VizNetwork Blog » techniques

Well-formed Data – Part 4: Missing Data

thcrawford — Wed, 10 Feb 2010 19:02:23 +0000

How often do you get data that needs to be charted or graphed only to find out some of the data’s missing? We see it all the time. However, just because it’s missing, doesn’t mean you can ignore it. You need to know how to make it consistent, how to deal with it if it is missing, or sometimes even how to recover it. In this 9:12 video, the fourth in our series on well-formed data for charts & graphs, we cover:

types of missing data
differences between incomplete, space, blank/null, and 0 data
dealing with missing text
dealing with missing numbers
dealing with implied data

Let me know what you think!

Also, be sure to check out the other 3 videos in this series:

Part 1: Consistency (9:43)
Part 2: Concatenation & Extraction (9:57)
Part 3: Normalization (8:52)
Part 4: Missing Data (9:12)

Next up, more tips on using PowerPoint to create a visual impact.

Well-formed Data – Part 3: Normalization

thcrawford — Mon, 08 Feb 2010 19:14:23 +0000

Our lives are made up almost exclusively of relationships and that includes data. In fact, it’s hard to think of examples of data that doesn’t involve a relationship. Customers (hopefully) have many Orders. Companies have multiple employees. People (hopefully) have multiple friends, who also have multiple friends, and may even share some of each others friends. People may have multiple phone numbers, e-mail addresses, and so on.

Admittedly, this may be the geekiest of topics on creating well-formed data. In part 3 of our 4 part series, we discuss one way to make sure your data follows the ”1 Concept per Row” rule. Normalization is used whenever one concept relates to another concept. Given that almost all data is about relationships and we want to analyze that data, we have to know a bit about normalization which is all about those relationships. There are whole college classes dedicated to this topic, so we won’t be able to cover it all in this 8:52 video, but hopefully you’ll understand enough that you can do the basics yourself or talk intelligently to your data warehouse person.

In the video we discuss, 2 types of relationships…1 to many (e.g. 1 person may have 1 or more orders) and many to many (e.g. people have many friends that can overlap). These require slightly different solutions. There is a 3rd type of relationship, 1 to 1, however, dealing with that one doesn’t require normalization. That data can just be stored in the same spreadsheet since it follows the “1 Concept per Row” rule. The examples use Excel and the vLookup() function to illustrate the concepts.

Let me know what you think!

In the final video, we’ll wrap up our tour of well-formed data with a discussion on how to deal with missing data.

Well-formed Data – Part 2: Concatenation & Extraction

thcrawford — Mon, 18 Jan 2010 23:54:20 +0000

Continuing our series on preparing data for analysis & visualization, we’ve just released the next video on concatenation and extraction. The ability to break data apart and put it back together in new ways is essential to preparing data. By storing data at the lowest sensible level, it can be used separately or combined with other data for interesting analysis and visualization. In this practical video, we discuss 6 key Excel functions for extraction including left(), mid(), right(), search(), len(), and trunc().

Check out this 9:57 video to learn how to take the next step in preparing your data for analysis and visualization:

In the next 2 videos, we’ll continue the discussion with Normalization (advanced ideas to support One Concept per Row), and Missing Data (how to deal with missing or implied data).

Well-formed Data – Part 1: Consistency

thcrawford — Sun, 17 Jan 2010 15:23:55 +0000

Data visualization and analysis are powerful tools for discovering and communicating stories held in your data. However, before most of today’s data visualization tools can be used effectively, the data must be cleaned, organized and prepared. Over the next 4 videos, I’ll be discussing how to prepare your data to be visualized. The first step in the process is consistency. Consistency is made up of 4 principles:

One Concept per Row
One Data Type per Column
One Format per Data Type
Using the Lowest Sensible Level

Check out this 9:43 video to begin the most important step in preparing your data:

In the next 3 videos, we’ll continue the discussion with Concatenation & Extraction (how to put things together and take them apart), Normalization (advanced ideas to support One Concept per Row), and Missing Data (how to deal with missing or implied data).

Non-Linear Navigation in PowerPoint

thcrawford — Wed, 16 Dec 2009 15:00:39 +0000

Do you ever have to give the presentation to the executive where you’ve only got 5 minutes so you boil your presentation down to the core message, but you know you’ll get asked a (seemingly random) question about a slide? So you debate, do you put the supporting slides (details, data, sources, etc) in the main flow? Do you add them at the end? Do you keep them in a separate presentation? Do you hide them? If you do, how do you get to them quickly, easily, and in a way that makes it look like you anticipated their needs?

