Jazz up your data with Gapminder

Leave a comment

One of the most common data visualization tools for research projects is the standard graph. Graphs can be visually instructive when explaining how one indicator relates to another, but let’s face it – we’ve all seen a million graphs before, and they can be extremely dull. Moreover, if your aim is to compare multiple countries, all those lines on the same graph can be terribly confusing!

Enter Gapminder, at http://www.gapminder.org.

Gapminder allows you to select the development indicators between which you want to establish a relationship and then plots them on a graph with circles that represent individual countries. These bubbles are color-coded according to their respective regions, and their sizes indicate the relative size of the population of that country. If you hover your mouse over a specific bubble, a label with that country’s name will appear.

To create your graph, simply click on where the label is and select the indicators that you want to see, and the graph appears! If you want to isolate countries from a specific region, click on that region on the map in the upper right hand corner, and all of the bubbles for countries not in that region will disappear from the graph. To isolate one country in particular, check the box next to it on the list below the map.

One other nifty feature of Gapminder is that you can see how countries have changed over time. To see how this works, move the slider on the timeline back to whatever year you’d like to start from and click “PLAY”. The bubbles will move around the graph, indicating how the relationship between the two indicators you selected has changed regionally or globally over the years.

This is an incredibly effective tool for making your data visualization more interesting and engaging, as well as for showing changes over time! If you’re confused about its use or want to make sure you’re applying this tool effectively, you can stop by the lab in Hurst 202/203 for further pointers!


Research Discussion Luncheon on October 20

Leave a comment

The SSRL Research Discussion/Luncheon series continues! We are celebrating the spirit of academic research this fall by promoting and sharing the research work of our fellows, staff, and visiting scholars. The second presentation for this semester is Dr. Molly Brown, Research Scientist from NASA Goddard Space Flight Center.

“Droughts and Decision making in Famine Early Warning Systems: The Role of Earth Science Data”

The Famine Early Warning System Network (FEWS NET), funded by the United States Agency for International Development (USAID), works to strengthen the abilities of countries and regional organizations to improve food security through the provision of timely and analytical early warning and vulnerability information.

This talk will describe the different types of remote sensing imagery that FEWS NET uses and the methodologies it employs to transform the data into information to inform decisions regarding food aid and other humanitarian responses.  A description of the products that are used by FEWS NET scientists to monitor food production inAfricawill be presented, along with graphics and explanations as to how the products are used to inform decision-making.

Dr. Brown will present research that provides evidence on the use of data of rainfall information, agricultural production, food prices and food access parameters by FEWS NET analysts. There are significant differences in the use of remote sensing and other technical information between East, West and Southern African country analysts, with satellite remote sensing of vegetation being used 28% of the time, rainfall imagery 84% and gridded crop models only 10% of the time.

The talk continues with a description of how econometric models can improve understanding of how changes in local production can affect access to food through market prices.   Finally, a few case studies will be given that illustrate how the remote sensing is used to provide early warning of food insecurity that is both policy relevant and specific enough to motivate response by humanitarian organizations.

We look forward to seeing you at the event! Please RSVP to the email address below.

Creating a Chart in Google Docs

Leave a comment

Step away from Excel for a few minutes and marvel at the chart building possibilities available on Google Docs.

If you have a Google account (and if you’re an AU student, of course you do), log into your Google account and get to Google Docs. You know that you can upload text files and spreadsheets there. Click on one of the spreadsheets you have (and if you don’t have one uploaded, you can take a spreadsheet you’ve been working on in Excel and upload it to Google Docs).  Up at the top is a little button that looks like a red and blue bar graph. That is where the magic happens.

Click on the “Chart button” and begin building your chart. You can choose from line, bar, pie, trends, map, and other graph options.

On your spreadsheet, select the data you want to include in your chart. Or, you can do that under the “Start” tab once you’ve opened the Chart box.  Google Docs will recommend a chart for your first, but if you decide that is not what you want, you can move on to the “Charts” tab and select another option.  (Google Docs will even let you know if a certain chart is not possible with the data you provided). You’ll be able to preview the chart once you’ve made a selection. Create the title and other labels for your chart in the the “Customize” tab.

