Content
Visualizations
Tables and graphs are generally the best way to communicate with data. They are two members of a larger family of display methods known as charts (Few, 2004). In this category, we also include maps and diagrams.
Tables
In this project, I’ll start with tables. They are particularly useful when we have a lot of variables with different units of measure. Tables usually contain all of the information ready to look up and display simple relationships between quantitative and categorical values.
I grouped the data into three main tables. Each of them describes the overall characteristics of the five Nordic countries.
1.Sociodemographic indicators
2. Population health indicators
3. COVID-19 results
I employed a set of visual design principles to organize the data and highlight what is important as follows:
The relationships are structured unidirectionally, with the category subdivision arranged across the columns.
The first column lists the countries in italics format to visually group them from the rest of the columns.
Every column represents a characteristic (variable).
By boldfacing specific values, I used the preattentive processing attribute to outstand the highest one.
The color intensity of every row varies to enhance sequence.
Consistency is present in the three tables for immediate comparison by keeping the same order of countries. The order is arbitrarily set based on the highest HDI to the lowest.
Table 1. Sociodemographic indicators per Nordic Country
| Sociodemographic indicators per Nordic Country | ||||||||
|---|---|---|---|---|---|---|---|---|
| During COVID-19 pandemic1 | ||||||||
| Country | Economic indicators | Demographic indicators | ||||||
| HDI2 | Extreme poverty | Life expectancy | Median age | Population aged 65 or more | Population aged 70 or more | Total population | Population density3 | |
| Norway | 0.957 | 0.2% | 82.4 | 39.7 | 16.8% | 10.8% | 5,421,242 | 14.5 |
| Iceland | 0.949 | 0.2% | 83.0 | 37.3 | 14.4% | 9.2% | 341,250 | 3.4 |
| Sweden | 0.945 | 0.5% | 82.8 | 41.0 | 20.0% | 13.4% | 10,099,270 | 24.7 |
| Denmark | 0.940 | 0.2% | 80.9 | 42.3 | 19.7% | 12.3% | 5,792,203 | 136.5 |
| Finland | 0.938 | 0.6% | 81.9 | 42.8 | 21.2% | 13.3% | 5,540,718 | 18.1 |
| Source: Data on COVID-19 by Our World in Data | ||||||||
|
1
From February 13, 2020 to March 24, 2021
2
Human Development Index
3
The number of people per square kilometer of land area
|
||||||||
Table 2. Health indicators per Nordic country
| Health indicators per Nordic country | |||||
|---|---|---|---|---|---|
| During COVID-19 pandemic1 | |||||
| Country | Hospital beds2 | Diabetes prevalence rate3 | Cardiovascular death rate4 | Female smokers percentage5 | Male smokers percentage5 |
| Norway | 3.60 | 5.3% | 14.3% | 19.6% | 20.7% |
| Iceland | 2.91 | 5.3% | 18.0% | 14.3% | 15.2% |
| Sweden | 2.22 | 4.8% | 34.0% | 18.8% | 18.9% |
| Denmark | 2.50 | 6.4% | 14.8% | 19.3% | 18.8% |
| Finland | 3.28 | 5.8% | 53.5% | 18.3% | 22.6% |
| Source: Data on COVID-19 by Our World in Data | |||||
|
1
From February 13, 2020 to March 24, 2021
2
Hospital beds per 1,000 people
3
Percentage of people ages 20-79 who have
type 1 or type 2 diabetes
4
Number of deaths per 100,000
individuals
5
Percentage of men and women, of all
ages, who smoke daily
|
|||||
Maps
In this type of chart, I include part of the information shown in the tables above with their geographical position. It is an interactive map that isolates information of one country at a time instead of directly comparing them. It is broadening the perspective of what was discussed by interacting with our visual system.
Graphs
Another way to communicate with data is the use of graphs. Graphs will typically get the information across more quickly than a well-design table. The visual processing that occurs when viewing graphs involves high-bandwidth, simultaneous input of multiple data, enabling us to perceive a great deal of quantitative information in a burst of recognition (Few, 2004).
There are many different graphs. Different types of quantitative relationships require different forms of graphics. For the subsequent visualizations, I will show amounts, distributions, proportions, associations, and trends. I will select the type of graph that carries the intended message for at least one COVID-19 indicator for the Nordic countries.
Visualizing amounts
Graph 1. Reported positive COVID-19 cases in Nordic countries
To understand a bar chart, we look at its length and the point at which it ends (omitting the width of the bar). I show a single series making the relative order of the category (country) important. We can appreciate which country had the highest sum of positive cases.
Graph 2. Reported COVID-19 deaths in Nordic countries
I used contrast to highlight that the position of Finland in the number of positive cases (fourth place) in comparison to the number of deaths (third place).
Visualizing distributions
Graph 3. Boxplot of hospital patients per million in Nordic Countries
A boxplot summarizes the distribution of a continuous variable. In this case, I use it to compare distribution of several groups. I added individual observations on top of boxes, shifting all dots by a random value ranging from 0 to size, avoiding overlaps
This plot displays the number of COVID-19 patients that required hospitalization. It also shows the dispersion of the variable We can compare the median of every country. And in general we can observe that in all cases, the distribution is right skewed.
Visualizing Proportions
Graph 4. Relative proportion of positive cases and deaths
Visualizing Associations
Graph 5. Association plot
Visualizing Trends and Uncertainty
Graph 6. COVID-19 cases
One of the main interest of the pandemic, is understand the evolution of positive cases over time. I took the new cases smoothed that eliminates outliers and statistical noise to make the
All countries report a reasonably low number of positive cases from left to right from March up to October. Norway shows a disproportionate number of positive cases during March that drops in the subsequent months. Except for Sweden, from May shows an increase that reaches the first maximum in July and later decreases. There is a sudden shift around November in all countries that slowly decreases in February 2021. I suspect an upward trend being present for the most recent period but with a much smaller magnitude than the past highest period.
As a final comment, data smoothing does not necessarily offer an interpretation of the patterns it helps to recognize. It can also contribute to certain data points being overlooked by focusing on others.