During the 1990s, Michael Merzenich and Paula Tallal worked together to create a computerized language intervention program called Fast ForWord-Language^TM (FFW-L, Scientific Learning Corporation, 1998). The design of the FFW-L program was based on scientific discoveries about auditory processing disorders (e.g., Tallal & Piercy, 1974) and training protocols that had resulted in neural changes in nonhuman primates (e.g., Jenkins, Merzenich, Ochs, Allard, & Guic-Robles, 1990; Recanzone, Merzenich, & Jenkins, 1992).

Merzenich and colleagues (2002) have suggested that these strategies should be used in intervention with humans because they have been proven to drive brain remodeling in animals, and neuroplasticity-based training principles are being applied in many areas of speech and language rehabilitation (e.g., Ludlow et al., 2009). The neuroplasticity-based training principles underlying FFW-L (Byl, Merzenich, & Jenkins, 1996; Fitch & Tallal, 2003) are summarized in Table 1 [PDF]. Whether or not the neuroplasticity-based training principles of FFW-L are necessary components of language intervention can now be assessed in light of new evidence from a randomized controlled trial.

Randomized Controlled Trial

We conducted a large-scale randomized controlled trial (RCT) that compared the language outcomes of FFW-L to another computer-assisted instruction intervention, clinician-directed language intervention, and a non-intervention condition that focused on academic enrichment (Gillam et al., 2008; Loeb et al., 2009). The study was designed to answer a theoretically driven question: Does an intervention that employs auditory-temporal training with neuroplasticity-based procedures result in unusual improvement on measures of auditory processing, language, and literacy skills in children with specific language impairments?

The RCT included 216 school-age children with language impairments from nine school districts in Texas and Kansas. The participants were randomly assigned to one of four conditions.

FFW-L consisted of seven computer games that target auditory discrimination, memory for speech, and grammatical comprehension. The activities in the FFW-L games followed the behavioral principles of discrete trials teaching (Brown-Chidsey & Steege, 2004). A signal to attend was followed by a discriminative stimulus that prompted a mouse-click response. Correct responses were rewarded by points, jingles, and extra animations on the computer screen. The auditory stimuli in FFW-L were modified in ways that prolonged segments and differentially amplified particular frequencies (Nagarian et al., 1998). The extent of the auditory modifications and the difficulty of the stimulus items were constantly adapted in response to the children's response accuracy.

The other three conditions were similar to FFW-L in that they were delivered on an intense daily schedule and children were required to attend carefully to the instructional activities, received feedback about the correctness of their responses, and received rewards for successful responses and participation. The three conditions also differed from FFW-L in several ways (see Table 2 [PDF]).

Individual language intervention (ILI) compared FFW-L to a functional communication approach. Clinicians provided one-to-one language intervention in a quiet room. The ILI activities were not repetitive. Instead, ILI intervention promoted multiple opportunities for therapeutic discourse (e.g., growth-relevant recasts, focused stimulation, scaffolding) during conversations about children's literature (Gillam & Ukrainetz, 2005; Hoggan & Strong, 1994; Loeb, 2003). Clinicians did not tally each correct response; they tracked children's progress by rating the level of clinician effort and the level of child responsiveness for each activity.

Computer-assisted language intervention (CALI) assessed the importance of modified speech. Like the FFW-L exercises, the CALI modules targeted auditory discrimination, memory for speech, and grammatical comprehension. The CALI games followed a discrete trials presentation pattern similar to the one used in the FFW-L games; none of the CALI modules, however, contained modified speech.

Academic Enrichment (AE) was designed to be an attention control. Children played academically oriented computer games that focused on reasoning, mathematics, social studies, and science but were not designed to teach language skills to children with language impairments.

Study Outcomes

The participants were tested before treatment, immediately after treatment ended, and three and six months post-treatment. Children received a battery of auditory processing, language, and literacy
measures that included a backward masking task (Marler, Champlin, & Gillam, 2001), the Comprehensive Assessment of Spoken Language (CASL; Carrow-Woolfolk, 1999), the Test of Narrative Language (TNL; Gillam & Pearson, 2004), the Comprehensive Test of Phonological Processing (CTOPP; Wagner, Torgesen, & Rashotte, 1999), and the Woodcock Reading Mastery Test-Revised (WRMT-R; Woodcock, 1987).

We hypothesized that FFW-L and/or CALI would promote auditory processing and phono-logical awareness, and that ILI would enhance abulary, grammar, and narration skills. We expected minimal change in the children in the AE group.

The results were surprising. Children in all four groups improved significantly on auditory processing, language, and phonological awareness measures immediately after treatment, and continued to improve for the next six months. Over the course of the study, the CASL scores of more than 70% of the children in each group either moved into the normal range or improved beyond the 95% confidence interval of their pretest scores (Gillam et al., 2008). Across all analyses, results suggest that very different kinds of language-learning experiences can offer a wide range of auditory processing, language, and phonological awareness benefits to children with specific language impairments.

Core Components

The researchers who created FFW-L argued that interventions incorporating neuroplasticity-based training principles (Table 1 [PDF]) cause brain changes that result in unusual gains on measures of auditory processing, language, and literacy (Byl, Merzenich, & Jenkins, 1996; Fitch & Tallal, 2003; Merzenich et al., 1996). The results of our RCT suggest that not all eight principles of neuroplasticity training are necessary.

The intervention in all four conditions was provided on an intense daily schedule, promoted active attention, provided immediate feedback, and rewarded effort and success. Even though we did not collect direct measures of the children's neurophysiological changes, it is realistic to assume that changes in behavior are tied to neurophysiological changes (Tallal & Gaab, 2006). It is possible that one or a combination of these principles led to the kinds of brain changes that influence improvement of the children's language skills.

