Karl Wheatley: Has Direct Instruction Banished Exploration? Not So Fast!
Guest post by Karl Wheatley
A recent guest post here by John Thompson, Neither Teacher-less nor Teacher-proof: Constructivism Meets Guided Instruction, led to a lively discussion in the comments. I asked one reader to expand on his thoughts, and this post is the result.
Here we go again ... How much should teachers provide direct instruction/guidance and how much should they allow for child-initiated learning and jointly-planned activities?
I believe that "big-picture" research provides strong support for substantial child-initiated and jointly-planned learning (e.g., play, projects, emergent curriculum) at every grade level. Some folks consider this crazy:
Is Direct Instruction Clearly Superior?
In the Spring 2012 American Educator, Clark, Kirschner, & Sweller claimed that this pedagogical debate was over and that direct instruction had won decisively:
Our goal in this article is to put an end to this debate. Decades of research clearly demonstrate that for novices (comprising virtually all students), direct, explicit instruction is more effective and more efficient than partial guidance. p. 6
They claimed that one area of memory research proves that fully guided instruction is the only thing that can work, and seemed frustrated with teacher educators such as me:
We wonder why many teacher educators who are committed to scholarship and research ignore the evidence and continue to encourage minimal guidance when they train new teachers. p. 11
Multiple Disciplines Say Otherwise
OK, OK, I'll bite. My doctorate is in educational psychology too, and I wonder why researchers ignore the evidence against a steady diet of direct instruction and ignore the evidence in favor of substantial child-initiated and jointly planned learning.
From attachment to motivation and from creativity to student conduct, there are at least a dozen areas of psychological research and theory arguing against a steady diet of teacher-directed instruction. One fundamental issue is that researchers exploring self-determination theory have found strong evidence that autonomy is crucial to human motivation, development, learning, identity formation, internalization of values, and to mental and physical health. Thus, meaningful choices are simply essential to healthy development and learning.
Views of Effectiveness Begin with Our Goals
Yes, there are times when direct instruction is useful, but as an all-purpose instructional model, its effectiveness has been vastly overstated, and misunderstood. In a nutshell, the claim that direct, explicit instruction is more effective and more efficient than partial guidance is absolutely true in one way and absolutely false in another way.
Door #1: If our goal is to accelerate short-term learning of predetermined and easily tested academic knowledge and skills (regardless of broad and long-term effects), then direct instruction would be judged to be clearly more effective.
Door #2: However, what if we want what works best in the long run for the range of goals we value most for children, including real-world competence in subject matter plus creativity, love of learning, initiative, problem-solving, independence, critical thinking, citizenship, good decision-making, communication skills, leadership, and to be caring, happy, and healthy? If we really want this, then education with substantial child-initiated and jointly-planned learning is clearly superior.
After decades of researching people's top goals for children, I know that parents and employers consistently pick Door #2 (see Wagner, 2008).
Why "Scientific" Educational Research Can't Solve This Debate
Under NCLB, our current definition of "scientifically based research" is so flawed that a teaching method can be judged to be "evidence-based" if it reliably improves some narrow academic knowledge or skill (e.g., decoding, math facts) in the short run, but reliably makes five important things worse in the long run. Even fad diets can be "proven" effective in the short run with large-scale well-executed studies, despite setting in motion the mechanisms (cravings, reduced metabolism) that lead to long-term failure. Similarly, a steady diet of directive teacher instruction sets in motion a range of developmental mechanisms that undermine long-term development and learning. Thus, the way we are currently judging "evidence-based" practices is not good applied science or "gold standard" anything: It is fodder for a Comedy Central skit. Given our fatally-flawed definition of "scientific" educational research, I tell teachers that many so-called "evidence-based" practices actually do more harm than good.
Now some evidence for child-initiated and jointly-planned learning:
The Power of Play
In the 1970s, Germany was debating what kind of kindergartens to have, so teams of researchers conducted a major experiment, and followed children from 50 academic kindergartens and 50 play-based kindergartens through age ten. Results? Children experiencing play-based kindergartens did better than children from the "academic" kindergartens on all outcomes--social, cognitive, and language outcomes; industriousness, creativity, etc.
The benefits of play are widely acknowledged among early childhood educators. Other studies have found various benefits of play (greater creativity, initiative, prosocial actions, self-control, later academic achievement; see Zigler, Singer, & Bishop-Josef, 2004) including a ten-nation study of 1800 children that found that language outcomes were better at age 7 for children whose preschool experiences had emphasized free choice activities (Montie, Claxton & Lockhart, 2007).
Stipek et al. (1995) found that compared to children from play-based programs, children from direct instruction programs had more negative motivation outcomes, lower expectations of their own academic success, were more dependent upon adult approval, and showed less pride in their work. These advantages for play-based programs held true at both the preschool and kindergarten levels, and for both middle-class and economically disadvantaged children (in Zigler, Singer, & Bishop-Josef, 2004, p. 70).
