analysis
Joshua Rosenberg
9/13/2018
Update on qualitative coding
Item #13 - Please explain the one or two technology competencies that you consider as being most important to your individual work as a teacher educator.
Loading, setting up
In this section, we start out by doing:
library(tidyverse)
library(ggridges)
library(corrr)
library(clipr)
library(psych)
library(lavaan)
library(broom)
usethis::use_git_ignore("*.csv")## ✔ Setting active project to '/Users/joshuarosenberg/Documents/tetc-analysis'Read the data:
# d <- read_csv("tetcs-survey-export.csv", skip = 0)
d <- read_csv("full-tetcs-dataset.csv")## Parsed with column specification:
## cols(
## .default = col_character()
## )## See spec(...) for full column specifications.d <- d %>% slice(-c(1:2))Check that the data loaded:
d## # A tibble: 337 x 65
## StartDate EndDate Status IPAddress Progress `Duration (in s… Finished
## <chr> <chr> <chr> <chr> <chr> <chr> <chr>
## 1 2018-09-… 2018-0… IP Ad… 152.33.5… 100 499 True
## 2 2018-09-… 2018-0… IP Ad… 152.33.6… 100 1569 True
## 3 2018-09-… 2018-0… IP Ad… 152.33.7… 100 667 True
## 4 2018-09-… 2018-0… IP Ad… 134.139.… 100 555 True
## 5 2018-09-… 2018-0… IP Ad… 152.33.1… 100 600 True
## 6 2018-09-… 2018-0… IP Ad… 49.199.2… 100 1006 True
## 7 2018-09-… 2018-0… IP Ad… 120.151.… 100 491 True
## 8 2018-09-… 2018-0… IP Ad… 1.128.10… 100 868 True
## 9 2018-09-… 2018-0… IP Ad… 121.222.… 100 674 True
## 10 2018-09-… 2018-0… IP Ad… 144.133.… 100 616 True
## # … with 327 more rows, and 58 more variables: RecordedDate <chr>,
## # ResponseId <chr>, RecipientLastName <chr>, RecipientFirstName <chr>,
## # RecipientEmail <chr>, ExternalReference <chr>, LocationLatitude <chr>,
## # LocationLongitude <chr>, DistributionChannel <chr>,
## # UserLanguage <chr>, Q2 <chr>, Q3 <chr>, Q3_8_TEXT <chr>, Q4 <chr>,
## # Q5_1 <chr>, Q5_2 <chr>, Q6 <chr>, Q6_6_TEXT <chr>, Q7 <chr>,
## # Q7_7_TEXT <chr>, Q8 <chr>, Q8_12_TEXT <chr>, Q9_4 <chr>, Q9_5 <chr>,
## # Q9_6 <chr>, Q9_7 <chr>, `Q10#1_1` <chr>, `Q10#1_2` <chr>,
## # `Q10#1_3` <chr>, `Q10#1_4` <chr>, `Q10#1_5` <chr>, Q11_1 <chr>,
## # Q11_2 <chr>, Q11_3 <chr>, Q11_4 <chr>, Q11_5 <chr>, Q11_6 <chr>,
## # Q11_7 <chr>, Q11_8 <chr>, Q11_9 <chr>, Q11_10 <chr>, Q11_11 <chr>,
## # Q11_12 <chr>, Q12 <chr>, Q13 <chr>, Q14 <chr>, Q15 <chr>, Q16 <chr>,
## # Q17 <chr>, Q18_1 <chr>, Q18_2 <chr>, Q18_3 <chr>, Q18_4 <chr>,
## # Q18_5 <chr>, Q18_6 <chr>, Q18_7 <chr>, Q18_8 <chr>, Q18_9 <chr>Process the TETCs variables:
ds <- d %>%
