Subject: Re: [AUDITORY] stats use in psychology and hearing science From: Iftikhar Nizami <nizamii2@xxxxxxxx> Date: Tue, 25 Jun 2013 09:40:15 -0700 List-Archive:<http://lists.mcgill.ca/scripts/wa.exe?LIST=AUDITORY>--1029272498-426888069-1372178415=:39194 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: quoted-printable Dr Oberfeld's post hits the nail squarely on the head; I refer readers back= to the original Keppel and Wickens text known to many psychology graduates= . For many people, a refresher from the textbook may be in order. - Lance N= izami PhD=0A=0A=0A=0A>________________________________=0A>From: Daniel Ober= feld <oberfeld@xxxxxxxx>=0A>To: AUDITORY@xxxxxxxx =0A>Sent: Tues= day, June 25, 2013 12:25 AM=0A>Subject: Re: [AUDITORY] stats use in psychol= ogy and hearing science=0A>=0A>=0A>Dear list,=0A>=0A>important topic! A rel= ated while probably less general problem is that data from repeated-measure= s / within-subjects designs (that is, when a subject is tested in more than= one experimental condition) are sometimes not analyzed properly. I've seen= quite a few papers in hearing science where the authors used an ANOVA for = completely randomized / between-subjects design for repeated-measures data = (you can spot this problem by examining the degrees-of freedom for the F-te= sts), or fail to use the appropriate repeated-measures ANOVA approach (e.g.= , apply no correction for the degrees-of-freedom when using a univariate ap= proach).=0A>=0A>Another point that should be considered is that while for b= etween-subjects designs the general linear model is very robust against non= -normality, this is unfortunately *not* the case for repeated-measures desi= gns. Therefore, when analyzing non-normal data like error rates, percent co= rrect or response times from a within-subjects design, the p-values you get= from the rmANOVA could be either on the conservative or on the liberal sid= e.=0A>=0A>You can find an in-depth discussion of these issues in a recent p= aper from my lab (http://link.springer.com/article/10.3758%2Fs13428-012-028= 1-2).=0A>=0A>Best=0A>=0A>Daniel=0A>=0A>-- PD Dr. Daniel Oberfeld-Twistel=0A= >Johannes Gutenberg - Universitaet Mainz=0A>Department of Psychology=0A>Exp= erimental Psychology=0A>Wallstrasse 3=0A>55122 Mainz=0A>Germany=0A>=0A>Phon= e ++49 (0) 6131 39 39274=0A>Fax=A0 ++49 (0) 6131 39 39268=0A>http://www.sta= ff.uni-mainz.de/oberfeld/=0A>=0A>http://www.facebook.com/daniel.oberfeldtwi= stel=0A>=0A>=0A> --1029272498-426888069-1372178415=:39194 Content-Type: text/html; charset=iso-8859-1 Content-Transfer-Encoding: quoted-printable <html><body><div style=3D"color:#000; background-color:#fff; font-family:ti= mes new roman, new york, times, serif;font-size:12pt"><div style=3D"RIGHT: = auto"><SPAN style=3D"RIGHT: auto">Dr Oberfeld's post hits the nail squarely= on the head; I refer readers back to the original Keppel and Wickens text = known to many psychology graduates. For many people, a refresher from the t= extbook <VAR id=3Dyui-ie-cursor></VAR>may be in order. - Lance Nizami PhD<B= R style=3D"RIGHT: auto" class=3Dyui-cursor></SPAN></div> <div style=3D"RIGHT: auto"><BR> <BLOCKQUOTE style=3D"BORDER-LEFT: rgb(16,16,255) 2px solid; MARGIN-TOP: 5px= ; PADDING-LEFT: 5px; MARGIN-LEFT: 5px; RIGHT: auto"> <DIV style=3D"FONT-FAMILY: times new roman, new york, times, serif; FONT-SI= ZE: 12pt"> <DIV style=3D"FONT-FAMILY: times new roman, new york, times, serif; FONT-SI= ZE: 12pt"> <DIV style=3D"RIGHT: auto" dir=3Dltr> <DIV style=3D"BORDER-BOTTOM: #ccc 1px solid; BORDER-LEFT: #ccc 1px solid; P= ADDING-BOTTOM: 0px; LINE-HEIGHT: 0; MARGIN: 5px 0px; PADDING-LEFT: 0px; PAD= DING-RIGHT: 0px; HEIGHT: 0px; FONT-SIZE: 0px; BORDER-TOP: #ccc 1px solid; B= ORDER-RIGHT: #ccc 1px solid; PADDING-TOP: 0px" class=3Dhr contentEditable= =3Dfalse readonly=3D"true"></DIV><FONT size=3D2 face=3DArial><B><SPAN style= =3D"FONT-WEIGHT: bold">From:</SPAN></B> Daniel Oberfeld <oberfeld@xxxxxxxx= INZ.DE><BR><B><SPAN style=3D"FONT-WEIGHT: bold">To:</SPAN></B> AUDITORY@xxxxxxxx= LISTS.MCGILL.CA <BR><B><SPAN style=3D"FONT-WEIGHT: bold">Sent:</SPAN></B> T= uesday, June 25, 2013 12:25 AM<BR><B><SPAN style=3D"FONT-WEIGHT: bold">Subj= ect:</SPAN></B> Re: [AUDITORY] stats use in psychology and hearing science<= BR></FONT></DIV> <DIV class=3Dy_msg_container><BR>Dear list,<BR><BR>important topic! A relat= ed while probably less general problem is that data from repeated-measures = / within-subjects designs (that is, when a subject is tested in more than o= ne experimental condition) are sometimes not analyzed properly. I've seen q= uite a few papers in hearing science where the authors used an ANOVA for co= mpletely randomized / between-subjects design for repeated-measures data (y= ou can spot this problem by examining the degrees-of freedom for the F-test= s), or fail to use the appropriate repeated-measures ANOVA approach (e.g., = apply no correction for the degrees-of-freedom when using a univariate appr= oach).<BR><BR>Another point that should be considered is that while for bet= ween-subjects designs the general linear model is very robust against non-n= ormality, this is unfortunately *not* the case for repeated-measures design= s. Therefore, when analyzing non-normal data like error rates, percent correct or response times from a within-subjects design, the p-values you = get from the rmANOVA could be either on the conservative or on the liberal = side.<BR><BR>You can find an in-depth discussion of these issues in a recen= t paper from my lab (http://link.springer.com/article/10.3758%2Fs13428-012-= 0281-2).<BR><BR>Best<BR><BR>Daniel<BR><BR>-- PD Dr. Daniel Oberfeld-Twistel= <BR>Johannes Gutenberg - Universitaet Mainz<BR>Department of Psychology<BR>= Experimental Psychology<BR>Wallstrasse 3<BR>55122 Mainz<BR>Germany<BR><BR>P= hone ++49 (0) 6131 39 39274<BR>Fax ++49 (0) 6131 39 39268<BR>http://w= ww.staff.uni-mainz.de/oberfeld/<BR><BR>http://www.facebook.com/daniel.oberf= eldtwistel<BR><BR><BR></DIV></DIV></DIV></BLOCKQUOTE></div></div></body></h= tml> --1029272498-426888069-1372178415=:39194--