Final Report on the Two-Year Cornerstone
Wilkes Elementary School
Reynolds School District
Charles Arthur, 1st grade teacher and project developer
The present challenge for our schools is
to educate all students to a higher level of achievement than
ever before with a student population that is becoming increasingly
academically diverse. The higher demands for educational quality
within this context has made the mission of the schools increasingly
difficult to accomplish. It calls for extraordinary efforts and
The reform efforts of the last 15 years have so far failed in meeting this challenge. According to the 1998 Goals Report, by the panel charged with tracking progress towards the Goals 2000, there has not even been an upward trend towards their attainment. Education Week's, Quality Counts, reports that measured student achievement has stayed essentially flat. There are isolated examples of significant improvement, but the broad picture is that the schools are "treading water."
This lack of improvement comes at a critical time. With technological advances and the information-based economy has come the increased importance of a higher quality of education for all. According to the 1998 National Research Council report, the increased urgency for teaching all children to read does not come from "declining absolute levels of literacy" but from the "rising demands for literacy" in our "increasingly competitive economy." This report also states that "in a technological society, the demands for higher literacy are ever increasing, creating more grievous consequences for those who fall short." Poor academic skills and knowledge helps to create a permanent and costly under-class, cut off from the rest of the world and from any opportunity to become productive contributors to or participants in society.
Improved achievement is particularly important in the early grades. There is consistent evidence that children with low academic achievement in these grades have greater eventual likelihood of continued school difficulties in later grades, and that they consequently face greater risks of negative social and economic outcomes. Deficits in vocabulary knowledge in basic subjects of children with learning problems exist as early as the first grade, and the gap between them and the average learner grows increasingly larger each year. This gap contributes to low self-esteem and school failure. If low achieving students can be brought up to grade level in the basic subjects of reading, language and math within the early grades of school, their performance tends not to revert but to stay at grade level thenceforth.
The nationally known, research-based curricular programs, Reading Mastery, Reasoning and Writing, and Connecting Math Concepts, all programs that were researched and developed by Siegfried Engelmann and his team at the University of Oregon, have proved to be effective in meeting this challenge. Teachers using these programs have been able to accelerate all learning and decrease the gap between children with learning problems and the average learner. Teaching all children in the early grades phonetic decoding skills in order to read accurately, easily, and with understanding contributes to this. These programs also teach all children how to begin to think and write clearly and how to compute and understand basic math operations and concepts which are then applied to solving common problems.
These programs are part of a teaching approach that is the most heavily researched and field-tested available. The American Federation of Teachers, the National Education Association, and three national administrators' organizations have recently acknowledged this fact. A study, sponsored by all five of these organizations, that was conducted by the American Institute for Research found that the instructional approach taken by these programs, was one of only two approaches at the elementary level, among 24 other identified reform programs, that had strong evidence of positive student effect.
Programs that derive from this approach are a result of over thirty years of research and development. This research includes taking part in the largest national experiment ever conducted, called Project Follow-Through. The model that these programs are based on far out-did all other competing models in the project. Not only did they produce higher achievement, but they also produced higher student satisfaction and self-esteem.
However, much of the research done on this approach has been done in special settings or under special conditions. In regular classrooms, this included the use of paid teaching assistants. Continuous work is needed to test the applicability and practicality of these programs in regular classrooms under normal conditions. The Cornerstone Project is a further test of the effectiveness of these programs in regular classrooms by two teachers, with experience and training in the programs, using no paid teaching assistants.
The over-all goal of this small project
To further test the hypothesis that children, taught with these programs in regular first and second grade classrooms, under normal conditions, can learn in two years what is usually learned in three.
There were three objectives that defined this goal, two that involved average progress of a whole class and one that concerned individual levels of progress.
The three objectives were:
1. To achieve in the first year with a first grade class, an average of one-half grade level higher than national norms, and
2. To achieve in the second year with the same group of children in second grade, a second grade class average of a full grade level higher than national norms.
