The Four Blocks framework was first implemented in one first grade classroom.  In the decade since then, the framework has been implemented in thousands of primary classrooms throughout the country.  Different schools and school district have different ways of assessing how children are growing and evaluating the instructional program.   To answer this question, we will summarize results from three schools or districts.  (For more details see Cunningham, Hall & Defee, Nonability Grouped, Multilevel Instruction: Eight Years Later,  The Reading Teacher, May, 1998.)

Clemmons Elementary School, the school in which the framework was originally implemented is a large suburban school with a diverse student population.  Some children come from homes surrounding the school and others are bussed from the inner city.  In any year, 20-25 percent of children qualify for free or reduced-priced lunches.  Approximately 25-30 percent of the children are African-American, Hispanic or Asian-Pacific Islander.  Since the program began, the student population has remained relatively stable, with approximately 10 percent of the children moving in and out each year.  All classes are heterogeneously grouped and contain an average of 22 children.

Throughout the year, teachers conduct assessment by observing and conferencing with children, taking running records and looking at writing samples.  At the end of the year, children are given the Basic Reading Inventory (Johns, 1994) by an assessment team headed by the curriculum coordinator.  Instructional levels are computed using the standard procedures and included measures of oral reading accuracy and comprehension as measured by responses to comprehension questions..  Because the IRI is administered at the end of the year, an instructional level of first or second grade is considered grade level at the end of first grade and an instructional level of second or third grade is considered grade level at the end of second grade.

Across five years, instructional level results have remained remarkably consistent.  At the end of first grade, 58-64 percent of the children read above grade level--third grade or above; 22-28 percent read on grade level; 10-17 percent read below grade level--preprimer or primer.  On average, one child each year is unable to meet the instructional level criteria on the preprimer passage.  At the end of second grade, the number at grade level is 14-25 percent. The number above grade level--fourth grade level or above--increases to 68-76 percent The number reading below grade level drops to 2-9 percent.  Standardized test data on these children collected in third, fourth and fifth grades each year indicates that 90% of the children are in the top two quartiles.  Most years, no children's scores fall in the bottom quartile.

The original school in which the framework was implemented does not do standardized testing until the end of third grade. Other districts, however,  do administer standardized reading tests  in the primary grades and one district devised an evaluation model, the results of which will be reported here.  Lexington One in Lexington, South Carolina is a suburban southeastern school district with eight elementary schools, in which 25% of the children qualify for free/reduced price lunch.  During the 1995-96 school year, first grade teachers in the district were given information about the four-blocks framework and allowed to choose whether or not they wanted to implement the framework in their classrooms.  Approximately half of the teachers chose to implement the framework and were provided with several workshops/ books and collegial support throughout the year in their classrooms.

In January 1996, 100 first graders in classrooms using the four-blocks framework and 100 first graders in classrooms not using the framework were randomly selected and were given the Word Recognition in Isolation and Word Recognition in Context sections of the Basic Reading Inventory (Johns, 1994).  Adjusted means for both measures favored students in the four-blocks classrooms.   For the word recognition in context means, the differences were statistically significant.  Students in the four-blocks classrooms were on average reading at the beginning of second grade level.  Students in the other first grades were on average at the first grade, second month level.

While these results were encouraging, district officials were concerned about lack of reliability on the IRI and about teacher bias, fearing that the enthusiasm of the teachers who chose to implement the model may have created a Hawthorne effect.  They then devised an experiment using cohort analysis and standardized test results.   In May of 1996, all 557 first graders in four-blocks classrooms were administered the Metropolitan Achievement Test.  Each child was matched with a first grader from the previous year  (1994-95) based on their scores  on the CSAB (Cognitive Skills Assessment Battery), a test of readiness given each year during the first week of school.  The total reading mean score for the four-blocks first graders was significantly better (.0001 level) than that of the previous years matched students.  In grade equivalent terms, the average four-blocks first grader's total reading was 2.0 while that of the 1994-95 student was 1.6.

This district then analyzed their data by dividing both groups of students into thirds according to their CSAB scores.  This analysis demonstrated that children of all ability levels profited from the multilevel four-block instruction.  There was a 15 point difference in total reading scores for the lower third, a 23 point difference for the middle third and a 28 point difference for the upper third.   The district concluded that organizing in this nonability grouped way had profited the struggling students and had been even more successful for students who would traditionally have been placed in the top group.
 

During the same year, a nearby school adopted the Four Blocks framework and mandated its use in all first and second grade classrooms.  Brockington Elementary School in Florence School District Four in Timmonsville, South Carolina is a small rural district in which 84% of students qualify for free/reduced price lunch.   Based on low achievement tests scores, the elementary school had been placed on the list of the state's worst schools and had tried a variety of approaches to improving reading and math test scores.  During the 1991-92 school year, the school was mandated by a new superintendent to "teach the basics."  A state-developed basic skills curriculum focused on "skill and drill" was implemented along with a computer-lab basic skills remediation program for Chapter 1 students.  End of the year achievement test scores showed no improvement.  During the 1992-93 school year, teachers took a year-long graduate course on whole language.  Again, the end-of-year test results failed to show improvement.

During the 1993-94 school year, another new superintendent arrived.  The district continued to emphasize whole language and teachers were trained in cooperative learning.  This year's test scores showed some improvement at grades two and three, though none at grade one.  During the 1994-95 school year, teachers were urged to continue to use whole language and cooperative learning and they were also trained in Learning Styles approach of Rita Dunn.  It is hard to compare test scores for this year because the state changed from the Stanford Achievement Test to the Metropolitan Achievement Test but scores were the worst they had ever been.  In grade one, only 20% of the students scored at or above the 50th percentile on total reading.  At the second grade level, only 9% scored at or above the 50th percentile.

During the 1995-96 school year, all ten teachers--six at first grade and four at second grade--were trained in and mandated to try the Four Blocks framework.  (It boggles the mind to imagine how enthusiastic and confident these teachers must have been to implement one more "miracle solution!").  These teachers were given workshops/books, state-department, central office support, etc. and in the opinion of those central office and state department facilitators who visited weekly in their classrooms, four of the six first-grade teachers and three of the four second-grade teachers implemented the framework. MAT total reading scores for all first and second graders in that school (including the three classes which did not really implement the framework) indicated that 30 percent of the first graders and 38 percent of the second graders had total reading scores at or above the 50th percentile.

The data from this school system are, of course, open to interpretation.  Since different children were tested in the 1994-95 group and we have no pretest data on these children, we cannot be sure that the huge jump in the number of children reading at or above grade level is due to the implementation of the four-blocks framework.  Officials in this school district, having tried literally "almost everything" in the previous five years, are convinced, however, that the differences are real and attributable to the balanced multilevel instruction which most of the 1995-96 first and second graders received on a daily basis.

This research can be found along with other research on the Four Blocks in the final chapter of "The Teacher's Guide to Four Blocks" (Cunningham, Hall, and Sigmon, 1999, Carson-Dellosa Publishing)