A few years ago I entered the entrepreneurial space as a “newly" minted applied data analyst. Although I had been working along analysts and data scientists in my work as a medical writer and outcomes professional, it was more of an observed curiosity than an immersive existence. Somewhere along the line I began asking questions. I hesitated at first not wanting to appear ignorant but quickly noticed vague responses to questions I would have thought had straightforward responses.
Curiosity morphed into agency as problems presented themselves and needed viable solutions. But whatever that sound is when a record scratches to halt forward movement in a movie--insert that here. Collaborative efforts to improve data collection and processes were not hailed as the “secret sauce” I had imagined. Here is the rub. Data literacy was lagging the needs by quite a significant gap. Most data departments (and I use the term loosely) consisted of finger pointing, a small measure of chest pounding, and gasps of “But we have always done it this way...”.
Fast forward to today and I am contractually obligated to author my first book on geospatial analytics. How I got here and why will be the subject of more than a few future posts. I want to first introduce you to an organization that does a great job illuminating the importance of thinking spatially (below).
We all respond to graphic images. Instinctually we are grounded in the what and where of an image. Any student of data visualization recalls pre-attentive attributes--the preliminary detection of the image. But we often don’t appreciate the attentive attributes as well. Attentive attributes call in to play the higher centers of the brain to make inferences following four principles as described by Eric Kandel--Nobel prize winner in Physiology or Medicine:
1. Disregarding details that are perceived as behaviorally irrelevant in a given context,
2. Searching for constancy,
3. Attempting to abstract the essential, constant features of objects, people, and landscapes,
4. Comparing the present image or graphic to images encountered in the past.
“Perception is the process whereby reflected light becomes linked to an image in the environment, is made enduring by the brain, and becomes coherent when the brain assigns it meaning, utility, and value.--Eric Kandel, Reductionism in Art and Brain Science."
What do we mean by “spatial literacy”? Let’s take a look. Location data looks at the environmental or first-order effects and the second-order or interaction effects. We can simplify processes into data-driven and model driven. But let’s not get ahead of ourselves. We are first looking to summarize the data. What are the characteristics of the data? The testing begins during model-driven analysis.
WorldPop data is the perfect place to start understanding the process of mapping and providing “high resolution, open and contemporary data on human population distributions, allowing accurate measurement of local population distributions, compositions, characteristics, growth and dynamics, across national and regional scales.” You won’t find a lot of US data here but I have successfully used the methods discovered here on US CENSUS data for example.
There really isn’t--at least not yet--a handbook to guide you through the insights needed to make granular assessments about poverty beyond quantitative assessments but applying even a few of these insights to our data questions can only improve our ability to provide a 360 perspective.
Spatial data reminds me of a useful definition I once heard of big data. It isn’t the volume that makes it big--it is the interactivity of blending different datasets to answer complex questions. This is visually appreciated when we look at layers of data integration to examine patterns in geographic regions.
The ability to explore characteristics that influence differences in population density. It isn’t enough to simply drop a pin onto a map to indicate populations--we need context.
Summaries of the workflows will continue to be posted here. If you are interested in the detailed "how-to"--subscribe below:
Graphics are from HDX Dataset Deep Dive on WorldPop’s Gridded Population Datasets.
HDX Humanitarian Data Exchange.