Study Overview

Improving the Reading-related Skills of Preadolescents with Learning Disabilities:  A Comparison of Four Marketplace Alternatives



Approximately 15% of children in the United States’ school system are dyslexic or experience so much difficulty learning to read that they are considered learning disabled with educational and occupational prospects that are much impaired.  To grapple with this deficit and the growing public awareness of its societal costs, several programs have emerged in the marketplace for promoting the development of reading-related skills and capabilities of preadolescents with learning disabilities. These programs vary considerably, from mode of intervention to level of adult involvement to expense. Yet, to date, little is known about their relative effectiveness.


To address this issue, this project will build on an already-conducted initial field test to mount a quasi-experimental pre-post evaluation designed to compare four such programs.  A large cohort of K-3rd grade children, each with a dyslexia or related learning disability diagnosis, will be recruited to participate in a funded course of instruction via one of these four programs. To maximize the level playing field nature of the comparisons, systematic efforts will be made to give each program “its best shot” on its own terms as an instructional medium over a normal academic school year. This evaluation will compare and contrast the effectiveness of the four programs in bringing the reading-related skills of these children up to or beyond grade level by the end of the study.


Four Programs for Reading-Related Skill Development

Orton-Gillingham (O-G) is the primary learning program today, used by 95% of schools for the learning disabled throughout the U.S. It is predicated – and dependent — on a highly personalized and customized connection between student and teacher that relies on the use of systematic phonics, applied linguistics, and a process of continuous feedback and positive reinforcement. Its approach is structured, language-based, multi-sensory, sequential and cumulative. While the heart of the curriculum is focused on grades 3, 4 and 5, 20 to 30% of pupils are tutored all the way throughout high school.


Lindamood-Bell (L-B) develops language processing by focusing on the underlying skills necessary for independence and self-correction. Although likewise dependent on interaction between students and teachers outside the home, this focus downplays the use of diagnostic labels with an emphasis on cultivating students’ self-confidence and trust in their own abilities. Central to this method are multi-sensory reading tools for developing phonemic awareness. These tools teach students to utilize oral-motor movements to attach sounds that phonemes make within words and to manipulate sounds within words.


Learning to Read (LtR) is predicated on two premises: 1) symbol confusion is the primary culprit for children with learning disabilities; and 2) parent/guardians and other non-professional trainers can materially assist these children in mastering this confusion. These children typically require the processing of text in concrete, three-dimensional terms for comprehension and often get hamstrung when encountering sight words (e.g., “the”) and symbols (e.g., punctuation marks such as periods) for which they have no such referents.  LtR techniques employ multi-sensory tools for mastering these sight words – what they look and sound like as well as what they mean. These tools are laid out in a set of 13 hardcopy volumes that retain the tutorial student/teacher mode.  They are formatted for use not only by teachers in school settings but also by parent/guardians and other non-professional trainers in the convenience and privacy of the home or other non-school settings.


Undoda-Discover Intensive Phonics (U-D): The U-D program likewise zeroes in on the mastery of abstract sight words, but as an internet mobile application that is also coupled with the Discover Intensive PhonicsTM computerized phonics course provided by Reading Horizons, LTD. The Undoda component combines the imagination of a narrated children’s storybook with the pure action of touch screen and tilt-based gaming to cultivate the concept of multi-mind learning without sacrificing fun. It consists of three different learning games which together provide 25 different levels of play. The Discover Phonics component teaches the 42 sounds of the alphabet, and moves sequentially through blends, word formation, and seven skills for decoding multi-syllable words. Students learn common sight words, sentence structure, word meaning, handwriting, and spelling, as well as listening and thinking skills. 


Of these four programs, the O-G and L-B programs hinge on participation in highly structured settings outside the home, whereas the LtR and U-D programs prioritize more flexible technologies equipped for use in the home. The four programs also differ in the extent of reliance of tutoring by paid professionals.  The O-G and L-B programs are virtually entirely reliant in this regard, whereas the LtR and U-D programs are explicitly formatted in ways that will accommodate tutoring by parents/custodians and other non-professional trainers. A third differentiating dimension is the extent to which the children being coached actually drive the instruction process. The U-D program stands out in this regard with its computer gaming component, which is largely paced by the children themselves.


