Sunday, July 28, 2013

Our Study Still Stands Tall!

One of our Baby Signs® Instructors called my attention last week to an article by a Speech Pathologist who was questioning our research findings that signing helps children learn to talk. She was basing her criticism on a study by researchers in England who claimed to be replicating our study but finding no positive effects when the verbal development of their signing group of babies was compared to that of their control group of babies.

Whenever such criticisms arise, the first thing I do is get my hands on the supposedly contradictory study and read it carefully. So that’s what I did, and just as I suspected, the authors didn’t have a leg to stand on! Let me outline a few of the more blatant differences between their study and ours, differences that explain why they didn’t find the language advantages we did. This post will be longer and a bit more “academic” than usual, but please bear with me. We want every parent who’s signing with a child or even contemplating doing so, to share the unwavering confidence we have in our own research.

Issue: Sample Size. The smaller the sample size (number of subjects within two groups), the harder it is to demonstrate a statistically significant difference between them, especially when the behavior in question varies dramatically from subject to subject as early language development does.
• Our study--Goodwyn, Acredolo, & Brown (2000): 32 in the Signing Group, 39 in the Control Group.
• English study: 10 in each Signing Group, 10 in the Control Group.
Conclusion: Goodwyn et al. was more likely to detect a facilitative effect, the English study more likely to miss a facilitative effect genuinely there.

Issue. Length of the study. Language development in its early stages is notoriously variable across children. The longer data collection continues, the more likely that trends will be visible.
• English study: Ceased data collection at 20 months.
• Goodwyn, Acredolo, & Brown (2000): Continued data collection well past 20 months (i.e., 24, 30, and 36 months).
Conclusion: Goodwyn et al. (2000) was more likely to discover a meaningful trend in the direction of a facilitative effect of signing.

Issue: Number of Signs Learned. To determine whether signing by infants facilitates learning to talk, the “signing group” has to have actually learned a meaningful number of signs. The larger the number of signs learned, the better equipped the study is to determine if signing makes a difference. If a baby learns only a handful of signs, one wouldn’t expect much of an effect on learning to talk.
• Goodwyn, Acredolo, & Brown (2000): Average number of signs learned = 20.3 (Range 9-61 signs)
• English study: Highest average number of signs learned across ages was 8.67 (Range 2-17 signs)
Conclusion: Goodwyn et al. was in a better position to test the effect of signing because the subjects in their study actually learned many more signs.

Why does the number of signs a baby uses matter to verbal development?

Signs pull language from adults. Babies learn words from listening to adults. When a baby uses a sign (e.g., BIRD) upon seeing a bird, the adult responds with lots of words (“Yes! That’s a birdie! We call that a robin. See, there’s another bird. Oh, the bird flew away.”) The more signs a baby knows, the more likely this is to happen.

Signs enable babies to pick the topic. Again, babies learn words from listening to adults. When a baby uses a sign, it starts a conversation about something the baby is interested in, thereby making it more likely the baby will listen to and learn from what the adult says. The more sign a baby knows, the more likely this is to happen.

Signing excites babies about communicating and motivates them to move on to an even better system--words. The more signs a baby is able to use successfully to communicate, the more motivated he/she is to get better at communicating—that is, to move on to words.

Signs increase a baby’s interest in books. Babies learn new vocabulary from reading books with parents. Because signs enable babies to be active participants in book-reading (naming pictures with signs), babies are drawn more strongly to books, thereby exposing them to more new vocabulary items. The more signs a baby knows, the more likely this is to happen.

Issue: Why did children in Goodwyn et al. (2000) learn more signs?

Instructions to Parents. The instructions to parents clearly can affect the number of signs learned.
o Goodwyn, Acredolo, & Brown (2000): Parents were encouraged to choose whatever signs they thought their child might be interested in learning and to add new signs at any time. This freedom addresses the fact that not all babies are interested in the same things.
o Kirk et al. (2013): Parents were told to stick to a specific set of first 10, and then 20 signs no matter what their child’s interests.

Amount of modeling. The more a sign is modeled by the parent, the more likely it is to be learned by the child.
o Kirk et al. (2013): Data indicate that parents on average modeled a sign “a few times a week to once a day.” This isn’t nearly enough!
o Goodwyn et al. (2000): Modelling rate was not reported. However, the likelihood that parents would, in fact, model signs was increased by the fact that they were given specific toys and books depicting objects represented by signs.

Given all the points I’ve outlined, I hope you can see why we disagree so strongly that this new study somehow negates our findings of a facilitative effect of signing on verbal development. And even if you hadn’t read this posting, all you would have had to do is talk to parents of preschoolers who used signs as babies to feel confident that signing is not just “okay,” but actually really good for language development.

Happy Signing! (and don’t forget to look for us on Facebook)


Linda Acredolo, Ph.D.
Co-Founder, the Baby Signs® Program
Professor Emeritus, UC Davis

No comments: