Immunization Coalition Survey Analysis Plan
Immunization Knowledge and Beliefs Chi-Square Example
- Descriptive statistics for the questions How much do you know about vaccines? and Where do you get information?
- Descriptive statistics and inferential analysis for the questions Vaccines are generally safe and Vaccines help prevent diseases
Vaccines are generally safe
Education Comparison
##
## Cell Contents
## |-------------------------|
## | Count |
## | Chi-square contribution |
## | Row Percent |
## | Column Percent |
## | Total Percent |
## |-------------------------|
##
## Total Observations in Table: 243
##
## | Q20Recode
## Q6_1 | High Education | Low Education | Row Total |
## ------------------|----------------|----------------|----------------|
## Strongly disagree | 17 | 14 | 31 |
## | 1.512 | 1.040 | |
## | 54.839% | 45.161% | 12.757% |
## | 17.172% | 9.722% | |
## | 6.996% | 5.761% | |
## ------------------|----------------|----------------|----------------|
## Disagree | 6 | 8 | 14 |
## | 0.015 | 0.011 | |
## | 42.857% | 57.143% | 5.761% |
## | 6.061% | 5.556% | |
## | 2.469% | 3.292% | |
## ------------------|----------------|----------------|----------------|
## Neutral | 12 | 27 | 39 |
## | 0.952 | 0.654 | |
## | 30.769% | 69.231% | 16.049% |
## | 12.121% | 18.750% | |
## | 4.938% | 11.111% | |
## ------------------|----------------|----------------|----------------|
## Agree | 32 | 55 | 87 |
## | 0.335 | 0.230 | |
## | 36.782% | 63.218% | 35.802% |
## | 32.323% | 38.194% | |
## | 13.169% | 22.634% | |
## ------------------|----------------|----------------|----------------|
## Strongly agree | 32 | 40 | 72 |
## | 0.242 | 0.167 | |
## | 44.444% | 55.556% | 29.630% |
## | 32.323% | 27.778% | |
## | 13.169% | 16.461% | |
## ------------------|----------------|----------------|----------------|
## Column Total | 99 | 144 | 243 |
## | 40.741% | 59.259% | |
## ------------------|----------------|----------------|----------------|
##
##
## Statistics for All Table Factors
##
##
## Pearson's Chi-squared test
## ------------------------------------------------------------
## Chi^2 = 5.158175 d.f. = 4 p = 0.2714494
##
##
##
## Minimum expected frequency: 5.703704## Q6_1 High Education Low Education
## Strongly disagree 55% 45%
## Disagree 43% 57%
## Neutral 31% 69%
## Agree 37% 63%
## Strongly agree 44% 56%Income Comparison
##
## Cell Contents
## |-------------------------|
## | Count |
## | Chi-square contribution |
## | Row Percent |
## | Column Percent |
## | Total Percent |
## |-------------------------|
##
## Total Observations in Table: 220
##
## | Q21Recode
## Q6_1 | Above Median Income | Below Median Income | Row Total |
## ------------------|---------------------|---------------------|---------------------|
## Strongly disagree | 15 | 10 | 25 |
## | 1.081 | 0.918 | |
## | 60.000% | 40.000% | 11.364% |
## | 14.851% | 8.403% | |
## | 6.818% | 4.545% | |
## ------------------|---------------------|---------------------|---------------------|
## Disagree | 3 | 10 | 13 |
## | 1.476 | 1.253 | |
## | 23.077% | 76.923% | 5.909% |
## | 2.970% | 8.403% | |
## | 1.364% | 4.545% | |
## ------------------|---------------------|---------------------|---------------------|
## Neutral | 10 | 23 | 33 |
## | 1.751 | 1.486 | |
## | 30.303% | 69.697% | 15.000% |
## | 9.901% | 19.328% | |
## | 4.545% | 10.455% | |
## ------------------|---------------------|---------------------|---------------------|
## Agree | 30 | 52 | 82 |
## | 1.553 | 1.318 | |
## | 36.585% | 63.415% | 