People often write to me asking how they should score the Stuttering Severity Instrument (SSI; Riley, 2009). They wonder if they should count the interjections or revisions as stuttering, or if they should only count the blocks and other stuttering-like behaviors.
In the SSI, your goal is to count the number of stutters that the person exhibits. The SSI does not count non-stuttered disfluencies, moments when the person is thinking about what he wants to say, or any other type of disfluency than stuttering. That means that, in general, you do not count disfluencies on the SSI.
Of course, this is not as simple as it sounds. What if the person is using the interjections as a way of hiding actual moments of stuttering that are felt under-the-surface? In that case, it might be appropriate to count an interjection as part of a moment of stuttering.
Interestingly, there is no way to determine with certainty what disfluencies are stuttered and what are not. That’s because the real determination of what is stuttered and what is not can only really be sensed by the speaker. When a person feels that they are stuck, that is, that they’ve lost control of their speech mechanism, then that is stuttering. They may feel this when they exhibit so-called stuttered (or stutter-like) behaviors like repetitions, prolongations, and blocks. They may also feel it when they exhibit so-called non-stuttered behaviors, like interjections and revisions. They may even feel this at times when they don’t exhibit any observable disfluencies in their speech at all!!
Many times, however, the disfluencies that people exhibit when they feel that underlying sensation of being stuck take the form of a repetition or a prolongation or a block—that’s why people tend to count these behaviors when counting stuttering. Ultimately, however, the only way we can know if a person is stuttering is to ask them.
I think of it this way: Most of the time, when people are stuttering, they exhibit certain types of disfluencies, and these are the ones that people count as stuttered or stutter-like; other times, they may still feel like they are stuttering but not exhibit those stuttered types of disfluencies. The best we can do, if we're trying to count stutters, is to count the behaviors that most commonly co-occur with the speaker's perception that he is stuttering. Again, this brings us to the part-word repetitions, prolongations, and blocks, rather than the interjections, revisions, and hesitations.
Going back to the SSI... Our task when using that test is to count stuttering—not typical disfluencies or any other types of disfluencies; just stuttering. Glyn Riley, the author of the SSI, used to say, "if it feels like stuttering, call it stuttering." He was not focused on counting specific types of disfluencies. Instead, he wanted to know specifically if a speech disruption felt like a moment of stuttering. So, most of the time, you will not count those interjections as part of stuttering on the SSI—unless it's clear that the person is really stuck when producing them.
To summarize: There's no definitive answer, but when you’re scoring the SSI, count a disfluency as a moment of stuttering if it seems that the person is stuck. If it seems that the person is just thinking, or if it’s some other type of disfluency, then consider a non-stuttered disfluency and don’t count it in the SSI.
Now, even more important question is this: do you even need to use the SSI? That’ll be a topic for another post. (My answer, not surprisingly, is "no.")