select(Q11_1:Q11_12) %>%
set_names(1:12) %>%
mutate_all(str_extract, "\\(?[0-9,.]+\\)?") %>%
mutate_all(as.integer) %>%
set_names(str_c("v", 1:12))
ds %>%
summarize_all(funs(mean, sd), na.rm = TRUE) %>%
gather(key, val) %>%
separate(key, into = c("var", "stat")) %>%
spread(stat, val) %>%
mutate(TETC = str_sub(var, start = 2),
TETC = as.integer(TETC)) %>%
arrange(TETC) %>%
mutate(TETC = str_c("TETC", TETC)) %>%
mutate(mean = round(mean, 2),
sd = round(sd, 2)) %>%
mutate(mean_sd = str_c(mean, " (", sd, ")")) %>%
select(mean_sd)## # A tibble: 12 x 1
## mean_sd
## <chr>
## 1 3.66 (1.11)
## 2 3.79 (1.03)
## 3 3.86 (0.98)
## 4 4.21 (0.89)
## 5 3.54 (1.11)
## 6 3.69 (1.05)
## 7 3.74 (1.15)
## 8 3.06 (1.31)
## 9 3.41 (1.17)
## 10 3.59 (1.2)
## 11 3.22 (1.37)
## 12 3.87 (1.15)# c <- cbind(ds)
#
# g <- ds %>%
# gather(key, val)
#
# summary(aov(val ~ key, g))rec_f <- function(x) {
recode(x,
`Major asset I experience / have experienced.` = 5,
`Asset I experience / have experienced.` = 4,
`Neutral / neither a barrier nor an asset.` = 3,
`Barrier I experience / have experienced.` = 2,
`Major barrier I experience / have experienced.` = 1
)
}
dd <- d %>%
select(Q18_1:Q18_9) %>%
mutate_all(rec_f)
ddd <- dd %>%
gather(key, val) %>%
mutate(key = str_replace(key, "v", "Asset/Barier")) %>%
group_by(key) %>%
summarize(mean_val = mean(val, na.rm = TRUE),
sd_val = sd(val, na.rm = TRUE)) %>%
arrange(desc(mean_val))
dd %>%
gather(key, val) %>%
ggplot() +
geom_jitter(aes(x = key, y = val),
alpha = .5,
color = "gray",
height = 0.2) +
geom_point(data = ddd,
aes(x = key, y = mean_val),
size = 2.5) +
geom_errorbar(data = ddd,
aes(x = key, y = mean_val,
ymin = mean_val-sd_val,
ymax = mean_val+sd_val)) +
theme_bw() +
ylab("Value") +
xlab(NULL) +
theme(text = element_text(size = 14),
axis.text.x = element_text(angle = 45, hjust = 1)) +
scale_x_discrete(limits = c("Q18_7", "Q18_9", "Q18_8", "Q18_6", "Q18_1", "Q18_3", "Q18_4", "Q18_5", "Q18_2"))
ggsave('assets-barriers-plot.png', width = 8, height = 6)EFA
scree(dd[complete.cases(dd), ])
f <- fac(dd[complete.cases(dd), ], nfactors = 2)## Loading required namespace: GPArotationprint(f)## Factor Analysis using method = minres
