“…This assumption has not been found to be robust (e.g., Lecluyse & Meddis, 2009), though Gu and Green (1994) argued this finding away, claiming that because the threshold estimates produced by the procedure are consistent with others reported in the literature, it is a moot point whether these assumptions are upheld. However, Lecluyse and Meddis reported that psychometric functions do differ across clinical groups, and a universal form should not be assumed; this stance was supported by Leek (2001). Amitay, Irwin, Hawkey, Cowan and Moore (2006) reported that with novice participants, the form of the underlying psychometric function is prone to change due to learning effects.…”