Finding the right comments

[liwenna]

Hi Gilis,

In by far you're not the first and will definitely also not be the
last person to struggle with preventing stimuli to 'double up' when a
list is repeated. As David wrote in your previous thread, searching
this group and the PST forums should give you a bunch of threads on
this topic.

The solution you propose seems reasonable but the main problem would
be that whenever a doubling is discovered, the doubled word should be
replaced (or the trial with the double skipped and added again later
in the program, causing a potential new doubling etc). This requires
interfering with the lists, after which a list should be reset and the
randomization up to that point is then lost. It is possible to create
a new list from skipped trials that runs after the initial lists have
run (if doubles have occurred) but this tampers with a blocked design
which is often not desired.

In by far the most easy solution therefore is to simply split the
stimuli (or whatever the to be randomized values are) in two separate
lists that are alternating. I.e. first run a randomized listA then a
randomized listB, followed by listA etc. This admittedly results in
pseudorandomization but for most designs I'd reckon this is a better
solution than to tamper with trialorder 'on the fly', breaking up the
blocked design, having a memory heavy comparison process running
throughout the test etc etc.

best,

liw

On Oct 18, 11:17 pm, gilis <gilads... at gmail.com> wrote:
> In continue to my former question I would like to update that I
> believe that I find the solution though the problem is that I've tough
> time to find the right commands in eprime. First, in short, this is
> what I want to do: I've one main list, in this list there are about 10
> nested blocks, each block contain an attribute with 10 words and from
> this attribute the slide object take the word to present each time, so
> the selection of words from this attribute is set to random. Each
> block run several times so though the selection of words is random, it
> could happen that the same word repeat itself twice in the same block
> (and I don't want it). The solution I was thinking about go like that:
>
> Dim blocknum, blockname as integer
>
> Blockname=c.GetAttrib(“block_name") ' each block has its own name,
> that's important
>
> do  ' starting loop statement
>
> Dim Comparison as Boolean 'setting another important variables
> Dim worda as string
> Dim wordb as string
> Dim n as integer
>
> N=n+1 ' counting the number of words that were presented, so each word
> has its own ordinal number
>
> If mod n/2=0 then 'well, I don't know the right command in eprime, but
> I try to see if there is remnant when dividing n by two to conclude if
> it's odd or even
>
> c. SetAttrib  “word1”, c.GetAttrib(“word”)  'in the case that n is
> even number keep the last word in attribute word1
> else
> c.SetAttrib “word2”, c.GetAttrib(“word”)   ' in the case that n is odd
> number, keep the last word in attribute word2
> end if
> c.GetAttrib(“word1”)=worda          'worda= the last word or the
> last-1 word
> c.GetAttrib(“word2”)=wordb          'wordb= the last-1 word or the
> last word
> if worda=wordb then                    'the Boolean variable is false
> in case that two consecutive words are actually the same
> Comparison= false
> Else
> Comparison=true                       'if the two consecutive words
> are different then the Boolean variable is set to "true"
> End if
> If Comparison = false then
> randomarray(List.&Blockname)  'well, I actually want to randomly
> select another word from the current block as long as Comparison value
> is        false but this is obviously not the right command-help is
> needed
> end if
> Loop until Comparison=true 'loop until different word is chosen
> N=0 'reset n value because there is no problem that two consecutive
> words will be the same if they are in different blocks and if n is not
> reset then the algorithm will prohibit such cases

-- 
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.