Archive for January, 2015

A journalist’s crash course in data visualization

Monday, January 26th, 2015

Last year, I was asked to produce interactives for CityNews, engaging and informative data visualizations that the TV shows could use to augment their coverage of stories and drive traffic to the website.

Neither I nor anyone else on the web desk knew how. So I hunted around and found a workshop sponsored by the Canadian Association of Journalists and jointly led by CBC’er David McKie last May to learn how to use Google fusion tables as well as the proprietary software product ArcGIS, which yields similar data-rich maps. But that was just scratching the surface.

Over the summer, I came across a tailor-made package of data journalism webinars by Poynter, the online training site for journalists and learned in my off-duty hours about creating interactive charts, tables and maps with Google spreadsheets and fusion tables. There was also a webinar on Tableau Public which in my opinion turned out to be even more powerful than fusion tables but somewhat unwieldly and harder to tame.

One of the bread-and-butter stories that CityNews often covers is murders. If it bleeds, it leads, right? So natuarlly one of my first data ideas was for an interactive map of Toronto homicides in 2014. Click here to view. (At some point, I’ll do a visualization of the data from the last 20 years from Toronto police.)

That was followed by many other interactive maps on Nuit Blanche exhibits, Ebola screening airports and commute times by city wards, as well as more sophisticated ones using shape files to show how Toronto mayoral candidates did by wards, monthly transit fares by city, global gas prices, homicide rates going back 50 years and city watermain breaks going back two decades. When maps weren’t “the right way” to show information, I used interactive charts to explore poll results, breakdown of Toronto crimes and most in demand GTA jobs, to name a few. When I didn’t have access to proprietary tools, such as a before/after slider, I co-created an in-house version. But I’ll devote another blog post on that later.

In addition to interactive maps and charts, I also experimented with infographics but never found a tool that I could really champion and recommend. I used Piktochart and but the former was limited in the number of photos you could upload when using the free version and the latter had software glitches that rendered it useless. I’d say they were all just meh. Thinglink, which makes images interactive, was neat but all infographics were too time consuming to build.

Along the way, I encountered some challenges. Aside from coming up with the questions, the No. 1 challenge was finding the datasets, ie., getting the information and in a format that’s easy to manipulate in spreadsheets. Another issue I faced was how to explore a huge dataset to find trends. One example was Toronto parking tickets. I parked that one aside, pardon the pun, because I didn’t know enough about databases to crunch the numbers for me. But I look forward to exploring some so I can write about which streets received the most tickets, etc.

In conclusion, this is only the beginning of my data journalism. I’m looking forward to learning from others at the upcoming IRE CAR conference in Atlanta in March and taking that knowledge to produce even more original and powerful data visualizations in the near future.

Top of page