Reading First = Kids First
by Siegfried Engelmann
for Oregon's Future, Volume VI, Number 1, Winter 2005
The best place to gain a perspective of what Reading First is all about is in kindergarten through third grade in a failed urban school. Consider a typical student in one of these schools, Alan Jones, an African American in a high-poverty neighborhood. It could be in Atlanta, Chicago, Philadelphia, New York, or even some neighborhoods in Portland. He entered kindergarten far behind his middle-class peers in vocabulary and information. That means, he didn't understand many of the words the teacher used to direct him. For example, his kindergarten teacher told him, "Alan sit between Amy and Sidney." To Alan this direction sounded something like this: "Alan, sit jumble Amy jumble Sidney." He didn't have any idea of what to do beyond sitting because he did not understand the words between or and.
His kindergarten teacher does not view children from an instructional standpoint. She doesn't say to herself, "I have to teach them what they don't know or they'll fail in school and life." In college, she learned that teaching children earlier than the first grade will stunt them and even cause them to have strong antisocial behaviors later. Of course, this is not true, but she, and thousands of other early childhood advocates, believe it is.
According to her view of Alan, he would go through stages of development, and her job was to set the stage for his natural learning, not try to force learning on him. She believed that he would learn when he was ready to learn. She would pique his curiosity, encourage him to express his interests and concerns, and promote general "growth" both through nutrition and lots of physical activities that she believed would accelerate the rate at which he internalizes the operations involved in physical manipulation. These activities, she believed, would set the stage for him to learn reading and math in grade 1. Yes, she also presented ample "readiness" activities, workbooks, coloring, discussions, word rhyming. She even did what she believes is teaching her children the alphabet. As part of every post-nap activity, she pointed to the letters of the alphabet as the children recited from A to Z.
The Reality for Alan
For Alan, this activity was a recitation, and he didn't attend much to the symbols. He learned most of the recitation, except the part around l, m, n, o, p and the part around t, u, v, w. Alan sort of learned to identify several letters, not because of this activity but because he printed them, especially A is for ant.
After his kindergarten experience, Alan entered the first grade. At the beginning of the year, he and his classmates took readiness tests. The children were tested on words that are spoken a sound at a time, on letter identification, and on rhyming with words like eat and lip. Alan performed at the 10th percentile. That means that if kids were arranged from the highest performing to the lowest performing, the average performers would be at the 50th percentile, 40 percentiles above Alan. The top performers would be about 90 percentile points above him. Alan identified only one letter and failed all the rhyming items.
His first grade teacher was strict, and it took Alan some time and tears to adjust to her demands. Alan didn't like her or first grade because there was no fun and freedom, and she always told him to do things he didn't know how to do. Fortunately, there was a lot of time spent coloring, matching, and following the dots. Alan knew that he was pretty good at these. Then there were books he could look at. Some had pretty pictures. Sometimes there was a TV show that was funny. And the teacher read a lot of stories, but Alan didn't always know what was happening in them. The stories Alan listened to were obviously written by people who base their assumptions of what kids know on kids quite different from Alan.
I once evaluated a program in New Jersey like the one Alan went through in first grade. After the teacher read a story that told about children finding a treasure at the top of a hill, I tested the children on some of the phrases that were in the story. One phrase was that the children, "were panting after their long climb." I asked the class, "What is panting? Who can show me what you do when your are panting?" One kid raised his hand. I had him stand up and show me what panting was. He posed something like a model on a runway and ran one hand up and down his hip. I asked the group, "What is that?" A few said panting. I asked one of them, "Is that panting?" He said, "Yes, panting hose."
Another phrase was that one of the children "started to lose her balance." I asked
the group, "Who can show me what you do when you lose your balance?"
One kid raised his hand. I had him come up to the front of the group and
show what you do when you are losing your balance. The kid sort of picked
at his belt. I asked the group, "What is that? What is he doing?"
One kid answered, "Losin his bel." (Loosen his belt.) I could
have done the same thing with at least 50 more phrases in the storyshowing
that not one child in the whole classroom knew what the words meant.
Alan's first-grade teacher knew what to expect from Alan. When she had started teaching seven years earlier, she had high expectations about what her students would learn. Seven waves of kids hammered her expectations down to what they were for Alan's class. Most kids would fail. Those on Alan's level would "struggle" and would certainly not learn to read in first grade. The teacher also believed that a teacher could not give the children any slack or the classroom would become bedlam (like it did during her first two years of teaching). Her motto was, "Don't ever show that you're soft." She didn't praise her charges very much, but then, she didn't feel that they did much that deserved praise.
