Analysis of Data Science Workshop Pre and Post Survey Responses
Loading, Setting Up
library(tidyverse)
library(janitor)Finding out how many people responded to each survey
pre <- read_csv("pre-survey.csv")
post <- read_csv("post-survey.csv")
pre <- pre %>%
filter(Progress == 100) %>%
clean_names() # pre n = 27
post <- post %>%
filter(Progress > 25) %>%
clean_names() # two who completed 69%; post n = 21; 19 complete responses
pre_id <- pre$q23
post_id <- post$q23[nchar(post$q23) >= 6 & !is.na(post$q23)] # this filters someone who typed 5 and this filters NA responses
pre_id## [1] "2eajon" "2ortro" "0masha" "2gryor" "1spban" "3orfor" "1knweo"
## [8] "1grhal" "2carea" "2sutim" "2mocum" "1grpin" "0206may" "3robr"
## [15] "1kibev" "9wiand" "0chlon" "1coyor" "2aista" "0miwyc" "2kisha"
## [22] "3anlan" "1ausil" "2basan" "1auowe" "17grbib" "2dafai"post_id## [1] "0masha" "1kibev" "2gryor" "0miawyc" "2aista" "1grhal" "1ausil"
## [8] "1spban" "2basan" "3carea" "1knweo" "9wiand" "2eajon" "2ortro"
## [15] "1AuOwe" "may0206" "1coyor" "2kisha"length(pre_id)## [1] 27length(post_id)## [1] 18c(pre_id, post_id) %>%
unique() %>%
length()## [1] 31pre_id %in% post_id %>% sum() # 14 of the respondents took both## [1] 14# 13 pre-survey respondents didn't take post
# 4 post-survey repsondents didn't take pre
pre$took_both_surveys <- pre$q23 %in% post$q23
post$took_post_surveys <- post$q23 %in% pre$q23Motivation for Attending Carpentries Workshops
pre %>%
select(contains("q10_")) %>%
select(-q10_7_text) %>%
pivot_longer(q10_1:q10_7) %>%
filter(!is.na(value)) %>%
count(name, value) %>%
arrange(desc(n)) %>%
mutate(prop = n / nrow(pre)) %>%
mutate(prop = prop * 100) %>%
mutate_if(is.numeric, round, 1) %>%
select(-name) %>%
set_names("Reason for Participating", "n", "%")## # A tibble: 7 x 3
## `Reason for Participating` n `%`
## <chr> <dbl> <dbl>
## 1 To learn new skills. 24 88.9
## 2 As a requirement for my degree program or current position. 23 85.2
## 3 To learn skills that I can apply to my current work. 22 81.5
## 4 To learn skills that I can apply to my work in the future. 20 74.1
## 5 To refresh or review my skills. 8 29.6
## 6 To learn skills that will help me get a job or a promotion. 8 29.6
## 7 Other (please describe) 1 3.7freq_use <- pre %>%
select(freq_use_1:freq_use_3) %>%
set_names(c("Statistical", "Spreadsheet", "Programming")) %>%
pivot_longer(Statistical:Programming) %>%
count(name, value) %>%
filter(!is.na(value)) %>%
spread(value, n, fill = 0)
freq_use %>%
mutate(name = factor(name, levels = c("Spreadsheet", "Statistical", "Programming"))) %>%
gather(key, val, -name) %>%
group_by(name) %>%
mutate(val = val/sum(val),
val = val * 100) %>%
ggplot(aes(x = key, y = val, color = val)) +
geom_point() +
theme_bw() +
scale_x_discrete("Frequency of Use", limits = rev(c("Daily", "Weekly", "Monthly", "Several times per year", "Less than once per year", "Never"))) +
ylab("% of Participants") +
xlab(NULL) +
facet_wrap(~name) +
coord_flip() +
theme(legend.position = "none",
text = element_text(size = 14, family = "Times"))
ggsave("use-of-software.png", width = 9, height = 5)prior experience - qual
pre$q18 %>%
knitr::kable()qual - concerns you have
pre$q12 %>%
knitr::kable()Workshop Type and Perception of Workshop Environment/Experience
experience <- post %>%
select(q21_1:q21_4, q11_4) %>%
set_names(c("Questions Were Answered",
"Comfortable With Instructor",
"Instructor Was Enthusiastic",
"Instructor Was Knowledgable",
"Comfortable Learning")) %>%
gather(name, value) %>%
mutate(value = case_when(
value == "Strongly disagree" ~ 1,
value == "Disagree" ~ 2,
value == "Somewhat disagree" ~ 2,
value == "Neither agree nor disagree" ~ 3,
value == "Somewhat agree" ~ 5,
value == "Agree" ~ 6,
value == "Strongly agree" ~ 7,
)) %>%
group_by(name) %>%
summarize(mean_val = mean(value, na.rm = TRUE),
sd_val = sd(value, na.rm = TRUE)/sqrt(21-1))
experience %>%
ggplot(aes(x = reorder(name, mean_val), y = mean_val)) +
geom_col() +
# geom_errorbar(aes(ymin = mean_val - sd_val,
# ymax = mean_val + sd_val))
theme_bw() +
ylab("Mean Response") +
xlab(NULL) +
coord_flip() +
theme(legend.position = "none",
text = element_text(size = 14, family = "Times"))
ggsave("experience-in-workshop.png", width = 6, height = 4)major strengths
post$q30 %>%
knitr::kable()ways session could be improved
post$q31 %>%
knitr::kable()effects of instructor
post$q32 %>%
knitr::kable()Effect of Workshops on Learners’ Self-Reported Perspectives, Skills, and Confidence
effects_pre <- pre %>%
select(q11_1:q11_3) %>%
set_names(c("Overcome Getting Stuck", "Search for Answers", "Confidently Work With Data")) %>%