37.273% |
## | 29.703% | 43.697% | |
## | 13.636% | 23.636% | |
## ------------------|---------------------|---------------------|---------------------|
## Strongly agree | 43 | 24 | 67 |
## | 4.871 | 4.135 | |
## | 64.179% | 35.821% | 30.455% |
## | 42.574% | 20.168% | |
## | 19.545% | 10.909% | |
## ------------------|---------------------|---------------------|---------------------|
## Column Total | 101 | 119 | 220 |
## | 45.909% | 54.091% | |
## ------------------|---------------------|---------------------|---------------------|
##
##
## Statistics for All Table Factors
##
##
## Pearson's Chi-squared test
## ------------------------------------------------------------
## Chi^2 = 19.84103 d.f. = 4 p = 0.0005368059
##
##
##
## Minimum expected frequency: 5.968182## Q6_1 Above Median Income Below Median Income
## Strongly disagree 60% 40%
## Disagree 23% 77%
## Neutral 30% 70%
## Agree 37% 63%
## Strongly agree 64% 36%Vaccines help prevent diseases
Education Comparison
##
## Cell Contents
## |-------------------------|
## | Count |
## | Chi-square contribution |
## | Row Percent |
## | Column Percent |
## | Total Percent |
## |-------------------------|
##
## Total Observations in Table: 239
##
## | Q20Recode
## Q6_2 | High Education | Low Education | Row Total |
## ------------------|----------------|----------------|----------------|
## Strongly disagree | 13 | 10 | 23 |
## | 1.439 | 0.983 | |
## | 56.522% | 43.478% | 9.623% |
## | 13.402% | 7.042% | |
## | 5.439% | 4.184% | |
## ------------------|----------------|----------------|----------------|
## Disagree | 7 | 7 | 14 |
## | 0.306 | 0.209 | |
## | 50.000% | 50.000% | 5.858% |
## | 7.216% | 4.930% | |
## | 2.929% | 2.929% | |
## ------------------|----------------|----------------|----------------|
## Neutral | 8 | 22 | 30 |
## | 1.432 | 0.978 | |
## | 26.667% | 73.333% | 12.552% |
## | 8.247% | 15.493% | |
## | 3.347% | 9.205% | |
## ------------------|----------------|----------------|----------------|
## Agree | 30 | 54 | 84 |
## | 0.491 | 0.336 | |
## | 35.714% | 64.286% | 35.146% |
## | 30.928% | 38.028% | |
## | 12.552% | 22.594% | |
## ------------------|----------------|----------------|----------------|
## Strongly agree | 39 | 49 | 88 |
## | 0.302 | 0.206 | |
## | 44.318% | 55.682% | 36.820% |
## | 40.206% | 34.507% | |
## | 16.318% | 20.502% | |
## ------------------|----------------|----------------|----------------|
## Column Total | 97 | 142 | 239 |
## | 40.586% | 59.414% | |
## ------------------|----------------|----------------|----------------|
##
##
## Statistics for All Table Factors
##
##
## Pearson's Chi-squared test
## ------------------------------------------------------------
## Chi^2 = 6.682233 d.f. = 4 p = 0.1536643
##
##
##
## Minimum expected frequency: 5.682008## Q6_2 High Education Low Education
## Strongly disagree 57% 43%
## Disagree 50% 50%
## Neutral 27% 73%
## Agree 36% 64%
## Strongly agree 44% 56%Income Comparison
##
## Cell Contents
## |-------------------------|
## | Count |
## | Chi-square contribution |
## | Row Percent |
## | Column Percent |
## | Total Percent |
## |-------------------------|
##
## Total Observations in Table: 217
##
## | Q21Recode
## Q6_2 | Above Median Income | Below Median Income | Row Total |
## ------------------|---------------------|---------------------|---------------------|
## Strongly disagree | 11 | 7 | 18 |
## | 0.882 | 0.754 | |
## | 61.111% | 38.889% | 8.295% |
## | 11.000% | 5.983% | |
## | 5.069% | 3.226% | |
