Visualisation

library(retention.helpers)
library(tidyverse)
#> Warning: package 'tidyverse' was built under R version 4.2.2
#> ── Attaching packages ─────────────────────────────────────── tidyverse 1.3.2 ──
#> ✔ ggplot2 3.4.0      ✔ purrr   1.0.0 
#> ✔ tibble  3.1.8      ✔ dplyr   1.0.10
#> ✔ tidyr   1.2.1      ✔ stringr 1.5.0 
#> ✔ readr   2.1.3      ✔ forcats 0.5.2
#> Warning: package 'ggplot2' was built under R version 4.2.2
#> Warning: package 'tibble' was built under R version 4.2.2
#> Warning: package 'tidyr' was built under R version 4.2.2
#> Warning: package 'readr' was built under R version 4.2.2
#> Warning: package 'purrr' was built under R version 4.2.2
#> Warning: package 'dplyr' was built under R version 4.2.2
#> Warning: package 'stringr' was built under R version 4.2.2
#> Warning: package 'forcats' was built under R version 4.2.2
#> ── Conflicts ────────────────────────────────────────── tidyverse_conflicts() ──
#> ✖ dplyr::filter() masks stats::filter()
#> ✖ dplyr::lag()    masks stats::lag()

Renaming relevant toy data to names from the data.csu.retention data package.

academic <- toy_academic_for_interventions
flags <- toy_flags
interventions <- toy_interventions

Visualising a grade distribution

We may like to display grade distributions of the academic data. The data is already set up in factor levels for the major grades, and there are also colour palettes to use; csu_colours, csu_colours_dark and csu_colours_light.

Basic grade distribution in Charles Sturt colours

academic |> 
    add_grade_helpers() |> 
    filter(grade_substantive) |> 
    ggplot(aes(x = grade, fill = grade_success)) +
    geom_histogram(stat = "count") +
    theme_minimal() +
    scale_fill_manual(values = csu_colours, guide = NULL)
#> Warning in geom_histogram(stat = "count"): Ignoring unknown parameters:
#> `binwidth`, `bins`, and `pad`

We may also want to compare grade distributions of different groups. For instance, the table flags and interventions include data on students flagged for some concern in the various campaigns, and the immediate results of the intervention. The toy data included in this package is a simulated sample based on a pre-census campaign were students are called if they are flagged as ‘at-risk’. Some answer the phone call, and some do not.

academic |> 
    inner_join( # only including subjects from the flags list
        flags |> 
            distinct(subject)
    ) |> 
    left_join(
        flags |> 
            mutate(flagged = TRUE)
    ) |> 
    mutate(flagged = replace_na(flagged, FALSE)) |> 
    mutate(grp = if_else(flagged, "Low engagement", "Engaged")) |> 
    add_grade_helpers() |> 
    filter(grade_substantive) |> 
    ggplot(aes(x = grade, fill = grade_success)) +
    geom_histogram(stat = "count") +
    theme_minimal() +
    scale_fill_manual(values = csu_colours, guide = NULL) +
    facet_wrap(~grp, scales = "free_y")
#> Joining, by = "subject"
#> Joining, by = c("id", "subject")
#> Warning in geom_histogram(stat = "count"): Ignoring unknown parameters:
#> `binwidth`, `bins`, and `pad`

