Millisecond Forums

Adjusting random selection from two sets of primes

https://forums.millisecond.com/Topic2877.aspx

By Msquared - 5/27/2009

Hello!

I have two set of primes, and each set contains 24 items. One set is the experimental set and one set is the neutral set. I would like to run 72 trials with these two sets of primes. Thus, I would Inquisit to randomly select with no replacement the 24 experimental primes two times, but I would like it to only use each of the neutral primes once (again determined by random selection). Is this possible?

By Dave - 5/27/2009

There are several ways to achieve this, I believe. One is to have a set of two <trial> / <text> etc. elements, one for the experimental primes and another one for the neutral primes.

<trial prime_expt>
/ stimulusframes = [1=prime_expt]
[...]
</trial>

<trial prime_neut>

/ stimulusframes = [1=prime_neut]

[...]

</trial>

<text prime_expt>
/ items = experimentalprimes
/ select = noreplace
[...]
</text>

<item experimentalprimes>
/1 = "Experimental Prime 1"
[...]
/ 24 = "Experimental Prime 24"
</item>

<text prime_neut>

/ items = neutralprimes
/ select = noreplace
[...]

</text>

<item neutralprimes>

/1 = "Neutral Prime 1"

[...]

/ 24 = "Neutral Prime 24"

</item>



Then put these in a <block> element in the desired proportions -- you want twice as many trials showing experimental primes than trials presenting neutral primes. Thus


   
<block primingblock>
/ trials = [1-72=noreplace(prime_expt, prime_expt, prime_neut]
[...]
</block>



should result in 48 experimental trials and 24 neutral prime trials. Since the respective <text> elements are set to 'noreplace', each of the 24 neutral stims should be shown exactly once while each of the experimental prime stims should be shown a second time after the first 24 experimental trials.

Hope this helps. Best wishes from a fellow Inquisit user,
~Dave

Note to self: Yay, 400th forum post.
By duschenhocker - 1/6/2011

Hello Dave!


I have a similar question about composing trials out of several item pools. There is the AMP procedure I have shaped to my requestion. Now it's about distibuting the 3 item pools to the blocks. There is a primeA pool = 20items , primeB pool= 20 items and primeNeutral pool= 32 items. I would like to run 72 items per block. Inquisit distribute the items equaly, each pool 24 times, therefore there is a repetition of items, which I don't intend.I would like the items distributed like the pools 20, 20, 32.


here is the scipt:


<block test2000>
/ bgstim = (attributeAlabel, attributeBlabel)
/ trials = [1-72 = replacenorepeat(primeA2000, primeB2000, primeNeutral2000)]
</block>


How can I get this right?


By Dave - 1/6/2011

Hi Michael,


see http://www.millisecond.com/forums/Topic3693.aspx#3693.


Regards,


~Dave

By duschenhocker - 1/6/2011

That's a great solution. Thank you!