## ------------------|---------------------|---------------------|---------------------|
## Disagree | 5 | 8 | 13 |
## | 0.164 | 0.140 | |
## | 38.462% | 61.538% | 5.991% |
## | 5.000% | 6.838% | |
## | 2.304% | 3.687% | |
## ------------------|---------------------|---------------------|---------------------|
## Neutral | 7 | 18 | 25 |
## | 1.774 | 1.516 | |
## | 28.000% | 72.000% | 11.521% |
## | 7.000% | 15.385% | |
## | 3.226% | 8.295% | |
## ------------------|---------------------|---------------------|---------------------|
## Agree | 29 | 52 | 81 |
## | 1.858 | 1.588 | |
## | 35.802% | 64.198% | 37.327% |
## | 29.000% | 44.444% | |
## | 13.364% | 23.963% | |
## ------------------|---------------------|---------------------|---------------------|
## Strongly agree | 48 | 32 | 80 |
## | 3.362 | 2.874 | |
## | 60.000% | 40.000% | 36.866% |
## | 48.000% | 27.350% | |
## | 22.120% | 14.747% | |
## ------------------|---------------------|---------------------|---------------------|
## Column Total | 100 | 117 | 217 |
## | 46.083% | 53.917% | |
## ------------------|---------------------|---------------------|---------------------|
##
##
## Statistics for All Table Factors
##
##
## Pearson's Chi-squared test
## ------------------------------------------------------------
## Chi^2 = 14.91178 d.f. = 4 p = 0.004887729
##
##
##
## Minimum expected frequency: 5.990783## Q6_1 Above Median Income Below Median Income
## Strongly disagree 60% 40%
## Disagree 23% 77%
## Neutral 30% 70%
## Agree 36% 64%
## Strongly agree 65% 35%Exploratory Factor Analysis
There are five questions from the Immunization Knowledge and Beliefs and Attitudes and Concerns sections with similar 5-point Likert scales that could be analyzed with an exploratory factor analysis (EFA).
Questions to include in the EFA:
- Vaccines are generally safe
- Vaccines help prevent disease
- How safe do you think vaccines are?
- To what extent do you agree vaccines have more benefits than risks?
- How much do you trust the information you receive about vaccines?
Cronbach’s Alpha
##
## Reliability analysis
## Call: psych::alpha(x = efa_items)
##
## raw_alpha std.alpha G6(smc) average_r S/N ase mean sd median_r
## 0.93 0.93 0.93 0.72 13 0.0075 3.7 1.1 0.72
##
## 95% confidence boundaries
## lower alpha upper
## Feldt 0.91 0.93 0.94
## Duhachek 0.91 0.93 0.94
##
## Reliability if an item is dropped:
## raw_alpha std.alpha G6(smc) average_r S/N alpha se var.r med.r
## Q6_1 0.90 0.90 0.88 0.70 9.2 0.0102 0.0093 0.71
## Q6_2 0.92 0.92 0.90 0.74 11.1 0.0087 0.0037 0.72
## Q13 0.91 0.91 0.90 0.71 9.6 0.0101 0.0124 0.71
## Q14 0.90 0.90 0.90 0.69 9.0 0.0108 0.0113 0.70
## Q15 0.93 0.93 0.92 0.76 12.6 0.0079 0.0054 0.74
##
## Item statistics
## n raw.r std.r r.cor r.drop mean sd
## Q6_1 246 0.91 0.91 0.90 0.85 3.6 1.3
## Q6_2 246 0.85 0.86 0.83 0.77 3.8 1.2
## Q13 246 0.90 0.90 0.87 0.83 3.8 1.3
## Q14 246 0.92 0.92 0.89 0.86 3.7 1.3
## Q15 246 0.83 0.82 0.76 0.73 3.6 1.3
##
## Non missing response frequency for each item
## 1 2 3 4 5 miss
## Q6_1 0.13 0.06 0.16 0.35 0.30 0
## Q6_2 0.09 0.06 0.13 0.36 0.36 0
## Q13 0.11 0.07 0.16 0.28 0.38 0
## Q14 0.11 0.05 0.22 0.28 0.33 0
## Q15 0.10 0.12 0.14 0.33 0.31 0EFA Results
Model
Includes all five questions in one analysis. Overall, the model accounts for over 80% of the variability, a very high value. ML1 indicates a stronger relationship between the questions Vaccines are generally safe and Vaccines help prevent disease, while ML2 shows a strong relationship between the other three questions.