## Call: fac(r = dd[complete.cases(dd), ], nfactors = 2)
## Standardized loadings (pattern matrix) based upon correlation matrix
## MR1 MR2 h2 u2 com
## Q18_1 0.22 0.56 0.46 0.54 1.3
## Q18_2 0.28 0.51 0.46 0.54 1.5
## Q18_3 0.16 0.61 0.48 0.52 1.1
## Q18_4 -0.11 0.83 0.62 0.38 1.0
## Q18_5 -0.10 0.71 0.46 0.54 1.0
## Q18_6 0.37 0.37 0.39 0.61 2.0
## Q18_7 0.67 0.11 0.52 0.48 1.1
## Q18_8 0.93 -0.03 0.85 0.15 1.0
## Q18_9 0.89 -0.01 0.79 0.21 1.0
##
## MR1 MR2
## SS loadings 2.58 2.44
## Proportion Var 0.29 0.27
## Cumulative Var 0.29 0.56
## Proportion Explained 0.51 0.49
## Cumulative Proportion 0.51 1.00
##
## With factor correlations of
## MR1 MR2
## MR1 1.00 0.41
## MR2 0.41 1.00
##
## Mean item complexity = 1.2
## Test of the hypothesis that 2 factors are sufficient.
##
## The degrees of freedom for the null model are 36 and the objective function was 4.53 with Chi Square of 1462.71
## The degrees of freedom for the model are 19 and the objective function was 0.44
##
## The root mean square of the residuals (RMSR) is 0.06
## The df corrected root mean square of the residuals is 0.08
##
## The harmonic number of observations is 328 with the empirical chi square 75.38 with prob < 1.1e-08
## The total number of observations was 328 with Likelihood Chi Square = 143.03 with prob < 4.8e-21
##
## Tucker Lewis Index of factoring reliability = 0.835
## RMSEA index = 0.143 and the 90 % confidence intervals are 0.12 0.163
## BIC = 32.97
## Fit based upon off diagonal values = 0.98
## Measures of factor score adequacy
## MR1 MR2
## Correlation of (regression) scores with factors 0.96 0.91
## Multiple R square of scores with factors 0.92 0.83
## Minimum correlation of possible factor scores 0.83 0.67CFA
mod <- 'access =~ Q18_1 + Q18_2 + Q18_3
vision =~ Q18_4 + Q18_5 + Q18_6
personal =~ Q18_7 + Q18_8 +Q18_9'
c2a <- cfa(mod, data = dd)
summary(c2a, fit.measures = TRUE)## lavaan 0.6-3 ended normally after 28 iterations
##
## Optimization method NLMINB
## Number of free parameters 21
##
## Used Total
## Number of observations 328 337
##
## Estimator ML
## Model Fit Test Statistic 127.682
## Degrees of freedom 24
## P-value (Chi-square) 0.000
##
## Model test baseline model:
##
## Minimum Function Test Statistic 1484.588
## Degrees of freedom 36
## P-value 0.000
##
## User model versus baseline model:
##
## Comparative Fit Index (CFI) 0.928
## Tucker-Lewis Index (TLI) 0.893
##
## Loglikelihood and Information Criteria:
##
## Loglikelihood user model (H0) -3813.350
## Loglikelihood unrestricted model (H1) -3749.508
##
## Number of free parameters 21
## Akaike (AIC) 7668.699
## Bayesian (BIC) 7748.352
## Sample-size adjusted Bayesian (BIC) 7681.741
##
## Root Mean Square Error of Approximation:
##
## RMSEA 0.115
## 90 Percent Confidence Interval 0.096 0.135
## P-value RMSEA <= 0.05 0.000
##
## Standardized Root Mean Square Residual:
##
## SRMR 0.098
##
## Parameter Estimates:
##
## Information Expected
## Information saturated (h1) model Structured
## Standard Errors Standard
##
## Latent Variables:
## Estimate Std.Err z-value P(>|z|)
## access =~
## Q18_1 1.000
## Q18_2 1.010 0.081 12.398 0.000
## Q18_3 0.886 0.071 12.413 0.000
## vision =~
## Q18_4 1.000
## Q18_5 0.872 0.073 11.904 0.000
## Q18_6 0.561 0.069 8.184 0.000
## personal =~
## Q18_7 1.000
## Q18_8 1.347 0.088 15.336 0.000
## Q18_9 1.308 0.086 15.157 0.000
##
## Covariances:
## Estimate Std.Err z-value P(>|z|)
## access ~~
## vision 0.608 0.078 7.808 0.000
## personal 0.385 0.056 6.864 0.000
## vision ~~
## personal 0.221 0.046 4.759 0.000
##
## Variances:
## Estimate Std.Err z-value P(>|z|)
## .Q18_1 0.655 0.073 9.035 0.000
## .Q18_2 0.742 0.079 9.414 0.000
## .Q18_3 0.567 0.060 9.391 0.000
## .Q18_4 0.336 0.065 5.214 0.000
## .Q18_5 0.481 0.059 8.137 0.000
## .Q18_6 0.884 0.074 11.945 0.000
## .Q18_7 0.515 0.044 11.744 0.000
## .Q18_8 0.123 0.030 4.111 0.000
## .Q18_9 0.210 0.031 6.682 0.000
## access 0.924 0.123 7.486 0.000
## vision 0.886 0.109 8.122 0.000
## personal 0.502 0.071 7.101 0.000modb <- 'access =~ Q18_1 + Q18_2 + Q18_3
vision =~ Q18_4 + Q18_5
