“Considerable work exists to describe the functions of yeah.” explain authors Chad D. Nilep and colleague Tamara Grivičić from the Department of Linguistics at the University of Colorado. And there has also been some academic research investigating the so-called ‘Creaky Voice’. But the team’s study is one of the very first to investigate the two together (i.e. Yeah + Creaky Voice).

As a dataset, the researchers used approximately twenty-five minutes of telephone conversations which they describe as “found data” – in that it was not recorded by the investigators themselves. Instead, the phone calls were recorded by men who were ‘teasing’ telemarketers (attempting to keep them on the line for as long as possible with no intention of buying the service advertised). The data contained a number of examples of oh, ok, really, mm, mhm, uh huh, yeah and creaky voice – sometimes concurrently. (A word was coded for creaky voice when creak was audible over at least one syllable of the word.)
Analysis of the recordings showed that creaky and non-creaky yeahs are not identical.

“We have demonstrated that creaky yeah is not identical in function to yeah with modal voicing or other types of glottal stricture. While Jefferson (1984) suggests that yeah generally signals high speaker incipiency, yeah in conjunction with creaky voice signals passive recipiency.”

The paper When Phonation Matters: The Use and Function of yeah and Creaky Voice can be read in Colorado Research in Linguistics. (CRIL) June 2004. Volume 17, Issue 1.

Also see 1 (yeah) Some Uses of Yeah, Drummond, Hopper, 1993

Also see 2 (creaky voice) Phonation types: a cross-linguistic overview, Gordon, Ladefoged, 2001