Millisecond Forums

If I understand you correctly, you'll simply want to pair your "Suchbild" <picture> elements to your "Zielreiz" <picture>:

https://www.millisecond.com/support/docs/v4/html/howto/howtopairs.htm

In a nutshell, I'd do something like this:

<values>
/ zielitem = 1
/ suchitem = 1
</values>

<block exampleblock>
/ trials = [1-5 = exampletrial]
</block>

<trial exampletrial>
/ ontrialbegin = [values.zielitem = list.zr_itemnumbers.nextindex;
values.suchitem = values.zielitem + list.version.nextvalue;]
/ stimulustimes = [0=zielreiz; 2000=suchbild]
/ validresponse = (57)
</trial>

<list zr_itemnumbers>
/ poolsize = 5
</list>

<list version>
/ items = (0,5)
/ replace = true
</list>

<item Zielreiz>
/1 = "ZR1.jpg"
/2 = "ZR2.jpg"
/3 = "ZR3.jpg"
/4 = "ZR4.jpg"
/5 = "ZR5.jpg"
</item>

<text Zielreiz>
/items = Zielreiz
/select = values.zielitem
/position = (50, 50)
</text>

<item Suchbild>
/1 = "SB1A.jpg"
/2 = "SB2A.jpg"
/3 = "SB3A.jpg"
/4 = "SB4A.jpg"
/5 = "SB5A.jpg"

/6 = "SB1B.jpg"
/7 = "SB2B.jpg"
/8 = "SB3B.jpg"
/9 = "SB4B.jpg"
/10 = "SB5B.jpg"
</item>

<text suchbild>
/ items = suchbild
/ select = values.suchitem
</text>

Or, if you'd rather keep things a bit simpler, you could set up two separate trials (one for version a, one for version b) and then sample from those trials randomly at the block-level:

<block exampleblock>
/ trials = [1-5 = replace(a,b)]
</block>

<trial a>
/ stimulustimes = [0=zielreiz; 2000=suchbilda]
/ validresponse = (57)
</trial>

<trial b>
/ stimulustimes = [0=zielreiz; 2000=suchbildb]
/ validresponse = (57)
</trial>

<item Zielreiz>
/1 = "ZR1.jpg"
/2 = "ZR2.jpg"
/3 = "ZR3.jpg"
/4 = "ZR4.jpg"
/5 = "ZR5.jpg"
</item>

<text Zielreiz>
/items = Zielreiz
/select = noreplace
/position = (50, 50)
</text>

<item SuchbildA>
/1 = "SB1A.jpg"
/2 = "SB2A.jpg"
/3 = "SB3A.jpg"
/4 = "SB4A.jpg"
/5 = "SB5A.jpg"
</item>

<item SuchbildB>
/1 = "SB1B.jpg"
/2 = "SB2B.jpg"
/3 = "SB3B.jpg"
/4 = "SB4B.jpg"
/5 = "SB5B.jpg"
</item>

<text suchbildA>
/ items = suchbildA
/ select = text.zielreiz.currentindex
</text>

<text suchbildB>
/ items = suchbildB
/ select = text.zielreiz.currentindex
</text>

And if you want some balance between versions (e.g. either 3 trials of version A and 2 trials of version B or vice versa), you can do

<block exampleblock>
/ trials = [1-4 = noreplace(a,b); 5 = noreplace(a,b)]
</block>

This helps a lot, thank you Dave!

Dear Dave,

We slightly changed our design and I am in need of your help once again.

I have two sets of stimuli, (itemlists) ZR and NA, with Zielreiz and Suchbild linked as above: If Zielreiz1 is selected, it should accordingly select Suchbild1. 
Now, however, we have two versions of each Zielreiz and Suchbild - ZR and NA. It should randomly select from either ZR or NA item but if it selects item 1 from ZR it should never select item 1 from NA.
It should work as a linked either-or condition so to say. Is that possible?

Thanks a lot in advance!

Best regards,

Rebsu

Please provide the code.

