Homework – is it an unnecessary evil or a sound and valuable pedagogical practice? The media coverage of the debate often zeroes in on these two seemingly polar opposite views, even though they may not be all that far apart. Homework can be good until – well, until it isn’t. Assign too much or the wrong kind (or both) and the law of diminishing returns kicks in, says Dr. Harris Cooper, professor of psychology and neuroscience at Duke University, resulting in undue stress for students, aggravation for parents and no academic pay-off.
But as Cooper, author of “The Battle Over Homework: Common Ground for Administrators, Teachers, and Parents,” recently told NEA Today, homework levels and parental attitudes haven’t really changed dramatically over the years. Cooper also concludes – perhaps a shock of those who are convinced that very little in our classrooms is working as it should – “the vast majority of educators have got it right.”
There’s a lot of focus on homework now, but has it been scrutinized so heavily in the past?
Harris Cooper: Throughout the 20th century, the public battle over homework was quite cyclical. You can go back to World War I or a little after, when it was considered important for kids to exercise their brain like a muscle and that homework was a way to do that. During the 1930s, opinions changed. In the 1950s, people were worried about falling behind the communists, so more homework was needed as a way to speed up our education and technology. During the 1960s, homework fell out of favor because many though it inflicted too much stress on kids. In the 1970s and 1980s, we needed more homework to keep up with the Japanese economically. More recently, as everything about education and teachers is being scrutinized, homework has come into question again.
What’s interesting is that the actual percentage of people who support or oppose homework has changed very little over the years. And the actual amount of homework kids are doing has changed very little over the last 65 years.
But haven’t we seen an uptick in the amount of homework assigned to elementary students?
HC: There is a little bit of an uptick in lower grades. But when you look at the actual numbers, we’re talking about the difference between an average of 20 minutes and 30 minutes. So you’ll find some people who say the amount of homework being given to 2nd graders, for example, has increased 50 percent. But If you look at the actual numbers, it’s ten more minutes per night.
And probably a driving force behind that is obviously end-of-grade testing and accountability issues. Perhaps more legitimately is the importance of early reading. As they say, in third grade you learn to read, and in fourth grade you read to learn. So this has led to more reading assignments.
While most high school students are still doing approximately the same amount of homework on average, there’s a great deal of variation. That’s due to choices some kids make about how rigorous an academic program to take and the increased competition over college admissions. So there are a lot of kids out there taking four or five advanced placement and honors classes now, which might not have been the case a while back.
According to the MetLife Foundation national homework survey, 3 out of 5 parents said their kids are getting just the right amount of homework. One said too much and one said too little. That survey is a few years old now but I doubt that’s changed.
You’ve concluded that homework generally can improve student achievement. At what grade levels do we usually see this effect?
HC: There’s very little correlation between homework and achievement in the early grades. As kids get older, the correlation gets stronger. But there are experimental studies even at the earliest grades that look at skills such as spelling, math facts, etc. where kids are randomly assigned to do homework and not do homework. They show that kids who did the homework performed better.
But we’re really talking about correlation here, so we have to be a little careful. It’s also worth noting that these correlations with older students are likely caused, not only by homework helping achievement, but also by kids who have higher achievement levels doing more homework.
But at a particular point more homework is not a good thing. You’ve heard of the “10-Minute Rule,” where you multiply a child’s grade by 10 to determine how many minutes you assign per night. This rule fits the data. So 20-minutes for a second grader is where you’d start. In middle schools, it’s between 60-90 mins for 6th through 9th graders, about two hours later in high school. When you assign more than these levels, the law of diminishing returns or even negative effects – stress especially – begin to appear.
Have school districts coalesced around the 10-minute rule?
HC: From my experience, I have never seen a school district that recommends anything that isn’t consistent with the 10-minute rule. They won’t use the term “10-minute rule” usually, but they’ll say, primary school grades will be assigned up to 30 mins., grades 4-6 up to an hour, things like that. But If you translate the policy to the 10-minute rule, it’ll be very similar. Nobody has a policy that says you can expect your second-graders to bring home two hours of homework. The only place you’ll see a warning about it is in high school: you can expect half an hour a night per academic subject. Again, if the kid is taking AP, expect more.
What don’t we know about homework? Where are the gaps in the research?
HC: We need to know more about the the differing impacts by subject matter. Regarding the 10-minute rule, one question I am frequently asked is, “Does that include reading?” Generally, the answer would be yes, but if we’re interested in kids’ stress level, for example, they are more likely to burn out quicker doing math worksheets and studying vocabulary than if they were doing high-interest reading. So we really need more work on subject matter, on homework quality, on the level of inquisitiveness that it engenders and the way it motivates. Also we need to know more about the use of the Internet, especially as it relates to potential disparities between rich and poor and the ability to research at home.
