Dataviz Makeover 1

Visualisation Makeover on MOM’s Labour Force in Singapore 2019 Report in Tableau

Chee Kah Wen Gerald https://public.tableau.com/profile/gerald.chee1204#!/vizhome/DataVizMakeover1_16118032802250/Final?publish=yes
01-28-2021

1. Introduction

Data collected by the Department of Statistics Singapore in the past 10 years (2009-2019) highlight a potential problem in our current society. It is revealed that Singapore’s population has begun to age, due to declining birth rates since 2010 . This correlates to having an aging working population where the labour force consists of a higher proportion of workers in the 50 and over age band as evident in the figure below.

FIGURE 1: RESIDENT LABOUR FORCE BY AGE

In the context of Singapore, our labour force comprises of people above the age of 15 years old who are working or seeking work (regardless of employment status or citizenship). Using the dataset, Figure 1 was visualised by the Ministry of Manpower (MOM), Manpower Research and Statistics Department. The purpose of the document explores how a makeover on the exact same dataset would allow for an alternative graphical representation and the advantageousness of said work.


2. Critque on Clarity and Aesthetics

Clarity Aesthetics
Data presented does not match storytelling. (16% in 2009 and 25% in 2019 represents a summation of the percentages values for age bands of 55 and above) Colour scheme is dull and is not impactful to the reader.
No y-axis presented; reader has to refer to table values to determine that the y-axis represents the percentage of labour force. Horizontal reference lines looks similar to a bar, this may mislead the reader to place emphasis on the age band 40-44.
No annotation for values in table. Reader is unable to tell if they represent absolute values or a percentage. No ‘spikes’. A bar chart might be better to visually represent the ‘small’ changes in percentage values.
Age bands mentioned are not visually shown.
Labour Force Participation Rate (LFPR) increase not shown visually
Title depicts the graph as the proportion of labour force by age, thus it is unable to show change in LFPR as stated.

2.1 Proposed Design

FIGURE 2: PROPOSED VISUALISATION IN TABLEAU

3. Preparation

Visualisation for Figure 1 uses derived values from Table 7: Resident Labour Force Aged Fifteen Years And Over By Age And Sex, 2009 - 2019 (June). For the makeover, we are tasked to use Table 5:Resident Labour Force Participation Rate by Age and Sex 2009 - 2019 (June).

Steps Data Preparation
Data in excel was cleaned by removing unecessary data rows and formatting issues Dataprep1
Data was examined using Tableau’s data viewer after arranging data in proposed visualisation. This step is repeated twice for both files. Data view was exported to .csv format Dataprep2
After checking both files, a new excel (final.xlsx) was created containing the pivoted versions of the Labour force and LFPR Dataprep3

Visualisation Process in Tableau is listed in the following section

Steps Visualisation
Data was imported into Tableau Tableau1
For the first bar chart titled ‘Labour Force by Years 2009,2019’, Category,Age Band and Year were placed into Columns while sum(value) was placed into rows. Bars were coloured according to Year Values Tableau2
Next, to hide unnecessary data, Category and Age Band was placed into filters to display current visualisation Tableau3
To highlight the change between the 2 years, a line chart was added, this was done by placing sum(value) in Row and creating a dual axis.Plot was coloured by sum(value), to make the line bar a single color, I edit the color values to be a 2-step color band and changed the minimum value to allow the bar to be colored dark red Tableau4
Finally, annotations were added to highlight derived insights. Tableau5
For the dumb bell chart, similar steps was used as above, the main difference is that the circle option was selected for the visualisation. Tableau6

4. Observations from Makeover Visualisation

FIGURE 3:FINAL DASHBOARD VISUALISATION

From the makeover done on the visualisation. We can derive clearer insights compared to the original graph.