Why Data Quality Still Matters, Now More than Ever

In my presentation, Why Data Quality Still Matters, Now More Than Ever, (that I created in Microsoft Sway–give it a try sometime!), I point out that while data quality impacts the quality of your maps, there is an even more important concern:  You are responsible for ethical, wise decision making, no matter what role or job you have.  Data quality affects your decisions and those around you.  In the presentation, I provide information about data quality standards, past and present.  But my goal is to extend the things that I and Jill Clark have writing in this blog and in our book, and help people to remember “best practices” while on the job working with data and tools, day by day.

I believe that geospatial technologies are often and rightly described as “powerful.”  With power comes the ability to cause harm – intentionally or unintentionally – as well as to do good.  In the context of GIS, practical ethics is the set of knowledge, skills and abilities needed to make reasoned decisions in light of the risks posed by geospatial technologies and methods in a wide variety of use cases.   I believe there are 7 key reasons why data quality still matters in this age of ubiquitous maps and data:

1.  Maps tend to be believed.   They still have the air of authenticity, even in this age when just about anyone can create a map, and with the spread of crowdsourcing.  In their recent blog in The Map Room, the authors state that bad maps (such as “favorite food in each state” – really?  Says who?) generate lots of web traffic, which is the chief reason they are created.  I have included many bad map examples in my presentation that I hope help people think carefully about maps as powerful tools that can lead–or mislead.
2.  Maps are easy to make.  We frequently discuss the advent of web GIS in this blog including the ability to quickly make maps in ArcGIS Online, for example.  I love this development, and wouldn’t want to go back to the challenging early days of GIS for anything.  But with ease of creation, shortcuts can be taken, and data may not be verified in someone’s haste in easily creating a map.
3. Maps are often attached to “fun” online posts.  I’m not trying to be Mr Grumpy Pants here, and some of these posts are truly funny and interesting.  But the danger with these posts is that it often reinforces point #1 above.  I include the example of the supposed voyage of the SS Warrimoo in 1899 in many data quality presentations I give.
4.  Maps are not reality, but only representations of reality.  Useful ones, to be sure, but they are still representations.  One example I give in the presentation is the deciduous-coniferous tree line.  Is it really a line?
5.  Be critical of the data–even when it is your own!   Even when YOU generated it.  My example in the presentation is a GPS track that I collected and mapped.
6.  Thoroughly check your data sources when you are using them.  This often requires extra homework, some hearkening back to “old school” methods of actually -gasp – calling the data creator on the phone.  See my example of mapping Lyme disease in Rhode Island.
7.  Scale still matters. When you zoom in, the accuracy does not increase as you zoom in. Sounds crazy, but I get the distinct impression from many people that they think that because you can view the data at 1:1000 scale, that the data are spatially accurate to that scale.  See my example of “walking on water” on a pier on Lake Michigan to drive home this point.

This presentation is the result of expanded research I did after Jill Clark and I wrote about this topic in our book, and after I wrote about this topic in Directions Magazine, here.  I look forward to hearing your reactions to the points I raise in this presentation.

My presentation available online, Why Data Quality Matters More Now Than Ever.