From the upcoming The Annual Condition of Iowa's Community Colleges 2009

The state of Iowa has established three targeted industry clusters: advanced manufacturing; bioscience; and information solutions. How have Iowa’s community colleges been supplying the industry? Above is an ideogram showing the path from program major (left) to targeted cluster (right). Obviously bioscience is being well-stocked, mostly from health science majors. I’m a bit surprised though that information technology majors go on to information solutions at about the same rate as other targeted clusters.

I’ve already previewed a few ideograms which show the transition from college majors to industry of employment. You can find the project page here, a presentation, and the poster.

Ever get a letter from your alma mater asking what you’ve been up to since you graduated? That data is used to see how well the college has been doing. It’s especially important for public colleges to see if most graduates stick around. Ever send one back? No? Well, most don’t.

But that’s a pain for researchers who really want to know. The response are unlikely to be random so we can’t infer much about the results. Iowa Department of Education and Iowa Workforce Development teamed up to track alumni from Iowa community colleges using administrative records. That is, we don’t ask you, we just look in pre-existing databases to find you.

The result is an infographic [PDF] that explains the methodology and results. It’s part of an upcoming series on Iowa community colleges and labor supply.

A One-Dimensional Approach

Data visualization is great because it can merges statistics–a largely positive science–with explanation. Thus, unacquainted individuals fully into a conversation without relying on complex structures. However, I think many writers on data visualization underestimate how easy it is to forget the positive statements from statistics and turns into rhetorical statistics without a good theoretical basis.

David McCandless and Stefanie Posavec from Information Is Beautiful created an infographic of the distinction between left versus right in the United States and the European Union. Aesthetically, it’s a nice piece that tries to capture familial, economic, and social aspects of ideologies–all appropriate for analysis. However, David and Stefanie try to fit it into an antiquated one-dimension (left and right) paradigm. Some items have been debunked–for both sides–by other authors, such as Andrew Gelman’s book and his blog.

Left versus right isn’t a sufficient paradigm. In particular, the right has been wedged as a small government, psuedo-libertarian, theist-based party. But that isn’t capturing the internal structure within that dimension. It doesn’t describe big-government conservatives, staunch libertarians.

The best model for political philosophies is a two-dimensional (instead of one-dimensional) system. The diagram below from Political Compass shows the stance of beliefs in the economic realm (horizontal axis) and the social realm (vertical axis) based on the 2008 elections.

A Two-Dimensional Approach

A similar pattern is available for the European Union. David and Stefanie’s take on EU political spectrum is here while Political Compass has their own take. Ultimately, this is tricky ground for infographics. David and Stefanie’s analysis looks great, but it’s based on a flawed model that has been adapted by others.

Via Cool Infographics via VizWorld.

Today the Iowa Department of Education and Iowa Workforce Development released “Visualizing Transitions into the Workforce” poster. In short, DE and IWD partnered to follow students from Iowa community colleges into the workforce. The poster shows how specific majors lead to employment in various sectors in Iowa’s economy. The findings are discussed in the full poster.

The main image was created using Circos, a program originally used to map the relationship between genomes of different species. However, it’s been used recently to replace tables in general–like we did with this implementation. I think the advantages this has over a table is readily noticeable in two ways: first, it’s apparent how many college parallel (e.g., transfer) students forgo transfer and go into the workforce. Also, the relationship between health science majors and their eventual employment in the health care sector is very clear.

Circos has a built-in table viewer, but I passed on that tool and programmed in the karyotype file manually. I wrote an R script to translate the data from a pivot table to a karyotype file which I’ll post soon. Karyotype files seem more flexible for future use. For instance, we know wages for each of these students which can be overlapped on the 2-D data tracks.

Click the image above for the full PDF poster.