From BRobinso at mprc.umaryland.edu Mon Sep 16 15:23:06 2002 From: BRobinso at mprc.umaryland.edu (Ben Robinson) Date: Mon, 16 Sep 2002 11:23:06 -0400 Subject: n-back test Message-ID: I've been working on an n-back test where letters are briefly presented to screen, one letter at a time. There can either be a match or no match, and subjects press one of two buttons in response. In the one-back test, a match occurs when the letter currently on screen is the same as the letter that was presented immediately prior to it. In the two- and three-back tests, a match occurs when the current letter is the same as the letter that appeared two or three letters back, respectively. I've been trying to guarantee that 1 in 3 stimuli presented is a match. To do this, I've created two Procedures: a MatchProc and a NoMatchProc, and I run the NoMatchProc twice as often as I run the MatchProc. Well, that's the idea, anyway. Some of the problems I'm having are: How to allow for random selection of the two procedures, MatchProc and NoMatchProc, with the only stipulation being that MatchProc can never be called in the first trial (or second or third trial in the two- and three- back conditions, respectively), since a match cannot occur until at least one stimulus has already been presented? I've set the List.Order to PermutationOrder(1), where NoMatchProc occupies Level 1, but since I don't fully understand what PermutationOrder is meant to accomplish I don't know if this method will work, and since my code currently won't compile (!) I can't yet test this. In the one-back condition I need to test to make sure one trial has already been run before saving the first letter to an attribute column I've titled, OneLetterBack. The code I'm trying to use for this task isn't working, and all I can guess is that it has a problem when I ask it to GetAttrib from one list and SetAttrib to another, higher-level list. It's in this little block of code that at least the compile error occurs ― SetAttrib: Missing parameter(s). If c.GetAttrib(c.GetAttrib("Running") & ".Sample") > 1 Then OneList.setAttrib "OneLetterBack", OneNoMTrialList.**GetPrevAttrib("Letters") **Here's where the compiler stops. If I change "OneNoMTrialList.GetPrevAttrib" to "c.GetPrevAttrib" I'm told that GetPrevAttrib is not a method of the object. Anyway, I know this may be a lot of information at once, but if anyone has any suggestions I'd surely appreciate it. Thanks for any help. Ben Robinson brobinso at mprc.umaryland.edu -------------- next part -------------- An HTML attachment was scrubbed... URL: From jmccabe at email.unc.edu Mon Sep 16 16:37:17 2002 From: jmccabe at email.unc.edu (jmccabe at email.unc.edu) Date: Mon, 16 Sep 2002 12:37:17 -0400 Subject: Letter 2-back task Message-ID: Hello, We are currently planning an experiment that would include a letter 2- back task, in which letters are presented one at a time on the screen and the participant presses one key for a 2-back match and another key when there is not a match. Our question: Does anyone have this type of letter 2-back test already programmed in E-Prime, and if so, would you be willing to share the program with us? Thanks so much, Jennifer McCabe jmccabe at email.unc.edu Graduate Student UNC-Chapel Hill From slacey at umich.edu Wed Sep 18 14:11:47 2002 From: slacey at umich.edu (Steve Lacey) Date: Wed, 18 Sep 2002 10:11:47 -0400 Subject: Displaying/Generating Graphs In E-Prime Message-ID: Hi, I am writing a script for a basic CRT experiment. After every block of trials I want to provide feedback in the form of graphs. There will be several graphs--one will plot RT by block number, another that will plot accuracy by block number, ect. Any ideas about how I might do this? Is there a method within e-prime to do this? Could I send the data to Excel, generate the plots there, and then import them back into separate images in a slide object? If either of these are possible, how would I do it? Thanks, Steve From devo0023 at tc.umn.edu Wed Sep 18 15:46:22 2002 From: devo0023 at tc.umn.edu (Cynthia J. DeVore) Date: Wed, 18 Sep 2002 10:46:22 -0500 Subject: Displaying/Generating Graphs In E-Prime Message-ID: Steve, So far as I know, there isn't a readily available method for generating graphs. I have, however, developed something that performs a similar function. The "trick" is to reorient the Canvas space because down is up (positive number increase in the y-direction moving from top to bottom, rather than from bottom to top as we typically see in coordinate space). The code