report.qmd

---
title: "R Familiarity Survey Report"
format: 
  html:
    df-print: paged
    toc: true
execute: 
  echo: false
  warning: false
  message: false
editor_options: 
  chunk_output_type: console
---


```{r}
# Load packages

library(tidyverse)
library(googlesheets4)
library(janitor)
library(lubridate)
library(tigris)
library(hrbrthemes)
library(gendercoder)
library(tidygeocoder)
library(leaflet)
library(scales)

gs4_auth("david@rfortherestofus.com")
```




```{r}
# Set plotting default theme
theme_set(
  theme_ipsum(
    base_family = "Inter",
    axis_title_family = "Inter"
  ) +
    theme(panel.grid.major.y = element_blank())
)
```


```{r}
# Import our data
survey_responses_today <-
  read_sheet("https://docs.google.com/spreadsheets/d/1lTjv7EYyGPKzjekBJ29xRqCUxhrMT0nuO4gsW6qEb9s/edit#gid=1882477504") |>
  # Make the variable names easy to work with
  clean_names() |>
  # Convert the timestamp into a date
  mutate(timestamp = as.Date(timestamp)) |>
  # Filter to only keep responses from today
  filter(timestamp == as_date(now()))
```

# About our Respondents

```{r}
number_of_responses <- nrow(survey_responses_today)

today_month <- month(
  now(),
  label = TRUE,
  abbr = FALSE
) |>
  as.character()

today_day <- day(now()) |>
  as.character()

today_year <- year(now()) |>
  as.character()

today_pretty <- str_glue("{today_month} {today_day}, {today_year}")
```

We conducted a survey on `r today_pretty`. We received responses from `r number_of_responses` people. 

## Age

```{r}
mean_age <-
  survey_responses_today |>
  summarize(mean_age = mean(
    how_old_are_you,
    na.rm = TRUE
  )) |>
  mutate(
    mean_age =
      round_half_up(mean_age, 1)
  ) |>
  pull(mean_age)

max_age <-
  survey_responses_today |>
  summarize(age = max(
    how_old_are_you,
    na.rm = TRUE
  )) |>
  pull(age)

min_age <-
  survey_responses_today |>
  summarize(age = min(
    how_old_are_you,
    na.rm = TRUE
  )) |>
  pull(age)
```

Respondents listed their age. The youngest respondent was `r min_age` years old and the oldest respondent was `r max_age`. The mean age was `r mean_age`.

The graphs below shows the age of respondents broken into groups.

```{r}
#| fig-height: 3
age_categorized <-
  survey_responses_today |>
  mutate(age_category = case_when(
    how_old_are_you < 30 ~ "Less than 30",
    how_old_are_you < 50 ~ "Between 30 and 50",
    how_old_are_you > 50 ~ "50 or older",
  )) |>
  select(how_old_are_you, age_category) |>
  count(age_category) |>
  mutate(
    age_category =
      fct(
        age_category,
        levels = c(
          "Less than 30",
          "Between 30 and 50",
          "50 or older"
        )
      )
  ) |>
  mutate(
    age_category =
      fct_rev(age_category)
  ) |>
  drop_na(age_category)

ggplot(
  age_categorized,
  aes(
    age_category,
    n
  )
) +
  geom_col(fill = "#6CABDD") +
  geom_text(
    aes(label = n),
    hjust = 2,
    color = "white"
  ) +
  coord_flip() +
  scale_y_continuous(breaks = pretty_breaks()) +
  labs(
    title = "Age of Respondents",
    x = NULL,
    y = NULL
  ) +
  theme(
    panel.grid.major.x = element_blank(),
    panel.grid.major.y = element_blank(),
    panel.grid.minor.y = element_blank(),
    panel.grid.minor.x = element_blank(),
    axis.text.x = element_blank()
  )
```

## Education

The education levels of respondents are below.

```{r}
#| fig-height: 4
survey_responses_today |>
  count(what_is_the_highest_level_of_education_you_have_completed) |>
  mutate(
    what_is_the_highest_level_of_education_you_have_completed =
      fct(
        what_is_the_highest_level_of_education_you_have_completed,
        levels = c("Bachelor's degree", "Master's degree", "Doctoral degree", "Other")
      )
  ) |>
  mutate(
    what_is_the_highest_level_of_education_you_have_completed =
      fct_rev(what_is_the_highest_level_of_education_you_have_completed)
  ) |>
  drop_na(what_is_the_highest_level_of_education_you_have_completed) |>
  ggplot(aes(
    what_is_the_highest_level_of_education_you_have_completed,
    n
  )) +
  geom_col(fill = "#6CABDD") +
  geom_text(
    aes(label = n),
    hjust = 2,
    color = "white"
  ) +
  coord_flip() +
  labs(
    title = "Education Levels of Respondents",
    x = NULL,
    y = NULL
  ) +
  scale_y_continuous(breaks = pretty_breaks()) +
  theme(
    panel.grid.major.x = element_blank(),
    panel.grid.major.y = element_blank(),
    panel.grid.minor.y = element_blank(),
    panel.grid.minor.x = element_blank(),
    axis.text.x = element_blank()
  )
```


## Locations

Respondents listed the location of their primary residence. The map below shows their locations. 

```{r}
#| fig-height: 6
respondent_locations <- survey_responses_today |>
  drop_na(what_is_the_city_state_and_country_of_your_primary_residence) |>
  geocode(
    address = what_is_the_city_state_and_country_of_your_primary_residence,
    method = "osm"
  ) |>
  select(what_is_the_city_state_and_country_of_your_primary_residence, long, lat)

leaflet(respondent_locations) |>
  addProviderTiles(providers$CartoDB.Positron) |>
  addCircleMarkers(
    ~long,
    ~lat,
    color = "#FF7400",
    stroke = FALSE,
    fillOpacity = 0.7,
    popup = ~what_is_the_city_state_and_country_of_your_primary_residence,
    clusterOptions = markerClusterOptions(showCoverageOnHover = FALSE)
  )
```

## Gender

Respondents were asked to list their gender, in whatever form they choose to identify it. 

The original responses are below.

```{r}
survey_responses_today |>
  drop_na(what_is_your_gender) |>
  select(what_is_your_gender) |>
  set_names("Gender (Original Response)")
```


We then used the [gendercoder package](https://github.com/ropenscilabs/gendercoder) to recode responses. The table below shows the original responses and the recoded responses.

```{r}
survey_responses_today |>
  drop_na(what_is_your_gender) |>
  select(what_is_your_gender) |>
  mutate(gender_recoded = recode_gender(what_is_your_gender)) |>
  mutate(gender_recoded = str_to_title(gender_recoded)) |>
  replace_na(list(gender_recoded = "Unable to recode")) |>
  set_names("Original Responses", "Recoded Gender")
```

We can then summarize the recoded responses, which we do in the following table.


```{r}
survey_responses_today |>
  drop_na(what_is_your_gender) |>
  select(what_is_your_gender) |>
  mutate(gender_recoded = recode_gender(what_is_your_gender)) |>
  mutate(gender_recoded = str_to_title(gender_recoded)) |>
  replace_na(list(gender_recoded = "Unable to recode")) |>
  count(gender_recoded) |>
  set_names("Recoded Gender", "Number of Responses")
```


# Familiarity with R

```{r}
mean_familiarity <-
  survey_responses_today |>
  summarize(mean_familiarity = mean(
    how_familiar_are_you_with_r,
    na.rm = TRUE
  )) |>
  mutate(
    mean_familiarity =
      round_half_up(
        mean_familiarity,
        digits = 1
      )
  ) |>
  pull(mean_familiarity)
```

On a 5-point scale, respondents listed their level of familiarity with R as `r number(mean_familiarity, .1)`. The figure below shows familiarity broken down by education level.

```{r}
#| fig-height: 4
survey_responses_today |>
  drop_na(what_is_the_highest_level_of_education_you_have_completed) |>
  group_by(what_is_the_highest_level_of_education_you_have_completed) |>
  summarize(mean_familiarity = mean(
    how_familiar_are_you_with_r,
    na.rm = TRUE
  )) |>
  mutate(mean_familiarity = round_half_up(
    mean_familiarity,
    digits = 1
  )) |>
  mutate(
    what_is_the_highest_level_of_education_you_have_completed =
      fct(
        what_is_the_highest_level_of_education_you_have_completed,
        levels = c("Bachelor's degree", "Master's degree", "Doctoral degree", "Other")
      )
  ) |>
  mutate(
    what_is_the_highest_level_of_education_you_have_completed =
      fct_rev(what_is_the_highest_level_of_education_you_have_completed)
  ) |>
  ggplot(aes(
    what_is_the_highest_level_of_education_you_have_completed,
    mean_familiarity
  )) +
  geom_col(fill = "#6CABDD") +
  geom_text(
    aes(label = number(mean_familiarity, .1)),
    hjust = 1.5,
    color = "white"
  ) +
  coord_flip() +
  scale_y_continuous(
    limits = c(0, 5),
    breaks = seq(0, 5, by = 1)
  ) +
  labs(
    title = "Familiarity with R by Education Level",
    x = NULL,
    y = NULL
  ) +
  theme(
    panel.grid.minor.y = element_blank(),
    panel.grid.minor.x = element_blank()
  )
```



# Interest in Learning R

```{r}
mean_interest <-
  survey_responses_today |>
  summarize(mean_interest = mean(
    how_interested_are_you_in_learning_r,
    na.rm = TRUE
  )) |>
  mutate(mean_interest = round_half_up(
    mean_interest,
    digits = 1
  )) |>
  pull(mean_interest)
```

On a 5-point scale, respondents listed their level of interest in learning R as `r number(mean_interest, .1)`.

We can do a similar breakdown, but adding age. With one line of code, we can make small multiples. 

```{r}
r_interest_by_ed_and_age <-
  survey_responses_today |>
  mutate(age_category = case_when(
    how_old_are_you < 30 ~ "Less than 30",
    how_old_are_you < 50 ~ "Between 30 and 50",
    how_old_are_you > 50 ~ "50 or older",
  )) |>
  group_by(
    what_is_the_highest_level_of_education_you_have_completed,
    age_category
  ) |>
  summarize(mean_interest = mean(
    how_interested_are_you_in_learning_r,
    na.rm = TRUE
  )) |>
  ungroup() |>
  mutate(mean_interest = round_half_up(mean_interest, 1)) |>
  mutate(
    what_is_the_highest_level_of_education_you_have_completed =
      fct(
        what_is_the_highest_level_of_education_you_have_completed,
        levels = c("Bachelor's degree", "Master's degree", "Doctoral degree", "Other")
      )
  ) |>
  mutate(age_category = fct(
    age_category,
    levels = c(
      "Less than 30",
      "Between 30 and 50",
      "50 or older"
    )
  )) |>
  mutate(age_category = fct_rev(age_category)) |>
  drop_na(
    age_category,
    what_is_the_highest_level_of_education_you_have_completed
  )
```

```{r}
#| fig-height: 8
ggplot(
  r_interest_by_ed_and_age,
  aes(
    age_category,
    mean_interest,
    fill = what_is_the_highest_level_of_education_you_have_completed
  )
) +
  geom_col() +
  geom_text(
    aes(label = number(mean_interest, .1)),
    hjust = 1.5,
    color = "white"
  ) +
  coord_flip() +
  scale_y_continuous(
    limits = c(0, 5),
    breaks = seq(0, 5, by = 1)
  ) +
  scale_fill_manual(values = c("#6CABDD", "#FF7400", "#FFC659", "red")) +
  labs(
    title = "Interest in Learning R by\nEducation Level and Age",
    x = NULL,
    y = NULL
  ) +
  facet_wrap(
    ~what_is_the_highest_level_of_education_you_have_completed,
    ncol = 1
  ) +
  theme(
    legend.position = "none",
    panel.grid.minor.y = element_blank(),
    panel.grid.minor.x = element_blank()
  )
```