##
## Loadings:
## ML2 ML1
## Q6_1 0.563 0.705
## Q6_2 0.358 0.931
## Q13 0.778 0.430
## Q14 0.820 0.433
## Q15 0.758
##
## ML2 ML1
## SS loadings 2.298 1.814
## Proportion Var 0.460 0.363
## Cumulative Var 0.460 0.822Factor Scores
Each respondent now has two standardized scores:
- Factor_1_SafetyTrust → Reflects how strongly the respondent believes vaccines are safe and trustworthy.
- Factor_2_ImportanceBelief → Reflects how strongly the respondent believes vaccines are important and beneficial.
These are z-scores:
- A score of 0 means the respondent is at the average level of that factor in the sample.
- Positive scores = stronger agreement/endorsement of the factor.
- Negative scores = lower agreement/endorsement.
## # A tibble: 6 × 2
## Factor_1_SafetyTrust Factor_2_ImportanceBelief
## <dbl> <dbl>
## 1 -1.98 -0.822
## 2 0.290 0.0309
## 3 -1.31 0.641
## 4 -0.219 0.227
## 5 0.195 0.933
## 6 2.54 -3.43## Min. 1st Qu. Median Mean 3rd Qu. Max.
## -2.2991 -0.6950 0.1062 0.0000 0.8474 2.7153## Min. 1st Qu. Median Mean 3rd Qu. Max.
## -3.4652 -0.3863 0.2098 0.0000 0.6560 1.6327
Composite Score
If the EFA results remain high and robust, we could consider a simplified approach to calculate a composite score between the five items.
The Composite_VaccineBeliefsAttitudes variable
represents the overall index score for each respondent, averaging their
responses of the 5 questions, which are all scaled in the same
direction.
## Min. 1st Qu. Median Mean 3rd Qu. Max. NA's
## 1.000 3.000 4.000 3.708 4.600 5.000 2
Logistic Regression
Model 1
A Logistic regression model could be created for the outcome variable Have you ever delayed or skipped a recommended vaccine for yourself or your child(ren)?
With data gathered so far, results could be summarized as:
Parents/guardians who expressed stronger trust in vaccines were significantly less likely to delay or skip a recommended vaccine for themselves or their children. For each 1-point increase in the composite vaccine attitude score, the odds of delaying/skipping dropped by 78% (OR = 0.23, 95% CI [0.14, 0.35], p < .001).
Additionally, parents with below-median household income were 2.83 times more likely to delay or skip vaccines than those with higher incomes (95% CI [1.33, 6.27], p = .014).
Educational attainment (low vs. high) was not a statistically significant predictor in this model (p = .32), suggesting that income-related barriers may play a more important role in vaccine decision-making than education alone.