below is the core of the logic used to draw a histogram in which the low value is 3 and the high value is 18. (In this task, 3 dice are rolled and the sum is graphed.) XConst is the width of each bar. xbase is the location of the leftmost portion of the graph. ybase is the bottom of the graph (with the labels beneath this "bottom"). DiceTotalArray is an array that holds the frequency for each dice roll (3 to 18). For i = 3 to 18 OffScreenCnvs.Text xbase+(i*XConst), ybase, i OffScreenCnvs.Text xbase+(i*XConst), ybase+20, DiceTotalArray(i) ' draw a rectangle or a line if DiceTotalArray(i) > 0 then OffScreenCnvs.Rectangle xbase+(i*XConst), ybase-(DiceTotalArray(i)*YConst), XConst, DiceTotalArray(i)*YConst end if Next i ' draw a line for "zero" OffScreenCnvs.Line xbase+(3*XConst), ybase, xbase+(19*XConst), ybase If you want the whole program to pull apart, I'll send it to you. Cynthia Cynthia DeVore I/O Grad Student and E-Prime Consultant University of Minnesota, Minneapolis devo0023 at tc.umn.edu 612-624-3601 Steve Lacey wrote: > Hi, > > I am writing a script for a basic CRT experiment. After every block of > trials I want to provide feedback in the form of graphs. There will be > several graphs--one will plot RT by block number, another that will plot > accuracy by block number, ect. > > Any ideas about how I might do this? Is there a method within e-prime to do > this? Could I send the data to Excel, generate the plots there, and then > import them back into separate images in a slide object? If either of these > are possible, how would I do it? > > Thanks, > Steve From Btomlow at chdr.nl Wed Sep 25 11:59:30 2002 From: Btomlow at chdr.nl (Benjamin Tomlow) Date: Wed, 25 Sep 2002 13:59:30 +0200 Subject: Devided attention test Message-ID: CAn anybody help me and tell me if there's the possibility to implement a divided attention test within E-prime? From slacey at umich.edu Wed Sep 25 19:27:31 2002 From: slacey at umich.edu (Steve Lacey) Date: Wed, 25 Sep 2002 15:27:31 -0400 Subject: Displaying/Generating Graphs In E-Prime In-Reply-To: <3D889FCE.7C8C9E3E@tc.umn.edu> Message-ID: Thanks Cynthia, It turns out that the easiest way to draw plots is to use the canvas function, as you have suggested. I also submitted the question to PSTNET and they said it was not possible to dynamically send data from eprime to excel, plot the graphs, and then import them back into images on a slide object during runtime. Steve At 10:46 AM 9/18/2002 -0500, Cynthia J. DeVore wrote: >Steve, > >So far as I know, there isn't a readily available method for generating >graphs. >I have, however, developed something that performs a similar function. The >"trick" is to reorient the Canvas space because down is up (positive number >increase in the y-direction moving from top to bottom, rather than from bottom >to top as we typically see in coordinate space). > >The code below is the core of the logic used to draw a histogram in which the >low value is 3 and the high value is 18. (In this task, 3 dice are rolled and >the sum is graphed.) XConst is the width of each bar. xbase is the location of >the leftmost portion of the graph. ybase is the bottom of the graph (with the >labels beneath this "bottom"). DiceTotalArray is an array that holds the >frequency for each dice roll (3 to 18). > >For i = 3 to 18 > OffScreenCnvs.Text xbase+(i*XConst), ybase, i > OffScreenCnvs.Text xbase+(i*XConst), ybase+20, DiceTotalArray(i) > ' draw a rectangle or a line > if DiceTotalArray(i) > 0 then > OffScreenCnvs.Rectangle xbase+(i*XConst), >ybase-(DiceTotalArray(i)*YConst), XConst, DiceTotalArray(i)*YConst > end if >Next i > >' draw a line for "zero" >OffScreenCnvs.Line xbase+(3*XConst), ybase, xbase+(19*XConst), ybase > >If you want the whole program to pull apart, I'll send it to you. > >Cynthia > >Cynthia DeVore >I/O Grad Student and E-Prime Consultant >University of Minnesota, Minneapolis >devo0023 at tc.umn.edu >612-624-3601 > >Steve Lacey wrote: > > > Hi, > > > > I am writing a script for a basic CRT experiment. After every block of > > trials I want to provide feedback in the form of graphs. There will be > > several graphs--one will plot RT by block number, another that will plot > > accuracy by block number, ect. > > > > Any ideas about how I might do this? Is there a method within e-prime to do > > this? Could I send the data to Excel, generate the plots there, and then > > import them back into separate images in a slide object? If either of these > > are possible, how would I do it? > > > > Thanks, > > Steve