Data at work: Predicting places (in the US) that are most vulnerable to disease.

This news story focuses on parts of the United States that are particularly vulnerable to outbreaks of of preventable diseases.

https://www.cnn.com/2018/06/13/health/us-hot-spots-from-opting-out-of-vaccines-study/index.html

I share this story not to open up the vaccination debate, but rather to show how public health officials track aggregate data about vaccination rates to see where disease exposure is most likely to lead to a serious outbreak. They use vaccination rates obtained from schools, because that's where vaccination rates can easily be tracked. Students in public schools must present proof of vaccinations or have some sort of alternate forms or approvals. The laws vary by state, but exemptions from vaccinations can be granted for religious reasons, medical reasons, and sometimes for personal choice.

The vaccination-autism link originally reported in a since retracted study led many parents to opt out of vaccines for their children. Even after the retraction, many continue to opt out and believe that vaccines will do more harm than good (the data suggest otherwise,* but as with anything there are some risks associated with vaccines).

We look to the data, and it shows that with the rise of vaccines we saw the fall of related diseases. Those diseases are less common now, but not eradicated. These data suggest that vaccines are effective at lowering rates of disease by both preventing/lessening severity of disease in vaccinated people and also making it less likely that unvaccinated people will encounter the disease. In that sense, vaccinated people not only protect themselves, but also fellow citizens. However, as rates of vaccination decline, more people are believed to be vulnerable to disease should they encounter it. And that is the data story related to vaccines.


*This is a tricky data story overall. One reason why the vaccine-autism link was accepted by so many people on the basis of a single study was because signs of autism emerged in many children at about the same time that they were receiving many vaccines. There is a correlation between the age when certain vaccines are given and the youngest age at which autism tends to be diagnosed. However, correlation does not equal causation.

Are we our data or our stories?

Back to a K-12 topic ...

Stumbled on this story in the Washington Post about testing and school grades.

N.C. teacher: Test score says the year was a dismal failure for my student — but it really was ‘a resounding success’
It's easy to just use test grades as a metric, but should progress count too? And should we consider other factors?

US Birth Rates, by the numbers

An interesting, data-based explanation for why the US birth rate has declined:

https://www.huffingtonpost.com/entry/key-reason-birth-rate-declining_us_5b0725cfe4b0568a88097feb?ncid=APPLENEWS00001

This is a heated topic, for sure, and one on which many people have an opinion. I think it's interesting because the article really shows birth rate as a systemic issue.

How much family togetherness can you handle amid the current political climate?

This morning I came across a super interesting story involving analytics in the NY Times:

Thanksgiving Got Shorter Following the 2016 Election, Study Says. You Can Guess Why.

The authors used cell phone location data and voting data to come to their conclusion. If you've got the time to read it, I'd be curious to hear what you think.

Tools. Tools. Tools.

Specifically, data visualization tools.

Thought this little review might be of interest to some of you:

http://bigdata-madesimple.com/review-of-20-best-big-data-visualization-tools/

As we wait for what is possibly the first named storm of the season ...

...the NY Times has come out with something super cool.

Check this out:
https://www.nytimes.com/interactive/2018/05/24/us/disasters-hurricanes-wildfires-storms.html

It’s a data-driven exploration of locations in the US that have experienced natural disasters.

What do you think? Do you feel safe in your current location based on these data? If not, where would you rather be?

Free shipping and returns are great until ...

Who else is an Amazon Prime member? Fess up!

You love the 2-day shipping? Maybe also the streaming video? Whole Foods discounts?

If you return a lot, be careful -- you may get banned as an Amazon customer.

Well, last night this story caught my eye:
https://www.wsj.com/articles/banned-from-amazon-the-shoppers-who-make-too-many-returns-1526981401

I don't subscribe to WSJ and could only read the lede of the article, but that was enough. Did some googling and found out that too many returns = canceled account.

It's been an issue with Best Buy, too.

But how many is too many? And is it a hard number? Based on percentage of purchases? Value of returns? There are so many possibilities, right?

Apparently there are companies (e.g., Retail Equation) that will run an algorithm on your customer database and decide who  falls outside of the norm. And those folks? Banned!

(This weighs on my mind because I have shoes to return to Zappos ...)