Table 1 summarizes the demographic characteristics of participants who took part in this study. A total of 1047 participants from three continents working in schools and universities took part in this study. In terms of gender, 68% were females compared to 32% who were males. On age, 41% were 26-35 years, 32% were 36-45 years, and 12% were at least 46 years. On educational qualification, while 45% had high school qualification, 24% had Bachelor’s degree, and 6% had at least master’s degree (see Table 1 for more details).
Interaction effect of vaccine type attitudes
A two-way between groups analysis of variance was conducted to explore whether available vaccine type in a given context will moderate the relationship between other variables and attitudes (see Table 2 for details). First, there was interaction effect between vaccine type and age on opinion about vaccination only, F (6, 1033) = 3.24, p = .004, small effect size, partial eta squared = .02. Post-hoc comparison using Tukey HSD test found no difference between the participants.
Second, there were interaction effect of available vaccine and gender of perceived hesitancy [F (2, 1039) = 5.83, p = .003, partial eta squared = .01] and attitude towards vaccination [F (2, 103) = 4.84, p = .008, partial eta squared = .009]. The results of the mean score showed that females were less hesitant and more positive towards the vaccination than males.
Third, available vaccine moderated the relationship between educational qualification of participants and, opinion [F (6, 1033) = 2.25, p = .04, partial eta squared = .01], perceived hesitancy [F (6, 1033) = 9.33, p = .001, with a moderate effect size, partial eta squared = .05] and attitudes toward vaccination [F (6, 1033) = 7.87, p = .001, partial eta squared = .04]. Post-hoc comparison using Tukey HSD test showed that the higher the education of participants, the more they would be less hesitant and positive towards receiving the vaccine than those with less qualification.
Table 2. Interaction effect of vaccine available on attitudes
Variable
|
df
|
MS
|
F
|
p
|
η2
|
Vaccine available X Age
OVS
PHS
ACVS
|
6
6
6
|
46.01
573.32
647.59
|
3.24
1.93
1.88
|
.004**
.07
.08
|
.02
.01
.01
|
Vaccine available X Gender
OVS
PHS
ACVS
|
2
2
2
|
34.96
1757.26
1705.06
|
2.45
5.83
4.84
|
.09
.003**
.008**
|
.005
.01
.009
|
Vaccine available X Qualification
OVS
PHS
ACVS
|
6
6
6
|
32.01
2681.37
2642.46
|
2.25
9.33
7.87
|
.04*
.001**
.001**
|
.01
.05
.04
|
Vaccine available X Place of residence
OVS
PHS
ACVS
|
3
3
3
|
10.31
1433.67
1228.62
|
.73
5.22
3.78
|
.53
.001*
.01*
|
.002
.02
.01
|
Vaccine available X Health Status
OVS
PHS
ACVS
|
2
2
2
|
53.97
713.63
479.78
|
3.77
2.37
1.37
|
.02*
.09
.26
|
.007
.005
.003
|
Vaccine available X COVID-19 acquisition status
OVS
PHS
ACVS
|
2
2
2
|
87.56
201.65
283.83
|
6.15
.661
.80
|
.002**
.52
.45
|
.01
.001
.002
|
Vaccine available X Vaccination status
OVS
PHS
ACVS
|
4
4
4
|
23.38
835.97
976.33
|
1.78
3.13
3.01
|
.13
.01**
.02*
|
.007
.01
.01
|
*p<.05; **p<.01; *p<.05; **p<.01; *p<.05; **p<.01; OVS = Opinion towards Vaccination Scale; PHS = Perceived Hesitancy Scale; ACCS = Attitude towards COVID-19 Vaccination Scale
Fourth, there was interaction effect between vaccine type and place of residence on perceived hesitancy [F (3, 1037) = 5.22, p = .001, partial eta squared = .02] and attitudes [F (3, 1037) = 3.78, p = .01, partial eta squared = .01]. Post-hoc comparison using Tukey HSD test showed a significant difference between participants, with those in Africa more hesitant and less positive towards vaccination than their counterparts in Asia and Europe.
There was also interaction effect of vaccine type available and health status on opinion towards vaccination only, F (3, 1037) = 2.56, p = .05, partial eta squared = .007. Similarly, there was interaction effect of vaccine type on COVID-19 acquisition status and opinion towards vaccination only, F (2, 1039) = 6.15, p = .002, partial eta squared = .01. In terms of health status, the mean score showed that those with minor/chronic health issues had position opinion towards the vaccination than the others without health issues. Related to this, those who had been diagnosed with COVID-19 had positive opinion about vaccination than those who indicated otherwise.
Furthermore, there was interaction effect between vaccine type and vaccination status on perceived hesitancy [F (4, 1036) = 3.13, p = .01, partial eta squared = .01] and attitudes towards vaccination, [F (4, 1036) = 3.01, p = .02, partial eta squared = .01]. Post-hoc comparison using Tukey HSD test showed that those who had been vaccinated and intend to vaccinate were less hesitant and more positive on vaccination than those who indicated they did not intend to vaccinate.