## # A tibble: 4 × 7
## term estimate std.error statistic p.value conf.low conf.high
## <chr> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
## 1 (Intercept) 191. 0.887 5.92 3.18e- 9 37.7 1232.
## 2 Composite_VaccineBel… 0.220 0.229 -6.63 3.32e-11 0.136 0.333
## 3 Q20Recode 0.680 0.385 -0.999 3.18e- 1 0.315 1.44
## 4 Q21Recode 2.83 0.394 2.65 8.16e- 3 1.33 6.27Model 2
Parents who received vaccine information from a doctor were significantly less likely to delay or skip vaccines (OR = 0.37, p = .008), while those who cited news media (OR = 5.79, p = .011) or websites/online searches (OR = 3.33, p =p < .001) were significantly more likely to delay or skip.
Information from family/friends showed a marginally significant effect in increasing the odds of skipping vaccines (OR = 3.52, p = .037), while sources like nurses, public health messages, and social media were not statistically significant predictors in this model.
## # A tibble: 8 × 7
## term estimate std.error statistic p.value conf.low conf.high
## <chr> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
## 1 (Intercept) 1.13 0.352 0.342 0.733 0.569 2.28
## 2 Q7_1 0.374 0.373 -2.63 0.00845 0.177 0.771
## 3 Q7_4 0.804 0.371 -0.589 0.556 0.384 1.65
## 4 Q7_5 3.52 0.603 2.08 0.0371 1.14 12.6
## 5 Q7_6 1.23 0.589 0.347 0.729 0.389 4.06
## 6 Q7_7 5.79 0.696 2.52 0.0116 1.59 25.5
## 7 Q7_8 3.33 0.321 3.75 0.000174 1.80 6.34
## 8 Q7_9 0.442 0.432 -1.89 0.0587 0.183 1.01The chart illustrates the relationship between various vaccine information sources and the likelihood that a parent delayed or skipped a vaccine for themselves or their child(ren).
Receiving information from a doctor significantly reduced the odds of skipping or delaying vaccination (OR = 0.37, p < .01), reinforcing the importance of provider-based communication in building vaccine confidence.
In contrast, obtaining vaccine information from websites or online searches (OR = 3.33, p < .001), news media sources (OR = 5.79, p < .05), and family or friends (OR = 3.52, p < .05) were all significantly associated with higher odds of delaying or skipping vaccination.
Other sources, such as social media, nurse/clinic staff, and public health messaging, were not statistically significant predictors in this model, although the trends suggest that the quality and trustworthiness of information sources may influence vaccination behavior.
Model 3
After adjusting for household income, the effects of vaccine information sources on parental decisions to delay or skip a vaccine remained largely consistent.
Receiving vaccine information from a doctor significantly reduced the odds of delaying or skipping a vaccine (OR = 0.44, 95% CI [0.20, 1.00], p = .055), although the association narrowly missed traditional levels of statistical significance.
Obtaining information from news media (OR = 7.10, 95% CI [1.90, 32.74], p = .006) and websites or online searches (OR = 3.12, 95% CI [1.61, 6.19], p < .001) remained significantly associated with higher odds of delaying or skipping vaccines.
Income level was a statistically significant predictor in this model (OR = 1.86, 95% CI [1.01, 3.45], p = .047), with parents from below-median income households more likely to delay or skip vaccinations compared to those from higher-income households.
## # A tibble: 9 × 7
## term estimate std.error statistic p.value conf.low conf.high
## <chr> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
## 1 (Intercept) 0.682 0.437 -0.875 0.382 0.288 1.62
## 2 Q21Recode 1.86 0.312 1.99 0.0471 1.01 3.45
## 3 Q7_1 0.444 0.419 -1.94 0.0526 0.192 1.00
## 4 Q7_4 1.06 0.387 0.150 0.881 0.491 2.26
## 5 Q7_5 2.93 0.632 1.70 0.0889 0.882 11.0
## 6 Q7_6 0.944 0.625 -0.0918 0.927 0.272 3.29
## 7 Q7_7 7.10 0.715 2.74 0.00609 1.90 32.7
## 8 Q7_8 3.12 0.342 3.33 0.000857 1.61 6.19
## 9 Q7_9 0.454 0.452 -1.75 0.0809 0.180 1.07
Model 4
An interaction model tested whether the protective effect of receiving vaccine information from a doctor varied by household income level. The interaction term was not statistically significant (p = .26), indicating that doctor-provided information remains similarly beneficial across both lower- and higher-income groups.
