Baby signs boost IQ by 12 points

[Adele Abrahamsen]

The claim of later IQ effects unfortunately distracts from the most
important impact of Acredolo and Goodwyn's work.

It was long assumed that manual symbols would compete, and hence
interfere, with spoken words.  In my own research on what is now called
baby sign, initially I could offer potential participants nothing more
than indirect evidence and reasoning to the contrary. So I was delighted
when Goodwyn and Acredolo (1993) provided compelling direct evidence that
baby signs do not interfere with acquisition of spoken words. Moreover,
regardless of whether baby signs facilitate spoken words, clearly a baby
who acquires them has more symbols available for use. (E.g., a reanalysis
of G&A's data in Abrahamsen 2000 showed a median of 5.8 gestures when the
median number of words was 5.9; for those toddlers with no overlap, that
doubled the number of meanings that could be expressed through symbols at
that time.) There also is a hard-to-predict subgroup of toddlers for whom
baby signs carry much more than half the communicative burden for several
months, which would seem to be advantageous.

It is much more challenging to determine the extent to which, if at all,
baby signing results in facilitation of spoken language or general
development. To their credit Acredolo and Goodwyn included longitudinal
comparison groups in their design, but as Brian and Kathy noted, the
results raised more questions than they answered. Ideally someone would
step up and do a replication study with a larger number of randomly
assigned or closely-matched participants.

Meanwhile, it would be sensible to base the decision to enhance gestural
input to a baby on practical considerations (e.g., is there a caretaker
who would enjoy doing it?) and on those outcomes for which there is ample
evidence (e.g., a modest, possibly temporary increase in the number of
symbols when both gestures and words are counted). There is no need to
reach beyond this to apparent effects that are large or of long duration
and for which there is no obvious mechanism, such as the claim of a
12-point IQ boost at 8 years.  (Evidence and plausibility of verbal
facilitation lies between these extremes.)

Isabelle Barriere made the interesting point that, even as parents of
hearing toddlers buy videos and take baby sign classes, the old worries
about interference live on recommendations to parents of children with
cochlear implants. I would add that the issues and relevant studies extend
far beyond the baby sign literature, especially with respect to
syntactically structured language.

> Dear Info-CHILDES,
>
>       During the flurry of discussion of the C/P contrast, there was
> an message from Mechthild Kiegelmann that seemed to slip under the radar
screen.  This message summarized replies to a query about Baby Signs.  I
spent some time tracing the various web links involved and I would like to
draw colleagues' attention to one issue in this
> research that troubles me.  This is the status of a report by
> Acredolo and Goodwyn, which is cited prominently at www.babysigns.com
and www.signingtime.com (STResearch_Summary.pdf).  This reports
> speaks of a 12 point "increase" in IQ measured at age 8 for children who
are taught Baby Signs when they are toddlers.  Interestingly,
Mechthild's links also point to an article from the Canadian Language and
Literacy Research Network by J Cyne Johnston, Andrée Durieux- Smith, and
Kathleen Bloom that challenges the claims of this study by noting that it
provides no description of subject recruitment
> provedures, attempts at random assignment, or evidence of any
> pretesting.  They conclude that, "The high accessibility of a wide range
of baby signing products is not matched by good quality
> evidence that would reinforce manufacturers' claims."
>        It is worth adding that the groups were already different when
> the Bailey was given at 24 months, but this is presented not as
> evidence of initial group differences, but rather as the result of the
initial effects of the treatment.  The relevant study was
> presented as a conference paper at ISIS in 2000, but has never been
published in a journal.
>     I have mixed feelings about the plausibility of this result.  I
> certainly do not view IQ as immutable and genetically-given.  I am also
quite convinced that Baby Signs provide an excellent method for
achieving early and rewarding communications with toddlers.  However, I
find it difficult to believe that a program in Baby Signs alone could
achieve a 12-point increase in IQ when several years of Head Start lead to
nothing measurably permanent.
>     I hope that academic researchers take these unpublished claims
> with a healthy grain of salt.  If there are newer studies supporting
these claimed gains in IQ, I would love to learn about them.
>
> --Brian MacWhinney, CMU
>