Teaching computing via visualization

The following is not a particularly controversial statement: "In order to create a data visualization using [software/language], you first need to learn that [software/language]." But if one of the more exciting things you can do with [software/language] is data visualization, why not start learning [software/language] by learning to build data visualizations? In this talk we present a data-centric approach to teaching and learning R through creating data visualizations, as opposed to starting with fundamentals of R as a programming language. The talk will feature examples of in class activities, details of a curriculum that introduces students to data science through data visualization, and sample student work.

Transcript

1. Teaching computing
   via visualization
   mine-cetinkaya-rundel
   mine@stat.duke.edu
   @minebocek
   bit.ly/teach-viz-comp
   

   View Slide

2. In order to create a data
   visualization using __________,
   you first need to learn __________.
   [software/language]
   [software/language]
   bit.ly/teach-viz-comp · @minebocek
   

   View Slide

3. One way to learn __________,
   is to create data visualizations
   using __________.
   [software/language]
   [software/language]
   bit.ly/teach-viz-comp · @minebocek
   

   View Slide

4. [software/language]
   bit.ly/teach-viz-comp · @minebocek
   

   View Slide

5. In which of the following two
   scenarios do you feel like you
   have a better handle on the
   final product?
   Q
   bit.ly/teach-viz-comp · @minebocek
   

   View Slide

6. View Slide

7. View Slide

8. (a) (b)
   

   View Slide

9. start
   with
   cake
   1
   

   View Slide

10. Which of the following two
    examples is more likely to be
    interesting for a wide range of
    students?
    Q
    bit.ly/teach-viz-comp · @minebocek
    

    View Slide

11. # Declare variables
    # of different types
    x !<- 8
    y !<- "monkey"
    z !<- FALSE
    # Check class of x
    class(x)
    #> [1] "numeric"
    # Check class of y
    class(y)
    #> [1] "character"
    # Check class of z
    class(z)
    #> [1] "logical"
    Declare the following variables
    Then, determine the class of each
    variable
    (a)
    bit.ly/teach-viz-comp · @minebocek
    

    View Slide

12. (b) Open today’s example project
    Knit the document and discuss the results with your neighbor
    Then, change Turkey to a different country, and plot again
    bit.ly/teach-viz-comp · @minebocek
    

    View Slide

13. # Declare variables
    # of different types
    x !<- 8
    y !<- "monkey"
    z !<- FALSE
    # Check class of x
    class(x)
    #> [1] "numeric"
    # Check class of y
    class(y)
    #> [1] "character"
    # Check class of z
    class(z)
    #> [1] "logical"
    (a) (b)
    bit.ly/teach-viz-comp · @minebocek
    

    View Slide

14. with great examples,
    comes a great amount of code…
    bit.ly/teach-viz-comp · @minebocek
    

    View Slide

15. but let’s focus on the task at hand…
    Open today’s example project
    Knit the document and discuss the results with your neighbor
    Then, change Turkey to a different country, and plot again
    bit.ly/teach-viz-comp · @minebocek
    

    View Slide

16. bit.ly/teach-viz-comp · @minebocek
    un_votes %>%
    filter(country %in% c("United States of America", "Turkey")) %>%
    inner_join(un_roll_calls, by = "rcid") %>%
    inner_join(un_roll_call_issues, by = "rcid") %>%
    group_by(country, year = year(date), issue) %>%
    summarize(
    votes = n(),
    percent_yes = mean(vote !== "yes")
    ) %>%
    filter(votes > 5) %>% # only use records where there are more than 5 votes
    ggplot(mapping = aes(x = year, y = percent_yes, color = country)) +
    geom_point() +
    geom_smooth(method = "loess", se = FALSE) +
    facet_wrap(~ issue) +
    labs(
    title = "Percentage of 'Yes' votes in the UN General Assembly",
    subtitle = "1946 to 2015",
    y = "% Yes",
    x = "Year",
    color = "Country"
    )
    

    View Slide

17. bit.ly/teach-viz-comp · @minebocek
    un_votes %>%
    filter(country %in% c("United States of America", "Turkey")) %>%
    inner_join(un_roll_calls, by = "rcid") %>%
    inner_join(un_roll_call_issues, by = "rcid") %>%
    group_by(country, year = year(date), issue) %>%
    summarize(
    votes = n(),
    percent_yes = mean(vote !== "yes")
    ) %>%
    filter(votes > 5) %>% # only use records where there are more than 5 votes
    ggplot(mapping = aes(x = year, y = percent_yes, color = country)) +
    geom_point() +
    geom_smooth(method = "loess", se = FALSE) +
    facet_wrap(~ issue) +
    labs(
    title = "Percentage of 'Yes' votes in the UN General Assembly",
    subtitle = "1946 to 2015",
    y = "% Yes",
    x = "Year",
    color = "Country"
    )
    

