Fast ForWord uses computerized drills in which children identify computer-generated speech sounds (although the latest versions of the product apparently includes others kinds of computerized training as well). Typically, a program of treatment involves 100 minutes of training a day, 5 days a week, for six weeks
In the speech-sound drills, the training program starts off with sounds that have been altered by computer processing. These processed sounds preserve the frequency content of normal speech sounds, but are slowed down and have artificially exaggerated differences. These changes make the task easier for children with slower than normal temporal processing, but paradoxically are more difficult to discriminate for temporal processing normals. As the subject progresses, these differences are reduced to make the games more challenging. The premise of this approach is that the drills help students with a wide range of language problems develop enhanced phonological awareness, and that this enhanced awareness will have numerous benefits for their language functioning, including especially reading. The method of utilizing exaggerated differences in training a person to tell two things apart is commonly referred to in psychology as "fading". Fading has been widely used beginning in the 1950s in many areas of behavioral research and treatment, including animal learning (e.g., Lawrence, 1952), behavioral therapy of retarded individuals (e.g., Irvin & Bellamy, 1977), and problems with perception of speech sounds (e.g., Jamieson & Morosan, 1989).
Although the Fast ForWord programs were originally demonstrated to show statistically significant improvements in phonological awareness for children with slower than normal temporal processing, much debate exists over the generalizability of these effects to a broader range of children who are sub-par in their language skills.
Fast ForWord evolved from computer games that were demonstrated to deliver a 1-2 year age equivalent improvement in phonological awareness to children who had abnormally slow temporal processing. The efficacy of Fast ForWord for the broad population of children with reading problems--for which it is currently being marketed--has not been well established, and the most methodologically rigorous studies seem to call the efficacy into question.
Randomized Controlled Trials
Drawing inferences about the effectiveness of an intervention is done most powerfully with randomized controlled trials with appropriate outcome measurements. In this kind of study, individuals are randomly assigned to receive an intervention, or to receive whatever would be the conventional treatment. The outcome measure must be a set of pre-defined measures that assess the severity of the complaints for which treatment is being administered. This is the basis of the FDA's evaluation of new drugs, and is generally viewed as the "gold standard" for any kind of intervention research. Optimally, such a trial should be run with the so-called intention to treat procedure, which helps to prevent drop-outs from obscuring the results of the study.
Several studies have been published that have evaluated Fast ForWord using randomized controlled trial designs. In 2004, Cecelia Rouse and Alan Krueger from Princeton University published a study of Fast ForWord in a large urban school in the Northeast. Their sample consisted of 374 students who scored in the bottom 20% on the state's reading test. They found that although certain aspects of the children's language skills were slightly improved, "it does not appear that these gains translate … into actual reading skills" (Rouse & Krueger, 2004, p. 2). A potential confound in the Rouse and Kruger study is the fraction of the children who may have had reading difficulties because English was not their primary language. The Fast ForWord program was conceived to ameliorate language difficulties related to temporal processing as measured with the Tallal Repetition task. The control and experimental groups in the Rouse and Kruger study were sampled from schools in which 56% of the children spoke a language other than English at home, and over 65% of the control and experimental group came from Hispanic families
A second, and larger, randomized study was carried out by Geoffrey D. Borman, now at the University of Wisconsin School of Education, and colleagues These investigators studied 415 second and seventh graders performing far below national reading standards. Students randomly assigned to receive Fast ForWord treatment did not show statistically significant improvement in most of the reading measures examined, although there were a few small gains for certain subgroups, and a significant fraction of students with the lowest language test scores dropped out. This paper is currently in press in Educational Evaluation and Policy Analysis.
While there have been several positive results reported with Fast ForWord from randomized designs, these do not appear to constitute randomized controlled trials assessing the effectiveness of Fast ForWord as a reading intervention for "garden-variety" reading problems encountered in a school.
Scientific Learning Corporation's website reports one study that used a randomized design involving 208 elementary school students. This study found that Fast ForWord treatment enhanced performance on the Test of Phonological Awareness (TOPA) The study included subjects at all reading levels, not just those deemed impaired, and used both first and second graders. The average improvement on the TOPA was half a standard deviation across the board. When the student year was entered into their statistical model, only the first graders showed an improvement. However, the study does not report whether these early reading skill improvements translate into changes in reading or other language-related functions.
