On one slide, show different images, loaded from a list

Wed Mar 10 23:16:30 UTC 2010

Hey Vera,

Yes you are totally right about the pseudorandomness of my proposed
solution.

There is a real random way (I use it for a task with 1 target and 15
distractors) that involves loading imagenames (I used external .txt
files as the list of distractors depends on the targetstimulus and
thus I had 15 lists of 15 distractorimages) into an array (you could
also fill the array in an inline though), randomising the array and
then filling the triallist.

Parts of the code needed can be found in this thread:
http://groups.google.com/group/e-prime/browse_thread/thread/b44513145914a543/9a190bab1a34fa8f

I didn't suggest this as I was under the impression that you were
seeking for a programming-free solution.

best,

liw

On Mar 10, 10:55 pm, Vera <vera.d... at googlemail.com> wrote:
> Hi Liwenna,
>
> first a big, huge thanks for replying!! :) It sure makes one want to
> rip out less hair just knowing that you're not alone. ;)
> Just a few comments on what you wrote:
>
> > One thing that is elemental to understand when using e-prime, is that
> > on each run of a procedure one level of each of the nested lists can
> > be used. If you need nine different pictures in one 'run of the
> > procedure' you can not put them in different rows as e-prime can only
> > acces one level of each list on each run. So... you wrote: " like
> > putting the 9 images in one line of the nested list, but then I get
> > other errors", and you were definitely on the right track!
>
> Yep, doesn't work, but it was a nice try! ;)
>
> > Do the nine images have the same positions on each trial? Or are they
> > placed in different 'arrangements' on every trial?
>
> For the moment they are at the same position every time, but thanks
> for the instructions below, we might be using that later. :)
>
> > What you could do is make one (nested or direct) list that has nine
> > attributes on every row containing the nine pictures used in a trial,
> > call them for instance stim1, stim2, stim3,...stim8 and targetstim. In
> > each level of this list you can make a different combination of
> > stimuluspictures. In your slide you'll have nine imageobjects: make
> > each one of them refer to a different stimulus attribute: [stim1]
> > [stim2] ... [targetstim].
>
> Yep, thought of that option too, but there is one problem to it: we
> want the images to appear completely at random (so in each trial with
> 9 images there is 1 target, 4 distractors(type1) and 4
> distractors(type2). Each item can be at 9 different positions. Which
> would make (ok, I might be wrong here, I don't remember very well the
> correct chance calculations) something like... a lot (!!!!) of
> combinations, no? Or is there some way to tell E-Prime to randomize 9
> attributes (9 columns basically)?
>
>
>
> > So far so good: now you'll have different collections of stimpictures
> > on each run (as many different 'collections' as you have made
> > different levels in your list). If you need them to be placed in
> > different spatial arrangement on each trial you should start out with
> > creating the number of different arrangements that you want to use.
> > For each arrangement write down the x and y values of each of the
> > imageobjects. (so: imageobject 1 showing [stim1] has x = 500 and y=68
> > in arrangement 1 and x = 950 and y =780 in arrangement 2  for
> > instance). Into the list with stimuluspictures nest a second list that
> > holds the different arrangements. This list will have 9x2=18
> > attributes: stim1x, stim1y, stim2x, stim2y, stim3x... targetstimy. Go
> > back to your slide and now tell the slideobjects to find their x and y
> > values in the corresponding attributes, just like they find their
> > pictures: [stim1x] [stim1y] [stim2x] etc etc.
> > Now... you'll have spent a lot of time but you will have a random
> > combination of a collection of target and distractorpictures with one
> > of the different spatial arrangements on each trial.
>
> :D Like you said, I will for sure have spent a lot of time! :) But it
> might be some kind of solution maybe, I will have to think of it. But
> then again, I think that
> there will be too many combinations possible.
>
> > Alternatively.... I say it should be possible to use 'complete array'
> > pictures that you make in for instance photoshop or even paint....
> > pictures that have the size of the screen and show all nine stimuli at
> > once... you could tell e-prime to show one of those pictures on each
> > trial and locate a single 'transparant empty' imagobject over the
> > location of the targetstimulus (you should fiddle around a bit with a
> > non-transparant coloured object to find the exact location needed for
> > each 'array picture' and actually place this imageobject UNDER the
> > arraypicture and not in front).This imageobject will then be the
> > object to which you point your dohittest script (assuming that that's
> > what you will use? have your subjects click the mouse on the target?).
> > Not sure if that will work.. think it should.
>
> Yep, have been thinking of that too, but it for sure isn't the most
> elegant solution, as the random again will be more or less
> "biased" (one can only make so many combinations).
>
> > I hope the above stuff will get you started again, if not, please let
> > me know.
>
> Yep, thanks for the reflexions! :) As I said, it sure helps to know
> that other people have thought of this too. :) For the moment I am
> working on a (although I am an absolute beginner as far as programming
> is concerned) solution with an inline script. I was thinking that
> maybe I can tell Eprime to load 3 numbers into this column where
> normally pictures would be, randomize the whole and then maybe
> according to the numbers Eprime could load some corresponding images?
>
> I however just still don't want to believe that this could be the only
> solution (especially because "my problem" is working fine when you
> replace the images on the slide by text and tell Eprime to randomly
> load strings (words) from a nested file). Ah well...
>
> Thanks a lot anyway for replying and if I find a solution, I will let
> you know. :)
>
> Greetings,
>
> Vera
>
>
>
> > Best
>
> > liw
>
> > On Mar 10, 5:45 pm, Vera <vera.d... at googlemail.com> wrote:
>
> > > Hi forum members!
>
> > > I sure hope I will get some help here! I am trying to construct a
> > > visual search task in which participants have to find let's say a
> > > "yellow triangle" within "blue triangles" and "yellow triangles".
>
> > > For this, I created a list with different conditions (like number of
> > > items on the grid, target present or not) and I was actually counting
> > > on using a slide with (let's say) 9 images, on which I would each of
> > > the 9 images randomly from a nested list.
>
> > > So I made my nested list with images (.png's) and I of course made
> > > sure to have the image-objects on the slide point first to the
> > > procedure list and then to the nested list (I think I read about all
> > > the messages concerning this topic, including this onehttp://groups.google.com/group/e-prime/browse_thread/thread/84c742b85...
> > > (thanks a lot for the work-around, it was a good thing for me to see
> > > if I did it right, which I did, but it still will not load the right
> > > pictures).
>
> > > So the problem is: It seems that the "point to the nested list
> > > function" is working, but then it seems impossible to load 9 DIFFERENT
> > > images (it will show me 9 times the same image, at least therewith
> > > confirming that it points to the correct nested list).
>
> > > And I tried every option I could think of (like putting the 9 images
> > > in one line of the nested list, but then I get other errors). :-( I am
> > > now believing that I should maybe really write to the developers,
> > > simply because this "functionality" seems to be absent.
>
> > > Now I will probably have to code it (which is were the real problem
> > > starts, because I am an absolute no-programmer ;-) ) and the institute
> > > where I am currently working actually bought E-Prime because it should
> > > be more or less "programming-free" (well ok, I understand you cannot
> > > rule it out completely, but the thing I described above should be
> > > possible in my opinion).
>
> > > I really hope some of you guys have some creative solution for me!
>
> > > With kind regards,
>
> > > Vera

-- 
You received this message because you are subscribed to the Google Groups "E-Prime" group.
To post to this group, send email to e-prime at googlegroups.com.
To unsubscribe from this group, send email to e-prime+unsubscribe at googlegroups.com.
For more options, visit this group at http://groups.google.com/group/e-prime?hl=en.