Framing the Evaluation Context 

These differences lead to several questions. To what extent can primarily home-based methods geared to accommodate the schedules of students and their parents supplant more highly structured courses of instruction that require travel to external school settings? To what degree can improvement for dyslexic and similarly learning disabled children be achieved when a relatively narrow culprit like symbol confusion is singled out as the primary target? To what extent is program effectiveness contingent on the active engagement of a paid adult professional? How do children’s outcomes compare when they take on more of a driver role in the instruction process? What impact does new media technology have on reading comprehension? The comparisons mounted in the evaluation will be framed to address such questions.


Project Design Overview

A total of 1000-2,500 K-3rd grade children will be recruited to participate in this study, which will be 12 months in duration.  Each of these children will be linked to one of five study conditions, each of which will be configured to operate within the context of the academic school year in real-world school settings. Four of these conditions will each receive a nine-month course of instruction via one of the four skill-development programs.  The fifth condition, selected to represent the “grade level” performance group, will consist of K-3rd grade school children drawn from “typical” school settings that do not expressly target dyslexia or related learning disability diagnoses.


Assessment of these five groups of children, their parents/guardians, trainers and teachers will be conducted at baseline and at regular intervals throughout the study period to monitor student change on reading comprehension, fluency, math skills, self-esteem and related measures.  To the extent practical, control for potentially confounding factors — achieved through research design and/or statistical analysis — will be implemented to enable attributions of program effectiveness that are credible. The study protocol for recruitment and participation will be reviewed and approved in advance by an institutional review board for compliance with guidelines for the protection of health information under the Health Insurance Portability and Accountability Act.



Children participants will be rising K-3rd grade students who either have received a doctor diagnosis of dyslexia or a related learning disability OR whose parents/custodians have elected, even in the absence of a doctor’s diagnosis, to place them into treatment for one or more of these disorders. Entry into the study will require three commitments upfront from participating families: 1) agreement to read orientation materials; 2) agreement to complete the full course of program instruction; and 3) agreement to complete each round of assessment as described below.


Eligibility will allow for histories of past exposure to, and participation in, other competing methods of instruction, including those to be compared in this evaluation.  Any such histories will be measured and eventually accounted for in the analyses and, as such, will not be treated as a basis for exclusion from the study.



Recruitment for each of the four program conditions will be a consultative process supervised by a team of research experts, whereby children participants will be placed into the condition based on the choice made by his or her family to engage the recommended program. Recruitment into the fifth condition (i.e., the grade level performance group) will be made through negotiation with several suitable metropolitan schools.


As currently envisioned, each of the five study conditions will be comprised of a total of 200-500 children. The number of subjects per condition ultimately selected within this range will be driven by both logistical and statistical power considerations.  The logistical criteria will depend on the arrangements negotiated with each “feeder” source – negotiations which are ongoing and yet to be finalized. The statistical power criteria, on the other hand, will be driven by calculations conducted on means and standard deviations on selected outcome measures drawn from the field test. The objective will be to maximize – within the constraints of what turns out to be logistically feasible in terms of school and subject recruitment — the power to detect group differences. For example, at the high end of the 200-500 range, assuming a 25% attrition rate during the course of the study and an alpha level of 5% for detecting a Type 1 error, these calculations suggest that the resulting sample size of 375 will have the capacity to detect group differences of three points on a 0-100 point scale with 90.7% statistical power. Details for these calculations are available on request.


As incentives to encourage participation, each student and/or his family will receive a minimum of $500 for completion of the entire study protocol.


“Feeder” Sources  

The recruitment of “feeder” sources for candidate students will vary for the four conditions.  It is anticipated that up to three “feeder” schools will be tapped as sources for each of the O-G and L-M study conditions. Overtures by the Charter Group will come from home school networks over a variety of states. It is further anticipated that students for the LtR and U-D conditions will be drawn from the population of children who are currently in private practice/treatment using these tools and/or are in families which have recently purchased either of these programs.



