Beware the Writing Assessment: Q&A with George Hillocks Jr.
Q: Why is "teaching to the test" especially harmful in writing instruction?
A: Examinations of all state assessments, and interviews with more than 400 teachers and administrators in six school districts in five states, found that large-scale writing assessments in most states have a negative effect on what students learn. Teachers and administrators restrict their writing programs to what they hope will achieve high scores.
Most state tests ... are conducive to formulaic writing. [They] provide a prompt to which students must respond in a limited period of time with no chance for finding information that can be used in the writing. Many teachers resort to formulaic writing [assignments], assuming that students will be able to write if they have the structure.
In one district in Illinois, for example, it is district policy for teachers to teach the five-paragraph theme from third on up to 10th grade. One third-grade teacher said, "We pound it, pound it, pound it!" They pound it so much in most schools that directors of college writing programs in Texas and Illinois have told me that one of their most difficult tasks is to dislodge such writing from their students' psyches. The vast majority of writing assessments are like those in Texas and Illinois.
Q: In what ways can teaching basic essay structures hurt student writing?
A: In our research, the most frequently taught "basic essay structure" is the five-paragraph theme. The problem is that it does not represent any real essay I have ever seen, and it is not basic in any way. Everything about it indicates that it short-circuits thinking. In fact, I think it is taught so that students do not have to think. [The structure is] limited to five paragraphs, three of which are supposed to elaborate the three points that appear in the first paragraph. A writer cannot deal with counter arguments or even recognize them for fear of losing points.
Q: What aspects of the judging process for standardized writing assessments concern you most?
A: Most states send the writing to independent contractors who score the papers. Their effort is to score the papers at the rate of one per minute or faster. All the rater can look at is the outward structure of the piece and the apparent elaboration of points. Further, the [judging] criteria ... does not begin to deal with the relevance of the logic of the supposed support, permitting any kind of nonsense.
In scoring guides from both Texas and Illinois, we found pieces with perfect scores that were written as parodies of the assignment. One writer responding to a prompt asking that he nominate someone for the "best-relative-of-the-year award" uses the fiction of an unreliable narrator, naeive about all of Grandpa Dulong's dishonesty and peccadilloes, suggesting they are worthy of great praise. The evidence, which we are meant to see through, cannot support the contention that Grandpa Dulong is generous, helpful and caring. On the contrary, it demonstrates that he is a rascal at best, not a person worthy of the award. The [judging] criteria have no way of evaluating evidence.
Q: What types of writing assessments do you advocate?
A: Some assessments are not plagued with these problems. Probably the best is one that encourages a great range of writing-the Kentucky portfolio assessment. While it has some problems, it is far superior to other assessments. Students don't have to learn to write without thinking in order to fill a page or two in 40 minutes. They have the year to produce several pieces of writing that go into a portfolio for judgment. The assessment is not conducive to formulaic writing.
If testing must be a matter of sitting down and writing in a set period, I recommend a test that provides some subject matter material for students to write about, as do the New York Regents exams. [They supply] literary passages or sets of data from social sciences.
Q: What advice can you offer administrators on implementing writing assessments?
A: Try to steer teachers away from formulaic writing ... and encourage breadth in the range of writing in your school curriculum. If you have the energy, examine your state's writing assessments. For ideas how to do that, see my article in English Journal (March 2003) called "Fighting Back: Assessing the Assessments."
George Hillocks Jr., author of The Testing Trap: How State Writing Assessments Control Learning (2002, Teachers College Press), is a professor of English and education at the University of Chicago.
Joining Hands to Keep History Alive
It has long been known that Thomas Jefferson owned slaves, which has certainly fueled many a classroom discussion about his phrase "all men are created equal." Until recently, it wasn't widely known that Jefferson had children with one of his slaves. While it's likely a topic discussed in far fewer classrooms, educators aware of that discovery have a deeper understanding of Jefferson. In many cases, recent historical research is essentially lost if teachers don't come across it on their own.
Thanks to a movement 10 years in the making, educators are working to abolish the re-teaching of myths and stereotypes in history classrooms. Partnerships are becoming more common between historians and districts to provide social studies teachers with opportunities to incorporate current research into their curriculum.
Making Connections, which partners City University of New York faculty and secondary school humanities teachers, is one such program. Mentors from CUNY's Center for Media and Learning American Social History Project Graduate Center work closely with teachers to help them integrate social history scholarship and methodologies into their lessons. Teachers attend year-long monthly seminars and leave with a set of print and multimedia resources. In one recent project, they did a close "reading" of a piece of artwork and juxtaposed it with primary sources that reflected multiple perspectives from the Westward Expansion period.
"There's often a disconnect between what happens with new scholarship that's developed in academic settings and the reality of what's happening in schools," says Eliza Fabillar, co-director of education at the center. "We help teachers integrate the social history approach, which ultimately helps students to connect to history in general."
In California, where K-12 teachers are tied to content-laden state history standards, partnerships with university faculty are even more important. At UC-Davis, one of the 11 regional sites of the California History-Social Science Project, a historian recently met with teachers to discuss his research on diversity within both Native American tribes and the U.S. Army during the Great Plains Indian wars. The new scholarship directly connects to standards that deal with the settlement of the Great Plains and the wars with American Indians.
"We have teachers teaching history without a strong background or degree in history," says Jana Flores, executive director of the California project. "They are not necessarily grounded in the content. Given what the standards are now, professional development programs that bring in university faculty are so important."
Fortunately, the U.S. Department of Education agrees. Through the Teaching American History Grant Program, www.ed.gov/programs/teachinghistory, a growing number of districts are getting funds to form K-16 partnerships. And that's good news for all.
Looking to form a history partnership? Educators and historians involved in these types of projects suggest that administrators:
Acknowledge and value the contributions and commitment of teachers to the project. Offer a stipend and release time for both program attendance and the development of new curriculum.
Consider partnering with any organization involving scholars. Many museums and historical sites have an outreach administrator, says Sandria Freitag, director of the Monterey Bay History and Cultures Project at the University of California-Santa Cruz. These sites may already have university relationships. "The pre-existing cooperation that a smaller site or museum has established with university scholars could be extended to the school district," she says.
Progress Made On Math Gap
Overall improvement is worth celebrating. But the 2003 National Assessment of Educational Progress, which has data for all 50 states for the first time, reveals that math educators can also be proud of evidence that the achievement gap is narrowing.
The percentage of fourth- and eighth-graders who performed at or above basic in 2003 was 77 percent and 68 percent, respectively. Those figures are higher than in all previous assessment years since 1990.
When white, black and Hispanic student data are segmented, each group had higher average scores in 2003 than in any year since 1990. Still, Asian/ Pacific Islander students (the highest scoring group) and white students had higher average scores than their black, Hispanic and American Indian peers.