Millisecond Forums

You create two <expt> elements in the script -- one that runs the "compatible" condition, one that runs the "incompatible" condition -- and assign participants to one of the two via the /subjects and /groupassignment attributes:

<expt>
/ subjects = (1 of 2)
/ groupassignment = groupnumber
/ blocks = [... blocks for the compatible condition ...]
</expt>

<expt>
/ subjects = (2 of 2)
/ groupassignment = groupnumber
/ blocks = [... blocks for the incompatible condition ...]
</expt>

If you have two scripts instead -- one compatible, one incompatible -- you can do the same with <batch>, i.e. assign participants to one of the two scripts.

<batch>
/ subjects = (1 of 2)
/ groupassignment = groupnumber
/ file = "compatible.iqx"
</batch>

<batch>
/ subjects = (2 of 2)
/ groupassignment = groupnumber
/ file = "incompatible.iqx"
</batch>

Under Inquisit Lab, you simply enter a group number: An odd number will run the 1st condition, and even number will run the 2nd condition. Under Inquisit Web, a random group number will be generated (by default), i.e. assignment to one of the conditions will be random.

Also see: https://www.millisecond.com/forums/Topic13856.aspx

Thank you very much for the quick reply. Actually, I have 4 different groups. First two groups are in compatible condition, while last two groups are in incompatible condition. I may define the groups as:

Group 1: Set 1 Trustworthy Faces & Trustworthy behaviors and Set 2 Untrustworthy Faces & Untrustworthy behaviors (compatible condition)
Group 2: Set 2 Trustworthy Faces & Trustworthy behaviors and Set 1 Untrustworthy Faces & Untrustworthy behaviors (compatible condition)
Group 3: Set 1 Trustworthy Faces & Untrustworthy behaviors and Set 2 Untrustworthy Faces & Trustworthy behaviors (incompatible condition)
Group 4: Set 2 Trustworthy Faces & Untrustworthy behaviors and Set 1 Untrustworthy Faces & Trustworthy behaviors (incompatible condition)

I have tried to define the groups by:

<variables>
/group= (1 of 4) (facestrustworthy=trustworthyset1; facesuntrustworthy=untrustworthyset2; trustworthybehaviors=trustworthybeh; untrustworthybehaviors=untrustworthybeh)
/group= (2 of 4) (facestrustworthy=trustworthyset2; facesuntrustworthy=untrustworthyset1; trustworthybehaviors=trustworthybeh; untrustworthybehaviors=untrustworthybeh)
/group= (3 of 4) (facestrustworthy=trustworthyset1; facesuntrustworthy=untrustworthyset2; trustworthybehaviors=trustworthybeh; untrustworthybehaviors=untrustworthybeh)
/group= (4 of 4) (facestrustworthy=trustworthyset2; facesuntrustworthy=untrustworthyset1; trustworthybehaviors=trustworthybeh; untrustworthybehaviors=untrustworthybeh)
/ groupassignment = subjectnumber
</variables>

However, here Group 1 and Group 3; and Group 2 and Group 4 appear to be the same. However, in Groups 1 and 2, trustworthy faces pair with trustworthy behaviors and untrustworthy faces pair with untrustworthy behaviors. And in Groups 3 and 4 the opposite (trustworthy faces pair with untrustworthy behaviors, etc.). How can I show this difference while defining the groups? To be more clear, how can I define these different pairings in different conditions?

And, in this case, do I need to define four different experiments instead of two?

Groups 1 and 3 are the same because you've defined them to be the same.

/group= (1 of 4) (facestrustworthy=trustworthyset1; facesuntrustworthy=untrustworthyset2; trustworthybehaviors=trustworthybeh; untrustworthybehaviors=untrustworthybeh)
...
/group= (3 of 4) (facestrustworthy=trustworthyset1; facesuntrustworthy=untrustworthyset2; trustworthybehaviors=trustworthybeh; untrustworthybehaviors=untrustworthybeh)

The same is true for groups 2 and 4.

So, I'm actually not sure what the question is. It's also completely unclear what kind of elements "trustworthyset1" etc. actually refer to.

Hi again, yes I know they are the same, I have defined them to be the same on purpose, to show you how I am unable to define these two groups differently. My question is; the only difference between Group 1 and 3 is that in Group 1 trustworthy faces will be paired with trustworthy behaviors, and in Group 3, trustworthy faces will be paired with untrustworthy behaviors, but they use the same set of faces and behaviors. However, I don't know how to define this two different pairings while defining Group 1 and 3. There are 100 trustworthy and 100 untrustworthy faces and we have divided these faces into two sets so there are trustworthy faces set 1, trustworthy faces set 2, untrustworthy faces set 1 and untrustworthy faces set 2. "trustworthyset1" refers to set 1 of trustworthy faces.

