3.1 Participants’ Sociodemographic Characteristics
The 205 participants were recruited through telephonic inquiries and resident referrals, and 68.33% of eligible residents agreed to take part in the study. Overall, 190 participants who matched the inclusion criteria completed the study (Fig. 1). Table 2 displays the characteristics of the enrolled participants. The average age of the study participants was 25.16 years. The most common participant categories were female (59.47%), single (60.94%), and first-year residents (41.58%). Most of the participants were from internal medicine (41.58%), followed by surgical (33.1%), neurology (13.68%), radiology (10.53%), gynaecology (5.79%), and paediatrics (2.11%). A large number of participants had 12–24 months’ work experience (48.95%). The major work-related concerns recorded by participants were workload (76.32%) and concern about workplace violence (89.47%). Nearly half (47.89%) of the participants had received workplace violence training, but only 63 (33.16%) participants had reported violence in the workplace. Witnesses of physical violence were found for 54 (28.42%) of participants’ cases. The participants who had been exposed to verbal violence, physical violence, and sexual harassment were 53.68%, 3.68%, and 22.63% respectively.
Table 2
Socio demographic characteristics of medical residents
Characteristics
|
Overall
n = 190
|
Group A
n = 64
|
Group B
n = 66
|
Group C
n = 60
|
F or χ2
|
p
|
|
mean ± sd
n, (%)
|
mean ± sd
n,(%)
|
mean ± sd
n,(%)
|
mean ± sd
n,(%)
|
|
|
Age
|
25.16 ± 1.34
|
25.31 ± 1.38
|
24.92 ± 1.29
|
25.28 ± 1.35
|
1.682
|
.189
|
Sex
|
|
|
|
|
|
|
Male
|
77(40.53)
|
25(39.06)
|
26(39.39)
|
26(43.33)
|
0.288
|
.865
|
Female
|
113(59.47)
|
39(60.94)
|
40(60.61)
|
34(56.67)
|
|
|
Marital status
|
|
|
|
|
|
|
Single
|
175(92.11)
|
59(92.19)
|
60(90.91)
|
56(93.33)
|
3.489
|
.811
|
Married
|
5(2.63)
|
3(4.69)
|
1(1.52)
|
1(1.67)
|
|
|
Unmarried cohabitation
|
7(3.68)
|
1(1.56)
|
4(6.05)
|
2(3.33)
|
|
|
Widowed
|
3(1.58)
|
1(1.56)
|
1(1.52)
|
1(1.67)
|
|
|
Resident year
|
|
|
|
|
|
|
PGY1
|
79(41.58)
|
24(37.5)
|
30(45.46)
|
25(41.67)
|
1.538
|
.819
|
PGY2
|
68(35.79)
|
24(37.5)
|
24(36.36)
|
20(33.33)
|
|
|
PGY3
|
43(22.63)
|
16(25)
|
12(18.18)
|
15(25)
|
|
|
Department
|
|
|
|
|
|
|
Pediatrics
|
4(2.11)
|
2(3.13)
|
2(3.03)
|
0(0)
|
3.990
|
.961
|
Obstetrics-Gynecology
|
11(5.79)
|
5(7.81)
|
2(3.03)
|
4(6.67)
|
|
|
Internal medicine
|
79(41.58)
|
25(39.06)
|
30(45.45)
|
24(40)
|
|
|
Neurology
|
26(13.68)
|
10(15.63)
|
7(10.61)
|
9(15)
|
|
|
Surgical
|
50(26.32)
|
16(25)
|
19(28.79)
|
15(25)
|
|
|
Radiology
|
20(10.53)
|
6(9.38)
|
6(9.09)
|
8(13.33)
|
|
|
Work experience
|
|
|
|
|
|
|
༜6 months
|
14(7.37)
|
5(7.81)
|
4(6.06)
|
5(8.33)
|
0.434
|
.517
|
6–12 months
|
65(34.21)
|
19(29.69)
|
26(39.39)
|
20(33.34)
|
|
|
12–24 months
|
