Exploring Data Foundations and Visualization

Introduction to Understanding Data. Some topics we will cover in this section before we start using Tableau are what is data, where to find data, foundations for building data visualizations, Tableau version comparison, Tableau workflow. We'll also talk about Tableau chart types.

There are many chart types, but just working with the four core chart types is probably going to cover 80 to 90 percent of your needs. All these other charts are specialty charts that, depending on your industry or who you're working for, may also be useful. And there's just a whole list of them.

I'm not going to read them for the sake of time. But first, what are some examples of data visualization? So I'm going to take a look at an example. Here's one related to life expectancy.

I know, so morbid a topic so early in the morning we're talking about life expectancy. This is showing you the life expectancy all around the world. And so this is an interactive, multi-line graph chart.

It shows life expectancy from 1960 to 2020. Now, I'm going to click on the chart. You can click on the chart as well, because there's a link in your PDF that is clickable.

And let's take a look at a live data visualization. This is the entire world. Where are the countries? Move your mouse across these little threads.

So just like life, life is as fragile as a thread. And so each thread represents a country. And when you hover over a particular line, you can see information listed on top.

In Thailand, for instance, on average, a life expectancy of 55. Oh, I'd be dead already. In 1960.

And 74 years in 2009, an increase of 33%. So that has gone up. And that's what the line indicates.

So better to be born recently than when you would be around 55 in 1960. So some people look at this like, oh, this is pretty cool. This is more than just like creating a chart in Excel, where you have a bar chart or pie chart or line chart.

You also get information here. What is this? Oh, it looks like this shows me certain regions. I'm going to click East Asia and Pacific.

I'm not sure what this is saying. I guess all the countries that are East Asia Pacific are now colored this color. OK, that's interesting.

I'm going to click South Asia, Europe, Central and Middle East. You know what we have? We have a mess. It was already complicated.

This further complicates stuff. Color doesn't really make that much of a difference. I wish there was a way I could sort by individual country or filter by individual country, but there's not.

And every time I click a certain region, it doesn't hide the other ones. It just adds the complexity. If I go back to our PowerPoint, our object lesson here is this is not the best visualization.

Why? This is not practical for finding country info. So our point here is simply building a visualization does not mean you're making something useful. What distinguishes Tableau from Excel? What distinguishes Tableau from a regular chart or graph in Excel is interactivity.

So there was some interactivity here, but the interactivity didn't help. This is the same information in a different visualization. It's not as visually appealing, but it's easier to understand.

I could actually see the countries down here. But guess what? This is also not the best data visualization. It can be confusing.

So just making info available does not make it useful. You can go to public.tableau.com to take a look at visualizations of the day and get inspiration and see what other people are doing and see how their visualizations might inspire you to create something that's a little more useful than what we just saw. Here are two examples of data visualization using dashboards.

When I say dashboards, we mean that you're not just seeing a single chart. You're seeing many different elements, like this is a banner at the top. This could be an image.

And these could be text boxes. And then we have the line chart. We have labels.

But we have multiple elements here. So the important thing to stress here is dashboards are composed of multiple elements, some of them interactive. This one is even fancier.

This one has a lot of color, much more color than this. This is pretty simple. This allows us to see the New York City median income and health insurance information.

So let's take a closer look at this. This is a Tableau visualization. The last one wasn't even a Tableau visualization.

It was just on some website. This one uses a custom control called a parameter. So let's actually take a look at this information.

We have the link, and you have the link in your PDF that you can click on. So I make sure I click on the first link. I'll click here.

So this last one, that's from flowingdata.com. This one, you can see it says public.tableau.com. So how is this information organized? Well, the blue represents the income, and the orange represents the health insurance info for New York City. All the regions are broken up into different segments. If I hover my mouse over different segments of the data, I can go, I guess, to particular zip codes.

Now visually, I can change what's colored blue and orange. There's a little circle up here, and if I click on it and drag over to the left, I can look at the visualization color that represents the health insurance information. Now the thing about this is the density of the color already gives me some information before I even hover my mouse over any of the areas.

