Evaluating an early intervention reading program

One of the most widely researched and successful early intervention programs for beginning readers is Reading Recovery. There are indications, however, that Reading Recovery may be somewhat less effective with the most disabled readers. Researchers have speculated that these children may need more explicit instruction in phonological awareness than this program provides. A 1993 study compared the standard Reading Recovery program to a modified version that included supplemental, explicit phonological skill training. Children who received Reading Recovery training with additional explicit phonics practice required fewer lessons to achieve grade level performance.

The Kalispell, Montana School District, after reviewing past research and looking at several early-intervention programs, concluded that they wanted a program that shared many of the features of Reading Recovery while providing direct, intensive training in phonological skills at more reasonable cost. Carol M. Santa, Kalispell, Montana School District, and Torleiv Hoien, Dyslexia Research Foundation, Stavanger, Norway, report on an experiment that the Kalispell district carried out with Early Steps, a little-known one-to-one tutorial program that had a philosophy similar to Reading Recovery but more intensive phonics instruction.

The lower cost enabled the district to train all the first-grade and Title 1 teachers in two elementary schools. Because Early Steps lacked the thorough research of Reading Recovery, the district determined that their experiment with Early Steps must include a control group and be carefully evaluated. Early Steps was philosophically similar to the instructional programs already used in the district.

In the Early Steps Program, students spend at least half of their daily 30-minute, one-to-one tutoring session reading and rereading connected text. The teachers use real books to give the students a sense of accomplishment and enjoyment. Instructional level and pace are carefully matched to each child’s current reading ability to provide challenge and to progress systematically to more advanced levels.

Daily rereading of familiar texts

Oral fluency is developed by daily rereading of familiar texts. Daily writing activities encourage children to translate the sounds they hear into written form using invented spelling. At the same time, children receive explicit instruction in phonological analysis, giving them the skills to learn to spell correctly. Word study centers around beginning sounds and rhyming word families.

Experimental study

The district’s study sought to determine whether Early Steps helped at-risk first-graders learn to read. And if so, were these achievement gains maintained over time? Researchers also sought to determine whether students’ benefit from the program is affected by their initial level of reading difficulty. Is the relatively expensive one-to-one tutoring of Early Steps necessary for all children at risk for reading failure, or do some learn just as well in the small-group teaching provided by Title 1?

Four schools in Kalispell participated in the study; two serving as controls while two used the experimental Early Steps program. All four schools qualified for Title 1 funding, and five years of achievement data showed no significant differences in the average achievement of children in these schools. All students in the study continued to participate in a two-hour language arts block plus 30 minutes of independent reading in their classroom each day.

All participating teachers were certified and had at least five years’ experience. At the beginning of the school year, teachers observed and informally evaluated their students. Children whom they judged to be performing
in the lower half of the class were given the Early Reading Screening Instrument to measure their letter knowledge, concept of word in text, spelling, and word recognition. On the basis of these results, 49 children were selected for the study; 23 children were placed in the experimental group and 26 in the control group. The pretests showed no significant differences between these two groups.

The study continued for a full year with 12 professionals (regular classroom teachers, Title 1 teachers, principals and a language-arts coordinator) teaching in the study. Each regular classroom teacher tutored one student in the experimental program outside the classroom for 30 minutes a day during their planning period. Other professionals without regular classroom assignments tutored up to four students individually each day.

All who taught the Early Steps program received ongoing inservice training during the year. The language-arts coordinator in each school was trained as an on-site teacher trainer. The author of the Early Steps program visited the site five times during the year, observing and providing feedback on tutoring sessions. Teachers working with students in the control group received inservice training in guiding reading development in small groups.

Experimental Instruction

Fast-paced and intensive 30-minute sessions included rereading familiar books, word study, writing and introduction of a new book. Students reread books for the first eight to ten minutes. The next five or six minutes were devoted to word study emphasizing the way words with common sounds often share spelling patterns. Once children learned to name and write the letters, they practiced discriminating initial consonant sounds by sorting picture cards. Next, they progressed to word sorting using both auditory and visual patterns. They began with rhyming word families and progressed through short- and long-vowel patterns.

