Beware of Data Bait
—Annie Daniel
Last week we were given a dataset on texting-and-driving ordinances from the Texas Legislative Council - a group that does research and a host of other things for the Texas Legislature and its agencies. They sent us an Excel file with city, population, URLs to each city’s texting ordinance, crash data for several years before and after the ordinance was implemented, and ticketing data.
But is it ever that simple? For datasets like these, the research is often done for a particular legislator who might have an agenda, and the data is rarely as simple as it looks. That’s not to say all data with an agenda is scary- just that it’s worth looking extra closely at data that falls into your lap for anything that could be baiting you into making something that proves their point but might not be journalistically accurate.
In this case we had to be careful with the crash data. It was tempting, but how do we know who they talked to and how exactly they came to those numbers? The data could have been initially compiled for a representative with a dog in the statewide texting-and-driving ban fight. Plus there were only 38 cities, and some of the ordinances were introduced within the past two years, so the three and four year post ordinance data didn’t even exist. Even though the crash data was cool, it had a ton of holes, and we had to leave it out.
Instead, we took a look at what we do have: a bunch of cities with texting-and-driving bans, the ordinance, the type of ban and the year it was introduced. Still interesting and useful and totally worth making a fun thing.
I settled on a select box with the list of cities that returned cards with the city, type of ban, penalty, year it was adopted, and a link to the city’s ordinance. I also included some filters for the types of bans as a way to compare the 38 cities. No charts, no maps, just a quick embed with easy to absorb information out of an interesting and potentially disastrous dataset.














