“Data visualization is an essential aspect of data analysis. It provides a front line attack, reveals the complex structure of data that could not be understood in any other way. It can help to discover unexpected results and challenge the expected conclusions”
The data itself, composed of bits and bytes stored in a file on a hard disk, is invisible. To see and understand these data, we need to view them. In this section, I will discuss visualization in the broad sense, including also purely textual data representations. For example, simply load a database in a spreadsheet can be considered a visualization. Invisible data suddenly turn into an “image” visible on our screen. So the question is not whether journalists should view the data or not, but what type of display is most useful depending on the situation.
In other words: when is it necessary to create more complex online data visualization tools than a table? The short answer is, almost always. Simple tables are clearly not sufficient to give a good overview of a database, and they fail to immediately identify trends in data. For example, geographic trends that can not be represented on a map.
Visualize for getting ideas
It is strange to think that the tools and data visualization techniques will magically appear all cooked a lot of stories. There are no rules, no “protocol” which will guarantee to find an angle. It seems more sensible to me to look for “clues” of information that a talented journalist will weave to shape stories.
Every new visualization is likely to provide us information on our data. Some may already be known (but perhaps not yet proven), while others may be completely new or surprising. Some of this information may give rise to an article, others will prove to be the erroneous data product visualizations that are likely to appear.
Learn how to visualize your data
Visualization provides a unique perspective on a database. There are many ways to view data.
The tables are very powerful when you have relatively little data to visualize. They present the headers and the amounts of the more structured and organized way possible and reveal their true potential when combined with the ability to sort and filter data. Furthermore, Edward Tufte suggests to always include small pieces of graphics in tables – for example, one bar per column or a trend graph . But the fact remains that the tables are limited. They are perfect for displaying one-dimensional variables, such as a top 10, but they are insufficient to compare multiple dimensions simultaneously (for example, changes in the population of several countries over time).
To create cool data visualization examples you can check out https://public.tableau.com/s/.