Working With Us | Products | Case Studies | FAQ | About Online Media

‘Raw’ng Data

Nov
7
2007

We’re all pretty much a group of right brained thinkers in my family. But there is a rogue element around here that manages to assign a numerical equation to just about any situation that exists. They do this for fun.
Heir 2’s idea of a fruitful weekend is when he prints out personal statistics for every member of his cross country team. My brother keeps a chart of what plays on his CD player when it’s set on random so that he can tell if the term “random” can be legally applied. And Dirtman – please don’t mention anything about water output unless you want to spend the next hour discussing onsite system sizes and depth of placement.
The World Series is their time to shine, not because of the drama of the competition, but because this is the first time since 1935 that this league has entered this competition with a .49859608 average in outfield errors during a first quarter moon in an odd-numbered year. On Astroturf. Or something like that.
Anyway, that was the gist of this weekend’s after-dinner conversation. I couldn’t really go into detail because I nodded off at some point when my son piped up with the information that somebody was 39.574 percent more likely to be drafted somewhere than somebody else. The next thing I knew they’d woken me up with the loud argument that,
“Before you can assign a number, you have to decide if, indeed, it is a negative or positive factor!” This is what passes as controversy in the land of left-brainers.
I raised my head up groggily and said, “Ah-hah! If you are ‘deciding’ something, you are admitting that there are elements in statistical decision-making that are subjective.”
Three sets of eyes roll as my brother attempted to explain the “Something Something” factor. But I was nodding off.
My family pretty much confines their use of statistics to boring each other to death. People in general, though, seem to think statistics can prove any case and I would totally agree. You can find statistics to prove just about any case you set out to prove. In fact, you can start with the premise and I guarantee somewhere will be a study with the statistics you desire. Or you can just pay for one.
Or you can pay for a study that reveals absolutely nothing and assign meaning later and no one is the wiser – unless you start stepping on toes. Or so one Johns Hopkins team of bean counters found out last week.
The JHU researchers used U.S. Department of Education data to prepare a report for the Associated Press on high school drop out rates across the country. Are there enough big, important institutions listed in that sentence to impress you enough not to question the results any further? Researcher Bob Balfanz released the report and labeled any school with a low enrollment to graduation ratio “drop-out factories.” We’ll ignore the fact that the researcher is now bestowing a value judgment on his own research.
Let’s just focus on the outrage experienced by angry tax payers in northern Frederick County, Virginia, to find out one of their premier schools, James Wood High School, appeared near the top of the list for the entire commonwealth. Board of education and school officials were dragged on the carpet, press conferences were called, legal counsel was consulted, and lawsuits were threatened.
Turns out that the reason not as many seniors graduated as enrolled is that during that four-year period, the county built a new high school. Almost half the enrolled freshmen graduated from a different high school.
Oops.
In the first swell of criticism of the report – both from Fredrick County and other jurisdictions — Johns Hopkins researcher Chris West defended the results, stating the data used were just “raw enrollment numbers.” But it was not releasing a report merely giving an overall failure rate of high schools across the country. They were stigmatizing specific schools as (insert dangerous music here) Drop Out Factories (echo, echo, echo. . .).
The data did not reflect factors like new growth, rezoning or change in local industry causing family relocation – all major reasons high school enrollment and graduation would show a disparity. The numbers separate from the communities they represent are useless, rendering the report – along with its conclusive catch-phrase – meaningless.
As it turns out, John Hopkins dropped James Wood High School from its list of Drop Out Factories.
Frederick County School Superintendent Patricia Taylor said in a press release, “I hope this case will result in individuals being more careful when working with information that’s not well understood and can be misreported.”
Or, as my analytical, statistic-loving Heir 2 sneered: “Duh…

Share  Posted by Jeanne Jackson at 9:57 AM | Permalink

<< Back to the Spotlight blog

Jeanne Jackson's bio
Email Jeanne Jackson




Get Our Weekly Email Newsletter




What We're Reading - Spot-On Books

Hot Spots - What's Hot Around the Web



Spot-on.com | Promote Your Page Too

Spot-on Main | Pinpoint Persuasion | Spotlight Blog | RSS Subscription | Spot-on Writers | Privacy Policy | Contact Us