The “treatment” phase, during which subjects in each of the four programs will participate in and complete the course of instruction laid out by the program. For the O-G and L-B conditions, the paid professional teacher in whose course the child is enrolled will function as the medium for instruction.  For the LtR condition, a parent/guardian or layperson trainer will be selected as the medium of instruction for each child. Similarly, for the U-D condition, a parent/guardian or layperson trainer will function as the medium of instruction for the phonics component, while the child him or herself will be relied upon to complete the gaming component.


For each condition, the interface between the medium of instruction and the data collection phase will be closely monitored by the evaluation team.


Data Collection

Three rounds of measurement will be conducted. The first, captured during the 90-day month prior to the initiation date of the intervention (i.e., the start of the academic school year), will comprise the baseline assessment. The second, captured between the fall and winter/spring semesters, will comprise the midpoint assessment. The third, captured during the 90-day period following the end of the academic school year, will comprise the final assessment.


Each round will require the completion of three instruments. One will consist of standardized assessments of student academic performance, social skills and self-esteem. A review of the literature, including various instruments in the public domain that assess such concepts (e.g., the Behavioral Academic Self-esteem Rating Scale developed by Coopersmith & Gilbert (1982), is now being conducted to assemble this instrument.   


A second instrument will be delivered to the teacher/trainer serving as the medium of instruction for the child to assess his/her perceptions of the child’s functioning. As currently envisioned, this second instrument will be patterned after the survey used in the above-mentioned field test.  This latter questionnaire consists of approximately 120 items depending on how respondents answered certain contingency questions. It includes sets of multiple items for monitoring the following:


For the child – Reading, Other Cognitive Skills, Communication, and Emotional-Social.


For the Adult Trainer — Outlook.  It is also expected that this second instrument will include some 30 items to assess actual exposure to the treatment and experiences in working with it.


A third instrument will be delivered to the child’s parent/custodian in those instances where the child’s trainer is a non-parent/custodian. At a minimum, the parent/guardians of each child in the O-G and L-B conditions with receive one administration of this third instrument per household, as will the parent/guardian of each child in the LtR and U-D conditions whose trainer is an individual other than a parent/guardian.  This third instrument will be patterned after the second instrument above, but formatted to embrace the fact that the respondent did not function as the medium of instruction.


To facilitate the analyses of change over time, the versions of each of these three instruments that are slated for the second (i.e., midpoint) and third (i.e., final) rounds of data collection will be closely modeled after the baseline version.  Likewise, it will be strongly recommended that the same individual complete each of the three rounds for a given instrument for a given child to minimize responder bias across rounds.



As currently envisioned, three reports based on the data collected and analyzed are slated as deliverables for this project.  The first will be based on the baseline assessment. It will primarily serve to profile the participant children, parent/custodians and teacher/trainers and to determine the extent of equivalence between study conditions in children functioning and performance prior to the treatment phase. The second will be based on the midpoint assessment and will primarily serve to track the extent of change from baseline to the midpoint by study condition. The final deliverable will incorporate all three rounds of assessment and structured to deliver concluding assessments of each program’s relative effectiveness in bringing its subjects up to grade level performance.


Timeline (Tentative)

The following is the tentative timeline that has been set for this project. This timeline is still a work-in-progress, subject to a variety of factors that have yet to be finalized.

  • March – 2014: Recruitment of children, parent/guardians, teachers and feeder sources for all five study conditions
  • August, 2014: Baseline round of assessment
  • October, 2014: Delivery of Baseline Report
  • Late August-December, 2014: First half of intervention treatment
  • June, 2015: Midpoint round of assessment
  • March 2014: Delivery of Midpoint Report
  • Late June – May, 2015: Second half of intervention treatment
  • June, 2015: Final round of assessment.
  • August, 2015: Delivery of Final Report


All results will be made available to the public on this website.