Alan's first grade teacher taught reading the way it has been done in many Portland schools since the 90's. She read and reread the same stories to the children. She called on different children to read the words. After a while, children were able to read these stories, or at least that's what she thought. Later in the year, she also presented selections from a basal reading program. If children could not identify a word, she would say something like, "Well now come on. What could that word be in this sentence?" On several occasions, she told Alan, "Look at those other words and take a guess." Once in a while Alan guessed right.
Later in the year, Alan felt pretty good about these stories because he figured out what "reading" is. The teacher holds up the book and talks about the picture. What's going on in the picture shows you what to read. It's kind of tricky, though, because for some reason, you don't always say the right words. But if a picture shows Bill running down the hill, the words will say, "Bill run down a hill," or something close to that.
Also, these stories were good because you know that they use the same word patterns over and over. One page showed a picture of Bill with a bird, and had these words:
Bill asked the bird, "Have you seen my clock?" The bird said, "No."
On the next page Bill is with a dog. So you know the words. Bill asked the dog, "Have you seen my clock?" The dog said "No." Reading the first page is really really hard, but the rest of the pages are easy (except the last one).
We once did a demonstration in a classroom like Alan's. At the end of the year, we took some of the stories of "shared experiences" that had been read and reread during the year, and switched texts and pictures. So the picture of the zoo now accompanied text for the fountain, and so forth. We told the kids, "I think somebody got these words and pictures mixed up so read carefully." We tested the kids individually. Over half of them pointed to each word and recited the story for the picture on that page. They did not have a clue about what reading was. Nor did Alan. He scored at the 14th percentile on the standardized reading achievement test at the end of grade 1.
Alan's second-grade teacher was more lenient with the children than the first-grade teacher. She knew that these children were not very smart and she accepted the fact that a lot of them could not read. Even if she worked hard, these kids were not going to make much progress. She believed that if the kindergarten teacher and the first grade teacher had done their jobs, she would be able to teach second-grade material. But the children were not prepared, and it was not her job to go back to the beginning and teach them everything they should know. Most of them couldn't even read first-grade material.
She did what the district said she was supposed to do. She read literature selections to the class. She presented some word lists and continued through the same reading series the first grade teacher had used. And she had the kids do lots and lots of worksheets. That kept them occupied and made the school day a little less stressful.
By the middle of the second grade, Alan had figured out that words and reading are more mysterious than he had thought earlier. When you spell words or write them, you have to know the letters (most of which Alan knew by now). And when you read words that are in a list you have to know the letters, but when you read a story, the words are different and you somehow have to know something before you recite the story. Sometimes, you'll recite it and the teacher will say you used the wrong word. But sometimes, it's the right word. Very complicated.
Fortunately, during the second half of the year, the teacher did not do much structured work with the group, but assigned a lot of independent reading. Alan did the best he could, given that he wasn't highly motivated. It was just as well because when he read silently, he simply strengthened the various word-reading mistakes he made when he read out loud and nobody was there to correct him. The material he read made little sense to him.
Years ago, when we were working with fifth-and sixth-graders who were failed readers, one kid revealed something about the kind of confusion he had about reading. The students were taking turns reading a simple passage composed of words that are "regular," which means that students could sound out the words. This student read a sentence and made a mistake. The teacher told him. "You missed that word. Sound it out."
The kid looked at the teacher and said, "Tell me the word and I'll sound it out." Translation: I have to understand the context and meaning before I can identify story words. Some version of this confusion is why these poor readers are often able to read words like this, that, what, the, an, a in a list, but in stories their word identification will be random. The most revealing fact is that the words these students tend to miss most are words they have been exposed to since first grade.
The third grade provided Alan with more evidence that he had reason for hating school. He had to sit through horrible lessons and pretend that he knew what the teacher was saying about those things called vowels and consonants and paragraphs and other stuff. But even though Alan rejected school, it hurt a lot to know that some of the other kids could read. Alan really wanted to read, but he just couldn't do it.
The third grade had a much stronger sense of urgency than the second grade because at the end of third grade, the class would be tested on the "state standards" and if you didn't pass the test you had to take it again, and you couldn't get out of third grade until you passed it. Teachers had to follow a district rule that all the material students read had to be third-grade material, because that's what they would be tested on. Alan read on the middle first-grade level, at best, and only three or four of his classmates could read third-grade material without making many mistakes.
