Data
Behold--the DataSnob
Sunday, May 25, 2008 10:55 AM (permalink)
I started a new blog last week about the day-to-day butchery of data, numbers and statistics on the web. As someone who makes his living telling stories with numbers, I am constantly amazed by what passes for research (and even more so what passes for reporting on research) on the web, so I have started a little journal on the topic called DataSnob.
The upside for readers of Quite Content will be to keep this blog blissfully free of numbers, statistics, or pesky 'facts.' You'll have to visit DataSnob if you miss that sort of thing. I hope you do!
Tom
Data Visualization
Wednesday, February 27, 2008 9:04 AM (permalink)
I am finally going to take Edward Tufte's course here in Raleigh Durham later next month. When I am done, I expect to be able to make data visualizations like this one from the New York Times this week. Take that, PowerPoint.
Chartjunk
Saturday, February 16, 2008 9:00 AM (permalink)
This is some first-class data manipulation here. Check out the maps presented as "data" in this article, and see if you spot the problem:
Russell Shaw: "Usury Maps" Show Hypocrisy of Many "Christian Conservative" Pols.
Hint: Do you think North Carolina, Georgia and West Virginia are not high on the "Christian Power List?" If you check the underlying data, you can see what went on here: if those states, where payday-lending is illegal, were colored correctly on the "Christian Power" map, that graphic is not nearly so compelling.
Economic Indicators
Friday, February 8, 2008 8:51 AM (permalink)
The Huffington Post reports that Wal-Mart customers are using their holiday gift cards to buy groceries. That can't be good.
Economic Worries Rise for Upscale Americans
Wednesday, January 30, 2008 12:00 AM (permalink)
Daily Number: 75% - Economic Worries Rise Among Upscale Americans. Three-quarters of Americans (75%), now rate strengthening the economy as a top priority for policymakers -- up from 68% a year ago, according to Pew Research. This isn't surprising, and is certainly echoed in the exit polls from the past 3 or 4 primaries. Florida, for instance, has a high percentage of retirees and others living off of income-based investments. If the market goes down 10%, that has a real impact on your financial situation--perhaps even delaying your retirement. That is an issue no matter what your personal income might be. As my old Equities Analysis prof in business school (the late Tony Wingler) taught me, the "economy" and Mr. Market are only loosely related. You might have a good job, own your home and be debt free, but the market doesn't necessarily react to the economy. If more people want to sell your stocks than buy them, you'd best delay your retirement party. [PewResearch.org | All Feeds]
CNN's Data Porn
Wednesday, January 30, 2008 12:00 AM (permalink)
Nick Harris [NewsGator Inbox] wants to know what is powering CNN's cool touch screen display, noting that "It acts a lot like an iPhone, but also reminds me of Microsoft Surface." Dunno about the technology, but the data is powered by your friends at Edison, of course!
Presenting Data Vs. Telling a Story
Monday, July 9, 2007 10:04 PM (permalink)
I have been giving a lot of presentations lately--some good, some bad. In my business, clarity is Job #1. I often scour the Internet for mentions of my presentation or the data I give not out of vanity (though I have no small amount of that) but out of the fear that the data will be misrepresented, miscopied, or otherwise butchered. I always get one or two of these, and those I chalk up to the writer either reading what they wanted to into the slide, no understanding research in general, or just making a typo. When I find more than one or two specific examples, however, I know that I was the butcher.
For instance, I gave two different presentations over the past couple of weeks that came out very differently-one pretty good (at least for me) and one not really up to my standards. In both, I was presenting data-slide after slide of data. But while I left the first one, a presentation to the Country Radio Broadcasters about the state of Country Radio partisans, with a good feeling, I left the second one, to the Corporate Podcast Summit, feeling like I hadn't done my best.
For both I was well prepared, knew the data cold, spoke confidently and had few nerves. But there was a big difference in how I felt afterward, and it taught me something. In the CRB presentation, I went through about 50 slides in an hour--with only 30 or so complex data graphs. In the podcast presentation, I used 40 slides in 25 minutes, and actually skipped 4 or 5 slides as I raced to leave myself enough time for my conclusions.
I have read lots of advice about how many slides you should have in a presentation, and most of the experts say that you should count on one slide for every :45 to 1:00 of your talk. That might work for a Tom Peters-esque slide that says nothing but "FOCUS" or "WOW!" but not so much for tracking the purchasing habits of two distinct lifegroups over 4-5 years. So I think I am going to make myself a new rule. The Internets love lists, so here is my "list of two":
- 1. Allow at least three minutes per slide for a data-rich chart or graph
- 2. If you can't talk for three compelling minutes about a data-rich chart, cut it.
The last bit will be the most difficult for me--after all, I have all this great data--why not tell it all to the world? But I have learned that even in a presentation filled with exciting, really new-to-the-world data, less is more. I found myself in the podcast presentation really racing through some slides (and, in hindsight, they weren't that important) while spending 4-5 minutes on others. It is the slides that I spent 4-5 minutes on that really told the story, and are the ones that no one got wrong, miscopied or otherwise corrupted when I read the write-ups later.
I think if I have to "fill" a certain amount of time I wildly overestimate the number of charts to use, and this is, I am sure, a crutch. After all, if I run out of things to say in a 30-minute presentation, it is comforting to know that I have 60 slides of data to blow through as a safety net. Unfortunately, this results in a less-than-memorable presentation, though I used to congratulate myself for "not even using these slides--that's how great I was!" Now I see excess material as a real failure--a lack of proper planning and foresight to what the story really is.
Looking back over the podcast presentation, I had 7 or 8 really great charts, and then a bunch of charts that were best left as conversation bits at the cocktail hour. Had I gone with that, I probably would have given a 22-minute talk that told the same story, only better. That is the best 30-minute presentation advice you'll ever get!