When you’re happy with the chart you built, click on “Insert” and it will appear on top of your spreadsheet.

It’s that easy!

For more instructions, read it from Google themselves.

SSRL No Longer Processes Scantrons

Leave a comment

You read it correctly. The Scantron machine has left the building.

Scantron services have been moved permanently to the third floor of the library (Room 321) in the Faculty Blackboard Support Center.

Now we have more time to help you with other important things!  Let a consultant know if you need more information on how to find the Scantron machine.

Tips for Data Visualization (Part 1)

Leave a comment


What’s more frustrating than trying to interpret lots of data?

Trying to interpret that data in a confusing graphic form.  Looking at a graphic is the supposedly easier way to understand the point, right? Whoever came up with the phrase “A picture is worth a thousand words” might also have been a statistician. You’ve got to make sure that the data you represent with charts, graphs, pies, bubbles, lines, rainbows, you name it, actually makes sense to your audience.

Here are few ways you can do that:
(from Prof. Jim Lee’s presentation: “How much is a picture worth?”) 

1. Encourage the eye to compare data.

This means avoid lots of white space, use the same color if you can, and try using shades to show intensity.

2. Follow some of Edward Tufte’s advice.

Edward Tufte is a statistician who knows a thing or two about data visualization. His fundamental rule of efficient graphical design is to minimize the ratio of ink-to-data.
He even has a formula for this:

data-ink ratio = data ink/ total graphic ink (the closer to 1.0 the better)

Basically, don’t add more to your graphical data than you need to.

3. Consider pie and bar charts – is it the shape or number that tells you information? Be wary of distorted shapes that can throw your audience off.

4. If you’re going to use a pie chart…make sure it adds up to 100%.  This is a more common mistake than you think.

5.  When adding graphics to your text, clearly indicate the graphic location. Provide explanations with appropriate citations that support the existence and details of the relationships.  Explain the overall behavior of the model or the themes in the graphic that add to the understanding of the discussion.

Don’t leave your audience hanging or trying to interpret your simple graphic like it’s an abstract painting. Be clear in what you’re trying to say! (Like the true pie graph above.)

More tips on data visualization to come!

Great Educational Resources Online: Academic Earth

Leave a comment

With the proliferation of information technologies the obstacles to obtaining good education are quickly disappearing. Some of the best American universities have decided to upload entire classes online including video lectures, lecture notes, exercises with solutions and reading materials. Among the universities taking the lead in this revolutionary approach to education are MIT, Stanford, Harvard, Yale and many others. Although most videos can be found on popular video sharing sites like YouTube a somewhat better first stop is the website www.academicearth.org. The site tries to organize and rate the educational resources available freely online. Links to the video lectures and other materials can be found on the site. The website contains resources on virtually all standard subjects taught at the undergraduate and graduate levels including statistics, mathematics, economics, computer science, the natural sciences and many more. Here is a brief overview of some of the materials available online.

  • Mathematics
  • Differential equations(http://academicearth.org/courses/differential-equations). A great introductory class on differential equations taught at MIT. The class includes the lecture notes, exams and solutions. It is taught at the undergraduate level but it can be useful even for some graduate students who need to brush up on some solution techniques.
  • Linear algebra: There are two wonderful resources available. The undergraduate MIT class on linear algebra taught by Gilbert Strang (http://academicearth.org/courses/linear-algebra) starts with the basic concepts but quickly builds to more complicated ideas and methods. On the other hand, Stanford’s graduate course on Linear Dynamical Systems (http://academicearth.org/courses/introduction-to-linear-dynamical-systems) may be challenging for undergraduate students but provides many real world examples and the lecturer, Stephen Boyd is very engaging and fun to watch. This course will be very helpful to graduate students in statistics and econometrics.
  • Computer Science
  • A great introductory course in computer science is provided by MIT (http://academicearth.org/courses/introduction-to-computer-science-and-programming). The class uses Python as a programming language so lectures 2 to 7 are a wonderful introduction to programming with Python. The class also covers major computer science topics such as sorting, efficient algorithms, memory management and others.

A list of all subjects can be found here: http://academicearth.org/subjects/

%d bloggers like this: