|
January 15, 2002
Accountability: Meeting The Challenge With Technology (con'td)
An Interview With Anthony Amato
Superintendent for the Hartford Public Schools in Hartford, Conn.
Q: When do you give the state-mandated tests and how go you get and use the results?
The official administration date is in September but, in order to offer another data point to see how each kid has grown, we administer a second version in the spring. This helps us see what needs to be done to prepare for next September.
The state sends us the digitized data from September tests in January. This gives us the 20,000 foot view. Then we use our student information system (SASI) to massage the data and drill in. How did each school, class, and kid perform? The teacher or principal can view reports tailored to individual students, schools, and teachers. What's the administrator most concerned about? Perhaps he or she wants to know how a certain population is doing in science or how a group did after using a new program. We can disaggregate based on what they want.
Q: Can you give some more examples of drilling in on the data?
Suppose, for argument sake, we find that Latinas are scoring low in a particular area. That's when we sit down in focus groups and begin to analyze why this is happening. What's the root cause? After taking a closer look at the numbers and disaggregating them further, we might discover that the girls who are doing the worst are ones with high mobility rates; ones whose parents are migrant workers. Knowing this helps us anticipate the problem and tackle it. For example, perhaps we need to give them Web browsers so they can access learning resources from anywhere, even if they're moving around.
Another group we have looked at is foster children, since we have a high percentage of foster children in Hartford. When disaggregating, we found that there was a clear pattern. They were doing poorly and we needed to figure out why. We realized that we needed to reintroduce ourselves to each foster family; if a kid moved around a lot, we needed to start from scratch each time. We might have gotten to know the previous family but this new one was seeing us for the first time. We also looked homeless kids and sent people out to work with them. In particular, we found it important to check on the oldest kid because a common pattern is to let the oldest kid go off on his own, to auntie's house or somewhere else, while the other kids go together somewhere else. By definition, that oldest kid is loose and more likely to engage in high-risk behavior.
Some people think of data as cold facts. They think paying attention to the numbers is unfeeling or mercenary. But that's not the way I see it. The data can be used to uncover all sorts of important information. If done correctly, data mining has high value--morally, ethically and educationally. It is the way to reach every kid, to make sure that nobody gets lost. Analyzing the data is a way to shine a light on students, to identify problems, and to get an understanding about how to help.
Q: Can't this sort of zeroing in backfire since it points a finger at particular groups as doing poorly?
No, because the idea is to look at what is happening and change it. It is not OK for any group of students to be doing poorly on test scores. We can't just accept it. Or give them more of what they've been getting already. If I accept that there's something inherent that keeps a student or group of students from performing well, then I don't belong in education. We need to see it as an anomaly, something that we're not doing right. We need to figure out what's wrong with the way we've been approaching the situation and do something to change it.
Until we take every failure as a personal loss, there will always be kids who are failing. As educators, we need to feel that failure in our hearts and our guts. The problem is not the kids--it is us! It is our inability to look at the data appropriately, to shine a light on each and every student. We need to focus on this constantly. We need to dig and probe until we figure it out. All kids can make it!
Q: What sorts of things have you done (in New York and Hartford) to help raise test scores?
We approach it in a variety of ways. For one thing, we have dramatically increased the number of days school is available to students. We've added "Super Saturdays" for K-8 and an after-school "power hour." There are also classes during winter break. Any student who took advantage of all the extra offerings could go to school 240 days a year.
I also set goals with principals for how much improvement is expected and then they sit down with teachers to figure out where those gains can be attained. Detailed reports help teachers set goals and objectives for their own students. If you break the math results down far enough, for example, you might notice that number facts and use are problems for a bunch of the students in your class-or in the whole school. Teachers need to meet in groups with others and discuss what to do to improve. We have numeracy and literacy teams at each school that get together and talk about curriculum and individual students. What's the particular issue with Jose? Is there something we should know to help him improve?
It's also helpful to take a close look at the numbers and where the most dramatic gains are possible. For example, if there are 20 kids in class, and two of them are "almost there" according to the test results, it makes a lot of sense to bring them up to the next level. (I call this the "golden band.")
Q: Couldn't this be viewed as manipulating the data for specific gains, rather than improving true achievement?
I don't see it that way. If I'm going to hold teachers accountable, I have to give them the tools to make those outcomes happen. It's a team effort. We can't expect all students to improve at the same rate and focusing on kids who need just a bit of help "moving across the line" is one logical and legitimate way to achieve the best outcome. This doesn't mean you pay less attention to other kids. You just pay different types of attention. The proof is that the first year we tried this, there was a12 point bump overall but the kids at the lowest level moved up 15 and 16 points. It is very gratifying when you see that all kids are being helped. It has the Hawthorne effect; they do better because they BELIEVE they can do it.
Q: Is it OK to study for the tests themselves?