In this adjusted model, obtaining vaccine information from news media (OR = 6.88, 95% CI [1.87, 31.21], p = .006) and websites or online searches (OR = 2.97, 95% CI [1.52, 5.94], p = .001) remained significantly associated with increased odds of delaying or skipping vaccination, reinforcing earlier findings about the role of information source quality in vaccine decision-making.
## # A tibble: 10 × 7
## term estimate std.error statistic p.value conf.low conf.high
## <chr> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
## 1 (Intercept) 0.469 0.551 -1.38 0.169 0.154 1.38
## 2 Q7_1 0.700 0.582 -0.614 0.539 0.223 2.24
## 3 Q21Recode 4.22 0.801 1.80 0.0723 0.917 22.1
## 4 Q7_4 1.13 0.390 0.308 0.758 0.520 2.42
## 5 Q7_5 2.97 0.631 1.73 0.0844 0.892 11.1
## 6 Q7_6 0.987 0.629 -0.0211 0.983 0.282 3.47
## 7 Q7_7 6.88 0.707 2.73 0.00636 1.87 31.2
## 8 Q7_8 2.97 0.346 3.14 0.00167 1.52 5.94
## 9 Q7_9 0.449 0.452 -1.77 0.0765 0.178 1.06
## 10 Q7_1:Q21Recode 0.376 0.870 -1.12 0.261 0.0638 2.00Inferential Analysis Examples with T-tests and ANOVA
T-Test – Compare Composite Vaccine Attitude Scores by Income
Vaccine confidence levels were consistent across income groups, with an average score of 3.74 for below-median income households and 3.81 for above-median income households (p = .63).
However, income remained a significant predictor of vaccine behavior. In regression models predicting whether parents delayed or skipped vaccination, parents from lower-income households had significantly higher odds of delaying or skipping vaccines compared to those from higher-income households.
These findings suggest that differences in vaccine behavior may be driven more by logistical or access barriers than by differences in vaccine confidence or attitudes. Future outreach strategies should continue to promote vaccine safety and benefits, while also addressing real-world access challenges faced by lower-income families.
##
## Welch Two Sample t-test
##
## data: Composite_VaccineBeliefsAttitudes by Q21Recode
## t = -0.47657, df = 180.6, p-value = 0.6342
## alternative hypothesis: true difference in means between group 0 and group 1 is not equal to 0
## 95 percent confidence interval:
## -0.3791860 0.2316529
## sample estimates:
## mean in group 0 mean in group 1
## 3.736134 3.809901ANOVA with Insurance Type
An ANOVA was conducted to examine differences in vaccine confidence based on insurance provider (excluding respondents who were unsure of their insurance status).
The analysis did not find statistically significant differences in vaccine confidence across insurance types (F(1, 137) = 0.001, p = .971), suggesting that insurance coverage alone may not explain variation in parents’ confidence in vaccines.
##
## 1 2 3
## 99 30 10## Df Sum Sq Mean Sq F value Pr(>F)
## Q23 1 0.0 0.0014 0.001 0.971
## Residuals 137 145.2 1.0602Additional Regression Analysis
Composite belief model with demographics and doctor information
Model what predicts higher vaccine confidence (the composite score), using:
- Demographics (income and education)
- Trusted information sources (e.g., received info from a doctor)
This flips the earlier model: instead of predicting vaccine behavior, now predicting attitudes.
What this could tell us: What kinds of parents are most likely to trust vaccines more—and are those who receive information from a doctor significantly more confident, even after controlling for income and education?