The four treatment conditions did not share the principles of repetitive stimuli, item-by-item tracking of correct responses, adapting the difficulty of the stimuli, or modified speech. Therefore, these components of intervention do not appear to be necessary parts of language intervention programs that result in significant changes in auditory processing and language. Highly repetitive teaching, tracking the correctness of every response, systematic adaptation of difficulty levels of activities, or auditory modification of stimuli may influence language learning; however, they do not appear to be necessary components of successful language intervention programs. Four components of language intervention were associated with successful auditory processing and language outcomes for children across the four conditions in our study: intensity, active attention, feedback, and rewards.

Intensity

All the RCT conditions were delivered on the same intense schedule: 100 minutes per day, five days per week for six weeks. To gain a general sense of the importance of the intensity component, we compared the pre-test and post-test CASL standard scores from the children in our study to pre-test and post-test standard scores on the Test of Language Development Primary (3rd edition) from a large group (n=156) of same-age children (Tomblin, personal communication) who had participated in a longitudinal epidemiological study in Iowa (Tomblin et al., 1997; Tomblin et al., 2003). On average, children in the Iowa study received speech and/or language intervention in public school settings twice each week for 20-minute sessions for two years (an approximate total of 48 hours). The children in our study received 1 hour and 40 minutes of intervention five days per week for six weeks (a total of 50 hours) followed by some intervention in school settings (12 hours on average).

Figure 1 [PDF] depicts the outcome comparisons. The improvement made by the children in our RCT over six months was approximately five times greater than the improvement made by children from the Iowa epidemiological study in two years. These comparisons suggest that intense intervention schedules may be more beneficial than less intense intervention schedules.

A recent survey by Brandel (2009) found that most school-age children with language impairments receive intervention twice a week for 20-minute sessions, regardless of severity or diagnosis. Only a small number of SLPs use block scheduling, which allows for more intensive intervention. Given the importance of intensive treatment, school clinicians may want to explore options for block scheduling in their districts. Alternative solutions such as providing language intervention before or after school and in summer school programs offer clinicians other means for providing intensive remediation.

Active Attention

Children learn more when their attention is directed at the learning experiences that are being presented (Erickson, 2008). LaBerge's (1995) model of attention involves preparatory, selective, and maintenance processes. FFW-L promoted all three kinds of attention. Children received an attention signal before every item to make sure that they were looking at the screen and ready to listen (preparatory attention). They were then rewarded for making accurate perceptual judgments (selective attention) and for completing a number of items in a specified period of time (maintenance of attention).

The other three conditions promoted primarily selective attention and maintenance of attention. The ILI clinicians and aides who were monitoring the CALI and AE conditions encouraged children to return to tasks if they stopped responding or began talking about topics that were not related to the immediate activities. Results suggest that procedures that focus on children's attention by delivering a pre-attentive signal and/or that promote selective attention and maintenance of attention more generally have the potential to yield similar beneficial effects.

Feedback

In all of the trial conditions, children received some form of feedback as to whether their responses were correct. In FFW-L, CALI, and AE, the feedback for correct answers was musical or visual rewards. The feedback for incorrect responses was either a mildly noxious sound (such as the "clunk" in the FFW-L circus sequence game) or an indication that the response was wrong followed by an opportunity to repeat the trial. The feedback from the ILI clinicians was less systematic. Sometimes clinicians told children if their answers were right or wrong and explained why (e.g., "That's right. You remembered to name the boy in your story."). Often clinicians provided indirect feedback that took the form of expansions or recasts (e.g., Child: "Him's going over there." Clinician: "Yes, he is going over there."). Although the type of feedback differed, children in each condition received information about the correctness of their responses.

Rewards

Most models of learning recognize the importance of internal motivation and the benefits of rewards for appropriate behavior. In the computer conditions, the children received animated reinforcement designed to keep them interested and motivated. In the ILI condition, the SLPs smiled at the children when they responded correctly and told them they were doing well. All of the children received tangible rewards as well. We kept charts for each child with daily stickers and prizes at the end of the week. In all four conditions, there were rewards for attending and responding correctly to questions or directions and for maintaining acceptable behavior during the course of an entire session.

Research to Practice

Merzenich and colleagues (2002) have shown that teaching that employs a set of neuroplasticity-based training strategies drives brain remodeling in animals. It has been suggested that these behaviorally based, discrete-trial training procedures are necessary to bring about brain-based changes that lead to language development in children with language impairments.

The results of our randomized controlled trial suggest otherwise. In our study, four different language-teaching conditions yielded similar gains in auditory processing and language skills. Four aspects of neuroplasticity-based intervention (highly repetitive stimuli, tracking correct responses, adapting difficulty levels, and use of modified speech stimuli) were part of one or two conditions—but not all four—and appear to be unnecessary elements of successful language intervention.

Four other aspects of intervention (a daily intense schedule, promoting active attention, providing immediate feedback, and rewarding effort and success) were part of each study intervention condition. Although more research needs to be done to identity the critical attributes of successful language intervention, it appears that there are sound theoretical and practical reasons for clinicians to consider incorporating aspects of intensity, attention, feedback, and rewards in their intervention programs.

We question the assumption that it is best to base intervention procedures for children on teaching principles that have been shown to result in neurological change in rats and monkeys. As Gillam (1999) noted earlier, there are real differences in the kinds of mental functions that children and monkeys perform. Intervention procedures should take advantage of children's capacity for symbolic representation, grammar, and reasoning through language. It makes sense to engage children in language-learning experiences that are matched to their interests, typical activities, linguistic and conceptual functioning, and information-processing abilities.

Our results suggest that neural reorganization that promotes language development can result from a variety of interventions in which highly motivated children have multiple opportunities to respond to challenging and accomplishable tasks within intensive programs.