Absolutely. In our district, 90 percent of the students who go to college are the first ones in their family ever to do so. Their parents cannot afford to hire special coaches or sign them up for SAT prep classes. We offer these classes as part of their regular course of study. We just brought in Achieva.com, a Web site with test prep and career planning help. We plan to lend the students computers to practice from home.
This sort of help isn't just for high school students. Other students need help learning how to take tests and set high goals for themselves. If it is a district where the families are wealthy enough to take care of poor test scores at home with tutors or extra help, fine, but otherwise it is important for us to help the kids. In Hartford and other lower-income districts, 95 percent of that is going to happen at school. It is morally wrong not to afford that advantage to kids. We can't make excuses about how "at risk" the kids are; we need to focus on what we CAN do for them. We need to hold the entire system accountable.
Competition plays an important role in all this. All students need to be encouraged to set high goals for themselves. My experience as a superintendent in New York City for 12 years showed me that healthy competition helped students achieve. Our entire nation thrives on competition-elections, Little League, Super Bowl, even cat and dog shows. Everybody is competing against each other at every level, but for some reason we turn around and say let's not have our students compete academically.
Q: To play devil's advocate, isn't there an assumption that such competition automatically assumes there will be losers?
It obviously needs to be a healthy competitive environment. Kids should be competing against themselves rather than other kids. They should set achievement goals and see progress reports that allow them to know how they are doing at reaching those goals. For example, we use Success for All in the elementary schools; student goal cards are used to set eight-week goals for SFA. We also ask "What are your standardized test scores and where do you want them to be? What do you need to work on?"
When you take this approach, schools invariably have tremendous outcomes and kids go on to do great things. I've seen it happen over and over again.
Q: At the Leadership Institute in Chicago you talked about data other than test scores. We have included some examples in the Technology & Learning article but can you summarize a few?
We track expulsions and disciplinary action. Our expulsion rate went way down in one year. That is because we made it clear that it was unacceptable simply to throw students out if they did anything wrong. One of my highest priorities when I arrived in Hartford was to lower the expulsion rate. Last year only seven students were expelled. That is great progress, but what we really should be asking what went wrong for those seven kids. Why did they get expelled? How did we fail them?
Another example is tracking attendance for school personnel. We look at who is there and how often. We found last year that the corridor supervisors (security staff) were absent the most. Why? Because they get 26 sick days a year. We're changing that in their new contract. We also found teachers took 18 days of sick time on average. We cross-referenced this with class attendance and found that low teacher attendance correlates with low average attendance of students. Students are mimicking the teacher. The test scores were correlated as well. So this made it clear we had to do something.
Q: What do you do with data about teacher performance? Once you identify which teachers are doing best and which are struggling, what do you do with that information?
Just before I arrived, the district had completed a teacher evaluation instrument that had been six years in the making. A teacher's objectives are defined in terms of student outcome ("My students will increase their test scores by X points" or "Attendance will improve by X percent"). The state uses the Hartford tool as a model for how to set actual data points--how will it be measured? It is no longer acceptable to have those undefined and unmeasurable goals schools used to set.
These goals are set annually. Since one of the big measurement points is September, with the state mandated tests, we've experimented with ways of keeping the teachers with their students till after the tests. In some schools, we loop so that teachers stay with their students for two years, or stay with the same class over the summer and into October before they move on.
When great teachers are identified, we make them coaches for others. Unfortunately, this often means moving them out of the classroom, which takes them away from the kids. Still, they're in a leadership position where their "karma" is spread out and can have more of an impact. We make them leaders in the school and district. We've also had success moving the best teachers from higher to lower functioning classes. That's not the usual way it is done but it makes sense. That way, we give the struggling kids the best we have. And the newer or struggling teachers get the chance to have an easier time because they have kids who are already doing well. They feel better about themselves.
When teacher evaluation is conducted in a calm, nurturing way it doesn't have to be threatening, even to teachers who haven't been meeting high goals. We need to help them see HOW to improve. Often they come away saying, "Oh, I can do that. It's not so tough!"
Q: At the Leadership Institute you talked about some city-wide intervention programs in development. Can you give an example?
We are creating a megasystem that links a lot of the data that has been collected separately. The district has SASI and a special ed system of its own. Every other community-based situation has its own legacy systems-social services, medical help at hospital, counseling services. We're trying to build a city-wide data warehouse so that we can "laser beam" students. All the libraries, firehouses, and police stations were wired with E-rate help and are already working off the school district system. Now we're trying to connect up other services.
It's important, for example, to get information on where there's a fire that's made somebody homeless or an arrest involving kids in the district. That way we can get right to the family, send social service providers. Eighty percent of all babies born in Hartford are to unwed mothers and there are all sorts of at-risk behaviors associated with that. How do we support those moms, get the right education for them?
< < < Back to previous page
Read other articles from the January Issue
Send a letter to the Editor in response to this article.
|