Ever have an audience member ask a question that is somewhat tangential, but still a good question that you have material prepared for, but didn’t want to put in the main flow?

This 5:30 video walks you through the steps of creating buttons, links and hidden buttons that let you navigate to any part of your presentation, quickly, easily, and professionally:

Be sure to let me know what you think and if there are other videos you would like to see!

Interactive Area Charts in Excel

thcrawford — Mon, 14 Dec 2009 21:00:04 +0000

In a recent post by Nathan from FlowingData, he does a great job of explaining how to interactive area charts. For the average person, though, his approach requires a lot of existing knowledge…Flash, Flex, coding, etc. So, I wanted to create a version that most people could create and use with NO coding required. For all the grief (often deserved) that Microsoft gets, there’s actually a lot of power in the Office suite, and Excel is no exception.

In the following 3 videos, I cover how to create an interactive area chart in Excel without coding. While I use Excel 2007, most of the features are available in at least the last two versions. As with everything, this approach has limitations…two big ones, in fact. First, it only works on PC. It takes advantage of the Active X controls and those are, sadly, PC only. Second, deals with a feature of charts in Excel that displays the label of lines/area with values of all zeros. There is a work around for the second issue, but since it requires code, I decided that I could live with the limitation. Generally, you won’t be displaying graph items with all zeros anyway, because why would you graph them? But that’s another post.

In the first video (6:29), I discuss how to set up the basic spreadsheet to make the interactive chart:

In the second video (8:12), I show you how to create the first set of controls:

In the final video (5:48), we finish off all the controls and talk about the limitations:

I hope you find this useful! Please let me know what you think and if you’d like to see other topics.

The Magic of Visualization

thcrawford — Mon, 09 Nov 2009 15:52:01 +0000

Some people would definitely think the process of creating a visual is some form of magic requiring some ability to conjure something from nothing. The myth of left brain/right brain reinforces the false notion that creativity is something that only belongs to a special, magical few. The reality is that, like magic, there is a science to visualization, that it can be learned, it requires practice, and anyone can do it!

Last week, I attended the presentation of Jamy Ian Swiss, magician extraordinaire, at the Penny W. Stamps Distinguished Visitor Series at the University of Michigan‘s School of Art & Design. Followers of visual thinking and information design & visualization will know him from co-writing a chapter of Edward Tufte‘s Visual Explanations. It’s no wonder that Tufte used him in his book. His explanation of the magic trick was more about information design than magic. From the intentional placement location of cards on the table to the narrative story arc to body language, every step of Jamy’s magic is an attempt to direct (or maybe more appropriately misdirect) attention.

Card Placement

When the card is thrown face down on the table, if it’s thrown in position 1, close to the audience member, they are more likely to pick it up or peak at it. If the card is thrown in position 2, the audience member is unlikely to pick it up, but also is expecting some sort of deception since it is placed in a defensive position. Either case pushes the participant out of the story and breaks the story arc. By placing the card in position 3 (just short of halfway between the magician and the audience member), a balance is achieved by keeping the audience member in the illusion, but not offering a peak ahead in the story.

What fascinates me about this is how similar it is to page or screen design. Placement and directing attention matter to the audiences perception of the piece.

Body Language

Shifting weight: Placing weight on one leg or another will guide people’s attention and expectations. Putting weight on the forward leg adds energy and almost an arrow implying direction. Placing weight on the back leg implies an anchor or station. Lifting the heal of the lighter foot further adds to that implication.

Muscle Tension: Letting a hand hang loosely implies that there is nothing in it, while holding one stiffly implies that it is holding something.

Why talk about body language on a visualization blog? Minimally because you may need to visualize people in your messages and how they appear will impact your story, but I think it’s much more than that. Body language is a visual story in it’s own right. How you use your body in daily conversations, meetings, and presentations influences how people perceive you. Not only being aware of the messages, but crafting the message your sending with your body can have a significant impact on your communications. In fact, much of people’s perceptions of you come from the message your body sends. On occasion it can even override your words and your other visuals.

Visualization doesn’t have to be magic. A few simple principles can help you improve the way you solve problems and sell ideas. Want to learn more? Contact me and I’ll help you find a solution that’s right for your needs.