The class spent one whole period every day working on test preparation. The teacher passed out sheets of "items." The students were to read them and circle the right letter. Then the teacher went over the test, and told the answers. She presented those same items again in later lessons. This period was not as bad as the daily reading period because you got a chance to memorize some of the items and get them right.
At the end of the third-grade year Alan passed the state reading test on only the second try. It was comforting to see that some of the items he had practiced were on the test, but some of the test items were new. For those, he did what his teacher taught him to do, guess. As his teacher put it, "If you don't guess and mark one of the choices, you have no chance of getting it right." He was pretty proud of his accomplishment because some kids had to take that test again and again.
Unfortunately, passing this test didn't accurately predict that Alan could read fourth-grade material, with its rich vocabulary and textbook sentences far different from those in stories. Alan actually read below the beginning second-grade level. Like the teachers in the other grades, the fourth-grade teacher presented fourth-grade material, and the fifth-grade teacher presented fifth-grade material. Alan slipped farther and farther behind his middle-class peers as the material became increasingly overwhelming and mysterious. He quit school after the ninth grade. By now, he had learned some more about reading, but he was definitely a corrective, or remedial, reader. Some of his classmates graduated, and a few even went on to community colleges. But they read on or about the fourth-grade level.
As the National Assessment of Educational Progress (NAEP) shows, Alan is close to the norm for at-risk students. When the NAEP tests are given, the scores for all students are much lower than they are on "state tests" and "standardized achievement tests." The reason is that NAEP scores have not been inflated through inappropriate "test prep." They show that about 66% of all students fail to reach "proficient" (grade level performance) in all grade levels tested by the NAEP, except grade 12, where only 60% fail. That's largely because kids like Alan don't stick around for the humiliation of grade 12.
The overall picture is sad. A large part of the pathos comes from the fact that the children learn what the teachers teach, but teachers typically don't know that they are inducing crippling misunderstandings or that they are failing to teach skills. Teachers taught Alan to guess, and he guessed. They suggested that you could look at the first letter of the word and figure it out by its shape (which is impossible), and that's what Alan tried to do. They presented material that was highly patterned and that could be "read" by following the pattern, not by reading the words. That's what Alan tried to do.
One of the more stable statistics is that if the children fail in grades K through 2, they fail in all the later grades and drop further and further behind. Another statistic that underscores the need for more effective programs in the early grades is that teaching failed students to read in the later grades requires far more effort than it does to teach them in K and 1. With careful instruction, which bears no resemblance to what Alan received, it might have taken something like 40 practice trials for Alan to learn to read words like what and that with more than 95% accuracy in story contexts. If we try to achieve the same level of accuracy when Alan is in the 6th grade, after he has practiced whatever superstitious rituals he performed for 5 years, it will take something like 500 trials. With a careful program all kids like Alan can read by the end of kindergarten and perform at or above the 50th percentile by the end of the first grade. Their performance will probably drop some but will remain within the average range throughout their school career (so long as the instruction is careful). With good instruction there are no "dyslexic kids".
Effective early interventions are not easy to implement. That's why they aren't accomplished very frequently by schools. The problems Alan encountered pretty much describe what has to change in failed schools if we're to save the lower half of the school population. We have to teach kids like Alan to manipulate spoken words before introducing reading. In kindergarten we have to teach them the language that the teacher will use when she directs the children to do something like, "Touch the next word." Quite commonly, the children don't know the meaning of touch or next.
Put simply, if the program is to be effective with Alan, it has to start at his level, teach what he doesn't know, and do so with great efficiency so that he is able to learn at an accelerated rate. We have to redesign the stories so that they don't encourage kids like Alan to "predict" words by referring to the picture or referring to the context. We have to monitor Alan's progress very closely starting in kindergarten and respond to it in a timely way. We have to train teachers to see to it that Alan receives the practice he needs to achieve mastery of everything an effective program teaches. If he requires more practice than some of the other students, so be it. He'll get the practice he needs. And it won't be punitive. His teachers will acknowledge his hard work, celebrate his successes, and use evidence of his performance to let him know that he is capable of learning whatever the teacher teaches.