3. To see no individual scores below grade level.
It these objectives were met, it would demonstrate how teachers can, on the one hand, prevent school failure, and, on the other, increase the acceleration of learning and progress of all children. Therefore, a secondary goal was to demonstrate the effectiveness of this model of teaching in order to encourage its use and further its examination by local school districts.
The Cornerstone Project at Wilkes Elementary School was a small, privately funded, two-year project that started in September 1997. In 1999, the project was extended one more year. In the original grant proposal, the over-view of the project was described as follows:
The Cornerstone Model Project is a special, accelerated pilot program for grades one and two in Wilkes Elementary School that will begin in September 1997. Students will be in the program for two years and will receive instruction in nationally known, research-based curricular programs. These programs teach core foundation skills in reading, math and language directly and explicitly and stress the need to master what they learn in order to serve as a firm foundation for the more advanced kinds of learning that will be needed in later years..
In this project, the progress of students in the first Cornerstone class with the same programs was tracked through the second grade. However, in doing so, two other classes became involved in the project. In the first year, a second grade class that had not had previous instruction in the Cornerstone programs was taught with the programs. This involved intensive catch-up instruction for at least ten students. The results from this class are a part of this report. In the second year, the programs were used again with a new class of first graders. Plans are to continue the Cornerstone Project so that this class will also be followed with the same programs and measurements, when they are in second grade. The results from this class are also included in this report. It was expected that there would be some improvements in instruction and student performance in each subsequent class, and that similar results would be consistent each year, thus adding credibility to the whole project.
Most of these objectives were measured with the group administered Stanford Achievement Tests, 9th edition. However, the individual reading tests, the Multilevel Academic Survey Tests (MAST) and the Woodcock Reading Mastery Tests (WRMT), were also given to collaborate the group tests. Whenever possible, Control Classes were involved in order to provide some comparison with other similarly composed classes that were not using the same curricular programs as well as comparisons to national norms.
The over-all progress of the students who stayed in this project for two years is shown in the graphs on Figure 1. Reading achievement progressed from the 17th percentile at the beginning of the first grade to the 68th percentile by the end of the second grade, a 51 point gain. This represents a grade-level progression from the K .4 to the 3.5 grade level, a 3.1 grade gain. Scores from the individual WRMT given in June 1999 are identical to the SAT-9 June scores. The math progress began at the 30th percentile and ended at the 79 percentile, a 49 point gain over the two-year period. The grade level equivalency for the math is from the K .6 to 4.0 grade level, a 3.4 grade gain. A gain of 3.1 grade levels in reading and a 3.4 gain in math means that the over-all goal of learning in two years what is normally learned in three was achieved.
Only one of the separate grade level objectives was attained. The objective of achieving an average of one-half grade level above the national norms with the first grade was attained. This was done in spite of the fact that the first grade class started from far below the expected grade for first graders. Two grade levels were gained during that year in order to meet this objective. This was done in both reading and math. However, the goal of extending this rate of progress to one full grade level above the norm by the end of second grade was only attained in math, not reading. This goal for reading was possibly too ambitious. This will be discussed further in the next section of this report.
The Control Group for this Cornerstone class, that was also tracked through the second grade, attained a very similar level of achievement, but the gains were smaller because the starting point at the beginning of first grade was higher. Their average reading level at the beginning was at the 35th percentile, only a little below average. A 31 point gain by the Control group is very respectable, however. The Cornerstone class was being compared to very effective classrooms. No Control groups in math were possible for the full two-year period because these classes began using the same math program as the Cornerstone class during the second year.