gather(name, value) %>%
mutate(value = case_when(
value == "Strongly disagree" ~ 1,
value == "Disagree" ~ 2,
value == "Somewhat disagree" ~ 2,
value == "Neither agree nor disagree" ~ 3,
value == "Somewhat agree" ~ 5,
value == "Agree" ~ 6,
value == "Strongly agree" ~ 7,
)) %>%
group_by(name) %>%
summarize(mean_val = mean(value, na.rm = TRUE),
sd_val = sd(value, na.rm = TRUE)/sqrt(27-1))
effects_post <- post %>%
select(q11_1:q11_3) %>%
set_names(c("Overcome Getting Stuck", "Search for Answers", "Confidently Work With Data")) %>%
gather(name, value) %>%
mutate(value = case_when(
value == "Strongly disagree" ~ 1,
value == "Disagree" ~ 2,
value == "Somewhat disagree" ~ 2,
value == "Neither agree nor disagree" ~ 3,
value == "Somewhat agree" ~ 5,
value == "Agree" ~ 6,
value == "Strongly agree" ~ 7,
)) %>%
group_by(name) %>%
summarize(mean_val = mean(value, na.rm = TRUE),
sd_val = sd(value, na.rm = TRUE)/sqrt(21-1))
effects_pre <- effects_pre %>%
mutate(time = "pre")
effects_post <- effects_post %>%
mutate(time = "post")
bind_rows(effects_pre, effects_post) %>%
mutate(time = factor(time, levels = c("pre", "post"))) %>%
ggplot(aes(x = reorder(name, mean_val), y = mean_val, fill = time)) +
geom_col(position = "dodge") +
geom_errorbar(aes(ymin = mean_val - sd_val,
ymax = mean_val + sd_val),
position = position_dodge(width = 0.9)) +
theme_bw() +
ylab("Mean Response") +
xlab(NULL) +
coord_flip() +
theme(legend.position = "none",
text = element_text(size = 14, family = "Times")) +
ylim(0, 7) +
scale_x_discrete("", limits = c("Overcome Getting Stuck", "Search for Answers", "Confidently Work With Data"))
ggsave("change-in-effects.png", width = 6, height = 4)t-test
pre_effects_to_join <- pre %>%
select(q11_1:q11_3, id = q23) %>%
set_names(c("Overcome Getting Stuck", "Search for Answers", "Confidently Work With Data", "id")) %>%
gather(name, value, -id) %>%
mutate(value = case_when(
value == "Strongly disagree" ~ 1,
value == "Disagree" ~ 2,
value == "Somewhat disagree" ~ 2,
value == "Neither agree nor disagree" ~ 3,
value == "Somewhat agree" ~ 5,
value == "Agree" ~ 6,
value == "Strongly agree" ~ 7,
))
post_effects_to_join <- post %>%
select(q11_1:q11_3, id = q23) %>%
set_names(c("Overcome Getting Stuck", "Search for Answers", "Confidently Work With Data", "id")) %>%
gather(name, value, -id) %>%
mutate(value = case_when(
value == "Strongly disagree" ~ 1,
value == "Disagree" ~ 2,
value == "Somewhat disagree" ~ 2,
value == "Neither agree nor disagree" ~ 3,
value == "Somewhat agree" ~ 5,
value == "Agree" ~ 6,
value == "Strongly agree" ~ 7,
)) %>%
rename(post_value = value)
for_t_test <- left_join(pre_effects_to_join, post_effects_to_join)
for_t_test <- for_t_test %>%
select(-id) %>%
gather(key, val, -name)
t.test(val ~ key, filter(for_t_test, name == "Overcome Getting Stuck"))##
## Welch Two Sample t-test
##
## data: val by key
## t = 0.31031, df = 23.85, p-value = 0.759
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
## -0.9870958 1.3363022
## sample estimates:
## mean in group post_value mean in group value
## 5.285714 5.111111t.test(val ~ key, filter(for_t_test, name == "Search for Answers"))##
## Welch Two Sample t-test
##
## data: val by key
## t = 2.0097, df = 26.987, p-value = 0.05456
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
## -0.01859124 1.79107802
## sample estimates:
## mean in group post_value mean in group value
## 6.071429 5.185185t.test(val ~ key, filter(for_t_test, name == "Confidently Work With Data"))##
## Welch Two Sample t-test
##
## data: val by key
## t = 2.5518, df = 34.487, p-value = 0.01531
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
## 0.2817322 2.4801725
## sample estimates:
## mean in group post_value mean in group value
## 4.714286 3.333333Ability to Perform Computing Tasks
immediately apply what was learned
post %>%
select(q11_4) %>%
mutate(q11_4 = case_when(
q11_4 == "Strongly disagree" ~ 1,
q11_4 == "Disagree" ~ 2,
q11_4 == "Somewhat disagree" ~ 2,
q11_4 == "Neither agree nor disagree" ~ 3,
q11_4 == "Somewhat agree" ~ 5,
q11_4 == "Agree" ~ 6,
q11_4 == "Strongly agree" ~ 7,
)) %>%
set_names("apply") %>%
summarize(mean_val = mean(apply, na.rm = TRUE),
sd_val = sd(apply, na.rm = TRUE)/sqrt(21-1))## # A tibble: 1 x 2
## mean_val sd_val
## <dbl> <dbl>
## 1 5.84 0.271Accessibility requirements and issues
post %>%
count(q27)## # A tibble: 3 x 2
## q27 n
## <chr> <int>
## 1 No 13
## 2 Yes 7
## 3 <NA> 1post %>%
count(q28)## # A tibble: 2 x 2
## q28 n
## <chr> <int>
## 1 No 7
## 2 <NA> 14Other comments or feedback
post$q7 %>%
knitr::kable()