academic |> 
    inner_join( # only including subjects from the flags list
        flags |> 
            distinct(subject)
    ) |> 
    left_join(
        flags |> 
            mutate(flagged = TRUE)
    ) |> 
    mutate(flagged = replace_na(flagged, FALSE)) |> 
    left_join(interventions) |> 
    mutate(grp = case_when(
        !flagged ~ "Engaged",
        intervention_result == "dialogue" ~ "Spoke with disengaged student",
        intervention_result == "no dialogue" ~ "Disengaged student did not answer") |> 
            fct_relevel("Engaged", "Spoke with disengaged student")) |> 
    add_grade_helpers() |> 
    filter(grade_substantive) |> 
    ggplot(aes(x = grade, fill = grp, alpha = grade_success)) +
    geom_histogram(stat = "count") +
    theme_minimal() +
    scale_fill_manual(values = csu_colours, guide = NULL) +
    scale_alpha_discrete(range = c(1, 0.5), guide = NULL) +
    facet_wrap(~grp, scales = "free_y")
#> Joining, by = "subject"
#> Joining, by = c("id", "subject")
#> Joining, by = "id"
#> Warning in geom_histogram(stat = "count"): Ignoring unknown parameters:
#> `binwidth`, `bins`, and `pad`
#> Warning: Using alpha for a discrete variable is not advised.

Aggregating grades for visualisation

The interventions (for the campaign in the toy data at least) are performed at a student-session level, but grades are reported at a student-session-subject level. As such it can be useful to aggregate grades from student-session-subject to student-session level.

# aggregating at student-session progress rate
academic_summary_1 <- 
    academic |> 
    mutate(session = "Session 1") |> # normally session would be included
    group_by(id, session) |>
    add_grade_helpers() |> 
    summarise(
        progress_rate = sum(grade_success) / sum(grade_substantive),
        gpa = gpa(grade)
    ) |> 
    ungroup()
#> `summarise()` has grouped output by 'id'. You can override using the `.groups`
#> argument.
    
academic_summary_1 |> 
    left_join(interventions) |>
    mutate(grp = case_when(
        is.na(intervention_result)  ~ "Engaged",
        intervention_result == "dialogue" ~ "Spoke with disengaged student",
        intervention_result == "no dialogue" ~ "Disengaged student did not answer") |> 
            fct_relevel("Engaged", "Spoke with disengaged student"))|> 
    ggplot(aes(x = gpa, colour = grp)) +
    stat_ecdf(size = 2, alpha = 0.8) +
    scale_y_continuous(labels = scales::percent_format(),
                       name = "Cumulative distribution") +
    theme_minimal() +
    scale_color_manual(values = csu_colours)
#> Joining, by = "id"
#> Warning: Using `size` aesthetic for lines was deprecated in ggplot2 3.4.0.
#> ℹ Please use `linewidth` instead.
#> Warning: Removed 333 rows containing non-finite values (`stat_ecdf()`).

# aggregating by the custom `summarise_academic` function
academic_summary_2 <- 
    academic |> 
    mutate(session = "Session 1") |> # normally session would be included
    group_by(id, session) |>
    summarise_academic()  
    
glimpse(academic_summary_2)
#> Rows: 1,876
#> Columns: 4
#> $ id          <chr> "Student10", "Student1001", "Student1003", "Student1005", …
#> $ session     <chr> "Session 1", "Session 1", "Session 1", "Session 1", "Sessi…
#> $ result_long <fct> Pass all, Pass all, Pass all, Passing (some non-zero fails…
#> $ result      <fct> Pass, Pass, Pass, Pass, Pass, Pass, Pass, Pass, Pass, Non-…

academic_summary_2 |> 
    left_join(interventions) |>
    mutate(grp = case_when(
        is.na(intervention_result)  ~ "Engaged",
        intervention_result == "dialogue" ~ "Spoke with disengaged student",
        intervention_result == "no dialogue" ~ "Disengaged student did not answer") |> 
            fct_relevel("Engaged", "Spoke with disengaged student"))|> 
    ggplot(aes(x = 1, fill = result)) +
    geom_bar(stat = "count", position = "stack") +
    facet_wrap(~grp, scales = "free_y") +
    theme_minimal() +
    theme(axis.text.x = element_blank(), axis.title.x = element_blank()) +
    scale_fill_manual(values = csu_colours)
#> Joining, by = "id"

Visualising a grade distribution

Basic grade distribution in Charles Sturt colours

Using facets to compare grade distributions

Aggregating grades for visualisation