relatives does not guarantee a genetic link to the proband. It is entirely possible that the stuttering observed amongst a number of stuttering relatives relates to nongenetic influences. For this reason and others the family history design requires large sample sizes to ensure that the results noted really do relate directly to the research issue, and cannot be explained by extraneous variables (or “confounding” variables, as statisticians call them.) One simula¬tion study in the area of psychiatry has speculated that in order to run a study with an 80 percent chance of detecting heterogeneity, around 430 pro¬bands would be needed, each with reliable evidence of stuttering or nonstutter¬ing from three relatives. In short, unless a very large scale study is undertaken, the chances of finding genetic links using the family history method are remote.
There are also problems in comparing the many research papers on the subject. A substantial number are flawed, some have not used control groups, whilst others may have unwittingly gathered a biased population for their study. Equally difficult is the problem of reliability of the reporting, and relatives’ recollections of their stuttering may well be inaccurate. Two rela¬tives, when asked independently at interview or by questionnaire as to whether Cousin John once had a stutter, may well come up with a different interpretation. And while those same two relatives may agree that Aunt Jane used to stutter, they may hold differing views as to whether she actually fully recovered from it. Indeed, as we have seen earlier in this chapter, the very notion of what constitutes recovery is a problem that even experts in the field struggle with, and we still await a consensus as to a definition. Felsenfeld (1997) argues for the development of a standard assessment battery for the diagnosis of stuttering across the lifespan to help reduce the number of these intrusive variables.
A further difficulty is that knowing there may be a genetic component to stuttering does not get us any closer to knowing what that component actually is. Does the genetic predisposition code for a problem with the development of well-coordinated motor speech control, or perhaps it is responsible for some fragility in the development of linguistic skills in the preschool child? Perhaps,