The Scientific Learning Corporation website lists many dozens of studies with positive results that do not use the "gold standard" randomized designs, but instead compare children's performance before and after treatment. For example, in a study on Australian children age 6-13, the average improvement was from the 14th to the 32nd percentile. These kinds of studies are difficult to interpret because children with reading problems may show some degree of improvement over time. Such improvements could result both from maturation and from other experiences in and out of the classroom (which might even include other treatments or tutoring provided on top of the therapy being evaluated). An untreated control group is the best way to isolate the impact of the treatment.
In early studies that pre-dated the commercial development of Fast ForWord, Merzenich et al. (1996) and Tallal et al. (1996) reported that 8-16 hours of training using Fast ForWord produced "1.5 to two years of progress in reading skills", according to Tallal, quoted in Newsweek (Begley & Check, 2000). These age-equivalent improvements are based on their published data expressed in the same format (Science. 1996 Jan 5;271(5245):81-4).
The Scientific Learning Corporation website does not appear to state what specific diagnostic indications the product is being marketed for. The search engine keywords listed on the website include "improve reading scores improve reading skills" which suggests that the key market includes children with virtually any sort of reading problem. However, the website also describes results involving "pervasive developmental disorders" (a term usually applied to children with problems in the autism spectrum) , attention deficit disorder and non-native English language learners , and quotes testimonials relating to the "broad range" of students supposedly helped by the product.
The Scientific Learning Corporation does not list the current cost of Fast ForWord on its website. One reseller's website in 2001 estimates the cost to a school district at $38,000 for 50 children, and Rouse and Krueger (2004) provide a similar estimate.
The Fast ForWord products evolved from the work of a number of scientists, including Michael Merzenich and Bill Jenkins at the University of California, San Francisco, and Paula Tallal and Steven Miller at Rutgers University. This team started Scientific Learning shortly after publishing two papers in Science (Science. 1996 Jan 5;271(5245):77-81, and 81-84). These papers demonstrated that children who had abnormal temporal processing could be trained on software (which later evolved into Fast ForWord). This training results in 1-2 years of age equivalent improvement in language reception measures. The magnitude of the improvement, subject by subject, was correlated with their improvement in temporal processing. In other words, these studies showed that software like Fast ForWord, when applied to subjects with abnormally poor temporal processing and reading skills, could remediate both their temporal processing and language reception powerfully, and further suggest that temporal processing abnormalities can form a perceptual bottleneck in learning to comprehend language. The studies also included control groups and found significant differences in language reception improvements between control and experimental groups.
The founders of the Scientific Learning Corporation have impressive scientific credentials. Michael Merzenich, who is a member of the National Academy of Siences and is also the Chief Scientific Officer for Posit Science, and Bill Jenkins (currently Senior Vice President of Product Development at Scientific Learning) are internationally known for their research on brain plasticity, which is the concept that the brain changes as we learn new skills. Paula Tallal is currently co-director of the Center for Molecular and Behavioral Neuroscience at Rutgers and an active participant on many scientific advisory boards and government committees for both developmental language disorders and learning disabilities. She has published over 150 papers on the topic of language and learning and is the recipient of national and international honors. Steven Miller, currently Senior Vice President of Research at Scientific Learning, has extensive experience in organizing clinical research studies and conducting longitudinal studies of children who have language and reading problems.
Begley, S., & Check, E. (2000). Rewiring your gray matter. Newsweek, Jan 1, 63.
Borman, G.D., & Benson, J. (in press). A randomized field trial of the Fast ForWord Language computer-based training program. Educational Evaluation and Policy Analysis.
Irvin, L. K., & Bellamy, G. T. ation of stimulus features in vocational-skill training of severely retarded (1977). Manipulindividuals. American Journal of Mental Deficiency, 81, 486-491.
Lawrence, D. H. (1952). The transfer of a discrimination along a continuum. Journal of Comparative and Physiological Psychology, 45, 511-516.
Merzenich, M. M., Jenkins, W. M., Johnston, P., Schreiner, C., Miller, S. L., & Tallal, P. (1996). Temporal processing deficits of language-learning impaired children ameliorated by training. Science, 271, 77-81.
Tallal, P., Miller, S. L., Bedi, G., Byma, G., Wang, X., Nagarajam, S. ., Schreiner, C., Jenkins, W. M., & Merzenich, M. M. (1996). Language comprehension in language-learning impaired children improved with acoustically modified speech. Science, 271, 81-84.