Fluid Intelligence, Part 1: Ryan Holmes

Guest blog by Ross T.

I recently got wind of a randomized control trial that Match Next—our blended learning pilot for 4th graders—will run in the fall with MIT post-doc Allyson Mackey. It’s a study of the effects of cognitive game play on fluid intelligence processes like working memory and reasoning. Match Next is already a burgeoning education laboratory, so I wasn’t altogether surprised until it came out that it wasn’t the scientist who proposed the study. It was our own Ryan Holmes.

Back when I used to play pre-dawn pickup basketball with Ryan, I wasn’t aware that he was on his way to becoming one of the best math teachers in Boston. We’re talking multiple years with top-5 student growth percentages in the state. (He credits his former colleagues at Excel Academy for much of that success.) Today, he works upstairs as the Math Director at Match Next.

To outsiders like me, the Match Next floor with its pods of kids towing Kindles to extended group tutorials initially looks plain weird. But if you give Ryan ten minutes to explain the innovative practices, weird starts to make good sense. And when he’s done rapping about impacting fluid intelligence, good sense starts to sound downright cool.

“While we know that a ton of brain formation happens before the age of five,” Ryan says,  “I didn’t look at my twelve year olds and see kids whose intelligence couldn’t be deepened in really meaningful ways.”

He explains that teaching math isn’t all about rote skills. There also needs to be a problem-solving piece, and problem solving relies on working memory—how we hold and manipulate multiple pieces of information—and processing speed. “A lot of people view these capacities as static,” he continues. “You have what you have and there’s really no way to improve them. But when you teach middle school you see kids develop so much, and it’s hard to believe you can’t move the needle in those areas.”

As if Ryan’s day job weren’t idiosyncratic enough—he’s not a traditional classroom teacher but the closest thing to one that Match Next employs—he spends a lot of his own time boning up on neuroscience and cognitive development. When I asked him if he could explain this fall’s RCT, he started rattling off titles like Hurley’s Smarter and Klinberg’s The Learning Brain, and pulling up the Scientific American Mind website on my laptop.

“What we’re not talking about is crystallized intelligence, facts and processes,” he explains. “We’re experimenting with cognitive exercises that may be able to promote the kind of fluid intelligence that allows kids to navigate new problems in new contexts.” 

The trial will ask a treatment group of students to participate in quick-recognition sorting games to improve processing speed and pattern recognition games like N-back and Set to improve working memory. There’s some evidence that games like these can influence fluid intelligence, and the investigators will try to test for this effect with before and after non-verbal intelligence tests for the treatment and control groups.

Ryan does also hope to see some measurable impact on the participating students’ ANET and MCAS scores. “But these tests, maybe 80 or 90% of them are measuring crystallized intelligence,” he cautions. “So while doing well on them is very important, they’re not the be-all and end-all. If we can make a difference on those assessments, I’m convinced we will also improve results on standardized tests like the SAT that better measure log and reasoning.”

Match Schools have become places of genuine social science research, and have already participated in a handful of randomized control trials, from the efficacy of parent phone calls in the Match High School to the impact of Match Teacher Coaching in New Orleans.  And if Ryan has his way—he’s already toying with the idea of an education podcast—Match Next will be no exception.

This blog will continue with Part II on the particulars of how the trial will be organized in the fall. Stay tuned!