personal =~ Q18_7 + Q18_8 +Q18_9'
c2b <- cfa(modb, data = dd)
summary(c2b, fit.measures = TRUE)## lavaan 0.6-3 ended normally after 34 iterations
##
## Optimization method NLMINB
## Number of free parameters 19
##
## Used Total
## Number of observations 328 337
##
## Estimator ML
## Model Fit Test Statistic 35.342
## Degrees of freedom 17
## P-value (Chi-square) 0.006
##
## Model test baseline model:
##
## Minimum Function Test Statistic 1323.569
## Degrees of freedom 28
## P-value 0.000
##
## User model versus baseline model:
##
## Comparative Fit Index (CFI) 0.986
## Tucker-Lewis Index (TLI) 0.977
##
## Loglikelihood and Information Criteria:
##
## Loglikelihood user model (H0) -3357.441
## Loglikelihood unrestricted model (H1) -3339.770
##
## Number of free parameters 19
## Akaike (AIC) 6752.882
## Bayesian (BIC) 6824.949
## Sample-size adjusted Bayesian (BIC) 6764.681
##
## Root Mean Square Error of Approximation:
##
## RMSEA 0.057
## 90 Percent Confidence Interval 0.030 0.084
## P-value RMSEA <= 0.05 0.296
##
## Standardized Root Mean Square Residual:
##
## SRMR 0.034
##
## Parameter Estimates:
##
## Information Expected
## Information saturated (h1) model Structured
## Standard Errors Standard
##
## Latent Variables:
## Estimate Std.Err z-value P(>|z|)
## access =~
## Q18_1 1.000
## Q18_2 1.010 0.082 12.331 0.000
## Q18_3 0.891 0.072 12.406 0.000
## vision =~
## Q18_4 1.000
## Q18_5 0.790 0.080 9.851 0.000
## personal =~
## Q18_7 1.000
## Q18_8 1.351 0.088 15.312 0.000
## Q18_9 1.308 0.086 15.127 0.000
##
## Covariances:
## Estimate Std.Err z-value P(>|z|)
## access ~~
## vision 0.601 0.078 7.678 0.000
## personal 0.383 0.056 6.850 0.000
## vision ~~
## personal 0.190 0.047 4.062 0.000
##
## Variances:
## Estimate Std.Err z-value P(>|z|)
## .Q18_1 0.659 0.073 9.046 0.000
## .Q18_2 0.746 0.079 9.419 0.000
## .Q18_3 0.561 0.060 9.298 0.000
## .Q18_4 0.217 0.089 2.427 0.015
## .Q18_5 0.527 0.069 7.686 0.000
## .Q18_7 0.517 0.044 11.753 0.000
## .Q18_8 0.121 0.030 4.022 0.000
## .Q18_9 0.212 0.032 6.730 0.000
## access 0.920 0.123 7.457 0.000
## vision 1.005 0.129 7.815 0.000
## personal 0.501 0.071 7.090 0.000fs <- lavaan::lavPredict(c2b, newdata = dd)
names(fs) <- str_c("fs", 1:3)
mod1 <- 'access_vision =~ Q18_1 + Q18_2 + Q18_3 + Q18_4 + Q18_5 + Q18_6
personal =~ Q18_7 + Q18_8 +Q18_9'
c3 <- cfa(mod1, data = dd)
summary(c3, fit.measures = TRUE)## lavaan 0.6-3 ended normally after 25 iterations
##
## Optimization method NLMINB
## Number of free parameters 19
##
## Used Total
## Number of observations 328 337
##
## Estimator ML
## Model Fit Test Statistic 206.917
## Degrees of freedom 26
## P-value (Chi-square) 0.000
##
## Model test baseline model:
##
## Minimum Function Test Statistic 1484.588
## Degrees of freedom 36
## P-value 0.000
##
## User model versus baseline model:
##
## Comparative Fit Index (CFI) 0.875
## Tucker-Lewis Index (TLI) 0.827
##
## Loglikelihood and Information Criteria:
##
## Loglikelihood user model (H0) -3852.967
## Loglikelihood unrestricted model (H1) -3749.508
##
## Number of free parameters 19
## Akaike (AIC) 7743.933
## Bayesian (BIC) 