Saturday, October 20, 2012

What Does Science Tell Us About Teaching Kids to Think?

This is the same title of an article that came out in The Atlantic early this month. It was written by University of Virginia psychology professor Daniel Willingham. The article was in fact a response to another piece entitled "The Writing Revolution" by Peg Tyre in The Atlantic that talked about a school in Staten Island called New Dorp:
For years, nothing seemed capable of turning around New Dorp High School’s dismal performance—not firing bad teachers, not flashy education technology, not after-school programs. So, faced with closure, the school’s principal went all-in on a very specific curriculum reform, placing an overwhelming focus on teaching the basics of analytic writing, every day, in virtually every class. What followed was an extraordinary blossoming of student potential, across nearly every subject—one that has made New Dorp a model for educational reform.
Willingham then writes:

The history of education is littered with flavor-of-the-month interventions, many of which began at a model school, but that, once implemented elsewhere, flopped. Dejected educators then begin scouting for the next "big thing." 
There is, but implementing it correctly is no small matter.In this instance, a heavy emphasis on writing seemed to make kids better writers, better readers, and perhaps, better thinkers. What does published research say? Is there any reason to expect that a writing curriculum, if implemented in other schools, would bring the same benefits?

Asking the above questions is very important. These are the questions that any Department of Education must ask before embarking on any large-scale reform. And the answers lie in performing well-controlled experiments and publishing the results in a peer-reviewed journal so that experts could evaluate the work. If the questions have been addressed before by others, then a careful examination of published studies can be useful. Willinghan then points us to an article published in the Journal of Educational Psychology to gain further insights on how teaching students how to write could improve learning:
Downloaded from
To understand the above paper, it is important to know what effect size (ES) means. In this particular work, ES measures the effect of a particular intervention on the performance of the students. For example, the average score of students who were exposed to the additional teaching component is compared to those who did not receive the treatment. This difference is then compared against the standard deviation of either population. An ES close to 1 indicates a highly significant effect. It means that students who have received a particular intervention are scoring one standard deviation higher than students who have not. One standard deviation in grading is usually the separation between "average" and "above average". An ES of 1 essentially means that the intervention brings the student's performance from mediocre to the top 20%. With these statistical analyses of previous studies, it then becomes possible to zero in on the interventions that have significant effect on student learning. The authors of the above paper are then able to make the following recommendations:
  • Teach students strategies for planning, drafting, or revising different types of text (ES=1.02)
  • Teach students procedures for regulating the writing strategies they are taught (ES=0.50)
  • Teach students how to form images and be more creative (ES=0.70)
  • Teach students how different types of text are structured and formed (ES=0.59) 
  • Teach students spelling, handwriting, and keyboarding (ES=0.55)
  • Develop instructional arrangements where children work together to plan, draft, revise, and edit their papers (ES=0.89)
  • Set clear and specific goals for what students are to accomplish when writing (ES=0.76)
  • Engage students in activities that help them gather and organize ideas for their papers before they write a first draft (ES=0.54)
  • Assess students’ writing and progress learning to write (ES=0.42)
  • Make it possible for students to use word processing as a primary tool for writing (ES=0.47)
  • Increase how much students write (ES=0.30)
  • Implement a comprehensive writing program (ES=0.42)
Still, the authors add the following cautionary statement:
Just because a writing practice was effective in multiple research studies does not guarantee that it will be effective in all other situations. Rarely, if ever, is there a perfect match between the conditions under which the writing practice was implemented in the research studies and the conditions in which it was subsequently put to use in classrooms. Even if there was a good match, the safest course of action is for teachers implementing the writing practice to monitor its effects to be sure it works in their classrooms with their students.
And Willingham ends his article with the following:
And this is why I say that the New Dorp results are likely replicable, but we must pay close attention to what happened there. It's easy to remember only that "they asked the kids to write a lot." But as Tyre describes, they placed an intense emphasis "on teaching the skills that underlie good analytical writing." There was explicit teaching of writing, and the emphasis was on analysis. I would add that this instruction must be paired with substantive content in order to pay dividends in critical thinking. 
To read more articles on how to teach writing visit

Downloaded from

No comments:

Post a Comment