Original


Source: SAP Community Blogs Section (2016).


Objective

The objective of this 3-D pie chart was to display the enrollments in Australian Universities of International students by nationality. This particular visualization is part of a blog written by a university student discussing the impact of international students living in Australia on various aspects. The target audience would be members involved in Australian education systems, board members, students, etc.

The visualisation chosen had the following three main issues:

  • One of the biggest issues this diagram has, is that even though pie chart is not a good visualization technique for large amount of data, using the 3-D feature to get an overview of the number of international students studying by comparing the depth of every slice is very confusing and inefficient. The depth of the slices do not correspond to the actual amount of students enrolled, making the diagram conceal and misinterpret the data.

  • The colors chosen in the chart are very similar across different nationalities. (For e.g: The greens, For China and oter nationalities), which can create confusion as to which slice represents which proportion tha lack of labelling also contirbutes to the ambiguity factor of the diagram.

  • Lastly,the slices (percentages) themselves are uneven to size, and due to the #-D effect, some nationalities appear to be missing. Even though China has the highest amount of students enrolled into Australian universities in 2014, it still has a proportion smaller in size, as opposed to other countires who have far less students studyin in AUstralia. The proportions are uneven, making it impossible to interpret.

Reference

Code

The following code was used to fix the issues identified in the original.

library(ggplot2)
library(readr)
library(tidyr)
library(dplyr)

pdf <- read_csv("data.csv")

df <- data.frame (Nationality = pdf$Nationality, 
                 No_of_students = pdf$`Student cohort (2014)`)
df <- df[-c(12),] 

par(las=2)
plot <-ggplot(df, aes(Nationality, No_of_students,fill= Nationality)) +
    geom_bar(stat = "identity") + coord_flip()
    labs(y = "Nationalities", x = "Student cohort (2014)");
## $y
## [1] "Nationalities"
## 
## $x
## [1] "Student cohort (2014)"
## 
## attr(,"class")
## [1] "labels"

Data Reference

[1] Microsoftword-totalstudentcountsto2014.docx.https://internationaleducation.gov.au/research/Research-Snapshots/Documents/Total (Accessed on 09/24/2019)

Reconstruction

The following plot fixes the main issues in the original.

By using a horizontal bar chart, the issues in the original visualisation have been fixed:

  • The actual number of students from every country can now be easily understood in the x-axis, from where we can determine a round figure of the number of students enrolled in Australian universities as opposed to some arbitrary size presented in the original pie chart. The horizontal bars can be used to do comparison among the different countries very easily.

  • The colour are very distinct, and the whole figure is in two dimensions, making it easy to interpret data. Since we are handling categories with large proportions, the bar chart proves to be a better model than a pie chart because it uses lengths to convey information rather than angles.