Re: identification task...negative d' values (silvadmr )


Subject: Re: identification task...negative d' values
From:    silvadmr  <silvadmr@xxxxxxxx>
Date:    Wed, 3 May 2006 21:54:50 -0300

--_=__=_XaM3_.1146704090.2A.352952.42.11786.52.42.007.794033487 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: quoted-printable Dear all, thank you very much for the help. Concerning my last e-mail about negative d' values, questions were made a= bout the fact that my identification task has 3 labels for the subjects t= o select when a vowel sound drawn from a continuum is presented (vowels w= ith F1 varying from 250 to 650 Hz; the labels are "u", "o" and "O"("O" is= similar to the "o" in "off"). I think d' is appropriate since there is almost no overlap between the ca= tegories in the extremes of the continuum - "u" and "O". I have imagined = two solutions: 1. take only the d's obtained from the probabilities of responding with t= he label corresponding to the category in the middle of the continuum - "= o". 2. simply split the results in two at an arbitrary point in the middle of= the '"o" category' (where we have a region of "100%" (AND THAT IS ANOTHE= R QUESTION TO WITCH I HAVE SOME ALTERNATIVES BUT NO CLEAR CRITERION TO CH= OOSE ONE! - up to now the 0,01-0,99 range is O.K. for my purposes). That = sounds a little arbitrary but I dont think that it is a great evil...but = there is still a question to be made: is it better either to discard the = extremely rare cases of overlap between "u" and "O" or calculate d' keepi= ng the data as they are? At the end, the differences between the alternatives are very small. To be honest, really arbitrary is my choice cause I was not able to decid= e whats the greater of two evils. What do you think? Daniel. --_=__=_XaM3_.1146704090.2A.352952.42.11786.52.42.007.794033487 Content-Type: text/html; charset=iso-8859-1 Content-Transfer-Encoding: quoted-printable <DIV > <DIV>Dear all, </DIV> <DIV></DIV> <DIV>thank you very much for the help. </DIV> <DIV></DIV> <DIV>Concerning my last e-mail about negative d' values, questions were m= ade about the fact that my identification task has 3 labels for the subje= cts to select when a vowel sound drawn from a continuum is presented (vow= els with F1 varying from 250 to 650 Hz; the labels are "u", "o" and "O"("= O" is similar to the "o" in "off"). </DIV> <DIV></DIV> <DIV>I think d' is appropriate since there is almost no overlap between t= he categories in the extremes of the continuum - "u" and "O". I have imag= ined two solutions: </DIV> <DIV>1. take only the d's obtained from the probabilities of responding w= ith the label corresponding to the category in the middle of the continuu= m - "o". </DIV> <DIV>2. simply split the results in two at an arbitrary point in the midd= le of the '"o" category' (where we have a region of "100%" (AND THAT IS A= NOTHER QUESTION TO WITCH I HAVE SOME ALTERNATIVES BUT NO CLEAR CRITERION = TO CHOOSE ONE! - up to now the 0,01-0,99 range is O.K. for my purposes). = That sounds a little arbitrary but I dont think that it is a great evil..= .but there is still a question to be made: is it better either to discard= the extremely rare cases of overlap between "u" and "O" or calculate d' = keeping the data as they are? </DIV> <DIV>&nbsp;</DIV> <DIV></DIV> <DIV>At the end, the differences between&nbsp;the alternatives&nbsp;are v= ery small. </DIV> <DIV>To be honest, really arbitrary is my choice cause I was not able to = decide whats the greater of two evils. </DIV> <DIV>What do you think? </DIV> <DIV>&nbsp;</DIV> <DIV></DIV> <DIV>Daniel. </DIV> <DIV></DIV> <DIV></DIV></DIV> --_=__=_XaM3_.1146704090.2A.352952.42.11786.52.42.007.794033487--


This message came from the mail archive
http://www.auditory.org/postings/2006/
maintained by:
DAn Ellis <dpwe@ee.columbia.edu>
Electrical Engineering Dept., Columbia University