grant_analyses

alphas

1. trust in science

psych::alpha (data.frame(
  d$sciTrust_1, 
  d$sciTrust_2, 
  d$sciTrust_3, 
  d$sciTrust_4, 
  d$sciTrust_5, 
  d$sciTrust_6,
  d$sciTrust_7, 
  d$sciTrust_8, 
  d$sciTrust_9, 
  d$sciTrust_10, 
  d$sciTrust_11, 
  d$sciTrust_12,
  d$sciTrust_13,
  d$sciTrust_14,
  d$sciTrust_15), cumulative = F, na.rm = T, delete = T)

## 
## Reliability analysis   
## Call: psych::alpha(x = data.frame(d$sciTrust_1, d$sciTrust_2, d$sciTrust_3, 
##     d$sciTrust_4, d$sciTrust_5, d$sciTrust_6, d$sciTrust_7, d$sciTrust_8, 
##     d$sciTrust_9, d$sciTrust_10, d$sciTrust_11, d$sciTrust_12, 
##     d$sciTrust_13, d$sciTrust_14, d$sciTrust_15), cumulative = F, 
##     na.rm = T, delete = T)
## 
##   raw_alpha std.alpha G6(smc) average_r S/N     ase mean   sd median_r
##       0.94      0.94    0.95       0.5  15 0.00078  3.5 0.82      0.5
## 
##  lower alpha upper     95% confidence boundaries
## 0.94 0.94 0.94 
## 
##  Reliability if an item is dropped:
##               raw_alpha std.alpha G6(smc) average_r S/N alpha se var.r med.r
## d.sciTrust_1       0.94      0.94    0.95      0.51  15  0.00080 0.021  0.51
## d.sciTrust_2       0.93      0.93    0.95      0.49  14  0.00087 0.022  0.49
## d.sciTrust_3       0.93      0.93    0.95      0.50  14  0.00085 0.022  0.49
## d.sciTrust_4       0.93      0.93    0.95      0.50  14  0.00083 0.022  0.49
## d.sciTrust_5       0.93      0.93    0.95      0.49  14  0.00087 0.022  0.49
## d.sciTrust_6       0.94      0.93    0.95      0.51  14  0.00081 0.021  0.53
## d.sciTrust_7       0.93      0.93    0.95      0.50  14  0.00082 0.021  0.51
## d.sciTrust_8       0.94      0.94    0.95      0.51  15  0.00079 0.020  0.53
## d.sciTrust_9       0.94      0.94    0.95      0.51  15  0.00080 0.019  0.53
## d.sciTrust_10      0.94      0.93    0.95      0.51  14  0.00081 0.021  0.53
## d.sciTrust_11      0.93      0.93    0.95      0.49  13  0.00088 0.023  0.48
## d.sciTrust_12      0.93      0.93    0.95      0.50  14  0.00085 0.022  0.49
## d.sciTrust_13      0.93      0.93    0.95      0.49  13  0.00088 0.021  0.49
## d.sciTrust_14      0.93      0.93    0.95      0.50  14  0.00084 0.022  0.49
## d.sciTrust_15      0.93      0.93    0.95      0.51  14  0.00083 0.022  0.49
## 
##  Item statistics 
##                  n raw.r std.r r.cor r.drop mean   sd
## d.sciTrust_1  1077  0.65  0.64  0.60   0.59  3.3 1.16
## d.sciTrust_2  1077  0.81  0.80  0.79   0.78  3.2 1.23
## d.sciTrust_3  1075  0.78  0.77  0.75   0.74  3.6 1.11
## d.sciTrust_4  1075  0.72  0.70  0.68   0.66  3.2 1.19
## d.sciTrust_5  1075  0.81  0.80  0.79   0.77  3.5 1.19
## d.sciTrust_6  1077  0.68  0.70  0.68   0.63  3.7 1.00
## d.sciTrust_7  1071  0.72  0.74  0.73   0.68  3.6 1.01
## d.sciTrust_8  1073  0.64  0.66  0.64   0.58  3.6 1.06
## d.sciTrust_9  1076  0.62  0.64  0.62   0.57  3.5 0.96
## d.sciTrust_10 1077  0.68  0.69  0.68   0.63  3.5 1.06
## d.sciTrust_11 1075  0.83  0.83  0.81   0.80  3.5 1.19
## d.sciTrust_12 1075  0.76  0.75  0.73   0.71  3.2 1.19
## d.sciTrust_13 1075  0.84  0.83  0.83   0.81  3.3 1.17
## d.sciTrust_14 1072  0.73  0.72  0.69   0.68  3.2 1.19
## d.sciTrust_15 1073  0.70  0.70  0.68   0.66  3.8 1.04
## 
## Non missing response frequency for each item
##                  1    2    3    4    5 miss
## d.sciTrust_1  0.08 0.16 0.30 0.30 0.16 0.92
## d.sciTrust_2  0.11 0.17 0.27 0.28 0.16 0.92
## d.sciTrust_3  0.05 0.10 0.28 0.32 0.24 0.92
## d.sciTrust_4  0.09 0.18 0.32 0.24 0.18 0.92
## d.sciTrust_5  0.07 0.12 0.27 0.29 0.25 0.92
## d.sciTrust_6  0.04 0.07 0.26 0.41 0.22 0.92
## d.sciTrust_7  0.04 0.07 0.29 0.40 0.19 0.92
## d.sciTrust_8  0.06 0.08 0.27 0.41 0.18 0.92
## d.sciTrust_9  0.04 0.06 0.39 0.36 0.14 0.92
## d.sciTrust_10 0.05 0.09 0.30 0.38 0.18 0.92
## d.sciTrust_11 0.07 0.13 0.25 0.31 0.23 0.92
## d.sciTrust_12 0.09 0.17 0.31 0.26 0.16 0.92
## d.sciTrust_13 0.08 0.16 0.31 0.27 0.18 0.92
## d.sciTrust_14 0.09 0.18 0.31 0.25 0.17 0.92
## d.sciTrust_15 0.04 0.05 0.26 0.33 0.32 0.92