From BRobinso at mprc.umaryland.edu Mon Sep 16 15:23:06 2002 From: BRobinso at mprc.umaryland.edu (Ben Robinson) Date: Mon, 16 Sep 2002 11:23:06 -0400 Subject: n-back test Message-ID: I've been working on an n-back test where letters are briefly presented to screen, one letter at a time. There can either be a match or no match, and subjects press one of two buttons in response. In the one-back test, a match occurs when the letter currently on screen is the same as the letter that was presented immediately prior to it. In the two- and three-back tests, a match occurs when the current letter is the same as the letter that appeared two or three letters back, respectively. I've been trying to guarantee that 1 in 3 stimuli presented is a match. To do this, I've created two Procedures: a MatchProc and a NoMatchProc, and I run the NoMatchProc twice as often as I run the MatchProc. Well, that's the idea, anyway. Some of the problems I'm having are: How to allow for random selection of the two procedures, MatchProc and NoMatchProc, with the only stipulation being that MatchProc can never be called in the first trial (or second or third trial in the two- and three- back conditions, respectively), since a match cannot occur until at least one stimulus has already been presented? I've set the List.Order to PermutationOrder(1), where NoMatchProc occupies Level 1, but since I don't fully understand what PermutationOrder is meant to accomplish I don't know if this method will work, and since my code currently won't compile (!) I can't yet test this. In the one-back condition I need to test to make sure one trial has already been run before saving the first letter to an attribute column I've titled, OneLetterBack. The code I'm trying to use for this task isn't working, and all I can guess is that it has a problem when I ask it to GetAttrib from one list and SetAttrib to another, higher-level list. It's in this little block of code that at least the compile error occurs ? SetAttrib: Missing parameter(s). If c.GetAttrib(c.GetAttrib("Running") & ".Sample") > 1 Then OneList.setAttrib "OneLetterBack", OneNoMTrialList.**GetPrevAttrib("Letters") **Here's where the compiler stops. If I change "OneNoMTrialList.GetPrevAttrib" to "c.GetPrevAttrib" I'm told that GetPrevAttrib is not a method of the object. Anyway, I know this may be a lot of information at once, but if anyone has any suggestions I'd surely appreciate it. Thanks for any help. Ben Robinson brobinso at mprc.umaryland.edu -------------- next part -------------- An HTML attachment was scrubbed... URL: From jmccabe at email.unc.edu Mon Sep 16 16:37:17 2002 From: jmccabe at email.unc.edu (jmccabe at email.unc.edu) Date: Mon, 16 Sep 2002 12:37:17 -0400 Subject: Letter 2-back task Message-ID: Hello, We are currently planning an experiment that would include a letter 2- back task, in which letters are presented one at a time on the screen and the participant presses one key for a 2-back match and another key when there is not a match. Our question: Does anyone have this type of letter 2-back test already programmed in E-Prime, and if so, would you be willing to share the program with us? Thanks so much, Jennifer McCabe jmccabe at email.unc.edu Graduate Student UNC-Chapel Hill From slacey at umich.edu Wed Sep 18 14:11:47 2002 From: slacey at umich.edu (Steve Lacey) Date: Wed, 18 Sep 2002 10:11:47 -0400 Subject: Displaying/Generating Graphs In E-Prime Message-ID: Hi, I am writing a script for a basic CRT experiment. After every block of trials I want to provide feedback in the form of graphs. There will be several graphs--one will plot RT by block number, another that will plot accuracy by block number, ect. Any ideas about how I might do this? Is there a method within e-prime to do this? Could I send the data to Excel, generate the plots there, and then import them back into separate images in a slide object? If either of these are possible, how would I do it? Thanks, Steve From devo0023 at tc.umn.edu Wed Sep 18 15:46:22 2002 From: devo0023 at tc.umn.edu (Cynthia J. DeVore) Date: Wed, 18 Sep 2002 10:46:22 -0500 Subject: Displaying/Generating Graphs In E-Prime Message-ID: Steve, So far as I know, there isn't a readily available method for generating graphs. I have, however, developed something that performs a similar function. The "trick" is to reorient the Canvas space because down is up (positive number increase in the y-direction moving from top to bottom, rather than from bottom to top as we typically see in coordinate space). The code