93(48.95)
|
31(48.44)
|
33(50)
|
29(48.33)
|
|
|
༞24 months
|
18(9.47)
|
9(14.06)
|
3(4.55)
|
6(10)
|
|
|
Working hours per day
|
|
|
|
|
|
|
༜8 hours
|
45(23.68)
|
16(25)
|
16(24.24)
|
13(21.67)
|
0.428
|
0.981
|
8–12 hours
|
128(67.37)
|
42(65.63)
|
45(68.18)
|
41(68.33)
|
|
|
༞12 hours
|
17(8.95)
|
6(9.37)
|
5(7.58)
|
6(10)
|
|
|
Workplace violence concern
|
|
|
|
|
|
|
Absolutely not worried
|
20(10.53)
|
8(12.5)
|
8(12.12)
|
4(6.67)
|
8.544
|
.382
|
A little worried
|
75(39.47)
|
31(48.43)
|
18(27.27)
|
26(43.33)
|
|
|
Moderately worried
|
48(25.26)
|
13(20.31)
|
20(30.3)
|
15(25)
|
|
|
Worried
|
26(13.68)
|
6(9.38)
|
11(16.67)
|
9(15)
|
|
|
Very worried
|
21(11.05)
|
6(9.38)
|
9(13.64)
|
6(10)
|
|
|
Workplace violence report
|
|
|
|
|
|
|
Yes
|
63(33.16)
|
25(39.06)
|
21(31.82)
|
17(28.33)
|
1.691
|
.429
|
No
|
127(66.84)
|
39(60.94)
|
45(68.18)
|
43(71.67)
|
|
|
Workplace violence training
|
|
|
|
|
|
|
Yes
|
91(47.89)
|
36(56.25)
|
34(51.52)
|
21(35)
|
6.134
|
.047
|
No
|
99(52.11)
|
28(43.75)
|
32(48.48)
|
39(65)
|
|
|
Physical violence witness
|
|
|
|
|
|
|
Yes
|
54(28.42)
|
18(28.13)
|
19(28.79)
|
17(28.33)
|
0.007
|
.996
|
No
|
136(71.58)
|
46(71.87)
|
47(71.21)
|
43(71.67)
|
|
|
Verbal violence exposure
|
|
|
|
|
|
|
Yes
|
104(53.68)
|
31(48.44)
|
38(57.58)
|
33(55)
|
1.152
|
.562
|
No
|
86(46.32)
|
33(51.56)
|
28(42.42)
|
27(45)
|
|
|
Physical violence exposure
|
|
|
|
|
|
|
Yes
|
7(3.68)
|
2(3.13)
|
2(3.03)
|
3(5)
|
0.549
|
.797
|
No
|
183(96.32)
|
62(96.87)
|
64(96.97)
|
57(95)
|
|
|
Sexual harassment exposure
|
|
|
|
|
|
|
Yes
|
43(22.63)
|
11(17.19)
|
18(27.27)
|
14(23.33)
|
1.912
|
.383
|
No
|
147(77.37)
|
53(82.81)
|
48(72.73)
|
46(76.67)
|
|
|
Preliminary results indicate that there were no significant differences among residents’ sociodemographic characteristics between groups A, B, and C. Although workplace violence training reflected significant differences among the three resident groups (p = .047), there were no significant differences in pairwise comparisons (p > .0167). The results show that there was homogeneity among the groups, which meets the basic condition of intervention.
To test for perception, attitude, and self-efficacy pertaining to workplace violence, all three groups received pre- and post-test questionnaires. The ANOVA test was used to analyse the comparison between the groups, whereas Bonferroni’s and Dunnett’s tests were used to analyse the pairwise comparisons. To analyse the effect of simulation training and behavioural economics teaching after the intervention, we measured perception, attitude, and self-efficacy related to WPV by using POAS, MAVAS, and GSES. These results indicated that there was no significant difference in the pre-test scores among the three groups for the perception, attitude, and self-efficacy of WPV, as p > .05 (see Table 3).
