Dprime and false alarm rates (Kala Lakshminarayanan )

Subject: Dprime and false alarm rates
From:    Kala Lakshminarayanan  <kala(at)PEGASUS.RUTGERS.EDU>
Date:    Fri, 14 Nov 2003 11:46:29 -0500

Hello, In analysis of data from speech perception experiments using the AX same-different roving model, the dprime is calculated using the differencing model (McMillan and Creelman, 1991). Has anyone had the problem of having false alarm rates that exceed the hit rate sometimes and not being able to calculate dprime? How does one get around this? Is it possible to do it, without compromising on the number of steps (currently ten ) in the stimulus continuum? I am sure one suggestion might be that the stimuli need to be redesigned. But the high false rates are seen in discrimination despite the fact that in identification using the same stimuli people have no trouble and are at ceiling for the endpoint stimuli. Also, I have seen a number of articles in literature that calculate the percent/proportion correct for AX discrimination studies. I need some help as to how this can be done. Is the percent correct calculated only for the trials that use the different stimuli and the catch trials (with 2 same stimuli) discarded? Thanks for any advice on this, Kala

This message came from the mail archive
maintained by:
DAn Ellis <dpwe@ee.columbia.edu>
Electrical Engineering Dept., Columbia University