GothamSchools — daily independent reporting on NYC public schools

Much ado is made every year about how students do on state tests. But are individual students’ test scores useful for them and their teachers?

Ruben, a Bronx teacher who blogs at Is our Children Learning?, says they might not be, because the scores come months after the tests are given and don’t give specific information about students’ skills. In a post about finding out his students’ reading scores yesterday (some improved; others showed a decline), he writes:

What’s most frustrating is how little the numbers tell me. We’re talking about a test that was taken in January. So the data doesn’t really even speak to the students I’m currently teaching. The data doesn’t really speak to anything at all, because it isn’t dissected in any way to show strengths in needs in specific areas such as vocabulary, drawing conclusions or writing. All I have are numbers, numbers that in many ways contradict what I know to be true about the reading and writing abilities of my students.

Of course all this is inconsequential, because even if the test was flawed, or too easy (the whole city went up 20%? Really?), those flaws apply to 4th grade students (and their teachers) universally. … Ultimately, I know my students should have done better, because pretty much everyone else did better. So, now it’s time to figure out what went wrong, so I can get it right next year.

Ruben’s objections to how the state delivers test scores represent two of the reasons Department of Education has offered to justify the introduction of more frequent, rapidly scored interim assessments in city classrooms.

I’ve posted about Ruben’s insights before, here, here, and here.

A few weeks ago, I passed on some readers’ questions about the Core Knowledge Reading Program to Matthew Davis, who is coordinating a pilot of the program in New York City elementary schools. He got back to me today with some answers.

Ira asked whether the program addresses syntax, since he finds that his students are very weak in understanding grammar and sentence structure.

Matthew Davis: In the Listening and Learning strand, the children will be hearing sentences with a lot of syntactical variety, including longer sentences than they would generally encounter in early reader type books they read on their own. We hope this oral experience of the language of books will help the students develop a sense of syntax. Also, beginning in grade 2, the Skills strand will address grammar and syntax explicitly. We expect to do some sentence-combining type of exercises to practice syntactic expansion. Details are being refined as I write.

Smith wanted to know how content is selected and sequenced, and how this program differs from what elementary teachers do already. (more…)

Yesterday, Michael Shaughnessy of EdNews interviewed Dr. Matthew Davis, who is leading the implementation of the Core Knowledge Reading Program pilot in New York City. Much of the interview covers basics of the program which we’ve discussed here already, including the two-strand approach to teaching reading and comprehension and the body of research supporting this method. What the interview highlighted for me are the contradictions of researching a program while trying to decide whether to continue using it, especially when real children are the subjects.

Davis says that the pilot will begin this year in kindergarten classes at 10 high-needs schools, then add grade 1 next year and grade 2 in 2010-11. But the continuation of the pilot “will be contingent on success in year one and a continuation of funding,” he says. Sounds fair: a program should prove itself before people (in this case, the Fund for Public Schools) invest further. Davis describes the plan for assessing the program:

Within the next several weeks, students in both sets of schools will be administered nationally standardized reading assessments in order to establish a baseline performance. These same tests will be administered again at the end of the kindergarten. In addition, there will be formal observation of all teachers in the pilot classrooms to ascertain any possible correlation between the level of implementation of the Core Knowledge program and the level of student achievement. In addition, specific case studies will be conducted by the NYCDOE in three pilot schools to provide additional qualitative information.

As far as the test are concerned, we hope to see a significant difference in word attack, word reading, decoding skills, and spelling by the end of the kindergarten year — because the program has what we think is a very strong way of teaching the mechanics of reading. Background knowledge and vocabulary take a bit longer to build, and gains don’t start to show up on some tests until later, but, by the end of the three-year period, we hope to see the front end of what we think will eventually be a very significant difference in vocabulary, oral comprehension, and reading comprehension.

So although the survival of the program may rest on a single year’s results, the promised impact of the program — increased vocabulary and content knowledge — may take three years to show up. At least three years: (more…)

Discussion of reading instruction — which started with a look at the Core Knowledge Reading Program (CKRP) being piloted in NYC this year — has really taken off, with commenters raising important questions: How does the content in CKRP differ from what’s being read now? What about helping children understand syntax? Does vocabulary development in Science differ from other subject areas?