Sorry, this doesn't make things any clearer. What <variables> does is *substitute* elements. It does not pair anything and it is unclear what "trustworthy faces will be paired with trustworthy behaviors" means concretely: "Paired" in what way? Please keep in mind that I have no knowledge of your experiment apart from what you shared here.

<variables> allows you to, say, switch different sets of items between-subjects. Suppose you have

<text mytext>
/ items = myitems
</text>

and two sets of items A and B

<item a>
/ 1 = "A1"
/ 2 = "A2"
/ 3 = "A3"
/ 4 = "A4"
</item>

<item b>
/ 1 = "B1"
/ 2 = "B2"
/ 3 = "B3"
/ 4 = "B4"
</item>

then

<variables>
/ group = (1 of 2) (myitems = a)
/ group = (2 of 2) (myitems = b)
</variables>

the 1st group will administer item set A, the 2nd group will administer item set B.

This does not only work with <item> elements, you can substitute other types of elements as well.

Suppose you have

<block myblock>
/ trials = [1-10 = mytrial]
</block>

and two different <trial> elements A and B (which can display totally different stimuli, etc.)

<trial atrial>
/ stimulusframes = [...]
...
</trial>

<trial btrial>
/ stimulusframes = [...]
...
</trial>

then 

<variables>
/ group = (1 of 2) (mytrial= atrial)
/ group = (2 of 2) (mytrial = btrial)
</variables>

the 1st group will result in running 10 x <trial atrial>, while the 2nd group will run <trial btrial> 10 times.

Sorry for the complicated explanation :) There are faces which seem as trustworthy and untrustworthy. There are also behavior statements which may be accepted as trustworthy and untrustworthy behaviors. For Groups 1 and 2, whenever a trustworthy face is represented on the screen, a trustworthy behavior statement will be shown below the picture, and whenever an untrustworthy face is represented on the screen, an untrustworthy behavior statement will be shown below the picture. For Groups 3 and 4, whenever a trustworthy face is represented on the screen, an untrustworthy behavior statement will be shown below the picture, and whenever an untrustworthy face is represented on the screen, a trustworthy behavior statement will be shown below the picture. This is what I called as pairing. So, will I use stimuluasframes attribute to define these two different conditions?

Suppose you have two <trial> elements, one for a displaying trustworthy face and associated behavioral statement, another one for displaying an untrustworthy face and associated behavioral statement:

<trial t_face_trial>
/ stimulusframes = [1=t_face, t_face_beh]
...
</trial>

<trial u_face_trial>
/ stimulusframes = [1=u_face, u_face_beh]
...
</trial>

with 

<picture t_face>
/ items = t_face_items
...
</picture>

<text t_face_beh>
/ items = t_face_beh_items
...
</text>


<picture u_face>
/ items = u_face_items
...
</picture>

<text u_face_beh>
/ items = u_face_beh_items
...
</text>

Then all you need to do is assign the <item> elements you want for each in the <variables> element's various groups.

I.e.

<variables>
/group= (1 of 4) (t_face_items=trustworthyset1; u_face_items=untrustworthyset2; t_face_beh_items=trustworthybeh; u_face_beh_items=untrustworthybeh)
/group= (2 of 4) (t_face_items=trustworthyset2; u_face_items=untrustworthyset1; t_face_beh_items=trustworthybeh; u_face_beh_items=untrustworthybeh)
/group= (3 of 4) (t_face_items=trustworthyset1; u_face_items=untrustworthyset2; t_face_beh_items=untrustworthybeh; u_face_beh_items=trustworthybeh)
/group= (4 of 4) (t_face_items=trustworthyset2; u_face_items=untrustworthyset1; t_face_beh_items=untrustworthybeh; u_face_beh_items=trustworthybeh)
/ groupassignment = subjectnumber
</variables>

Does that make sense?

Thank ytou very much. That makes perfect sense. And one little question. What is the function of these teo elements in a model?

<values>
/ currentpic = 0
/ currentbehavior = 0
</values>

<counter tracker>
/ items = (1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50)
/ select = values.currentpic
</counter>


Thank you very much!

<values> are global variables -- they store things put there via logic in /ontrialbegin, /ontrialend, etc. attributes.

A <counter> is just an element that stores and selects items (can be numbers, strings, etc.), in this case the selection of the item is determined by the value of values.currentpic.