Subject: d' calculation From: Mitchell Sommers <msommers(at)ARTSCI.WUSTL.EDU> Date: Thu, 19 Apr 2001 13:47:19 -0500This is a multi-part message in MIME format. ------=_NextPart_000_01CE_01C0C8D7.41E626E0 Content-Type: text/plain; charset="Windows-1252" Content-Transfer-Encoding: quoted-printable I'm trying to calculate d' from an experiment that used a modified = method of constant stimuli. Essentially, it is a same-different = paradigm in which S1 is presented in the first interval of a trial. On = half the trials the second interval is also S1 (same trials). The = remainder of the trials are equally distributed S1-S2, S1-S3, etc (in = the current instantiation, we have five different values S2-S6), with S2 = representing the smallest change and S6 representing the largest change = (the change is a increment in the level of 2 harmonics in 5-harmonic = complexes). This seems like a clear example of what McMillan and = Creelman refer to as a reminder task but it wasn't clear (at least to = me) what the most appropriate way of calculating d' is from this type of = paradigm. Fitting a logistic function (Green, 1993) seems like a = reasonable solution but I wonder if there were other possibilities. = Thanks for any suggestions Mitch Sommers Mitchell S. Sommers, Ph.D. Associate Professor Dept. of Psychology Washington University Campus Box 1125 St. Louis, MO 63130 Phone: 314-935-6561 Fax: 314-935-7588 e-mail: msommers(at)artsci.wustl.edu ------=_NextPart_000_01CE_01C0C8D7.41E626E0 Content-Type: text/html; charset="Windows-1252" Content-Transfer-Encoding: quoted-printable <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN"> <HTML><HEAD> <META http-equiv=3DContent-Type content=3D"text/html; = charset=3Dwindows-1252"> <META content=3D"MSHTML 5.50.4611.1300" name=3DGENERATOR> <STYLE></STYLE> </HEAD> <BODY bgColor=3D#ffff00> <DIV>I'm trying to calculate d' from an experiment that used a modified = method=20 of constant stimuli. Essentially, it is a same-different paradigm = in which=20 S1 is presented in the first interval of a trial. On half the = trials the=20 second interval is also S1 (same trials). The remainder of the = trials are=20 equally distributed S1-S2, S1-S3, etc (in the current instantiation, we = have=20 five different values S2-S6), with S2 representing the smallest change = and S6=20 representing the largest change (the change is a increment in the level = of 2=20 harmonics in 5-harmonic complexes). This seems like a clear = example of=20 what McMillan and Creelman refer to as a reminder task but it wasn't = clear (at=20 least to me) what the most appropriate way of calculating d' is from = this type=20 of paradigm. Fitting a logistic function (Green, 1993) seems like = a=20 reasonable solution but I wonder if there were other possibilities. = Thanks for=20 any suggestions</DIV> <DIV> </DIV> <DIV>Mitch Sommers</DIV> <DIV> </DIV> <DIV> </DIV> <DIV> </DIV> <DIV>Mitchell S. Sommers, Ph.D.<BR>Associate Professor<BR>Dept. of=20 Psychology<BR>Washington University<BR>Campus Box 1125<BR>St. Louis, MO=20 63130</DIV> <DIV> </DIV> <DIV>Phone: 314-935-6561<BR>Fax: 314-935-7588<BR>e-mail: <A=20 href=3D"mailto:msommers(at)artsci.wustl.edu">msommers(at)artsci.wustl.edu</A></= DIV></BODY></HTML> ------=_NextPart_000_01CE_01C0C8D7.41E626E0--