Parental involvement is a huge homework-related issue. How can educators work with parents to keep their role constructive?
HC: Parental involvement is more important in the earlier grades and teachers should try to make sure that parents have the skills to teach the material so to avoid any instructional confusion. Educators should also remind parents to not place great pressure on their child and to model behaviors, especially with young children. For example, when the child is doing math homework, a parent could balance the checkbook to demonstrate how the skill can be used in adult life, or they can they read their own book while their child is reading.
Homework also keeps parents aware of what their child is learning. I’ve had some very emotional parents come to me about having been told by teachers that their child is struggling, that there might be a learning disability. The parents don’t necessarily see it until they see their child work on homework.
If homework is going to have its intended affects, teachers should ask parents to take part less often as kids get older. If support from parents is withdrawn slowly, it can promote autonomous learning – teaching kids that they can learn on their own and they can learn anywhere.
Do you think overall the current debate or controversy over homework has been helpful and what, if anything, should educators take from it?
HC: Well, I recognize that the debate will always be there, but I generally choose to ignore it, or at least the people who, as the old saying goes, use science the same way a drunkard uses a lamp post – more for support than for illumination.
Homework is probably the most complicated pedagogical strategy teachers use because it’s open to variations due to child individual differences and the home context. But the vast majority of educators have got it right. They’re not going to satisfy everyone, because kids take homework home to different environments and to parents with different expectations. But, like I said before, three in five parents are satisfied and there’s one in each direction – too much homework or too little. That probably means teachers are doing their job properly.
Photo: Associated Press
The pros and cons of NCLB often seem to cancel each other out in the debate over this controversial law. In a recent issue of Applied Measurement in Education, Lihshing Wang and a team of researchers from the University of Cincinnati bring a third evidence-based perspective to the pros and cons of NCLB by examining the research on the four following issues:
1. assessment-driven reform;
2. standards-based assessment;
3. assessment-centered accountability; and
4. high-stakes consequences.
“Only a handful of scholars and practitioners have argued in defense of standardized tests,” write Wang and fellow researchers Gulbahar H. Beckette and Lionel Brown. However, there is emerging evidence that high-stakes assessment is a potent force for bringing about improvements in student learning.
The researchers present the pros and cons of NCLB for each of the four interrelated issues and then offer a critical synthesis based on their review of the research.
1. Assessment-driven reform
Assessment-driven reform is needed to counter declining trends in SAT and ACT scores and the mediocre performance of U.S. students in international rankings such as Trends in International Mathematics and Science Study (TIMSS), proponents argue. Assessment-driven reform can have a powerful influence on school curriculum and reform, if tests are carefully designed to be consistent with the kinds of learning desired in the classroom and if there is a tight connection between cognitive learning theory, the curriculum, classroom activities and assessment items.
SAT scores declined during the 1970s and 1980s because more students aspired to go to college and took the tests, not because of performance factors. There has been an upward trend in the 1990s and into the 2000s. The Department of Education statistics show improvement in areas such as a decrease in dropout rates and an increase in high school students taking advanced courses and Advanced Placement examinations. Standardized tests undervalue the “sensitive interaction between teachers and their students in the complex, social system of the classroom.” The real problem with the education system is the fundamental misdesign of schools, lack of qualified teachers and the instability of families and communities.
While there are encouraging statistics on domestic educational performance, American school children do not seem to perform well in international rankings. “It seems clear that in the world of increasing globalization, the U.S. educational system can and should do better,” the researchers conclude. The goal of using tests is not just to measure performance but also to drive changes in alternative instructional materials, learning models and staff development.
2. Standards-based assessment
It is desirable to agree on a common core of knowledge that teachers should teach and students should learn. Without common standards, it is difficult to compare grades across teachers and schools because of local norms.
All students, regardless of socioeconomic status, race or disability, should be expected to meet common standards that challenge them to acquire content and skills that are more than just minimum requirements. Neuroplasticity research in the past decade has shown that “the critical period for learning is now considered regulatable through environmental enrichment and mental force throughout life.”
In a nationwide survey by the National Board on Educational Testing and Public Policy, a majority of teachers supported their state content standards and more than one half reported that their state-mandated test is based on a curriculum that all teachers should follow. In public polls, there is wide support for assessment.
By imposing standards on students’ minds we are, in effect, depriving them of their fundamental intellectual freedom by applying one standard set of knowledge. Standardized tests oversimplify knowledge and do not test higher-order thinking skills. State standards are externally imposed on local teachers.
