Data Research and Development Center and IERI Research Cmmunity header

The IERI Research Community :: Projects

Testing the Effectiveness, Sustainability and Scalability of an Individualized Reading Program for African-American, Latino and Euro-American Inner-City Children

William Labov


Background: The project confronts the long-standing gap in reading achievement between mainstream and minority groups of students, concentrated in low-income schools in the large cities of the United States. It draws upon the findings summarized in the 1998 NRC report Preventing Reading Difficulties in Young Children and the 2000 National Reading Panel, emphasizing the importance of direct instruction in phoneme/grapheme relations, but recognizing that none of the methods developed so far have been effective with struggling readers in the minority communities.

Purpose: The project tests the effectiveness of an intervention method, the Individualized Reading Program [IRP], which takes into account differences in the linguistic and cultural backgrounds of African-Americans, Latinos and Whites in low-income schools. The IRP is compared to another phonologically oriented intervention built upon the SRA method, based on a uniform process of behavioral modification.

Intervention: The IRP intervention program calls for 40 hours of tutorial instruction with the Individualized Reading Manual with a 1:4 tutor-to-student ratio. It begins with a computer-driven analysis of students' knowledge of combinatory alphabetic relations. Direct instruction is concentrated on those graphemic/phonemic relations where the individual's knowledge is weakest, followed by practice with narrative readings with high self-to-text identification.

Setting: In the period 2001-2004, intervention and assessment was carried out in low-income public schools in Philadelphia, Atlanta, the bay area of northern California, and in the Long Beach area of southern California.

Research Design: Subjects were selected in grades 2 through 4 from schools where over 65 percent of the population was eligible for the federal free or reduced lunch program, who scored below the 35th percentile or below on the Woodcock Johnson III basic reading skills inventory of word identification and word attack. All subject groups were approximately evenly divided into African-American, Euro-American, Latinos who learned to read in English first, and Latinos who learned to read in Spanish first. In 2001-2, reading error analyses were carried out on a total of 738 subjects; complete data before and after intervention were obtained for a cohort of 212. In 2002-3, before and after intervention studies were carried out for 214 subjects. In 2003-4, a controlled comparison study was completed for 258 subjects in 2003-4.

The comparison program (DIR) was developed at Georgia State University, based upon the cognitive-behavioral approach to direct instruction of S. Englemann of SRA. It was originally designed for dyslexic children, is non-specific in cultural orientation and is focused upon attention management.

Progress of the experimental and control method is assessed by a range of standardized tests including Woodcock-Johnson III word-identification, word-attack, passage-comprehension, fluency, Gort Oral Reading. The RX program was developed to analyze 20 categories of syllable decoding errors in a diagnostic reading, and applied before and after intervention.


  • Effect sizes above 1.00 on standardized word attack and word identification tests were achieved for both the experimental (IRP) and control (DIR) programs. The RX analyses of specific decoding skills showed a significant advantage (p< .05) for the IRP over the DIR program for mean improvement and for 13 out of 20 individual grapheme/phoneme relations, with no significant effects in the reverse direction.
  • Latino subjects who learned to read in Spanish first showed several advantages in decoding patterns, particularly in the automatic application of the soft-C rule in oral reading, in the Woodcock-Johnson word-attack tests, and the low level of don't know responses in oral reading.
  • Subjects from California showed significant advantage over other regions in decoding skills. Struggling readers from that region are more likely to have relatively greater difficulty in comprehension than in decoding.

W. Labov. Applying our knowledge of African American English to the problem of raising reading levels n inner-city schools. In Sonja Lanehart (ed.) African American English : State of the Art. Philadelphia: Benjamins. Pp. 299-318.

W. Labov. When ordinary children fail to read. Reading Research Quarterly 38:131-133.

W. Labov and B. Baker. What is a reading error?

You can learn more about this project by visiting