counterbalancing blocks/tasks, randomizing the words over the blocks and masks

Tue Dec 12 19:55:06 UTC 2006

Thanks so much Ben, you've been a tremendous help!
I have a few questions, I adjusted some things in the file you sent me. The
background masks are in a bmp format, so I changed the link to the pictures,
I think that went ok.

Weight Nested    Procedure word   contrast    correctage correctcontrast
4         oldwords  trialproc    [old]    laag.bmp    o                  l
4         oldwords  trialproc    [old]    hoog.bmp   o                  h
4         newwords trialproc    [new]  laag.bmp    n                  l
4         newwords trialproc    [new]  hoog.bmp   n                  h

And I made a slide to put the picture in and the text [word], in the
'filename' box of the picture I put [contrast], it's all working. I'm so
happy.

My question is, can I counterbalance the block tasks
(contrasttask-oldnewtask-contrasttask etc.) so that the subject never gets
the same task as in the previous block?
And once a word (from the total of 144 words) has been presented it must not
be presented ever again(not in another block). Is it possible to change the
settings so that there will be no repetition of words and still be a random
selection (this is per subject of course)?

When that's fixed, the computer task is ready for experimentation :)

Thanks so much for all the help!
Greetings,
Suzanne Bakkum

On 12/11/06, Ben Robinson <brobinso at mprc.umaryland.edu> wrote:
>
> make a list with two procedures - identify age, identify contrast
> select from the list at random (to shuffle which procedure comes first)
> exit the list after 3 cycles (3 x identify age block, 3 x identify
> contrast block, for a total of six blocks)
>
> in both the identify age and identify contrast blocks you can use the same
> "eobjects:"
> have a list, "trailtypes:"
> weight  nested     procedure    age
> contrast        correctage      correctcontrast
> 3       oldwords           trialproc    [old]   LOW
> o                       l
> 3       oldwords           trialproc    [old]
> HIGH    o                       h
> 3       newwords   trialproc    [new]   LOW     n                       l
> 3       newwords   trialproc    [new]   HIGH    n                       h
>
> select from trialtypes at random
> exit the list after 1 cycle (12 trials)
>
> in the nested list, "oldwords," add a column, "old," and fill this with
> your 72 old words
> in the nested list, "newwords," add a column, "new," and fill with 72 new
> words
> select from each nested list at random, and exit each after 1 *sample*
>
> in your "trialproc" procedure add two TextDisplays.
> TextDisplay1:
> text = [word]
> background color = gray
> forecolor = [contrast]
> duration = infinite
>
> to this add a keyboard device.  under allowable responses type the letters
> o and n (old and new).  for correct response type [correctage].
> add a *second* keyboard device.  under allowable type the letters l and h
> (low and high).  for correct response type [correctcontrast].
>
> now add a second text display to your trial proc, set the background to
> gray, and you should be good to go.  the second text display will simply add
> an interstimulus interval, so the words don't come immediately after one
> another.
>
> now for the scripting.  there is very little.
>
> i'm not exactly sure how you want to adjust the contrast, but as an
> example you could switch between black and white lettering on a gray
> background.  to make the black/white example work, go to your "script"
> window, and type the following:
> CONST LOW = "black"
> CONST HIGH = "white"
> then add an inline at the very beginning of your SessionProc, and type:
> trialtypes.SetAttrib 1, "contrast", LOW
> trialtypes.SetAttrib 2, "contrast", HIGH
> trialtypes.SetAttrib 3, "contrast", LOW
> trialtypes.SetAttrib 4, "contrast", HIGH
> trialtypes.Reset
>
> that should do it.
>
> ben
>
> >>> "Suzanne Bakkum" <sbakkum at gmail.com> 12/10/2006 8:59 AM >>>
> Hi all,
>
> I'm having trouble programming the test phase of my experiment. I hope
> someone can help me here, I've tried so many things but I can't figure the
> script codes (InLine) out, it's abracadabra for me.
>
> This is what the experiment looks like:
> There's a list of old words (72) and a list of new words (72). Subjests
> will
> de shown the words on a background mask, there are to types of
> backgroundmasks (one in high contrast with the word shown and one in low
> contrast).
>
> The words will be given in blocks of 12 words (6 old, 6 new). The subject
> has to decide whether the word is old or new, and in the second block
> whether the word is in high contrast with the mask or in low contrast. I
> want the words to be random over the total of 72 words shown and i want
> the
> 2 types of blocks (old/new and high/lowcontrast) counterbalanced.
>
> So one block looks like this:
> o = old, n = new, h = high contrast mask, l = low contrast mask
> o - h
> o - h
> o - h
> o - l
> o - l
> o - l
> n - h
> n - h
> n - h
> n - l
> n - l
> n - l
>
> I want the words to be randomized over all the blocks so the words don't
> stay in the same block all the time and I want the words to be randomized
> over the backgroundmasks, so for example I don't want the word 'apple' on
> the high mask all the time.
>
> The blocks look all alike only the task is different, first block decide
> word is old or new, second block decide word is high contrast or low
> contrast, third block old/ new again, and so on. But I want one subject to
> begin with the old/new block and another with the contrasttask.
>
> I hope I was clear anough, English is not my first language. I hope
> someone
> knows how to program this. I'm doing an intership and my tutor never ever
> worked with e-prime, so I'm gonna have to fix this..
>
> Thanks very much in advance,
>
> Suzanne Bakkum
> Psychonomics
> University of Amsterdam
> sbakkum at gmail.com
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.linguistlist.org/pipermail/eprime/attachments/20061212/1507708d/attachment.htm>