1. correlation plots

i. all DVs

x <- cbind.data.frame(
  d$avgCRT, 
  d$avgSciLit,
  d$avgSymbBelief,
  d$vaxxBehavior, 
  d$vaxxAttitudes_w1, 
  d$vaxxAttitudes_w2,
  d$avgVaxxAttitudes, #wave 1 + wave 2
  d$SESladder, 
  d$education, 
  d$expertAtt_w1, 
  d$expertAtt_w2, 
  d$avgExpertAtt, #wave 1 + wave 2
  d$index_ANexp_w1, 
  d$index_ANexp_w2, 
  d$index_ANexp_w3)

colnames(x) <- c(
  'avgCRT', 
  'avgSciLit',
  'symbolicBelief',
  'vaxxBehavior', 
  'vaxxAttitudes_w1', 
  'vaxxAttitudes_w2',
  'avgVaxxAttitudes', 
  'SES', 
  'education', 
  'expertAttitude_w1', 
  'expertAttitude_w2', 
  'avgExpertAttitude', 
  'mediaIndex_w1', 
  'mediaIndex_w2', 
  'mediaIndex_w3')

cor2 <- cor(x, use = "complete.obs")

ggcorrplot(cor2, type = "lower",
   lab = TRUE, 
   title = "correlations", 
   show.legend = F, 
   insig = "blank", 
   lab_size = 2.5, 
   digits = 2,
   pch.cex = 7,
   tl.cex = 10) +
 theme(axis.text.x = element_text(margin = margin(-2, 0, 0, 0)),
        axis.text.y = element_text(margin = margin(0, -2, 0, 0)),
        panel.grid.minor = element_line(size = 7))

ii. CRT vs. media exposure wave 1

x <- cbind.data.frame(
  d$avgCRT, 
  d$ABC_exp_w1, 
  d$CBS_exp_w1, 
  d$CNN_exp_w1, 
  d$Fox_exp_w1,
  d$MSNBC_exp_w1,
  d$NBC_exp_w1, 
  d$NPR_exp_w1, 
  d$NYT_exp_w1, 
  d$PBS_exp_w1, 
  d$USAT_exp_w1,
  d$WSJ_exp_w1, 
  d$AOL_exp_w1,
  d$prop.media.exp_w1)