Finally, while substantial childhood play is a commonly reported characteristic of highly creative individuals, Brown (2009) found that a common feature of murderers in Texas prisons was an absence of play in their childhoods. Play is simply one of the main pillars of healthy learning and a healthy life.
Whole Language and Free Voluntary Reading Work Better
The poster child of being misled about educational effectiveness came from the National Reading Panel (NRP), whose summary report made strong claims that direct and de-contextualized skills instruction is scientifically proven to work better. However, after spending roughly $6 billion dollars, kids in classrooms that faithfully implemented this "scientific reading instruction" showed zero advantage in reading comprehension when compared to kids from control classrooms. Contradicting assumptions of direct instruction advocates, better reading subskills gained through direct instruction did not translate into better long-term reading comprehension.
Furthermore, direct, de-contextualized reading instruction has been found to yield less positive attitudes towards reading and poorer writing than more child-centered alternatives such as free voluntary reading and whole language (see Coles, 2004; Krashen, 2004). Direct reading instruction has also been a curriculum bully, pushing aside other important subjects and recess. Thus, a more complete accounting of the effects of direct reading instruction reveals a wide swath of collateral damage. A closer look at the research actually reveals that whole language and free voluntary reading are broadly superior in the long run for the range of goals we value most, while yielding equal or better reading comprehension. The NRP made a crucial error that obscured this reality--assuming that narrow and short-term test score gains translate into broad and long-term educational effectiveness. Most educational efficacy research makes the same error.
Open Education: Better than "They" Said It Was
Despite the myths, the progressive form of education called open education--with substantial child-initiated learning and jointly planned projects/investigations--seems to have been broadly superior to traditional direct instruction. That is, Walberg's 1986 meta-analysis of over forty studies on open education versus traditional education found open education to be superior in many ways (students' creativity, problem-solving, motivation to learn, critical thinking, and acceptance of diversity) while yielding only slightly lower test scores.
We know from various areas of research (e.g., the reading research above, math research) that when direct instruction yields superior test scores to more child-centered methods, these advantages often fade or disappear, or do not translate into superior long-term, real-world competence. Indeed, in one study, low-level academic skills (de-coding) gained through intensive direct instruction overestimated real-world competence (comprehension) by three full grade levels. Thus, the likely meaning of test scores depends heavily upon the conditions in which they were achieved. Given this, the only slightly lower test scores of open education students, coupled with their better problem-solving, critical thinking, and creativity means that open education achieved both a broad advantage on cognitive outcomes and a decisive advantage in student motivation. This success is especially remarkable because open education teachers were usually new to this approach and often didn't take learning and content as seriously as progressive educators do today.
Making High School Not Just "Another Brick in the Wall"
How should we prepare kids for college and adulthood? Give them real choices about their learning every day. They are almost adults, so treat them that way. What might happen?
In the Eight-Year Study (Chamberlin, Chamberlin, Draught, & Scott, 1942), thirty high schools around the country turned traditional education "on its head," particularly for college-prep students. Education became interdisciplinary, conceptual, experiential, collaborative, often un-graded, and fashioned jointly by teachers and students (e.g., "projects"). Researchers followed the high school-to-college trajectory of 1500 experimental students, compared to an equal number of matched students at traditional high schools. As Kohn (1999) summarized: "Experimental students did just as well as, and often better than their traditionally-educated counterparts on all counts: grades, extracurricular participation, ... drop-out rates, intellectual curiosity, resourcefulness."
Significantly, it was the most non-traditional high schools whose students did the very best in college. Students simply need more freedom and responsibility for their learning.
How Directiveness De-rails Later Learning and Development
Young children are naturally good learners, so it's essential to understand how consistently directive teaching derails healthy motivation, initiative, creativity, and more.
A pair of experimental studies in 2011 showed how, compared to a discovery approach to a new toy, a traditional instructional approach immediately made young children become more passive, less creative, and develop greater dependence on others to tell them what to do (see Gopnik, 2011). As Gopnik noted, "In other words, direct instruction made the children less curious and less likely to discover new information."
While motivation to learn steadily declines with traditional instruction (Lepper, Corpus, & Iyengar, 2005; Walberg, 1986; Wheatley, 2012), the attitudes towards learning of children in open education classrooms became increasingly positive (Walberg, 1986). Besides steadily eroding intrinsic motivation to learn, highly controlling teaching interferes with internalization of values, and other features of traditional instruction and assessment cause motivation problems-- ranging from a fixed view of ability, poor efficacy beliefs, and a focus on performance goals rather than on learning itself.
