The test of materials was conducted. The ratings of shame about 3 kinds of neutral mood materials were 1.58 (SD =1.15), 1.60 (SD =1.07), and 1.62 (SD =1.05), respectively. The average ratings of shame about 3 kinds of shameful materials were 6.18 (SD =1.05), 5.85 (SD =1.00), and 4.61 (SD =1.61), respectively. The t-tests of ratings found significant differences (ps < 0.05) between neutral and shame materials.
Analysis of variance (ANOVA) was conducted on the ratings of shame intensity. The results showed a significant difference among the three groups, F (2, 117) = 196.39, p < 0.001, η2 = 0.77. Post-hoc tests found the ratings of the neutral mood group (M = 1.19, SD = 0.31) were significantly smaller than those of the public (M = 3.11, SD = 0.51) and private shame groups (M = 2.97, SD = 0.59), and there was no significant difference between the public and private shame groups, indicating a high validity of the emotion-inducement.
Analysis of everyday helping behavior
A repeated-measures ANOVA on willingness of everyday helping showed a significant main effect of shame types, F (2, 117) = 12.64, p < 0.001, η2 = 0.18, with willingness of everyday helping in the public shame group (M = 4.05, SE = 0.18) being significantly higher than that in the neutral mood (M = 2.84, SE = 0.17) and private shame groups (M = 3.32, SE = 0.17), and higher in private shame group than that in the neutral mood. The main effect of help-seekers’ roles also was significant, F (1, 117) = 63.65, p < 0.001, η2 = 0.35, with willingness to help the benefactor (M = 3.97, SD = 1.20) being significantly higher than willingness to help the stranger (M = 2.80, SD = 1.62). A significant interaction of shame types and help-seekers’ role was found, F (2, 117) = 6.61, p < 0.01, η2=0.10. Simple effects analysis showed no significant difference between shame types when participants were benefactors, F (2, 117) = 2.88, p > 0.05. In contrast, participants in the public shame group were more willing to perform everyday helping than those in the neutral mood and private shame groups, and participants in the private shame group were more willing to perform everyday helping than those in the neutral mood group when help-seekers were in the stranger roles condition (see Table 1, Figure 2a), F (2, 117) = 16.07, p < 0.001, η2 = 0.22.
The decision-time was taken into consideration for speed-accuracy trade-off. A repeated-measures ANOVA on decision-time for everyday helping found no main effect of shame types, F (2, 117) = 0.45, p > 0.05 or help-seekers’ roles, F (1, 117) = 1.26, p > 0.05. Nor was there a significant interaction between the two variables, F (2, 117) = 0.09, p > 0.05.
Table 1. Willingness and decision-time of participants in different conditions
Shame types
|
Help-seekers’
roles
|
Willingness of everyday helping
|
Decision time of everyday helping (ms)
|
Willingness of donation
|
Decision time of donation (ms)
|
M
|
SD
|
M
|
SD
|
M
|
SD
|
M
|
SD
|
Neutral mood
|
Benefactor
|
3.73
|
1.50
|
6007.32
|
4584.78
|
3.71
|
1.35
|
6915.07
|
3757.84
|
Stranger
|
1.95
|
1.53
|
6987.56
|
5652.09
|
1.93
|
1.54
|
8085.88
|
5472.42
|
Public
shame
|
Benefactor
|
4.30
|
0.85
|
6975.81
|
6162.39
|
3.81
|
1.29
|
8338.70
|
5864.31
|
Stranger
|
3.81
|
1.24
|
7586.00
|
7873.87
|
3.16
|
1.07
|
7132.14
|
6130.81
|
Private
shame
|
Benefactor
|
3.91
|
1.10
|
6105.24
|
4877.75
|
3.98
|
0.95
|
6338.67
|
3591.95
|
Stranger
|
2.74
|
1.53
|
6487.41
|
5328.96
|
1.95
|
1.46
|
6622.86
|
4594.41
|
Analysis of donating money
A repeated-measures ANOVA on willingness to donate money showed a significant main effect of shame types, F (2, 117) = 4.49, p < 0.05, η2 = 0.07. Willingness to donate money in the public shame group (M = 3.49, SE = 0.17) was significantly higher compared to the neutral mood (M = 2.82, SE = 0.16) and private shames groups (M = 2.96, SE = 0.16). The main effect of help-seekers’ roles also was significant, F (1, 117) = 106.19, p < 0.001, η2 = 0.48. Willingness to donate money to the benefactor (M = 3.83, SD = 1.20) was significantly higher than willingness to donate to the stranger (M = 2.32, SD = 1.48). The interaction of shame types and help-seekers’ role was significant, F (2, 117) = 8.37, p < 0.001, η2 = 0.13. A simple effects analysis found no significant difference of shame types when participants were benefactors, F (2, 117) = 0.53, p > 0.05; while participants in the public shame group were more willing to donate money than were those in the neutral mood and privaye shame groups, when help-seekers were strangers (see Table 1, Figure 2b), F (2, 117) = 10.00, p < 0.001, η2 = 0.15.
A repeated-measures ANOVA on decision-time for donating money showed no significant main effect of shame types, F (2, 117) = 0.95, p > 0.05. There was also no significant main effect of help-seekers’ roles, F (1,117) = 0.04, p > 0.05, or significant interaction between shame types and help-seekers’ roles, F (2, 117) = 2.49, p > 0.05.