colnames(x) <- c(
  'avgCRT', 
  'ABC_exp', 
  'CBS_exp', 
  'CNN_exp', 
  'Fox_exp',
  'MSNBC_exp',
  'NBC_exp', 
  'NPR_exp', 
  'NYT_exp', 
  'PBS_exp', 
  'USAT_exp',
  'WSJ_exp', 
  'AOL_exp',
  'prop.media.exp')

cor2 <- cor(x, use = "complete.obs")

ggcorrplot(cor2, type = "lower",
   lab = TRUE, 
   title = "wave 1", 
   show.legend = F, 
   insig = "blank", 
   lab_size = 2.5, 
   digits = 2,
   pch.cex = 7,
   tl.cex = 10) +
 theme(axis.text.x = element_text(margin = margin(-2, 0, 0, 0)),
        axis.text.y = element_text(margin = margin(0, -2, 0, 0)),
        panel.grid.minor = element_line(size = 7))

iii. CRT vs. media exposure wave 2

x <- cbind.data.frame(
  d$avgCRT.c, 
  d$ABC_exp_w2, 
  d$CBS_exp_w2, 
  d$CNN_exp_w2, 
  d$Fox_exp_w2,
  d$MSNBC_exp_w2,
  d$NBC_exp_w2, 
  d$NPR_exp_w2, 
  d$NYT_exp_w2, 
  d$PBS_exp_w2, 
  d$USAT_exp_w2,
  d$WSJ_exp_w2, 
  d$AOL_exp_w2,
  d$prop.media.exp_w2)

colnames(x) <- c(
  'avgCRT', 
  'ABC_exp', 
  'CBS_exp', 
  'CNN_exp', 
  'Fox_exp',
  'MSNBC_exp',
  'NBC_exp', 
  'NPR_exp', 
  'NYT_exp', 
  'PBS_exp', 
  'USAT_exp',
  'WSJ_exp', 
  'AOL_exp',
  'prop.media.exp')

cor2 <- cor(x, use = "complete.obs")


ggcorrplot(cor2, type = "lower",
   lab = TRUE, 
   title = "wave 2", 
   show.legend = F, 
   insig = "blank", 
   lab_size = 2.5, 
   digits = 2,
   pch.cex = 7,
   tl.cex = 10) +
 theme(axis.text.x = element_text(margin = margin(-2, 0, 0, 0)),
        axis.text.y = element_text(margin = margin(0, -2, 0, 0)),
        panel.grid.minor = element_line(size = 7))

iv. CRT vs. media exposure wave 3

x <- cbind.data.frame(
  d$avgCRT.c, 
  d$ABC_exp_w3, 
  d$CBS_exp_w3, 
  d$CNN_exp_w3, 
  d$Fox_exp_w3,
  d$MSNBC_exp_w3,
  d$NBC_exp_w3, 
  d$NPR_exp_w3, 
  d$NYT_exp_w3, 
  d$PBS_exp_w3, 
  d$USAT_exp_w3,
  d$WSJ_exp_w3, 
  d$AOL_exp_w3,
  d$prop.media.exp_w3)

colnames(x) <- c(
  'avgCRT', 
  'ABC_exp', 
  'CBS_exp', 
  'CNN_exp', 
  'Fox_exp',
  'MSNBC_exp',
  'NBC_exp', 
  'NPR_exp', 
  'NYT_exp', 
  'PBS_exp', 
  'USAT_exp',
  'WSJ_exp', 
  'AOL_exp',
  'prop.media.exp')

cor2 <- cor(x, use = "complete.obs")

ggcorrplot(cor2, type = "lower",
   lab = TRUE, 
   title = "wave 3", 
   show.legend = F, 
   insig = "blank", 
   lab_size = 2.5, 
   digits = 2,
   pch.cex = 7,
   tl.cex = 10) +
 theme(axis.text.x = element_text(margin = margin(-2, 0, 0, 0)),
        axis.text.y = element_text(margin = margin(0, -2, 0, 0)),
        panel.grid.minor = element_line(size = 7))