What's really nice about this, let's say I hover over an area, not only does it give me the health insurance information, it also gives me the median household income. So I don't necessarily have to drag the blue part over to see that. So for Queens County, I can see the median household income is 55,000, and people with health insurance is 83.8%. So this is something that's definitely more useful than the life expectancy that we saw before, and it's very interactive, and it's based on a map, and this is something that you can create yourself using Tableau.

It's not that difficult. You can even download this as a Tableau workbook. Some people allow you to download it, and then when you download it, you can open it up in Tableau and get some sense of how this is constructed.

Now you won't necessarily get the raw data. You'll just get the data that's used to create the visualization. So I know you have, well, I want the raw data.

You mean they're going to give me? No, no, no. You probably don't want people to have access to all the data that you created. They could just duplicate your stuff and steal it and say they created it.

But you can sort of see what goes on behind the scenes. So I'm going to go back. Data visualizations can be found on public.tableau.com. Definition of data.

Data is a collection of facts such as numbers, words, measurements, observations, or even just descriptions of things. There's two types of data, two primary types of data. Qualitative.

Qualitative data is descriptive information. Categorical data like cities. So qualitative data doesn't mean the quality of the data, but it's categorical information.

You can just think of categories. Then we have quantitative data. When you think of quantitative data, think of the word quantity.

When you think of quantity, you're thinking of numerical information. It counts something. So population would be an example of quantitative data.

There's a numerical amount for the population. Quantitative data can be discrete or it can be continuous. This is important because this will determine how you can work with your visualizations.

I'll show you some examples of that. So what's an example of discrete? 12 months in a year. 12 months in a year.

How is that discrete? Well, if you're talking about a year, a year is not 11 months. It's not 13 months. It's 12 exactly.

What's an example of continuous? Time. What do you mean? Well, time can be measured and you can think of Neil deGrasse Tyson or Carl Sagan or some scientists talking about how time is on a continuum. When you talk about time, you can talk about centuries, decades, years, months, weeks, days, hours, minutes, seconds, milliseconds.

And so that is continuous data and it's continuously being broken down. It's not like 12 months. So discrete data can take only certain values.

So you can think of like integers, whereas continuous can split that data into decimals within a range. Tableau treats each differently. Where can you find datasets? You can query many databases online and download the results of these searches.

Some public datasets can be downloaded in their entirety. We have many sample databases that you can pull information from, websites that have sample datasets. The first three are more official.

So they're likely to be vetted information. We give you those links here. You can click on them and play around with them in Tableau.

The three below are crowdsourced. So they might not be as reliable or vetted. So Kaggle and Data. World are very popular.

Tableau even has sample data sources. But from my understanding, they're pulling information from Kaggle and Data. World. Things to keep in mind. When searching databases online, it's important to verify sources.

You don't want to create misinformation and share it out online. There may be inconsistencies in datasets that must be corrected. Datasets may not be properly formatted for analysis and visualizations.

So there is something called the CRAAP test. And some people can pronounce it a different way than I just did. And I think it's made to be a humorous way of describing how you can evaluate your sources.

Because either it's good or not. So the initials that are used for this test are actually pretty good. If you click on the link that we have in the PowerPoint, here are what those letters stand for.

Currency. The timeliness of the information. How recent is the information? Is it relevant based on the timeliness? Relevance.

Is it relevant to what you're researching or what you want to show? Maybe it's not relevant. Maybe there's no like they say in statistics, causation. What is it? Correlation does not mean causation.

So it may look like it's related and it's relevant, but it might not be. The authority. Where are you getting your source from? Is the source of the information biased? Accuracy.

The reliability and truthfulness. And purpose. The reason the information exists.

Again, someone may have an agenda. And if you're unaware, you may be advancing an agenda that is for some ulterior purpose or motive. So something to think about.

It's not something we're going to go into in the class, but these are just outside resources. Again, the stuff that you may think is boring before we actually get to working with Tableau.