 6.783****
Student-Team: 5.827****
|
Prof-Student: ns
Prof-Student: 5.666****
Student-Team: ns
|
|
Oxford, Professional
|
3.96 (0.08)
|
4.26 (0.08)
|
ns
|
Oxford, Student
|
3.80 (0.08)
|
4.01 (0.08)
|
ns
|
Oxford, Team
|
3.23 (0.08)
|
3.89 (0.08)
|
5.982****
|
t-ratio
|
Prof-Student: ns
Prof-Team: 10.554****
Student-Team: 8.242****
|
Prof-Student: ns
Prof-Team: 5.236****
Student-Team: ns
|
|
Ingroup Ties
|
Estimated marginal means (se)
|
t-ratio
|
|
Medical student
|
Nursing student
|
|
Newcastle, Professional
|
3.55 (0.08)
|
4.23 (0.12)
|
4.633***
|
Newcastle, Student
|
3.80 (0.08)
|
4.26 (0.12)
|
ns
|
Newcastle, Team
|
3.65 (0.08)
|
3.84 (0.12)
|
ns
|
t-ratio
|
Prof-Student: 3.597*
Prof-Team: ns
Student-Team: ns
|
Prof-Student: ns
Prof-Team: 3.754*
Student-Team: 3.994**
|
|
Oxford, Professional
|
4.02 (0.08)
|
4.28 (0.08)
|
ns
|
Oxford, Student
|
4.31 (0.08)
|
4.26 (0.08)
|
ns
|
Oxford, Team
|
3.79 (0.08)
|
3.90 (0.08)
|
ns
|
t-ratio
|
Prof-Student: 4.391**
Prof-Team: ns
Student-Team: 7.889****
|
Prof-Student: ns
Prof-Team: 5.752****
Student-Team: 5.286****
|
|
*p<0.05, **p<0.01, ***p<0.001, ****p<0.0001
p-values adjusted by Tukey method for 18 tests