Best Practices and Reality: Can We Do Language Sample Analysis (LSA) Of Children's Natural Conversational Speech Faster and Better Than We Currently Do? The Answer Is YES

I am very excited to host this issue, one of the last I have the privilege of “shepherding” as Editor in Chief of Seminars. As both disclosure and enthusiastic endorsement, I have been involved in the Child Language Data Exchange System (CHILDES) Project, one of the “parent” projects of TalkBank, for several years. Even as a young(er) researcher, I was amazed by the free computational (and collegial) benefits of CHILDES. As a clinician, I learned to appreciate these programs even more, and I have often been both surprised and dismayed by the fact that few practicing clinicians seem to know of these wonderful, absolutely free utilities and resources, which have expanded to enable analysis of many of the world's most frequently spoken languages in addition to English. In the past few years, the ability to link our patients' audio or video files to our language sample analyses (LSAs) have additionally speeded the process of transcription and made it more accurate.

Recently, TalkBank has developed specific “omnibus” programs to enable clinicians working with adult patients (our previous issue of Seminars) and child clients with delayed language and speech, which are the focus of the current issue. In addition to language analysis, we also include contributions about programs such as Phon to more quickly and insightfully analyze child phonology, about using computerized language analysis (CLAN) utilities to analyze fluency differences/disorder, and about a new horizon for language sample analysis—HomeBank, which will collect and analyze records from naturalistic daylong interactions among children and their family members.

We also have two contributions on new ways of approaching language sample analysis to better identify children who merit intervention. My article (cowritten with Brian MacWhinney) introduces discussion of which traditional measures of child language appear to do a better job of tracking maturation of language skill over early childhood in typical children. Eisenberg and Guo continue this conversation and offer some updated metrics for appraising child language that differ from traditional counts based on utterance length and complexity to include measures of grammatical accuracy and finite verb morphology. They explore the degree to which two new, fairly easily computed measures can more accurately identify children with language impairment, a continuous challenge for the practicing clinician, for whom few language sample analysis measures clearly distinguish typical from delayed or impaired language use for age.

We have worked for so long with Seminars to help the practicing clinician (and the clinician in training) to make clinical practice easier and smarter. It is my fervent belief that the free programs and utilities outlined in this issue have the potential to both ease and improve your everyday work. The majority of the contributing authors are also committed to helping you master these exciting assets. Please feel free to contact any of us for more information.