After looking at what was available in terms of datasets in the recent release of diabetes-related deaths in NYC, I saw an opportunity to do some experimenting.

It  resulted in the blog posts:

  1. Quick Look at Diabetes in Queens 
  2. Diabetes Related Death Rate Rise Throughout the City

I saw with the different datasets I could test out new tools and new features in tools I’ve already been using.

Also this would be an opportunity to try to attract eyes using social media by using multiple charts and factoids to lead them into the piece.

For example while the lede statistic and chart might be the citywide numbers, I would also use social media to publicize the analysis of the charts in the body of the blog like the demographic numbers and possibly hook them into the entire piece from there.

Also, the idea was to create simple charts with analysis throughout the week, leading up to the citywide map. So I would be moving from a focus analysis with charts of Queens to analysis and charts of the city as a whole and hopefully getting people in between.

About the Social Media Strategy

While I think the strategy has merit, it didn’t work this week in part because of big news events like the ruling on DOMA case and the votes on Immigration bill.

This might be a great strategy for this topic on Diabetes Awareness Month in November or other times with health related hooks.

In addition, while I used HootSuite to schedule the social media post in the time of day when there was the most social media engagement I didn’t use tools like Twubs to look for the best hastags to join into already existing conversations.

Also, I only used Twitter, Facebook, and LinkedIn. I should have tried to use Pinterests as well to spread the news.

It may have been better to start out Monday but the data analysis and testing out the new tools did take up most of the early parts of this week.

Some Insights About Tools

Google Pivot Table
While I’ve used it to sort out my personal budget information, this was the first time I did it for a study with tables having multiple columns. It has an amazing way of being easy and fast in filtering data.

However, it is also an easy way to get lost in the data, sorting out different combinations, adding up different categories.

While those are great ways of discovering new things it is also a great way to drain time. So you do need to be more focused than ever while using this tool.
While I knew of, this was the first time I got a chance to actually use it on a project. I found the tool to be easy to use and quick in creating beautiful interactive graphs with a few quirks and customization limits.

But like any tool once you’re aware of those quirks and limitations you know when to use the tool and when not to.

Some of them include an inability to adjust chart widths, add grids to line charts and use greater than or less than symbols in the description text.

Google Fusion Tables:
I originally thought of making two separate google fusion tables generated maps, one of the CDs with an increase in the percent of diabetes-related deaths and another of the CDs that had a decrease.

I realized it was better to actually have them both in the same map but having already created two maps the fastest way I saw to do this was to use Fusion Tables Layer Wizard to combine the two.

The problem was that the fusion table generated legends didn’t come along with the end result. So I screen grabbed the original legends and posted them along with the map instead.

It was a quick way around the problem and nice to know the limits of this tool now as opposed to later.

Other Thoughts

Aside from picking a better time to hook a specific week long data project like this, I think reversing the order of what is focused may be better. So instead of going from borough to citywide perspective, maybe go from citywide to a specific borough, possibly multiple boroughs. It might be able to draw more people by casting a wider net in the beginning.

This will mean making the map first, the more time consuming tasks. This might also mean starting a week before with the data analysis. But having done the hardest part first, it might be easier to concentrate on the other data-visuals and topics.

You might even have time to work on multiple boroughs. You can even create separat maps for specific borough since all you would need to do is filter the original citywide map.