7816.001
## Sample-size adjusted Bayesian (BIC) 7755.733
##
## Root Mean Square Error of Approximation:
##
## RMSEA 0.146
## 90 Percent Confidence Interval 0.128 0.164
## P-value RMSEA <= 0.05 0.000
##
## Standardized Root Mean Square Residual:
##
## SRMR 0.085
##
## Parameter Estimates:
##
## Information Expected
## Information saturated (h1) model Structured
## Standard Errors Standard
##
## Latent Variables:
## Estimate Std.Err z-value P(>|z|)
## access_vision =~
## Q18_1 1.000
## Q18_2 1.006 0.085 11.831 0.000
## Q18_3 0.906 0.075 12.134 0.000
## Q18_4 0.763 0.072 10.591 0.000
## Q18_5 0.650 0.070 9.302 0.000
## Q18_6 0.668 0.070 9.518 0.000
## personal =~
## Q18_7 1.000
## Q18_8 1.332 0.086 15.435 0.000
## Q18_9 1.307 0.086 15.284 0.000
##
## Covariances:
## Estimate Std.Err z-value P(>|z|)
## access_vision ~~
## personal 0.371 0.054 6.857 0.000
##
## Variances:
## Estimate Std.Err z-value P(>|z|)
## .Q18_1 0.724 0.073 9.984 0.000
## .Q18_2 0.818 0.080 10.278 0.000
## .Q18_3 0.589 0.059 9.956 0.000
## .Q18_4 0.724 0.065 11.175 0.000
## .Q18_5 0.793 0.068 11.731 0.000
## .Q18_6 0.782 0.067 11.655 0.000
## .Q18_7 0.510 0.044 11.707 0.000
## .Q18_8 0.134 0.030 4.500 0.000
## .Q18_9 0.203 0.031 6.484 0.000
## access_vision 0.855 0.119 7.209 0.000
## personal 0.508 0.071 7.149 0.000Items:
- access: Access to educational technologies (1)
- access: Time to try out and use educational technologies (2)
- access: Technical and instructional design support to use educational technologies (3)
- vision: Leadership around my department / program’s use of educational technology (4)
- vision: A clear vision or plan related to my department / program’s use of educational technology (5)
- CUT: The role of technology in the subject matter(s) I focus on (6)
- personal: My beliefs about the role of educational technologies in teaching and learning (7)
- personal: Confidence in my ability to use educational technology (8)
- personal: Confidence in my ability to integrate educational technologies with my own pedagogical strategies (9)
dx <- ds[complete.cases(dd), ] %>%
mutate(id = 1:nrow(.)) %>%
gather(key, val, -id) %>%
group_by(id) %>%
summarize(mean_tetc = mean(val)) %>%
cbind(fs) %>%
as_tibble()
dx %>%
corrr::correlate() %>%
slice(3:5) %>%
select(-id)##
## Correlation method: 'pearson'
## Missing treated using: 'pairwise.complete.obs'## # A tibble: 3 x 5
## rowname mean_tetc access vision personal
## <chr> <dbl> <dbl> <dbl> <dbl>
## 1 access 0.429 NA 0.707 0.629
## 2 vision 0.244 0.707 NA 0.299
## 3 personal 0.614 0.629 0.299 NAdx %>%
as.matrix() %>%
Hmisc::rcorr()## id mean_tetc access vision personal
## id 1.00 -0.02 -0.01 -0.04 0.03
## mean_tetc -0.02 1.00 0.43 0.24 0.61
## access -0.01 0.43 1.00 0.71 0.63
## vision -0.04 0.24 0.71 1.00 0.30
## personal 0.03 0.61 0.63 0.30 1.00
##
## n= 328
##
##
## P
## id mean_tetc access vision personal
## id 0.6838 0.8301 0.4769 0.6307
## mean_tetc 0.6838 0.0000 0.0000 0.0000
## access 0.8301 0.0000 0.0000 0.0000
## vision 0.4769 0.0000 0.0000 0.0000
## personal 0.6307 0.0000 0.0000 0.0000Demographics - RQ3
Grade level
m <- read_csv("grade-levels-m.csv")## Parsed with column specification:
## cols(
## Q7 = col_character(),
## n = col_double(),
## X3 = col_character()
## )m <- m %>% set_names(c("Q7", "Q7n", "Q7_code"))
dg <- d %>%
left_join(m, by = "Q7") %>%
filter(complete.cases(dd)) %>%
cbind(dx) %>%
as_tibble()
summary(aov(access ~ Q7_code, data = dg))## Df Sum Sq Mean Sq F value Pr(>F)