Impact of health status on attitudes
Table 3 summarizes the result of a two-way analysis of variance computed to ascertain the relationship between the interaction effect of health status on attitudes toward vaccination. First, there was interaction effect of health status on age and opinion towards vaccination only, F (3, 1038) = 8.26, p = .001, partial eta squared = .02. Post-hoc comparison using Tukey HSD test showed no relationship between the participants.
Second, there was interaction effect of health status and education qualification on perceived hesitancy [F (3, 1038) = 3.43, p = .02, partial eta squared = .01] and overall attitude towards vaccination [F (3, 1038) = 3.04, partial eta squared = .009]. Post-hoc comparison using Tukey HSD test showed that those with lower qualification were more hesitant and appears to have negative attitude towards vaccination than those who had higher qualifications.
Table 3. Interaction effect of health status on attitudes
Variable
|
df
|
MS
|
F
|
p
|
η2
|
Health Status X Age
OVS
PHS
ACVS
|
3
3
3
|
116.51
123.64
343.23
|
8.26
.42
.57
|
.001**
.74
.63
|
.02
.001
.002
|
Health Status X Gender
OVS
PHS
ACVS
|
1
1
1
|
50.44
21.59
6.03
|
3.53
.07
.02
|
.06
.79
.90
|
.003
.001
.001
|
Health Status X Qualification
OVS
PHS
ACVS
|
3
3
3
|
14.55
1014.06
1042.32
|
1.02
3.43
3.04
|
.38
.02*
.03*
|
.003
.01
.009
|
Health Status X Place of residence
OVS
PHS
ACVS
|
1
1
1
|
78.07
920.27
462.27
|
5.44
3.09
1.33
|
.02*
.08
.25
|
.005
.003
.001
|
Health Status X COVID-19 acquisition status
OVS
PHS
ACVS
|
1
1
1
|
4.62
11857.51
12330.28
|
.32
40.67
36.28
|
.57
.001**
.001**
|
.001
.04
.03
|
Health Status X Vaccination status
OVS
PHS
ACVS
|
2
2
2
|
65.61
740.96
699.46
|
5.02
2.78
2.16
|
.007**
.06
.12
|
.01
.005
.004
|
*p<.05; **p<.01; *p<.05; **p<.01; *p<.05; **p<.01; OVS = Opinion towards Vaccination Scale; PHS = Perceived Hesitancy Scale; ACVS = Attitude towards COVID-19 Vaccination Scale
Third, there was interaction effect of health status on place of residence on opinion towards vaccination only, F (3, 1042) = 5.44, p = .02, partial eta squared = .005. Post-hoc comparison using Tukey HSD test showed that those who indicated they were residing in Africa were less positive on opinion than those who indicated otherwise.
Fourth, there was interaction effect of health status on COVID-19 acquisition status and perceived hesitancy [F (1, 1042) = 40.67, p = .001, partial eta squared = .04] and attitude, F (1, 1042) = 36.28, p = .001, partial eta squared = .03. The mean score showed that on perceived hesitancy, those who had been diagnosed with COVID-19 were less hesitant toward the vaccination than those who had not been diagnosed. However, on overall attitude, those who had acquired COVID-19 were less positive toward the vaccination compared to those who indicated otherwise.
Last, there was interaction effect of health status on vaccination status and opinion only, [F (2, 1040) = 5.02, p = .007, partial eta squared = .01]. Post-hoc comparison showed that those difference between those who had vaccinated and those who did not intend to vaccinate, with the former having unfavorable opinion than the latter.
Influence of health status and vaccine type on attitudes
A hierarchical multiple regression was computed to assess the ability of health status and available vaccination type on attitude towards overall vaccination while controlling for other demographics (see Table 4 for details). In step 1, health status and vaccination type were entered in the model. The two variables made a significant contribution of only 2% in the variance in attitude, F (2, 1041) = 7.05, p = .001. however, only health status (beta = .11, p = .001) made significant contribution in the variance in attitude.
Table 4. Summary of the results of hierarchical multiple regression
Category
|
B
|
Beta
|
t
|
p
|
Step 1
Health Status
Vaccination available
|
4.45
.96
|
.11
.04
|
3.43
1.23
|
.001**
.22
|
Step 2
Health Status
Vaccination available
Age
Gender
Place of residence
Qualification
COVID-19 acquisition status
Vaccination status
|
4.07
.25
-.83
-.52
-2.67
-.48
.73
5.98
|
.10
.01
-.04
-.01
-.08
-.02
.01
.22
|
3.26
.32
-1.25
-.39
-2.10
-.70
.48
6.53
|
.001**
.75
.21
.70
.04**
.48
.63
.001**
|
In step 2, the addition of six more demographic variables contributed to 9% in the variance in attitudes. The overall model made contribution of 11% in the variance in attitudes, F (8, 1034) = 13.34, p = .001. In the second model, the largest contribution of the variance was vaccination status (beta = .22, p = .001). Once again, health status (beta = .10, p = .001) made a significant contribution in the variance in attitude. Also, place of residence made significant contribution in the variance in attitudes.