    View Slide

18. bit.ly/teach-viz-comp · @minebocek
    

    View Slide

19. Source: edx.org/course/introduction-r-data-science-1
    bit.ly/teach-viz-comp · @minebocek
    

    View Slide

20. Source: www.coursera.org/specializations/jhu-data-science#courses
    bit.ly/teach-viz-comp · @minebocek
    

    View Slide

21. Source: datasciencebox.org/hello/topics
    Making
    rigorous
    conclusions
    Looking
    forward
    Fundamentals of
    data & data viz,
    revision exercises,
    confounding
    variables and
    Simpson’s paradox
    Rmd + Git + GitHub
    Tidy data, data
    frames vs. summary
    tables, recoding and
    transforming
    variables, web
    scraping and
    iteration
    Building and
    selecting models,
    visualizing
    interactions,
    prediction and model
    validity, inference via
    simulation &
    discussion of CLT
    Interactive
    visualization and
    reporting,
    Bayesian inference,
    text analysis,
    …
    Exploring
    data
    Visualize Wrangle
    bit.ly/teach-viz-comp · @minebocek
    

    View Slide

22. skip
    baby
    steps
    2
    

    View Slide

23. Which of the following two
    visualizations is more likely to be
    motivating for a wide range of
    students?
    Q
    bit.ly/teach-viz-comp · @minebocek
    

    View Slide

24. ggplot(data = un_roll_calls, mapping = aes(x = amend)) +
    geom_bar()
    (a)
    bit.ly/teach-viz-comp · @minebocek
    

    View Slide

25. ggplot(data = un_votes_joined,
    mapping = aes(x = year, y = percent_yes, color = country)) +
    geom_point() +
    geom_smooth(method = "loess", se = FALSE) +
    facet_wrap(~ issue) +
    labs(
    title = "Percentage of 'Yes' votes in the UN General Assembly",
    subtitle = "1946 to 2015",
    y = "% Yes",
    x = "Year",
    color = "Country"
    )
    (b)
    bit.ly/teach-viz-comp · @minebocek
    

    View Slide

26. (a) (b)
    bit.ly/teach-viz-comp · @minebocek
    

    View Slide

27. non-trivial examples can be motivating,
    but need to avoid !
    bit.ly/teach-viz-comp · @minebocek
    

    View Slide

28. ggplot(data = un_votes_joined,
    mapping = aes(x = year, y = percent_yes))
    bit.ly/teach-viz-comp · @minebocek
    

    View Slide

29. ggplot(data = un_votes_joined,
    mapping = aes(x = year, y = percent_yes))
    bit.ly/teach-viz-comp · @minebocek
    function( arguments )
    often a verb what to apply that
    Verb to
    

    View Slide

30. ggplot(data = un_votes_joined,
    mapping = aes(x = year, y = percent_yes))
    bit.ly/teach-viz-comp · @minebocek
    rows =
    observations
    columns =
    variables
    “tidy”
    data frame
    

    View Slide

31. ggplot(data = un_votes_joined,
    mapping = aes(x = year, y = percent_yes)) +
    geom_point()
    bit.ly/teach-viz-comp · @minebocek
    

    View Slide

32. ggplot(data = un_votes_joined,
    mapping = aes(x = year, y = percent_yes, color = country)) +
    geom_point()
    bit.ly/teach-viz-comp · @minebocek
    

    View Slide

33. ggplot(data = un_votes_joined,
    mapping = aes(x = year, y = percent_yes, color = country)) +
    geom_point() +
    geom_smooth(method = "loess", se = FALSE)
    bit.ly/teach-viz-comp · @minebocek
    

    View Slide

34. ggplot(data = un_votes_joined,
    mapping = aes(x = year, y = percent_yes, color = country)) +
    geom_point() +
    geom_smooth(method = "loess", se = FALSE) +
    facet_wrap(~ issue)
    bit.ly/teach-viz-comp · @minebocek
    

    View Slide

35. ggplot(data = un_votes_joined,
    mapping = aes(x = year, y = percent_yes, color = country)) +
    geom_point() +
    geom_smooth(method = "loess", se = FALSE) +
    facet_wrap(~ issue) +
    labs(
    title = "Percentage of 'Yes' votes in the UN General Assembly",
    subtitle = "1946 to 2015",
    y = "% Yes",
    x = "Year",
    color = "Country"
    )
    bit.ly/teach-viz-comp · @minebocek
    

    View Slide

36. cherish
    day
    one
    3
    

    View Slide

37. Which of the following two
    tasks is more likely to be
    welcoming for a wide range of
    students?
    Q
    bit.ly/teach-viz-comp · @minebocek
    

    View Slide

38. (a) Install R
    Install RStudio
    Install the following
    packages:
    tidyverse
    rmarkdown
    …
    Load these packages
    Install git
    (b) Go to vm-
    manage.oit.duke.edu or
    rstudio.cloud (or some other
    server based solution)
    Log in with your ID & pass
    > hello R!
    bit.ly/teach-viz-comp · @minebocek
    

    View Slide

39. method of delivery,
    and medium of interaction matters
    bit.ly/teach-viz-comp · @minebocek
    

    View Slide

40. → →
    →
    →
    bit.ly/teach-viz-comp · @minebocek
    

    View Slide

41. tl;dr
    bit.ly/teach-viz-comp · @minebocek
    

    View Slide

42. 1
    2
    3
    start with cake
    skip baby steps
    cherish day one
    

    View Slide

43. Why = ?
    Q
    bit.ly/teach-viz-comp · @minebocek
    

    View Slide

44. Let them eat cake (first)!
    You can tell them all about the ingredients later…
    mine-cetinkaya-rundel
    mine@stat.duke.edu
    @minebocek
    bit.ly/teach-viz-comp
    

    View Slide