below is the core of the logic used to draw a histogram in which the low value is 3 and the high value is 18. (In this task, 3 dice are rolled and the sum is graphed.) XConst is the width of each bar. xbase is the location of the leftmost portion of the graph. ybase is the bottom of the graph (with the labels beneath this "bottom"). DiceTotalArray is an array that holds the frequency for each dice roll (3 to 18). For i = 3 to 18 OffScreenCnvs.Text xbase+(i*XConst), ybase, i OffScreenCnvs.Text xbase+(i*XConst), ybase+20, DiceTotalArray(i) ' draw a rectangle or a line if DiceTotalArray(i) > 0 then OffScreenCnvs.Rectangle xbase+(i*XConst), ybase-(DiceTotalArray(i)*YConst), XConst, DiceTotalArray(i)*YConst end if Next i ' draw a line for "zero" OffScreenCnvs.Line xbase+(3*XConst), ybase, xbase+(19*XConst), ybase If you want the whole program to pull apart, I'll send it to you. Cynthia Cynthia DeVore I/O Grad Student and E-Prime Consultant University of Minnesota, Minneapolis devo0023 at tc.umn.edu 612-624-3601 Steve Lacey wrote: > Hi, > > I am writing a script for a basic CRT experiment. After every block of > trials I want to provide feedback in the form of graphs. There will be > several graphs--one will plot RT by block number, another that will plot > accuracy by block number, ect. > > Any ideas about how I might do this? Is there a method within e-prime to do > this? Could I send the data to Excel, generate the plots there, and then > import them back into separate images in a slide object? If either of these > are possible, how would I do it? > > Thanks, > Steve From Btomlow at chdr.nl Wed Sep 25 11:59:30 2002 From: Btomlow at chdr.nl (Benjamin Tomlow) Date: Wed, 25 Sep 2002 13:59:30 +0200 Subject: Devided attention test Message-ID: CAn anybody help me and tell me if there's the possibility to implement a divided attention test within E-prime? From slacey at umich.edu Wed Sep 25 19:27:31 2002 From: slacey at umich.edu (Steve Lacey) Date: Wed, 25 Sep 2002 15:27:31 -0400 Subject: Displaying/Generating Graphs In E-Prime In-Reply-To: <3D889FCE.7C8C9E3E@tc.umn.edu> Message-ID: Thanks Cynthia, It turns out that the easiest way to draw plots is to use the canvas function, as you have suggested. I also submitted the question to PSTNET and they said it was not possible to dynamically send data from eprime to excel, plot the graphs, and then import them back into images on a slide object during runtime. Steve At 10:46 AM 9/18/2002 -0500, Cynthia J. DeVore wrote: >Steve, > >So far as I know, there isn't a readily available method for generating >graphs. >I have, however, developed something that performs a similar function. The >"trick" is to reorient the Canvas space because down is up (positive number >increase in the y-direction moving from top to bottom, rather than from bottom >to top as we typically see in coordinate space). > >The code below is the core of the logic used to draw a histogram in which the >low value is 3 and the high value is 18. (In this task, 3 dice are rolled and >the sum is graphed.) XConst is the width of each bar. xbase is the location of >the leftmost portion of the graph. ybase is the bottom of the graph (with the >labels beneath this "bottom"). DiceTotalArray is an array that holds the >frequency for each dice roll (3 to 18). > >For i = 3 to 18 > OffScreenCnvs.Text xbase+(i*XConst), ybase, i > OffScreenCnvs.Text xbase+(i*XConst), ybase+20, DiceTotalArray(i) > ' draw a rectangle or a line > if DiceTotalArray(i) > 0 then > OffScreenCnvs.Rectangle xbase+(i*XConst), >ybase-(DiceTotalArray(i)*YConst), XConst, DiceTotalArray(i)*YConst > end if >Next i > >' draw a line for "zero" >OffScreenCnvs.Line xbase+(3*XConst), ybase, xbase+(19*XConst), ybase > >If you want the whole program to pull apart, I'll send it to you. > >Cynthia > >Cynthia DeVore >I/O Grad Student and E-Prime Consultant >University of Minnesota, Minneapolis >devo0023 at tc.umn.edu >612-624-3601 > >Steve Lacey wrote: > > > Hi, > > > > I am writing a script for a basic CRT experiment. After every block of > > trials I want to provide feedback in the form of graphs. There will be > > several graphs--one will plot RT by block number, another that will plot > > accuracy by block number, ect. > > > > Any ideas about how I might do this? Is there a method within e-prime to do > > this? Could I send the data to Excel, generate the plots there, and then > > import them back into separate images in a slide object? If either of these > > are possible, how would I do it? > > > > Thanks, > > Steve