Table 3
Comparison of pre- and post-test of POAS, MAVAS and GSES
Item
|
Pre-Test
|
F
|
p
|
Post-Test
|
F
|
p
|
POAS
|
|
|
|
|
|
|
Group A
|
37.59 ± 5.31
|
0.221
|
.801
|
44.49 ± 6.52*¶
|
6.731
|
.002
|
Group B
|
37.08 ± 7.05
|
|
|
41.75 ± 7.18#
|
|
|
Group C
|
37.75 ± 5.38
|
|
|
38.18 ± 5.43
|
|
|
MAVAS
|
|
|
|
|
|
|
Group A
|
88.47 ± 12.64
|
0.087
|
.916
|
96.49 ± 12.91*
|
5.618
|
.004
|
Group B
|
87.71 ± 14.35
|
|
|
98.48 ± 11.71#
|
|
|
Group C
|
88.67 ± 13.74
|
|
|
91.8 ± 8.85
|
|
|
GSES
|
|
|
|
|
|
|
Group A
|
20.63 ± 6.34
|
0.062
|
.939
|
24.79 ± 7.98*
|
3.705
|
.026
|
Group B
|
20.76 ± 6.48
|
|
|
24.33 ± 7.20#
|
|
|
Group C
|
21.02 ± 5.92
|
|
|
21.13 ± 5.54
|
|
|
Note: p values are based on ANOVA or cross-tabs with Fisher’s exact tests for comparisons across the three resident groups and do not reflect any pairwise comparisons. |
* represents group A vs. group C p < .05, # represents group B vs. group C p < .05, ¶represents group A vs. group B p < .05. |
In contrast, when comparing group C with group A (44.49 ± 6.52 vs. 38.18 ± 5.43, p < .05) and group B (41.75 ± 7.18 vs. 38.18 ± 5.43, p < .05), significant increases in the post-test scores for perception were found. Moreover, group A’s score was higher than group B’s (44.49 ± 6.52 vs. .41.75 ± 7.18, p < .05). After the intervention, when compared with group C, it was revealed that group A (96.49 ± 12.91 vs. 91.8 ± 8.85, p < .05) and group B (98.48 ± 11.71 vs. 91.8 ± 8.85, p < .05) had significantly increased in the post-test scores for attitude. Although the average score of group B was higher than that of group A (98.48 ± 11.71 vs. 96.49 ± 12.91, p > .05), the difference was not significant. Similar results were found in the post-tests for self-efficacy, with significant differences in group A (24.79 ± 7.98 vs. 21.13 ± 5.54, p < .05) and group B (24.33 ± 7.20 vs. 21.13 ± 5.54, p < .05) in comparison with group C; however, no significant difference was found between groups A and B (p > .05).
The results show that both the SBME + BE and SBME interventions on WPV could improve residents' perception, attitude, and self-efficacy towards violence in the workplace. One notable exception was that the average score of group A was higher than group B’s in the post-test scores for perception, which can probably be attributed to the BE effect.
Univariate and multivariate generalised estimating equation (GEE) was used to analyse the variables affecting the cognitive, attitude, and self-efficacy differences of workplace violence among the three groups of subjects before and after they participated in the intervention. Multivariate generalised estimating equation analyses of factors significant in the univariate analysis was performed, as depicted in Table 4.
Table 4
Generalised estimating equations (GEE) analysis
ITEM
|
POAS
|
|
|
MAVAS
|
|
|
GSES
|
|
|
|
Wald
|
B
|
p
|
Wald
|
B
|
p
|
Wald
|
B
|
p
|
TIME (after vs. before)
|
46.147
|
0.433
|
< .001
|
33.039
|
3.133
|
< .001
|
19.057
|
0.117
|
< .001
|
Group
|
17.709
|
-
|
<. 001
|
6.042
|
-
|
.049
|
3.19
|
-
|
.203
|
Group A vs. Group C
|
39.144
|
6.537
|
<. 001
|
7.661
|
5.608
|
.005
|
7.124
|
3.272
|
.004
|
Group B vs. Group C
|
12.619
|
3.841
|
<. 001
|
16.523
|
7.365
|
< .001
|
16.066
|
3.139
|
.006
|
TIME * group
|
27.928
|
-
|
<. 001
|
6.919
|
-
|
.031
|
11.547
|
-
|
.003
|
sex (female vs. male)
|
-
|
-
|
-
|
8.38
|
3.769
|
.004
|
3.932
|
-1.478
|
.047
|
Marital status
|
9.129
|
-
|
.028
|
55.955
|
-
|
< .001
|
13.657
|
-
|
.003
|
Single vs. widowed
|
7.208
|
3.65
|
.007
|
53.584
|
-12.135
|
< .001
|
13.515
|
-3.875
|
< .001
1
11
|
Unmarried cohabitation vs. widowed
Widowed Single
|
3.047
|
3.114
|
.081
|
12.689
|
-13.805
|
< .001
|
0.808
|
-2.433