Education Reform and the Development of Reading First
Systems that are designed to reference instruction to Alan and carefully teach him all the skills he should know are not new. They are simply rejected by educators. In the 60s and 70s, the facts about what works and what doesn't were revealed through the largest educational experiment ever conducted, Project Follow Through. It involved 180 communities and over 500,000 students in K through 3. It rigorously compared the performance of 18 models of education and cost $500 million. The models spanned the full range of "beliefs" about how children are best taught, from permissive child development models, to a highly structured approach that controls all instructional variables that had been demonstrated to make a difference in student performance.
Parent groups for at-risk children in each participating community selected one of the instructional models. Participating districts received $750 per student in addition to the regular per-student cost. The sponsor of each participating model received ample funds to assure that the sites were implementing its model with fidelity. The Office of Education even assigned liaisons that monitored the sites and reported problems of inadequate implementation.
One model, the highly structured Direct Instruction model from the University of Oregon, outperformed all the other models in every subject testedreading, language, math and spelling. The third graders who went through DI programs also had the highest self-esteem. The reason was that they had a lot of evidence to show that they were smart kids.
Don't feel surprised that you've never heard of the study. In a poll, most educators reported that they had never heard of it or of Direct Instruction. The reason is that the results of Project Follow Through, by model, were never released. The only official statement the Office of Education released was that Follow Through failed. Indeed it did. The overall performance of Follow Through students was not even as high as that of the traditional title-one comparison children. All this proved, however, was how uninformed all but one of the models were about how to induce skills and self esteem in low performers. The Direct Instruction model outperformed the other models by an average of more than 25 percentile points in the various subjects.
The information on the results were not disseminated simply because policy makers didn't like the results. The Office of Education was influenced by a spirited effort from the Ford Foundation to place a gag order on releasing the results by model.
In any case, for 30 years there has been scientific knowledge about how to prevent failure of at-risk students. Schools haven't been using this knowledge until recently because there had been no incentives or requirement for states and districts to take a serious look at Alan and his classmates and do something about it. I know that this sounds like science fiction, but according to Dr. Grover Whitehurst, Assistant Secretary of Education and Director of the Institute for Education Sciences, only about 10 percent of the decisions made about curricula and teaching practices by districts and states are based on data of what is effective. In other words, districts and states generally don't use practices that have been demonstrated to be effective, and they don't respond to poor student performance with remedies that have been demonstrated to be effective.
Alan, his teachers, and the practices for teaching at-risk students form the backdrop for Reading First. The goal of the initiative is to assure that states and districts learn about kids like Alan and what it takes to make them successful students. The initiative focuses on grades K through 3 because that's where the action is. However, Reading First is far from perfect, it has provisions that unfortunately allow districts and states wiggle room to use practices that are questionable. But it moves states and districts in a more responsible direction by preventing them from using failed approaches like Whole Language and Reading Recovery. The legislation requires districts to apply to the state for Reading First funds. The state is required to accept only those proposals that address all five skill areas that have been identified through research to improve reading instruction: phonemic awareness, phonics, fluency, vocabulary, and comprehension. Proposals will not pass if they designate reading programs that "lack a scientific research base," that rely primarily on instructional strategies that "engage students in independent silent reading with a minimum of guidance and feedback," or that "use context or picture cues for word identification."
Another good feature of Reading First is that it requires districts to "have a clear plan to make decisions related to their Reading First programs based on evaluation outcomes." The district must also have a plan for dropping schools that do not achieve adequate progress. Also, the state must spell out how it will monitor the districts. All this is new stuff for both states and districts.
Some state and district officials complain about Reading First on the grounds that it requires evaluation. Evaluations have been around since the 1940s. The difference is that now the district and the state are expected to do something about poor evaluations. Districts are not to be selected if their proposal "does not demonstrate an adequate plan for making decisions based on evaluation outcomes."
The idea is for districts to use sensible evaluations. You wouldn't use a standardized achievement test or a state test to determine whether the kids are achieving oral-reading fluency. You need a fluency-accuracy test. In the same way you would test whether kids are mastering what the teachers teach with items that test what the teachers have just taught. Also you would always place children at their skill level, not their grade level. You would never place a student who reads at the second-grade level in a fourth-grade program. Furthermore, you would include provisions for spot-checking the performance of children to assure that they are actually reading in a way that is consistent with the evaluations.
States and districts can do nothing but benefit if they use serious measures of student progress, learn how to teach so that at-risk students master the material, and respond to data not with lame reforms that demand higher standards or that increase teacher prep time, but with reform based on the unwavering understanding that kids like Alan can learn if we have proper instructional material, strategies, and the will necessary to teach them.