Data permeates my conversations (and thoughts), as I bet they do yours (you just might not make the conscious connection). Over the past month there have been many conversations in this regard. For instance, I recently introduced a colleague to the ideas of creative and different data displays of Edward Tufte and his thoughts about “visually vibrant displays.” Those ideas lead me right into what I learned from Bruce Wellman and Laura Lipton (PDF) about using data as a “third point of reference” in a conversation or to start a conversation. Part of that thinking is that when it is just data, it isn’t mine or yours. Then some of us were brainstorming about data displays and the need for them to be self-explanatory – no legends or keys – just clear labels, color-coded so that no one needed to explain the data. Then there has been the data my girls and I have discussed – data collection about tasks at work to help build a business case or clarify information. And finally a new project started and data analysis was part of the initiation of the project.
For me, any data should be a conversation starter and more data, while it may help you answer some questions, should lead to more questions. That said what sparked this post was a recent article in JSD by Alyson Adams et al, entitled Talking Points (registration required). The focus of this article is about using data as a tool in instructional coaching. There were many gems about data that seemed applicable beyond that focus.
For example, the authors offered six questions to help determine if your data display is high-quality. I’ve adapted four of them below:
- Is the data display responsive to your questions? Sometimes we try to use data to answer questions that the data might not be able to answer. Sometimes how we display the data causes us to lose sight of our original question or make it difficult to “see” an answer.
- Are data non-evaluative, descriptive and easy to understand? Do you need to be a psychometrician or statistician to read the data? Could your grandmother make sense of the information, or at least be able to read it?
- Is the data display brief – no more than two pages long? I love this one. Sometimes we try to have data conversations with a binder’s worth of data. What’s our focus? Can we see it easily?
- Does the data display focus on a key area? When we talk about triangulating data, we have multiple data points we are looking at. Even in that specific instance it is important to keep the data focused. Go back to question #1 to keep your focus.
Here are some of the gems from the article:
…the quality of the data display dramatically influences the conversations that occur.
Clear, descriptive data enable to [owner] to feel ownership of successes and challenges…
The data display leads the conversation.
…developing a high-quality data display is a new and challenging skill.
The data display must specifically provide data that match the guiding question.
For years I’ve subscribed to a couple of blogs about data displays (probably Nathan Yau’s for the longest time) and have enjoyed watching Hans Rosling’s videos. All of these cause me to think about how I can display data differently so that it leads the conversation…the data itself get people started looking at things that are interesting, asking questions, noticing points that surprise them, seeing patterns or trends, and being open to the conversation. One way to start all this is with “visually, vibrant” displays.
The second part is having data as the “third point,” which ideally translates a couple of ways – one large data display that everyone views or by having folks share (groups of 2-3) data. Sharing data rather than each individual having the data engenders conversation from the beginning.
I’m going to borrow a display idea from Adams’ article; it was about wait time. A few years ago I shadowed my eldest through a day in her last semester in college. There were multiple purposes and part of the data I collected was about wait time. Here is what I observed in an advanced math class.
What questions would this data generate for you? What patterns might you identify? Where might your conversation go?