Methods for controlling amount of talk: Difficulties, considerations and recommendations

The aim of this article is to focus researchers’ attention on some central methodological issues involving control of amount of talk (AOT). First, appropriate methods for standardizing the size of language samples are considered. Second, the common practice of deriving linguistic rates by dividing the frequency of some linguistic variable of interest by a frequency tapping AOT is challenged. This operation does not adequately control for AOT and may result in spurious findings. Third, appropriate contexts for achieving control of AOT through statistical partialling are discussed. Some recommendations for research are offered.