v. CRT vs. media trust wave 1

x <- cbind.data.frame(
  d$avgCRT, 
  d$ABC_trust_w1, 
  d$CBS_trust_w1, 
  d$CNN_trust_w1, 
  d$Fox_trust_w1,
  d$MSNBC_trust_w1,
  d$NBC_trust_w1, 
  d$NPR_trust_w1, 
  d$NYT_trust_w1, 
  d$PBS_trust_w1, 
  d$USAT_trust_w1,
  d$WSJ_trust_w1, 
  d$AOL_trust_w1)

colnames(x) <- c(
  'avgCRT', 
  'ABC_trust', 
  'CBS_trust', 
  'CNN_trust', 
  'Fox_trust',
  'MSNBC_trust',
  'NBC_trust', 
  'NPR_trust', 
  'NYT_trust', 
  'PBS_trust', 
  'USAT_trust',
  'WSJ_trust', 
  'AOL_trust')

cor2 <- cor(x, use = "complete.obs")

ggcorrplot(cor2, type = "lower",
   lab = TRUE, 
   title = "wave 1", 
   show.legend = F, 
   insig = "blank", 
   lab_size = 2.5, 
   digits = 2,
   pch.cex = 7,
   tl.cex = 10) +
 theme(axis.text.x = element_text(margin = margin(-2, 0, 0, 0)),
        axis.text.y = element_text(margin = margin(0, -2, 0, 0)),
        panel.grid.minor = element_line(size = 7))

vi. CRT vs. media trust wave 2

x <- cbind.data.frame(
  d$avgCRT, 
  d$ABC_trust_w2, 
  d$CBS_trust_w2, 
  d$CNN_trust_w2, 
  d$Fox_trust_w2,
  d$MSNBC_trust_w2,
  d$NBC_trust_w2, 
  d$NPR_trust_w2, 
  d$NYT_trust_w2, 
  d$PBS_trust_w2, 
  d$USAT_trust_w2,
  d$WSJ_trust_w2, 
  d$AOL_trust_w2)

colnames(x) <- c(
  'avgCRT', 
  'ABC_trust', 
  'CBS_trust', 
  'CNN_trust', 
  'Fox_trust',
  'MSNBC_trust',
  'NBC_trust', 
  'NPR_trust', 
  'NYT_trust', 
  'PBS_trust', 
  'USAT_trust',
  'WSJ_trust', 
  'AOL_trust')

cor2 <- cor(x, use = "complete.obs")

ggcorrplot(cor2, type = "lower",
   lab = TRUE, 
   title = "wave 2", 
   show.legend = F, 
   insig = "blank", 
   lab_size = 2.5, 
   digits = 2,
   pch.cex = 7,
   tl.cex = 10) +
 theme(axis.text.x = element_text(margin = margin(-2, 0, 0, 0)),
        axis.text.y = element_text(margin = margin(0, -2, 0, 0)),
        panel.grid.minor = element_line(size = 7))

vii. CRT vs. media trust wave 3

x <- cbind.data.frame(
  d$avgCRT, 
  d$ABC_trust_w3, 
  d$CBS_trust_w3, 
  d$CNN_trust_w3, 
  d$Fox_trust_w3,
  d$MSNBC_trust_w3,
  d$NBC_trust_w3, 
  d$NPR_trust_w3, 
  d$NYT_trust_w3, 
  d$PBS_trust_w3, 
  d$USAT_trust_w3,
  d$WSJ_trust_w3, 
  d$AOL_trust_w3)

colnames(x) <- c(
  'avgCRT', 
  'ABC_trust', 
  'CBS_trust', 
  'CNN_trust', 
  'Fox_trust',
  'MSNBC_trust',
  'NBC_trust', 
  'NPR_trust', 
  'NYT_trust', 
  'PBS_trust', 
  'USAT_trust',
  'WSJ_trust', 
  'AOL_trust')

cor2 <- cor(x, use = "complete.obs")

ggcorrplot(cor2, type = "lower",
   lab = TRUE, 
   title = "wave 3", 
   show.legend = F, 
   insig = "blank", 
   lab_size = 2.5, 
   digits = 2,
   pch.cex = 7,
   tl.cex = 10) +
 theme(axis.text.x = element_text(margin = margin(-2, 0, 0, 0)),
        axis.text.y = element_text(margin = margin(0, -2, 0, 0)),
        panel.grid.minor = element_line(size = 7))

2. correlation tests

i. edu vs. SES

cor.test(d$education, d$SESladder)

## 
##  Pearson's product-moment correlation
## 
## data:  d$education and d$SESladder
## t = -16.939, df = 10722, p-value < 2.2e-16
## alternative hypothesis: true correlation is not equal to 0
## 95 percent confidence interval:
##  -0.1798177 -0.1429503
## sample estimates:
##        cor 
## -0.1614403

ii. CRT vs. vaxx attitudes W1

cor.test(d$avgCRT, d$vaxxAttitudes_w1)

## 
##  Pearson's product-moment correlation
## 
## data:  d$avgCRT and d$vaxxAttitudes_w1
## t = 2.4332, df = 999, p-value = 0.01514
## alternative hypothesis: true correlation is not equal to 0
## 95 percent confidence interval:
##  0.01486345 0.13806035
## sample estimates:
##        cor 
## 0.07675486

iii. CRT vs. vaxx attitudes W2

cor.test(d$avgCRT, d$vaxxAttitudes_w2)

## 
##  Pearson's product-moment correlation
## 
## data:  d$avgCRT and d$vaxxAttitudes_w2
## t = 3.8604, df = 930, p-value = 0.000121
## alternative hypothesis: true correlation is not equal to 0
## 95 percent confidence interval:
##  0.0618688 0.1882831
## sample estimates:
##       cor 
## 0.1255857

vi. CRT vs. vaxx behavior

cor.test(d$avgCRT, d$vaxxBehavior)

## 
##  Pearson's product-moment correlation
## 
## data:  d$avgCRT and d$vaxxBehavior
## t = 2.1255, df = 1058, p-value = 0.03378
## alternative hypothesis: true correlation is not equal to 0
## 95 percent confidence interval:
##  0.005014027 0.124928361
## sample estimates:
##       cor 
## 0.0652066

KEY PREDICTORS = CRT + AnalyticalMedia + TotalMedia

3. VaxxBehavior = symbolic belief + KEY PREDICTORS

m3 <- lm(vaxxBehavior ~ avgSymbBelief.c + 
           avgCRT.c + index_ANexp_w3 + prop.media.exp_w3, data = d)
summary(m3)

## 
## Call:
## lm(formula = vaxxBehavior ~ avgSymbBelief.c + avgCRT.c + index_ANexp_w3 + 
##     prop.media.exp_w3, data = d)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -3.0577 -0.3716  0.3391  0.7940  1.6699 
## 
## Coefficients:
##                   Estimate Std. Error t value Pr(>|t|)    
## (Intercept)        2.93860    0.05570  52.757  < 2e-16 ***
## avgSymbBelief.c   -0.14993    0.02090  -7.175  1.4e-12 ***
## avgCRT.c           0.41999    0.11292   3.719 0.000211 ***
## index_ANexp_w3    -0.01624    0.01045  -1.555 0.120361    
## prop.media.exp_w3  6.48914    3.35225   1.936 0.053178 .  
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 1.095 on 1004 degrees of freedom
##   (12488 observations deleted due to missingness)
## Multiple R-squared:  0.1356, Adjusted R-squared:  0.1321 
## F-statistic: 39.36 on 4 and 1004 DF,  p-value: < 2.2e-16

4. VaxxBehavior = SES + KEY PREDICTORS

m4 <- lm(vaxxBehavior ~ SESladder.c + 
           avgCRT.c + index_ANexp_w3 + prop.media.exp_w3, data = d)
summary(m4)

## 
## Call:
## lm(formula = vaxxBehavior ~ SESladder.c + avgCRT.c + index_ANexp_w3 + 
##     prop.media.exp_w3, data = d)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -3.3632 -0.3098  0.3961  0.7984  1.4520 
## 
## Coefficients:
##                   Estimate Std. Error t value Pr(>|t|)    
## (Intercept)        2.84720    0.05517  51.610  < 2e-16 ***
## SESladder.c       -0.05553    0.01813  -3.063 0.002249 ** 
## avgCRT.c           0.43303    0.11574   3.742 0.000193 ***
## index_ANexp_w3    -0.01343    0.01069  -1.256 0.209310    
## prop.media.exp_w3  5.87278    3.43185   1.711 0.087347 .  
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 1.115 on 992 degrees of freedom
##   (12500 observations deleted due to missingness)
## Multiple R-squared:  0.09756,    Adjusted R-squared:  0.09392 
## F-statistic: 26.81 on 4 and 992 DF,  p-value: < 2.2e-16

