Austin, Texas -- This is the Austin in Austin blog. Im in the Texas capital after being selected by the Education Writers of America for a seminar called Diving into Data. Speakers and workshops on standard deviation, regression to the mean and bell curves, big research names and government program acronyms can I have a good time or what?
But along with the stats class refresher , these speakers also talk about why numbers matter to policy makers and voters, what they can tell us and ways they can be twisted. Heres a first-day sampling of ideas and insights from various speakers:
-- Data gives news context. Simply presenting he said-she said stories, without the weight of evidence, gives readers no solid frame of reference. University of Texas at Austin Professor Julian Vasquez Heilig noted strong stories educate communities, but too often mega-media stories get it wrong.
-- Old saying: Statistics means never having to say youre certain.
-- Pearson researcher Kristen DiCerbo studies children at play, evaluating how games can teach what shes testing. Her job is to seek out ideas for what might become something to use in schools three to four years from now.
-- High stakes testing is too little, too late to help in the classroom. DiCerbo said frequent spot checks of how well students are learning will replace or at least supplement the annual state tests (most of our districts already do this). We cant keep counting on this one big data set at the end of the year, she said.
-- At the Dallas Morning News, every big reporting project gets a humility graph, a chart laying out the information they dont have. Dallas number-crunching reporter Holly Hacker said the exercise helps everyone see the holes in stories and address them.
--States that won competitive federal education grants are tying information to student ID numbers from preschool to age 20 schools, jobs, college or jail. The information will give researchers a powerful tool to gauge what helps kids succeed in life.
What helps kids succeed in life, what helps voters make better decisions, what improves classroom instruction, is well worth taking the time to study. Diving in the data without understanding statistical standards and recognizing its limitations just wouldnt add up.
Now, Im off to buy tee-shirts and coffee mugs. Austin in Austin, out.