Type to search

Decatur superintendent says AJC article example of ‘how to lie with statistics’

Decatur slideshow

Decatur superintendent says AJC article example of ‘how to lie with statistics’

Superintendent David Dude
David Dude. Photo provided by City Schools of Decatur

David Dude. Photo provided by City Schools of Decatur

City Schools of Decatur Superintendent David Dude is taking issue with a recent Atlanta Journal Constitution article that said two of the district’s elementary schools were “ranked among the worst in the state.”

The article states that two schools – Winnona Park Elementary and Oakhurst Elementary – had some of the “highest proportional increases” of third graders failing state standardized English tests. The article compares the schools’ scores to the previous year’s scores. The article says Oakhurst had a 355 percent increase in students flunking the test, and Winnona had a 169 percent increase.

The article does not include any response from CSD officials.

[adsanity id=27331 align=aligncenter /]

Dude blasted the article on his Facebook page, calling its conclusions “extremely misleading.”

“There was a great example of ‘how to lie with statistics’ in a recent AJC piece,” Dude said. “They stated, ‘Oakhurst Elementary School saw an increase of 355 percent, as the percentage of students who scored as ‘beginning’ learners on their English Language Arts Milestones test climbed from 2.4 in 2015 to 10.9 this spring. It was the largest proportional gain in Georgia. Winnona Park Elementary had the sixth largest such gain, with 3.4 percent scoring as beginners in 2015 and 9.3 percent doing so in the spring, a 169 percent increase.’ These values may be accurate, but they are EXTREMELY misleading.

“In 2015, Oakhurst had 3 out of 125 third graders score at the ‘beginning learners’ level in English (that’s 2.4 percent); in 2016, they had 12 out of 110 (that’s 10.9 percent). Is 10.9 percent a 355 percent increase over 2.4 percent, as the AJC claims? Yes, it certainly is (okay, it’s actually 354, but who’s counting). However, the reason this is ‘lying’ with statistics is because the 2015 count of ‘beginning’ students was so incredibly low (good job, Oakhurst!) that ANY increase in the count of students would appear extreme. Remember as well that the third graders in 2015 and the third graders in 2016 are not the same third graders. Two different cohorts are being compared so there will naturally be variation between the groups.”

He said the same is true for Winnona Park, which went from three out of 87 students failing the test to nine out of 97.

“We don’t want to see ANY of ‪#‎OurKids‬ scoring at the ‘beginning’ level, and will continue our focused work to help those students grow, but don’t be misled by statistics that make the issue out to be MUCH larger than it actually is,” Dude said.

[adsanity id=32721 align=aligncenter /]