These two-year scores were based on the scores of 9 Cornerstone students and 9 students in each Control group. The range of the beginning scores of the nine Cornerstone students do represent the ranges of the class as a whole, but the figures on the graphs are still based on a very small number. The Cornerstone class and each Control class started with 22 students. The eventual small number that stayed in the school was due to an unexpected high rate of student turnover. During the two-year project the mobility rate of the school in these grades increased from 30% to more than 50%. This increased rate of mobility may be due to the fact that there was a school zoning change just before the project began. The mobility rate did increase for the whole school. It is also possible that the rate of mobility is greater for the lower grades than the higher grades. Many times families of younger children are still in transition, which creates a higher rate of student turnover.
Because of the eventual small number that stayed in the program for two years, the yearly scores are more significant. These scores are presented on Figure 2 and 3. The scores for both the first and second grade classes are shown for each year. In the first year, there were two Control classes for reading and one for math in each grade. In the second year, there was only one Control class for both math and reading in the second grade. The classes that served as Controls during the first year could no longer continue because they began using the same programs as the Cornerstone class the following year. The school district was unable to supply the project with any other Control classes from other schools using different programs.
In the first year of the program, in both subjects of math and reading, for both grades, the Cornerstone classes began below the Control classes and finished above them. This was especially the case for the second grade class.
A further breakdown of the first year scores shows that the middle and high level readers made more gains than the lower level readers. In the first grade Cornerstone class, the middle and high level readers made over 50 points gain while the lower group gained 30 points. In the Control classes each level made similar gains ranging from 25 to 32. In the second grade Cornerstone class, the higher groups made over 40 points gain and the lower group made 29. The Cornerstone second grade class started with a group of 10 non-readers. In the Control class for this grade, the higher groups made 20 points, but the lower group made only 10.
The scores at the end of the second year for the new first-grade Cornerstone class were very similar to the first year. The only difference is that this class started at a lower point, so the gains were a little more. The scores on the individual tests were also very similar. What is interesting about the WRMT, given to each second year first grader and to each two-year Cornerstone second grader, is that the score for the Basic Reading Skills portion of this test were significantly higher than the Full portion of this test.
The gains of the second grade Cornerstone class for this year were less because the class started at a higher point. This was due to the presence of the 9 students that had stayed in the program for two years. Yet, with some loss over the summer by some of the lower performers in this group, and with 10 new students in the class with no previous Cornerstone experience, the beginning of the year average reading score was still lower than the previous spring for this group. The fall score was at the 36th percentile. The Control class began at the higher score of the 51st percentile. In spite of this gap, the Cornerstone class exceeded the Control class' end of the year score of 61. Similar results were obtained in math, except that the Control class made a larger gain in math than in reading.
At the end of the extended third year of the project, a summary of gains for all three years was made.
The Cornerstone programs were able to take a very low achieving group of students at the beginning of first grade and raise their level of achievement to above average by the end of that grade and sustain that same level until the end of second grade. It was unrealistic to expect that another half grade level could be gained in reading during the second grade from where they had ended up at the end of first grade. But maintaining this level is significant for these children. Not only were these students low academic achievers entering first grade, ninety percent of the students were eligible for free lunch. Also, the gains made by the whole second grade class in the second year of the project, made up of new and old Cornerstone students, equal to more than one and a half grade during that year.
The classes that served as Controls came from effective classrooms. There was very little difference in over-all educational objectives or teaching experience between these classrooms and the Cornerstone classes. The major difference was in curriculum materials and methods. In fact, when the first grade Control classes began the use the Cornerstone program in the next year, similar increases were seen in their student levels of achievement. In reading, their students progressed from the 20th percentile to the 68th in one class and from the 36th percentile to the 70th in the other. In math, in which they had achieved very small gains in the previous year, gains were from the 35th percentile to the 65th in one class and from the 44th to the 72nd in the other. This further demonstrates the added effectiveness of the programs.
Because the differences in teaching approach between the Cornerstone programs and prevailing methods in other schools are more striking, comparisons between these two approaches may have been more significant. This is the kind of comparison that the future Cornerstone projects would like to make.