Instruction emphasized developing metacognitive knowledge and strategies for reading and writing. When a student came to an unknown word during book reading, the teacher used prompts such as:

  • What can you do to figure out that word by yourself? What are some of the strategies that might help you?
  • Read that line again. What does the word start with?
  • I think the word is similar to some words that we have done in the word sorts. What word pattern do you know that might help you?
  • Read through the sentence. Leave the word out. What word makes sense there?

Five to eight minutes were spent on writing each day. The child begins by writing a sentence from her own experience. She says each word as she writes it. The teacher encourages her to write each sound she hears, calling attention to specific letter sounds and words. The child then rereads the sentence. The teacher rewrites the sentence, the child reads it once again and the teacher cuts it apart for the the child to put together and reread.

Lastly, a new book is introduced. Student and teacher look through the book, discussing the story and vocabulary. Teachers found coaching students during their initial reading of a new book the most difficult part of the Early Steps program. Children need time to figure out words on their own and correct their own errors. Teachers had to learn how long to let the student struggle, and when and how to prompt and model effective strategies.

Control group instruction

Children serving as controls were given an extra 30 minutes of reading practice a day in small groups of two to four students. Books were carefully selected to provide challenge and gradually become more difficult. Teachers often read a page followed by the students chorally rereading the same page. Students also practiced reading in pairs. No writing or word study activities were included in these sessions.

Year-End Results

At the end of the year, all experimental and control students were testing in spelling, word recognition, and passage reading. Students who received one-to-one tutoring in the experimental Early Steps program significantly outperformed control students in all three areas. No gender differences were found.

To analyze these results further, at-risk students in both experimental and control conditions were identified as either at high risk or low risk of reading failure. When year-end results were compared using these high- and low-risk groups, the high-risk experimental group significantly outperformed the high-risk controls, but there was no statistically significant difference between the mean performances of low-risk experimental and low-risk control groups.

Therefore, the difference between the total experimental and control groups was caused solely by the progress made by the children at highest risk (those that scored lowest on pretests). The results are dramatic when comparing end-of-year reading levels. More than half of the experimental group (12 of 23 students) were reading at or above grade level, and only four of these 23 students scored at a low first-grade level.

In contrast, only 24 percent of the control students (6 of 25) completed the year at grade level or above, while 10 of them were at a beginning first-grade level. In summary, the experimental group significantly outperformed the control group on all tests at the end of the year. However, the higher achievement of experimental students in the Early Steps program was due to the progress made by the highest-risk students.

A follow-up evaluation was done at the beginning of the next school year to further validate the end-of-the-year results and to test the stability of the achievement gains. The follow-up evaluation included tests of word recognition, passage reading and, for the first time, nonword reading. Results revealed that experimental students still outperformed control students. The size of the differences favoring experimental students was particularly great for high-risk students in nonword- and passage-reading tests.

Concluding Discussion

Santa and Hoien conclude that this study shows that Early Steps led to accelerated growth, particularly for children most at risk for not learning to read. These results were validated on a variety of independent measures and were maintained over the summer. Teaching phonological processing skills seems particularly beneficial for students most at risk. However, extra time each day, reading books on an appropriate level of difficulty in a small group, may be enough to accelerate the performance of students with less risk of failure.

These researchers found the results of nonword reading particularly interesting because it is among the purest measures of phonological competency and, therefore, indicates that Early Steps is successful in teaching phonological skills. It is important because nonword reading has been shown to be a good predictor of future reading achievement.

These researchers caution that the generalizability of these results is limited because the constraints of the schools did not allow for completely random assignment of conditions, students and teachers. In addition, it is not possible to determine the exact cause of the achievement differences because the two programs differed in more than one variable. They not only used different curricula, but the experimental group had individual teaching while the control group used a small-group format.

Santa and Hoien also point out that it would have been more informative to continue this study with the same students for another year or two. However, the school district felt it would be unethical for the poorest readers in the control group not to have the benefit of the Early Steps program.

“An Assessment of Early Steps: A Program for Early Intervention of Reading Problems” Reading Research Quarterly Volume 34, Number 1, March 1999 pp. 54-73.

Published in ERN April 1999 Volume 12 Number 4

Leave a Reply

  • (will not be published)