A Research Study
This research was conducted during the 1995-1996 academic year. Note the gains in percentile scores made by those using The Caldwell Reading Program.
Students in two first grade classrooms, who had been randomly assigned to their respective teachers, participated in the study. One teacher used the Treasury of Literature reading program by Harcourt Brace and Company, along with a considerable amount of her own "supplemental phonics" materials. The second teacher, at her request, used The Caldwell Reading Program in its entirety, and minimal portions of the Treasury of Literature program. Students in both classes were given three sub-scales (Word Identification, Word Attack, and Passage Comprehension) of the Woodcock Reading Mastery Tests--Revised in September 1995 and again in May 1996. Participants were individually tested in random order by psychologists who were unaware with which reading group the participants were being instructed. While both groups showed gains from pretest to the post test, those using The Caldwell Reading Program showed significantly greater gains on all three sub-scales, both statistically and practically. The following tables below show the results.
Pretest and Post-test Means of Age-Standardized Scores, Standard Deviations, Z-Scores, and Percentile Scores for the Control Group and the Caldwell Group
1A percentile score is the percent of the population one is equal to or better than. Thus the 50th percentile is
2"NS" means that no significant gains were made in percentile scores. The students in the control group made average improvement from pre to post-test, i.e. they maintained about the same position relative to other students in their grade.
**Significant increase between pretest and post-test and significantly greater than Control group on the post-test.
Note: There were no significant differences in mean sub-scale pretest scores between the control group and Caldwell group.
As can be seen from the table above, in Word Identification students went from slightly above average to the top one-fourth; in Word Attack from about the lower 40 percent to the top one-third, and in Passage Comprehension from about the lower one-third to nearly the upper one-third. All gains for the Caldwell group were statistically significant. No significant gains were made by the control group.
Number and Percent of Students Whose Scores Increased from Pretest to Post-test for the Control Group and the Caldwell Reading Group
The greater percentage of students whose scores increased on all three sub-scales for the Caldwell Group is quite apparent. Therefore the greater test scores for those using the Caldwell Reading Program did not occur because only a few students made unusually large gains. Rather, nearly all of those using the Caldwell Reading Program benefited from it.Because of the results of this research and the very favorable reports by teachers using the CRP, during the 1997-98 academic year it was used in all of the elementary schools in the county district as a supplemental program. The county was awarded honors for showing the most improvement in the entire state on the Stanford Achievement Test/9th Edition for the year.
The following survey was conducted in Marion County, West Virginia, by Jane Reynolds, Curriculum Specialist. The twelve second grade teachers had use the program for approximately six months.
Caldwell Reading Program
Statement Yes No Maybe "?"
* One marked "NA".
Selected Teacher Comments from Those Responding to the Survey:
"On the whole, the program is a wonderful and rewarding program both for the teacher and the students."
"It was repetitious and fun for the students. After they caught on their confidence grew to the point that they wanted to participate." "I have seen the most progress in my students since using the Caldwell program. Their vocabulary has gown as well."
"It built vocabulary for all students, not just students with reading difficulties. Students were able to use what they learned in all other subjects." "I liked all of it. In the beginning it seemed difficult for students but they bean flying through the sounds and activity pages in no time." "I noticed that Caldwell really helped most of my students. I still have one student that has shown no growth at all. I even sent the sounds home on cards for her to study and it didn't help."
Using The Caldwell Reading Program in Kindergarten
This research was conducted during the 1997-1999 academic years. Students in one of four typical kindergarten classes in one elementary school were given instruction in The Caldwell Reading Program and completed the first level of the five level program ( KDG Group). Those students instructed with the program were kept together as a class and given extensive instruction in the program the following year in first grade. The other students were given less extensive instruction with The Caldwell Reading Program and more instruction with the more traditional Harcourt Brace series (TRD Group). Although The Caldwell Reading Program was used with all of the students, those who began with the program in kindergarten and continued with extensive instruction in it in first grade performed much better on the Stanford Early School Achievement Tests (Stanford 9). In fact the lowest score in the class on the Total Reading sub test was at the 59th percentile, compared to the 4th percentile in the other classes, i.e., those with no kindergarten experience with the program and more limited use of it in first grade. The following results are from the Stanford Early School Achievement Tests (Stanford 9) given in the spring of 1999.
Mean Standard Scores, Standard Deviations, and Percentile Scores for the TRD (Traditional) and KDG (Used CRP in Kindergarten) Groups*
Stanford Early Achievement Test (Stanford
Note that all reading score differences are statistically significant with the exception of the comprehension sub test. When considered individually in another analysis, the KDG group was significantly better than one of the other three classes (p.<0196). None of the other classes was significantly better than any other class on comprehension. One might argue that the students in the KDG group were superior in general ability and this superiority accounted for the better performance rather than the reading materials used. If that were the case, one would expect their performance on the math sub test to be superior. As can be seen in the table, there was very little difference in the math scores, means of 578.16 and 578.38 that yielded a p. value of <.5883. i.e., it could have occurred by chance nearly 6 out of 10 times. When considered individually, the KDG group was significantly better in math than one of the other three classes (p<.0311.) But the other two classes also attained scores in math that were significantly better than that same class (p.<.0048 and p.<.0262). Thus it is most unlikely that the KDG group was superior in general ability.
*Four of the students in the KDG Group were considered "LD" and were given the test under non-standard conditions, as were all of the "LD" students in the other three classes. Further analyses of the data for all of the "LD" students was recently initiated.