While I look into those issues, here’s a technique one Queens teacher uses to help her students learn new words. Katie Kurjakovic, an English as a Second Language (ESL) teacher at P.S. 11 in Queens, illustrates the problem with an anecdote:

A second-grade teacher was preparing to read a story about George Washington’s wife, Martha, to her class. She anticipated all the unfamiliar vocabulary she thought they would encounter. She told them what colonies and colonists were. She spoke of the American Revolution and the Declaration of Independence. Then, shortly after she began reading, a girl raised her hand with a puzzled look on her face. “What’s a wife?” she asked.

Kurjakovic uses a six-step process to explicitly teach vocabulary to her English Language Learners. Before reading a text, she identifies and introduces (“previews”) new vocabulary for her students, then she reads the text, uses the words in the context of the text and then in a new context, and finally gives her students an opportunity to use the words. (more…)

In response to yesterday’s post about the Core Knowledge Reading Program, reader Smith asks,

Is he saying their is a core set of content that would prepare a student to understand a randomly selected reading passage on a standardized test? Could someone explain this idea to a non-ELA teacher? I’ve always assumed those reading passages could range from “The Mysteries of Ancient Egpyt” to “Sally’s Bad Day at School” to “Roger’s Time Machine Adventure”. How is content selected?

Great question. It’s true that the content of test reading passages varies, and I don’t think anyone believes that a child can be prepared with content knowledge specific to every possible topic.

Rather, some children enter school knowing thousands more words than others, and this difference compounds over years of schooling in a “rich get richer” scenario called the “Matthew Effect” by researchers. (Don’t take my word for it: this study, one of many, found that by age 3, children of parents with smaller vocabularies not only knew fewer words, used fewer words per hour, and used a smaller variety of words per hour, “but they were also adding words more slowly.”)

Hirsch summarized this effect in a 2006 article in American Educator:

Many specialists estimate that a child (or an adult) needs to understand a minimum of 90 percent of the words in a passage in order to understand the passage and thus begin to learn the other 10 percent of the words. Moreover, it’s not just the words that the student has to grasp the meaning of—it’s also the kind of reality that the words are referring to…. When a child doesn’t understand those word meanings and those referred-to realities, being good at sounding out words is a dead end. Reading becomes a kind of Catch-22: In order to become better at reading with understanding, you already have to be able to read with understanding.

(more…)

I was very interested to learn from last week’s New Yorker that some of the first public libraries for children were right here in New York City; the first in 1896 at the Pratt Institute in Brooklyn, followed in the early 1900s by a Central Children’s Room at the New York Public Library and children’s programs at the NYPL branch libraries. Anne Carroll Moore, who founded the Children’s Library at Pratt and went on to run the Department of Works for Children at NYPL, also reviewed children’s books, playing a decisive role in creating and shaping the field of children’s literature. E.B. White and his wife, Katharine White, who wrote reviews for the New Yorker, tussled with Moore over what was appropriate for and appealling to children.

(more…)

Yesterday, I took an initial look at the Manhattan Institute’s study, “Building on the Basics.” Today, I want to look at Florida’s state science exam, the focus of the study. A common criticism of standardized tests is that they all, to some degree, test reading ability. What does the Science FCAT look like? What skills would you need to perform well on it? I’ve only seen the NYS Science exams, so I decided to download a Florida sample test and take a look. The first thing that surprised me about this test was the reading level, which seemed high. Many of New York City’s fifth graders would (for better or for worse) stumble over sentences like, “Florida has many limestone caves containing formations called stalactites.” I tracked down a site of readability analyzers and entered text from test items.

Question 1: Melissa’s school rings a bell to alert students that it is time to start class. When the bell rings, it vibrates. The use of vibrations to send messages is an example
of which type of energy?

This one ranged from 4.72 to 10.07 in estimated US grade level required to understand it, which certainly calls into question the reliability of the readability analyzers, but also the ability of average 5th graders to understand this question.

(more…)