In reading, these differences center around the question of what is the best way of identifying the words in a given text. The options are: Do good readers read words well 1.) because they are good at deciphering the alphabetic code or 2.) because they use only some of the alphabetic code combined with other contextual information to identify the words? A closely related question centers around the importance of being able to read accurately and easily in the early grades.
The Cornerstone approach places a high value in learning how to read words well, in the early grades, which gives children a sense of success right away and enables them to understand what is read soon. As these skills begin to develop, the young reader is taught how to apply these skills to understand increasingly more complicated texts. For the Cornerstone classes, reading words well is accomplished by explicit and careful sequenced teaching of how to decipher the alphabetic decoding system. It is believed that the best way to learn how to read words accurately and easily is by learning the phonetic basis of the alphabetic writing system. This enables children to learn to read early, and it produces a high rate of success in the early grades, as was demonstrated in the Cornerstone project.
In more widely used programs, phonetic knowledge is only considered to be one of several sources of information that the young reader uses to know what the words are. The way in which words are read with this approach gives a small role to the phonetic basis of the alphabetic writing system. Therefore, phonetic knowledge does not need to be thoroughly taught. Neither is there a need to teach it systematically. Whatever knowledge of phonics is needed can be learned as needed from reading children's literature.
This approach also does not consider a high level of word accuracy to be important in the early grades. For example, the State of Oregon standard for word accuracy by the end of third grade is very low. This approach assumes that children can understand what is read without reading every word accurately. Increased accuracy with this approach becomes more important in the later grades and is assumed to be the result of using a combination of phonetic and contextual cues in the text. Even then, it is assumed that good readers do not need to read every word exactly right in order to comprehend meanings. This approach has resulted in slow progress by most students and a high rate of failure. As it turns out, poor readers are more likely to refer to contextual cues in the attempted of identifying words than good readers. With this approach, serious gaps are created in reading ability and in vocabulary development that handicap children throughout their schooling and beyond.
Math instruction also contains a similar difference in approach. The Cornerstone approach systematically teaches skills and knowledge of the number system and basic operations and concepts. As this knowledge is learned, it is applied to solving common problems. The prevailing methods work somewhat in reverse. With these methods, attempts to teach skills and knowledge of basic operations and concepts are done as the children attempt to solve problems.
In both reading and math, the prevailing approach counts on children discovering many of the basic skills and knowledge on there own, as they are occupied with interesting literature or with interesting problems to solve. The Cornerstone programs incrementally teach all that children need to know and do in each subject from the simplest and easiest to the more difficult and complex. Independent mastery is sought each step of the way.
The question of whether or not the prevailing programs can produce similar results as the Cornerstone classes remains unanswered? Schools using these programs have so far been reluctant to measure their results at these grade levels in the same way the Cornerstone classes were measured.
The high mobility rate of early primary age children illustrates how difficult it is to conduct a study in the early grades. It also illustrates how difficult it is to attain grade level achievements in the benchmark grades of 3 and 5. Children often do suffer academically from changing schools. High mobility also can indicate a lower economic level of the families. This lower economic level often is accompanied with lower educational levels. The largest numbers of children that move into the schools in later grades are far below grade level. In spite of these conditions, the yearly results of the Cornerstone project indicate that these students can catch-up to grade level and meet these standards. Yet, the further behind the students are, the harder it is to make these standards in one year. Children that are allowed to get behind by two years are not likely to catch-up in one year.
In the second year of the Cornerstone project, only two students in the second grade fell below the 40th percentile, which is considered within the average range. One was a Russian speaking child whose attendance was poor. Yet, he was able to read second grade books. His scores in the Basic Skills portion of the Woodcock Reading individual test was at the 43rd percentile. His lack of English continues to slow his progress. Yet, it can hardly be said that he is a non-reader. The second child who received a score lower than the 40th percentile missed several weeks of school because of an accident in which he lost his sight in one eye. Even with this, he gained 15 points in reading. In the second year first-grade class, only one student fell below the 40th percentile on the SAT-9. However, on the individually administered WRMT no one from this class scored below this mark. In math, only one student fell below this mark in each grade during the second year. This demonstrates that failure can be prevented.