When teachers in the early grades are highly controlling and provide no time for child-initiated and jointly planned learning, the broad competencies underlying decision-making and self-regulation atrophy or never develop well. Thus, when a 10th grade teacher feels that their students can't handle project-based learning or any choices, it is because their students didn't get much (or any) practice self-regulating and jointly-managing their learning in the ten previous grades. Students need frequent and meaningful choices about their learning, so they learn to make good choices, under supervision, while the stakes are lower. If education in the early grades embraced this approach, the possibilities and realities in upper grade teaching would be very different.
When we ask the wrong questions in life or in research, we get the wrong answers. Given what we know about children, education, and life, "How can we improve 'student achievement' fastest" is clearly the wrong question. However, we too often use this as our guiding question, and the result is education that tries to squeeze maximal academic gains out of the child today, but does so in a way that undermines the child's learning capacity for thousands of tomorrows.
How will the Common Core affect this?
The common core standards are likely to make this problem far worse, because they are likely to give rise to even more standardized curriculum that children find even more disconnected from their interests and abilities, and thus, even more alienating. Given that student engagement is a central problem in American education, it is also troubling that some advocates of common core standards seem to be delighting in the fact that the common core standards focus very little on students' personal and emotional responses to text. This is the disturbing perspective of people who confuse education with manufacturing or who have spent too little time with real children in the process of learning. Sadly, education has been taken over by people who simply do not understand that when we disconnect the curriculum from the children, the children often then decide to disconnect from the curriculum--and from learning itself. Instead of engaged learning, we often get numb and half-hearted compliance ... or fierce defiance.
The common core standards may be a useful reminder to teachers of some student outcomes they might otherwise overlook, but it is already clear that many of the outcomes in these standards either should never be taught directly or don't need to be learned at all, and many of the student outcomes that matter the most were left out of the common core standards. More fundamentally, the approach to education that works the best in the long run for the goals that we value most does not always begin with pre-determined objectives, but rather, frequently takes students' interests as the starting point, and features a substantial amount of emergent and highly individualized curriculum. Standardization may work wonders for making cars but standardization is a disaster for educating children.
The Main Question: Door #1 Effectiveness or Door #2 Effectiveness?
The main question for parents, educators, and all Americans is whether we want Door #1 effectiveness or Door #2 effectiveness:
Door #1 Effectiveness: Faster short-term gains on testable academic outcomes, but no better (and often worse) real-world competence in subject matter in the long run, and poorer outcomes on a broad range of the goals we value most.
Door #2 Effectiveness: Slower short-term gains on testable academic outcomes, but similar real-world competence in subject matter in the long run, and better outcomes on a broad range of the goals we value most. Oh yes, it's also a lot more fun and fits democracy better!
If we truly want Door #2 effectiveness, then education with substantial child-initiated learning and jointly planned learning is consistently superior. This is a blog, so I was brief, but you can read the rest of the argument, and how we work towards "big-picture" education, when I find a publisher for my book.
For now, here's the real question: Should we keep on doing what feels effective in the short run for boosting test scores or ... would it be better to start doing what's actually most effective in the long run for the goals we value most for the whole child?
What do you think?
Karl Wheatley is a veteran early childhood (PK-3) teacher educator at Cleveland State University. He has worked with children from infancy through adolescence and has been a teacher, director, and co-founder of educational programs. Karl is passionate about healthy motivation, meaningful curriculum, de-bunking education myths, and promoting wise education policies. He is currently working on a book, tentatively titled, Big-Picture Education: De-bunking Education Myths, Reframing the Debates, and Pursuing our Highest Ideals.
Brown, S. (2009). Play: How it shapes the brain, opens the imagination, and invigorates the soul. New York: Penguin.
Chamberlin, D., Chamberlin, E. S. Draught, N. E. & Scott, W. E. (1942). Did they succeed in college? The follow-up study of the thirty schools. New York: Harper & Brothers.
Hirsch-Pasek, K., & Golinkoff, R.M. (2003). Einstein never used flashcards: How our children really learn, and why they need to play more and memorize less. United States: Rodale.
Kohn, A. (1999). The schools our children deserve: Moving beyond traditional classrooms and "tougher standards." Boston, MA: Houghton Mifflin.
Wagner, T. (2008). The global achievement gap: Why even our best schools don't teach the new survival skills our children need--and what we can do about it. New York: Basic Books.
Walberg, H. J. (1986). Synthesis of research on teaching. In Handbook of research on teaching, 3d ed., ed. M.C. Wittrock, 214-229. New York: Macmillan.
Wheatley, K. F. (2012). How "healthy motivation" can help transform education. Encounter: Education for Meaning and Social Justice, 25(1), retrieved here.
Zigler, E. F., Singer, D. G., & Bishop-Josef, S. J. (eds.). (2004). Children's play: The roots of reading. Washington, DC: Zero to Three.