These mandatory assessments cannot work unless teachers understand and accept the philosophical underpinnings of standards. One-size-fits-all standards either dumb down instruction to the lowest common denominator or condemn low-ability students to frequent failure
Few would argue against the noble goal of helping all children meet the same set of high standards. Neurocognitive research provides strong evidence that the human brain is adaptable well into adulthood. However, genetic signals play a large role in the initial structuring of the brain and there is a a limit to how much and how quickly cells enlarge and add synapses. This suggests that the human mind may lose its plasticity in learning after reaching a certain age. There is a learning cap determined by genetic as well as socioeconomic factors that determines how far and fast a student can develop during their school years. The NCLB’s requirement that all children must reach the same set of standards at the same time fails to acknowledge this.
3. Assessment-centered accountability
Standardized testing is the best alternative for comparing student performance across different education systems because human judgment is error-prone. Decades of evidence show that the quality of teachers’ tests pales compared with more rigorously developed large-scale tests. When used for purposes of accountability, standardized tests can provide more objective and less ambiguous evidence. In one international study that looked at the effects of dropping and reintroducing standardized tests in 29 industrialized countries, academic standards declined, students studied less, curricula became incoherent and selection and promotion became arbitrary after standardized tests were dropped.
Important learning outcomes are not measured by standards testing. Only self-generated professional responsibility can sustain fundamental school and student improvement. To guide instruction, teachers should constantly look for evidence from a variety of sources to make sense of what is happening in their classrooms.
Standardized tests measure little more than socioeconomic status, and teachers and administrators should not be held responsible for that or should a fourth-grade teacher be held accountable for her students’ test scores when those scores reflect all that has happened to the children before. Standardized tests fail to differentiate instruction for different kinds of kids without condemning low-achieving students to boring and unproductive schooling.
Educators have the duty to help students break hereditary and environmental barriers. A well-established accountability system must make sure that the process of accountability is legal. Without adequate funding for test development and personnel training, the accountability mandate is likely to be challenged on legal grounds. There needs to be an evaluation mechanism that captures the individual contribution of a teacher and recognizes the preexisting differences in students. The current NCLB goal of bringing all children to a level of proficiency by 2014 has been projected to be unattainable. Holding students, teachers and administrators accountable for reaching an unattainable goal will lead to unintended negative consequences.
4. High-stakes consequences
Assessment-based accountability is possible only when high stakes are associated with the results. Educators must inform themselves about their content, construction and consequences. There is a “trickle-down effect” on teachers in that they must become more reflective and critical of their classroom instruction. One reason the American educational system has failed is because there have not been high stakes for failure. Realistically, students will only read a play by Shakespeare if they will be tested on it in a final exam. High-stakes testing has the unintended consequences of improving professional development. A number of studies have found a strong positive relation between high-stakes consequences and performance on assessments.
The behaviorist theory underlying high-stakes accountability oversimplifies how human behavior is conditioned by rewards and punishments. Decades of research has shown that extrinsic sources of motivation such as stars, stickers and grades actually undermine natural curiosity and a student’s enjoyment of learning. Punitive consequences achieve temporary compliance at the cost of demoralizing teachers and students.
The fundamental criticism of high-stakes accountability systems is that they rely excessively on extrinsic motivation at the expense of intrinsic motivation. Some of the negative consequences of high-stakes accountability systems include higher dropout and retention rates, lower motivation, teaching to the test, unethical test preparation, etc. Some reports of gains have been discredited as test-polluting practices such as excluding students or higher dropout rates.
There is emerging evidence that high-stakes state assessment is a potent policy for bringing about positive changes in student learning. In a re-analysis of the gain comparison between state assessment and National Assessment of Educational Progress (NAEP), average NAEP increases were much higher in high-stakes schools compared with no-stakes schools.
In response to criticism that gains in those school could be due to high dropout and exclusion rates, the authors cite a study (Phelps 2003) that reanalyzed the dropout rates and found that they were below the national average and that the exclusion rates were the same as the rest of the nation. “Whether such extrinsically motivated score improvement can sustain life-long learning and whether such positive effects offset the negative consequences, however, remain to be seen,” the researchers conclude.
Recommended research agenda
The debate about standardized testing will continue and the pendulum will continue to swing. The researchers recommend the following action research agenda:
- Develop classroom-level diagnostic tests for evaluation aligned with state-level standardized tests.
- Include classroom teachers and cognitive-developmental and social psychologists in state assessment panels to achieve meaningful alignment of content standards and classroom curriculum.
- Offer computerized adaptive testing so that students of diverse ability levels can meet learning goals that are tailored to their current ability level.
- Conduct research in accountability with value-added methodology which measures residual gain or loss between a student’s achievement score and his or her projected score to better isolate school and teacher effects.
“Controversies of Standardized Assessment in School Accountability Reform: A Critical Synthesis of Multidisciplinary Research Evidence,” Lishing Wang, Gulbahar Beckett and Lionel Brown, Applied Measurement in Education Volume 19 Number 4 2006 pps. 305-328.
Published in ERN November 2006 Volume 19 Number 8