What have we learned as a result of working in this project? In order to achieve these kinds of results, the following conclusions have been made.
1. For most students, learning how to read is a hard task. This is particularly true for students in the lower 40% of the population. We know that difficulties in learning to read do not necessarily mean difficulties in all learning, but we also know that not learning to read well in the early grades results in difficulties learning other subjects. It is sometimes expressed that first we learn to read and then we read to learn.
2. In order to teach all students well,
two factors must be in place. a.) effective programs must be in
the hands of the teachers, programs that have demonstrated this
b.) teachers are needed who can skillfully use these programs and be able to motivate students to apply their full effort, i.e., try very hard, to learn what is difficult. Neither factor alone is sufficient.
3. In addition to learning academic skills and knowledge, children need to learn how to learn. This means learning strategies and learning how to apply high levels of effort when they are in school. This implies that they need to learn that learning at school is serious business. Teachers need to know how to teach this approach to learning as well as know how to teach the academic content.
4. Once children experience the success that effort, serious concentration, smart strategies and good teaching brings, most enjoy school and become very proud of their accomplishments.
5. Even with the best teaching tools in their hands, teaching all students successfully is a very difficult task. But without good teaching tools, it is probably impossible.
6. Programs and teachers do exist that can successfully teach all students. The programs used in the Cornerstone Project proved to be effective tools that enable teachers to be successful if they are willing to learn to use them skillfully and if they are willing to work very hard.
7. Skillful use of these programs means applying them to appropriate learning levels of each child. In these programs, learning is seen as a progression of skills and knowledge. Children enter school at different levels and learn and progress at different rates. For the most part, the math and language programs accommodate for these differences. However, in reading, this is not possible. In order to assure that all children learn to read, it has been learned that there must be three levels of small group instruction in each class.
8. Part of the objectives for this project was to test the applicability and practicality of these programs without the use of teaching assistants during reading instruction. Because of the need to teach three small reading groups, this did not prove successful. Although the project was successful without them, after experiencing two years without an assistant for reading, it is the judgement of this teacher that the consequences of not having teaching assistants in the first grade, at least, is too costly. It is costly in terms of a teacher's energy, effort, time and well being.
Although there are some teachers who are willing to do this, it is inappropriate to ask a teacher to teach three reading groups consecutively every day and manage a full classroom with appropriate activities at the same time. The seatwork time, during small group instruction, is beyond the capabilities of many of the children in first grade to successfully manage by themselves. Given the difficulties in teaching all students successfully, teaching each group requires too much concentration to simultaneously attend to the rest of the class at their desks. An assistant is needed to teach one of those groups while the teacher is teaching one group. The third group can be taught by the teacher. This shortens the whole reading time, which allows more time for other activities, and shortens the time in which one group is being taught with the rest of the class doing seatwork.
In many schools, the teaching assistants can be found from existing programs. Assigning these assistants to the first grade for one hour is possible if priorities are weighed. The cost of additional assistants for each first grade classroom in other schools is not an unreasonable possibility and is well worth the cost if it means that all children can definitely be successful. Without this assistance, a teacher, who knows what can be accomplished with these programs, is placed in a very difficult situation. The teacher must choose between success for all and other important activities and considerations.
9. The programs that are successful may be less costly than most, initially, but more costly on a yearly basis. The total expense over a 5 or 6 year period is about the same. Because of the results, any added expense is considered worth it. It can save money in the long run. The Cornerstone project teachers did not have to look for extra special education help during these two years. If this can be sustained, it should result in large savings.
10. Given the success that this project has shown so far, it is important that information of the results be shared with other schools and teachers, and that teachers be encouraged and given the means to make use of these programs and this approach if they show an interest and an ability to do so. It is also important that the project continue in order to continually test its effectiveness.