Complaints force revisions in 'school climate survey'
Monday, May 6, 2002
Faced with a teacher boycott and concerns from parents and privacy advocates, Seattle Public Schools has backed off an annual "school climate survey" that required students to list their names and identification numbers while answering sensitive questions about cheating, race, weapons and other facets of student life.
"Not only sound policy reasons, but also well-established rules support conducting surveys such as this on an anonymous basis only," the American Civil Liberties Union wrote in a letter to the district last week.
The Seattle teachers union voted to boycott what it called a "poorly designed" survey, saying staff members were concerned about privacy issues and found some of the questions offensive. Others questioned how accurate the results would be with students required to give their names.
"Definitely, doing it that way hurts the outcome of the survey," said public-opinion research expert Howard Fienberg, a senior analyst with the Statistical Assessment Service in Washington, D.C., after hearing some of the sample questions.
"It means you are probably not going to get good data out of it. It will be junk."
The district has administered the student survey since 1994 and has consistently required student names and identification numbers without receiving complaints, spokeswoman Lynn Steinberg said.
But the survey was expanded this year to include questions on potentially more sensitive areas, including bullying and race. It was given this year to most students in grades three, four, five, seven, nine and 11. (Most students in grades six, eight, 10 and 12 received a different, anonymous survey designed with the University of Washington.)
The school climate surveys -- which focus on such issues as whether students feel comfortable at school, know their teachers and believe their school work is meaningful -- are meant to figure out where problems might exist in schools and to help them improve, Steinberg said. They have never been used to track individual students, she said.
"I would remind folks that we've been doing this since 1994, and that's never been an issue before, and we've never used it on an individual basis. ... I think our track record speaks for itself."
Still, Chief Academic Officer June Rimmer decided to scale back this year's survey after hearing complaints, and this year will be considered a "pilot year," said Mike O'Connell, the district's director of research, evaluation and assessment.
Rather than compiling all the survey results, the district will give schools the option to have the results compiled or to have the survey sheets destroyed.
Even schools that choose to have the results compiled will destroy the students' individual sheets afterward, as is typically done with such information, O'Connell said.
For schools that choose to compile the information this year, "In no case will it be used for principal evaluations or school evaluations or accountability. It will be essentially for schools to use for their own internal planning," O'Connell said.
Student names originally were required on the survey because the district began administering them along with standardized tests that required an ID, O'Connell said. The district designed a separate survey around 1998, when the introduction of a new standardized test made it harder to give the survey at all grade levels.
When the district expanded the survey this year, it sent it to its attorney's office for review, eliminating a number of questions because of privacy concerns and also asked whether to continue requiring names and identification numbers.
The logic behind requiring identification was that "it allows us to easily link the data to certain demographics," such as the child's gender, school, race and grade level, O'Connell said.
The question of reduced accuracy is certainly "one that could be asked," he said, although any survey is subject to some degree of error or bias.
"All of the issues that have been raised are certainly points for discussion and reconsideration next year," O'Connell said.
Steinberg said the district had received a handful of direct complaints from parents. The union said staff members have received many complaints, and the ACLU said it had received several.
The survey serves a valuable purpose, Steinberg said, and the expanded areas were significant.
"One of the things we needed to know is, to what extent do individual schools have some issues around either bullying or race? ... Does the program adequately address some of these issues? And if not, we need to know that so we can make it better."
For parent Jay Stansell, though, the methodology meant the survey was useless as a research tool and offensive from a civil liberties standpoint. He was concerned that parents were not notified before the survey was given and were not told how the information would be tracked -- issues the ACLU also raised.
"As a criminal defense attorney, I couldn't help but reflect on the fact that the survey asks these kids, really, to incriminate themselves," Stansell wrote the district.
"The message that they need to openly express their feelings -- including bad feelings -- with their names attached is an inherently coercive approach to the task of information gathering."
Beyond the question of accuracy, he said in an interview, requiring students to answer personal questions also raised broader questions about the tension between privacy and security in our society.
"It's a junior version of a coerced confession," he said. "We live in a country with a constitution, and we all have the opportunity to decline invasions of our privacy."
Steinberg noted that the survey directions told staff members to let students know they could skip any questions they did not want to answer. But it's not clear that those instructions were always followed.
Franklin High School junior Amanda Baldwin, for instance, said a boy in her class did not want to complete the survey because it required a name, and a teacher told him he was required to do so.
Even questions that appear less sensitive on the surface, Baldwin said -- such as asking how students get along with their teachers -- can be dicey when students are handing their survey in to those very teachers.
"It just didn't really seem fair. ... If it was an anonymous thing, maybe it would be better," she said.
Fienberg, the expert on public-opinion research, said requiring names is a good way to make sure students don't fill out more than one survey apiece, which can improve overall accuracy. But students might use fake names or skew their answers.
"Even if the subject matter were not that personal, most people value their privacy too much to make this method at all useful," he said. "They're not going to stand up and say, 'I bullied five people last week and it was great,' or, 'Cheating is OK; I do it all the time.'"
Sample questions from the student survey given to grades seven, nine and 11 in Seattle Public Schools:
1. I think it's OK to cheat at school. (Yes/No)
2. I think it is OK to take something without asking if you can get away with it. (Yes/No)
3. It is all right to beat up people if they start a fight. (Yes/No)
4. How wrong do you think it is for someone your age to take a handgun to school? (Not at all wrong/Sometimes wrong/Very wrong)
5. How wrong do you think it is for someone your age to steal anything? (Not at all wrong/Sometimes wrong/Very wrong)
6. How wrong do you think it is for someone your age to attack someone with the idea of seriously hurting them? (Not at all wrong/Sometimes wrong/Very wrong)
7. Students at this school bully me. (Responses to check range from "never" to "often")
8. How often have you yourself bullied another student at school in the past couple of months? (Responses to check range from "I haven't bullied another student at school in the past couple of months" to "several times a week")
9. I get along with/feel comfortable with teachers who are a different color from me. (Responses to check range from "never" to "often")
10. I can do well in school without being called a credit to my race. (Responses to check range from "never" to "often")
In accordance with Title 17 U.S.C. Section 107, any copyrighted work in this message is distributed under fair use without profit or payment for non-profit research and educational purposes only. [Ref. http://www.law.cornell.edu/uscode/17/107.shtml]