Complex (for me) randomization

[ccbaye]

Ok David, thank you for the time you took for me. I think I will use a
fixed random order for the problem. This order will be same across
experimental conditions, so it will not have a large influence I
guess.

> Ok..I finally managed to have the prog selecting one face and holding
> it for the rest of the block trials. yeepee.

>You did not say how you did this, that would help us.

You are right, so...for beginners who get the same problem. I simply
present the face picture with an imagedisplay object at the
Blockprocess level, with a "0" in duration (position centered). Then I
put my Triallist on its right, on the blockprocess line. From there I
start a new Trialprocess, and put the second picture (bottom position)
with Background set to "transparency". As the second picture is
smaller than the face picture, they fit both nicely simultaneously on
the screen. Now as I go through the block trials, only the second
picture changes, not the face picture. When the block is over, a new
face is selected, and the trials with the second pictures go on again,
and so on for the other blocks.

I really thank you for your advice David.

Cheers,
Christophe.





On May 19, 4:16 pm, David McFarlane <mcfar... at msu.edu> wrote:
> Christophe,
>
> > Ok..I finally managed to have the prog selecting one face and holding
> > it for the rest of the block trials. yeepee.
>
> You did not say how you did this, that would help us.  Just to make sure
> we are all talking about the same thing, let me lay out what I had in mind.
>
> First, the general rule, as explained in some detail in Chapter 4 of the
> User's Guide (and I will try to word this very carefully):  The current
> value of any attribute set at one level is available to each
> successively activated lower level.
>
> So, suppose you have a structure like
>
> - BlockList
>      - BlockProc
>          - TrialList
>              - TrialProc
>
> And suppose in BlockList you have an attribute (i.e., column) named
> "Face", and that has two levels (i.e., rows) with Face set to "Bob" on
> level 1, and "Alice" for level 2.
>
> Now, when your BlockList runs level (row) 1, all the trials in TrialList
> will use the face "Bob", and when your BlockList runs level 2 all the
> trials in TrialList will use the face "Alice".  So you see it really is
> quite simple, as you originally posed the problem.  And the rule applies
> whether you run BlockList sequentially, randomly, randomly with
> replacement, or any other way.  The key is to understand the use of
> blocks vs. trials, etc., and how attribute values propogate.
>
> > The problem now is that
> > for each new block, the prog select a face at random in the
> > corresponding face list, including the face that has been presented in
> > the preceding block..So I can't yet manage to have a randomized
> > selection of faces without replacement through the different blocks of
> > my task.
>
> But now you add a complication.  Now you want each block to use both a
> randomly selected new face, *plus* reuse the face from the previous
> block.  Well, in general E-Prime is just not very good at remembering
> the past unless you help it along with some inline code.  The only way I
> know how to do what you now ask is to define some global variables or
> arrays and then to manage these in inline code within the program.  I
> cannot give you any more detail than that.  If you are not already a
> competent computer programmer then I advise that you now either take
> some time out to learn some real computer programming skills, or hire
> someone else for this work.
>
> -- David McFarlane, Professional Faultfinder
>
> --
> 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 athttp://groups.google.com/group/e-prime?hl=en.

-- 
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.