5. VaxxBehavior = science literacy + KEY PREDICTORS

m5 <- lm(vaxxBehavior ~ avgSciLit.c + 
           avgCRT.c + index_ANexp_w3 + prop.media.exp_w3, data = d)
summary(m5)

## 
## Call:
## lm(formula = vaxxBehavior ~ avgSciLit.c + avgCRT.c + index_ANexp_w3 + 
##     prop.media.exp_w3, data = d)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -3.2596 -0.3257  0.3974  0.8317  1.4032 
## 
## Coefficients:
##                   Estimate Std. Error t value Pr(>|t|)    
## (Intercept)        2.82807    0.05399  52.385  < 2e-16 ***
## avgSciLit.c        0.11919    0.27186   0.438   0.6612    
## avgCRT.c           0.50211    0.11630   4.317 1.73e-05 ***
## index_ANexp_w3    -0.01409    0.01065  -1.322   0.1863    
## prop.media.exp_w3  6.14453    3.41970   1.797   0.0727 .  
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 1.127 on 1050 degrees of freedom
##   (12442 observations deleted due to missingness)
## Multiple R-squared:  0.09234,    Adjusted R-squared:  0.08888 
## F-statistic:  26.7 on 4 and 1050 DF,  p-value: < 2.2e-16

6. VaxxBehavior = age + KEY PREDICTORS

m6 <- lm(vaxxBehavior ~ age.c + 
           avgCRT.c + index_ANexp_w3 + prop.media.exp_w3, data = d)
summary(m6)

## 
## Call:
## lm(formula = vaxxBehavior ~ age.c + avgCRT.c + index_ANexp_w3 + 
##     prop.media.exp_w3, data = d)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -3.0495 -0.3290  0.4038  0.7697  1.7623 
## 
## Coefficients:
##                    Estimate Std. Error t value Pr(>|t|)    
## (Intercept)        2.744669   0.057786  47.498  < 2e-16 ***
## age.c              0.013437   0.002468   5.445 6.54e-08 ***
## avgCRT.c           0.462023   0.115604   3.997 6.91e-05 ***
## index_ANexp_w3    -0.005966   0.010730  -0.556    0.578    
## prop.media.exp_w3  3.577857   3.442759   1.039    0.299    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 1.104 on 976 degrees of freedom
##   (12516 observations deleted due to missingness)
## Multiple R-squared:  0.1158, Adjusted R-squared:  0.1122 
## F-statistic: 31.95 on 4 and 976 DF,  p-value: < 2.2e-16

7. VaxxBehavior = sciTrust + KEY PREDICTORS

m7 <- lm(vaxxBehavior ~ avgTrustSci.c + 
           avgCRT.c + index_ANexp_w3 + prop.media.exp_w3, data = d)
summary(m7)

## 
## Call:
## lm(formula = vaxxBehavior ~ avgTrustSci.c + avgCRT.c + index_ANexp_w3 + 
##     prop.media.exp_w3, data = d)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -3.3060 -0.4854  0.3058  0.7270  2.3410 
## 
## Coefficients:
##                    Estimate Std. Error t value Pr(>|t|)    
## (Intercept)        2.918640   0.050690  57.579   <2e-16 ***
## avgTrustSci.c      0.518995   0.040523  12.807   <2e-16 ***
## avgCRT.c           0.251502   0.107518   2.339   0.0195 *  
## index_ANexp_w3    -0.008340   0.009916  -0.841   0.4005    
## prop.media.exp_w3  3.929577   3.184105   1.234   0.2174    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 1.048 on 1050 degrees of freedom
##   (12442 observations deleted due to missingness)
## Multiple R-squared:  0.2148, Adjusted R-squared:  0.2118 
## F-statistic: 71.82 on 4 and 1050 DF,  p-value: < 2.2e-16