## # A tibble: 4 × 7
## term estimate std.error statistic p.value conf.low conf.high
## <chr> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
## 1 (Intercept) 2.97 0.195 15.2 5.96e-36 2.58 3.35
## 2 Q20Recode -0.0677 0.163 -0.416 6.78e- 1 -0.389 0.253
## 3 Q21Recode 0.118 0.159 0.740 4.60e- 1 -0.196 0.432
## 4 Q7_1 0.920 0.195 4.72 4.23e- 6 0.536 1.30Potential summary:
A linear regression model examined the predictors of parental vaccine confidence (scored 1–5). After controlling for income and education:
- Receiving vaccine information from a doctor was significantly associated with higher vaccine confidence scores (b = 0.95, 95% CI [0.53, 1.37], p < .001).
- In contrast, neither income nor education level were statistically significant predictors of vaccine confidence (p = .97 and p = .86, respectively).
These results suggest that trust in vaccines is more strongly influenced by receiving information from healthcare providers than by demographic factors like income or education.
Composite belief model with all information sources
All information sources tossed into one model to predict beliefs and attitude composite score.
##
## Call:
## lm(formula = Composite_VaccineBeliefsAttitudes ~ Q7_1 + Q7_4 +
## Q7_5 + Q7_6 + Q7_7 + Q7_8 + Q7_9, data = info_model)
##
## Residuals:
## Min 1Q Median 3Q Max
## -3.0847 -0.5793 0.1483 0.7196 2.1005
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 3.35557 0.15173 22.115 < 0.0000000000000002 ***
## Q7_1 0.72914 0.15843 4.602 0.00000672 ***
## Q7_4 -0.04406 0.16137 -0.273 0.785039
## Q7_5 -0.87581 0.22546 -3.885 0.000132 ***
## Q7_6 -0.28805 0.23592 -1.221 0.223280
## Q7_7 -0.59508 0.24261 -2.453 0.014875 *
## Q7_8 -0.45605 0.13521 -3.373 0.000865 ***
## Q7_9 0.67807 0.17851 3.799 0.000184 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 0.9837 on 244 degrees of freedom
## Multiple R-squared: 0.2807, Adjusted R-squared: 0.26
## F-statistic: 13.6 on 7 and 244 DF, p-value: 0.000000000000007895## # A tibble: 8 × 7
## term estimate std.error statistic p.value conf.low conf.high
## <chr> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
## 1 (Intercept) 3.36 0.152 22.1 3.22e-60 3.06 3.65
## 2 Q7_1 0.729 0.158 4.60 6.72e- 6 0.417 1.04
## 3 Q7_4 -0.0441 0.161 -0.273 7.85e- 1 -0.362 0.274
## 4 Q7_5 -0.876 0.225 -3.88 1.32e- 4 -1.32 -0.432
## 5 Q7_6 -0.288 0.236 -1.22 2.23e- 1 -0.753 0.177
## 6 Q7_7 -0.595 0.243 -2.45 1.49e- 2 -1.07 -0.117
## 7 Q7_8 -0.456 0.135 -3.37 8.65e- 4 -0.722 -0.190
## 8 Q7_9 0.678 0.179 3.80 1.84e- 4 0.326 1.03Interpretation:
A linear regression model was used to explore the relationship between various information sources and vaccine confidence scores. The results revealed that:
- Receiving vaccine information from a doctor was the strongest positive predictor of higher vaccine confidence (b = +0.73, 95% CI [0.42, 1.04], p < .001).
- Receiving information from public health messages was also positively associated with confidence (b = +0.68, 95% CI [0.33, 1.03], p < .001).
- In contrast, receiving vaccine information from family or friends (b = –0.88, p < .001), news sources (b = –0.60, p = .015), and websites/search engines (b = –0.46, p < .001) were all significantly associated with lower vaccine confidence.
- Social media and nurse/clinic staff were not statistically significant predictors in this model (p > .05).
These findings underscore the importance of trusted medical and public health sources in building vaccine confidence, and suggest that countering misinformation from informal or online sources remains a critical challenge.
