« Mix percent metaphors, add average confusion, and serve | Main | Bad charts can happen to good people »


Andy W

Thanks for taking the time to respond, exactly which chart is scary? Part of the reason I used the pre-processed data from the Minn. Fed is that I don't know what tables from BLS you are talking about! (this isn't my field, so I'm largely ignorant, and gave up after looking for a few minutes on the BLS website)

As long as you can define in numerical terms what a recession is you can do the data processing you suggest. The spreadsheet I used determined the start of recessions as defined by NBER. I don't know what definition you or the calculated risk blog are using.

I feel conflicted about the flexibility remark. Flexibility is good for ad-hoc analysis, I'm not sure if recessions should be defined in an ad-hoc manner though!

Knowing SPSS code really well, I don't feel that SPSS is inferior to Excel for this particular task or really any data manipulation task (ditto for the majority of major stat packages). I can fully admit though it took probably 4 years of using SPSS on a daily basis to get that saavy. I can confidently say I'm a much more productive analyst for it though. There are some things Excel is better at (such as a well designed interactive dashboard), but this IMO isn't one of them.

Thanks again for taking the time to respond!

PS - Having the lines terminate when they reach zero values is pretty trivial, so that certainly isn't a reason to prefer Excel. The "more code" you talk about is two lines in SPSS. I would roughly guess any data manipulation you perform in Excel amounts to between 1-5 lines of code in SPSS.


In software development, good programmers know that there are different tools for different jobs. Just because you can do something in language X doesn't mean you should. Those of us in data analysis should try to look at software the same way.

Excel gets a bad rap because lots of people misuse it to create bad graphs or use it to just store text information in a grid. But when used properly, Excel has some great features. I work in R and Excel almost every day. I find that Excel is great at reshaping, sorting, and subsetting data. It's also where I create most simple bar or line graphs.

I'm happy to know I'm not alone! You, Nate Silver and I'm sure many, many other good data analysts use Excel too.


Andy: Great to see you here. The "scary" moniker is applied to the jobs losses chart. It's just a name you see in the blogosphere. I was not referring to your charts.

The raw data from BLS comes from the CES survey. Here is the link for anyone interested.

The comments to this entry are closed.

Kaiser Fung. Business analytics and data visualization expert. Author and Speaker.
Visit my website. Follow my Twitter. See my articles at Daily Beast, 538, HBR.

See my Youtube and Flickr.

Book Blog

Link to junkcharts

Graphics design by Amanda Lee

The Read

Keep in Touch

follow me on Twitter