|
.369
|
Married vs. widowed
|
1.410
|
2.441
|
.235
|
6.472
|
-11.773
|
.011
|
7.328
|
-5.384
|
.007
|
Resident
|
-
|
-
|
-
|
-
|
-
|
-
|
5.187
|
-
|
.075
|
PGY3 vs. PGY1
|
-
|
-
|
-
|
-
|
-
|
-
|
5.163
|
2.164
|
.023
|
PGY2 vs. PGY1
|
-
|
-
|
-
|
-
|
-
|
-
|
0.104
|
0.262
|
.746
|
Work experience
|
8.914
|
-
|
.03
|
-
|
-
|
-
|
-
|
-
|
-
|
༞24 months vs. ༜6 months
|
4.059
|
1.801
|
.015
|
-
|
-
|
-
|
-
|
-
|
-
|
12–24 months vs. ༜6 months
|
1.194
|
-1.1
|
.275
|
-
|
-
|
-
|
-
|
-
|
-
|
6–12 months vs. ༜6 months
|
0.403
|
0.625
|
.526
|
-
|
-
|
-
|
-
|
-
|
-
|
Workplace violence report (yes vs. no)
|
3.581
|
1.321
|
.058
|
-
|
-
|
-
|
-
|
-
|
-
|
Workplace violence training (yes vs. no)
|
4.802
|
1.461
|
.028
|
2.111
|
1.842
|
.146
|
3.383
|
1.333
|
.066
|
Workplace violence exposure (yes vs. no)
|
5.656
|
1.717
|
.017
|
0.407
|
0.829
|
.524
|
-
|
-
|
-
|
Physical violence witness (yes vs. no)
|
-
|
-
|
-
|
1.632
|
1.694
|
.201
|
-
|
-
|
-
|
Workplace violence concern
|
5.677
|
-
|
.225
|
-
|
-
|
-
|
-
|
-
|
-
|
In the GEE analyses of WPV perceptions, significant differences were found in group, time, group*time interaction, marital status, work experience, WPV training, and WPV exposure (p < .05). The results showed that the changes of workplace violence perceptions in the SBME + BE intervention (the average score increased by 6.9, p < .01) and SBME intervention (the average score increased by 4.67, p < .01) were better than those in group C, and the SBME + BE intervention recorded better results than the SBME intervention only (the average score increased by 2.23, p < .05). The main differences stem from marital status, work experience, workplace violence training, and WPV exposure. The scores of the participants who were single (B = 3.65, p < .05) and those who had more than 24 months of work experience had the highest scores (B = 1.801, p < .05). The scores of the participants who had received WPV training were higher than those without such training (B = 1.461, p < .05) and those who had experienced WPV higher than those who had not (B = 1.717, p < .05).
In the GEE analyses of attitude, significant differences were found in group, time, group*time interaction, sex, and marital status (p < .05). The results indicate that the changes in attitude in the SBME + BE intervention (the average score increased by 8.02, p < .01) and SBME intervention (the average score increased by 10.77, p < .01) were better than those in the control group; however, there were no significant differences between group A and group B. The main differences arose from sex and marital status; the scores of the participants who were widowed had the highest scores (B = -12.135, p < .01), and females had better scores than males (B = 3.769, p < .01).
In the GEE analyses of self-efficacy, significant differences were found in time, group*time interaction, sex, marital status, and year of residence. The results indicate that the changes of self-efficacy in group A (the average score increased by 4.16, p < .01) and group B (the average score increased by 3.57, p < .01), were better than those in the group C. The main differences were related to sex, marital status, and year of residence. The scores of the participants who were widowed were the highest (B = -3.875, p < .01), males had better scores than females (B = -1.478, p < .05), and PGY3 residents had better scores than PGY1 residents (B = 2.164, p < .05).
After the SBME and SBME + BE interventions, the perception, attitude, and self-efficacy pertaining to workplace violence improved significantly. Furthermore, the SBME + BE intervention recorded better scores than the other two groups in the perception of workplace violence.