Laureate Software Associated with Clinically Significant Gains in MLU and Use of Grammar.

Gillam, R.B., Crofford, J.A., Gale, M.A., & Hoffman, L.M. (2001). Language change following computer-assisted language instruction with Fast ForWord or Laureate Learning Systems software. American Journal of Speech-Language Pathology, 10, 231-247.

Background
Gillam and colleagues observed that computer-assisted instruction has become increasingly popular for many good reasons. This approach provides, for example, a useful means for teachers to deliver more individualized instruction to students, and do so in a manner that can be highly motivating and allows students to work at their own pace with as much practice and repetition as needed. As well, there is a growing body of research showing that computer-assisted instruction is a highly effective means of instruction for children with special needs, including those with language-learning impairments. The authors also noted, however, that research in this area typically entails measuring language outcomes in contrived settings using formal test instruments. Of greater interest to clinicians and parents is the extent to which an intervention improves functional language in authentic contexts. The present study was therefore designed to include spontaneous language samples collected in natural settings as an additional measure of potential language gains.
A second goal of this study was to compare two sets of programs: seven program modules from Fast ForWord TM (FFW; Scientific Learning Corp, 1997), which uses acoustically modified speech; and seven programs published by Laureate Learning Systems (Concentrate; Following Directions; Microcomputer Language Assessment and Development System (Micro-LADS); Sentence Master; Swim, Swam, Swum or Adjectives and Opposites; and Twenty Categories).

Methods
Subjects were four children with diagnosed language impairment, ages 6;11 to 7;6, two of whom were identical twins. The intervention schedule involved 1 hour and 40 minutes of computer use per day for 20 days over 4 weeks. Two children used the FFW software, and two used the Laureate software. One child in each group began the intervention immediately and the other two, the twins, received delayed treatment to accommodate their family schedule. Twenty-minute spontaneous language samples were collected weekly, before, during, and after the intervention period. Baseline and post-intervention assessments using the Oral and Written Language Scales (OWLS; Carrow-Woolfolk, 1995) were also completed.

Results
After treatment, clinically significant gains in language ability were seen in all four subjects, as defined by post-intervention scores on the OWLS that exceeded the 95% confidence interval calculated using the pre-intervention baseline scores. Analyses of the spontaneous language samples revealed that both children who used Laureate software, and one who used FFW software, exhibited clinically significant gains in the mean length of their utterances (MLU), a general measure of language complexity, and used more grammatical morphemes after treatment. One of the two children who used Laureate software also made reliably fewer grammatical errors post-intervention.

Discussion
Formal testing using a standardized assessment showed that language intervention using either FFW or Laureate software over a four week period produced clinically significant gains in language ability. Moreover, language samples revealed that intervention with Laureate software produced clinically significant gains in MLU and morpheme use, and in one of the two children, a reliable reduction in grammatical errors. Gains in MLU and morpheme use were also seen in one subject who used FFW software. These language gains measured using spontaneous language samples are especially encouraging since they actually represent observable and pragmatically relevant improvements in spontaneous language use in authentic contexts.