## Q7_code 3 1.83 0.6116 0.794 0.498
## Residuals 324 249.53 0.7702summary(aov(vision ~ Q7_code, data = dg))## Df Sum Sq Mean Sq F value Pr(>F)
## Q7_code 3 3.03 1.0107 1.164 0.323
## Residuals 324 281.27 0.8681summary(aov(personal ~ Q7_code, data = dg))## Df Sum Sq Mean Sq F value Pr(>F)
## Q7_code 3 5.49 1.8309 4.035 0.00772 **
## Residuals 324 147.01 0.4537
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1TukeyHSD(aov(personal ~ Q7_code, data = dg))## Tukey multiple comparisons of means
## 95% family-wise confidence level
##
## Fit: aov(formula = personal ~ Q7_code, data = dg)
##
## $Q7_code
## diff lwr upr p adj
## Elem-All -0.28563405 -0.5536261 -0.01764203 0.0315911
## Other-All -0.08059171 -0.3779460 0.21676256 0.8970633
## Sec-All -0.26980336 -0.5181486 -0.02145809 0.0271995
## Other-Elem 0.20504234 -0.1427244 0.55280907 0.4251311
## Sec-Elem 0.01583069 -0.2910854 0.32274682 0.9991549
## Sec-Other -0.18921165 -0.5220740 0.14365071 0.4581361dg %>%
select(Q7_code, access, vision, personal) %>%
gather(key, val, -Q7_code) %>%
group_by(Q7_code, key) %>%
summarize(mean_val = mean(val)) %>%
ggplot(aes(x = key, y = mean_val, fill = Q7_code)) +
geom_col(position = 'dodge')
Content area/subject
m <- read_csv("subject-m.csv")## Parsed with column specification:
## cols(
## Q8 = col_character(),
## nn = col_double(),
## code = col_character()
## )m <- m %>% set_names(c("Q8", "Q8n", "Q8_code")) %>% mutate(Q8_code = ifelse(Q8_code == "STEM", "Science and Math", Q8_code))
dg <- d %>%
left_join(m, by = "Q8") %>%
filter(complete.cases(dd)) %>%
cbind(dx) %>%
as_tibble()
summary(aov(access ~ Q8_code, data = dg))## Df Sum Sq Mean Sq F value Pr(>F)
## Q8_code 3 7.25 2.418 3.241 0.0225 *
## Residuals 280 208.89 0.746
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 44 observations deleted due to missingnessTukeyHSD(aov(access ~ Q8_code, data = dg))## Tukey multiple comparisons of means
## 95% family-wise confidence level
##
## Fit: aov(formula = access ~ Q8_code, data = dg)
##
## $Q8_code
## diff lwr upr p adj
## Other-Non-STEM 0.2663706 -0.07798734 0.6107285 0.1907295
## Science and Math-Non-STEM 0.1639589 -0.18428961 0.5122074 0.6166822
## Technology-Non-STEM 0.4773361 0.05953710 0.8951351 0.0178631
## Science and Math-Other -0.1024117 -0.46824336 0.2634199 0.8876558
## Technology-Other 0.2109655 -0.22159869 0.6435297 0.5888775
## Technology-Science and Math 0.3133772 -0.12229054 0.7490450 0.2482323summary(aov(vision ~ Q8_code, data = dg))## Df Sum Sq Mean Sq F value Pr(>F)
## Q8_code 3 2.82 0.9389 1.084 0.356
## Residuals 280 242.48 0.8660
## 44 observations deleted due to missingnesssummary(aov(personal ~ Q8_code, data = dg))## Df Sum Sq Mean Sq F value Pr(>F)
## Q8_code 3 8.38 2.7939 6.353 0.000352 ***
## Residuals 280 123.14 0.4398
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 44 observations deleted due to missingnessTukeyHSD(aov(personal ~ Q8_code, data = dg))## Tukey multiple comparisons of means
## 95% family-wise confidence level
##
## Fit: aov(formula = personal ~ Q8_code, data = dg)
##
## $Q8_code
## diff lwr upr p adj
## Other-Non-STEM 0.22940442 -0.03498935 0.4937982 0.1144596
## Science and Math-Non-STEM 0.24597695 -0.02140394 0.5133578 0.0837736
## Technology-Non-STEM 0.52766367 0.20688276 0.8484446 0.0001693
## Science and Math-Other 0.01657253 -0.26430849 0.2974535 0.9987341
## Technology-Other 0.29825925 -0.03385820 0.6303767 0.0957032
## Technology-Science and Math 0.28168672 -0.05281361 0.6161871 0.1323431dg %>%
select(Q8_code, access, vision, personal) %>%
gather(key, val, -Q8_code) %>%
group_by(Q8_code, key) %>%
summarize(mean_val = mean(val)) %>%
ggplot(aes(x = key, y = mean_val, fill = Q8_code)) +
geom_col(position = 'dodge')