If taking a %SS measure, a sample of about 300 syllables is suggested. However, there are occasions where the clinician might decide that a longer sample is warranted if they do not think that they have a valid measure. On the other hand, if it takes too long to gather a 300 syllable sample e.g. for a quiet child, the clinician might choose to rely more on severity ratings or recordings from home.
If a child does this a few times in one conversation, and it is not the habitual way that s/he talks in the clinic, then it’s not a problem and there is no need to do anything about it. On the other hand, if this happens often, subtly intervene. The aim is to gently move the child on to talking about something different, so that a more representative sample of speech is heard. For example, try to move the “Counting Kid” on to talking about functions, and the “Nursery Rhyme Kid” on to playing a matching game. Then you are more likely to obtain a baseline speech measure that is valid.