<item Zielreiz>
/1 = "ZR1.jpg"
/2 = "ZR1k.jpg"
/3 = "ZR2.jpg"
/4 = "ZR2k.jpg"
/5 = "ZR3.jpg"
/6 = "ZR3k.jpg"
/7 = "ZR4.jpg"
/8 = "ZR4k.jpg"
/9 = "ZR5.jpg"
/10 = "ZR5k.jpg"
/11 = "ZR6.jpg"
/12 = "ZR6k.jpg"
/13 = "ZR7.jpg"
/14 = "ZR7k.jpg"
/15 = "ZR8.jpg"
/16 = "ZR8k.jpg"
/17 = "ZR9.jpg"
/18 = "ZR9k.jpg"
/19 = "ZR10.jpg"
/20 = "ZR10k.jpg"
</item>

<picture Zielreiz>
/items = Zielreiz
/select = noreplace
/position = (50, 50)
</picture>

<item Alternative>
/1 = "NA1.jpg"
/2 = "NA1k.jpg"
/3 = "NA2.jpg"
/4 = "NA2k.jpg"
/5 = "NA3.jpg"
/6 = "NA3k.jpg"
/7 = "NA4.jpg"
/8 = "NA4k.jpg"
/9 = "NA5.jpg"
/10 = "NA5k.jpg"
/11 = "NA6.jpg"
/12 = "NA6k.jpg"
/13 = "NA7.jpg"
/14 = "NA7k.jpg"
/15 = "NA8.jpg"
/16 = "NA8k.jpg"
/17 = "NA9.jpg"
/18 = "NA9k.jpg"
/19 = "NA10.jpg"
/20 = "NA10k.jpg"
</item>

<picture Alternative>
/items = Alternative
/select = noreplace
/position = (50, 50)
</picture>

<picture Alternative_2>
/items = Alternative
/select = current (alternative)
/ position = (50, 50)
</picture>

<item SuchbildZ>
/1 = "AZ1.jpg"
/2 = "BZ1.jpg"
/3 = "AZ2.jpg"
/4 = "BZ2.jpg"
/5 = "AZ3.jpg"
/6 = "BZ3.jpg"
/7 = "AZ4.jpg"
/8 = "BZ4.jpg"
/9 = "AZ5.jpg"
/10 = "BZ5.jpg"
/11 = "AZ6.jpg"
/12 = "BZ6.jpg"
/13 = "AZ7.jpg"
/14 = "BZ7.jpg"
/15 = "AZ8.jpg"
/16 = "BZ8.jpg"
/17 = "AZ9.jpg"
/18 = "BZ9.jpg"
/19 = "AZ10.jpg"
/20 = "BZ10.jpg"
</item>

<picture SuchbildZ>
/Items = SuchbildZ
/select = picture.zielreiz.currentindex
/size = (95%, 95%)
/position = (54, 50)
</picture>

<item SuchbildN>
/1 = "AN1.jpg"
/2 = "BN1.jpg"
/3 = "AN2.jpg"
/4 = "BN2.jpg"
/5 = "AN3.jpg"
/6 = "BN3.jpg"
/7 = "AN4.jpg"
/8 = "BN4.jpg"
/9 = "AN5.jpg"
/10 = "BN5.jpg"
/11 = "AN6.jpg"
/12 = "BN6.jpg"
/13 = "AN7.jpg"
/14 = "BN7.jpg"
/15 = "AN8.jpg"
/16 = "BN8.jpg"
/17 = "AN9.jpg"
/18 = "BN9.jpg"
/19 = "AN10.jpg"
/20 = "BN10.jpg"
</item>

<picture SuchbildN>
/Items = SuchbildN
/select = picture.alternative.currentindex
/size = (95%, 95%)
/position = (54, 50)
</picture>


<item ZrWahl>
/1 = "ZR1.jpg"
/2 = "ZR2.jpg"
/3 = "ZR3.jpg"
/4 = "ZR4.jpg"
/5 = "ZR5.jpg"
/6 = "ZR6.jpg"
/7 = "ZR7.jpg"
/8 = "ZR8.jpg"
/9 = "ZR9.jpg"
/10 = "ZR10.jpg"
</item>

<item NaWahl>
/1 = "NA1.jpg"
/2 = "NA2.jpg"
/3 = "NA3.jpg"
/4 = "NA4.jpg"
/5 = "NA5.jpg"
/6 = "NA6.jpg"
/7 = "NA7.jpg"
/8 = "NA8.jpg"
/9 = "NA9.jpg"
/10 = "NA10.jpg"
</item>

<picture ZrWahl>
/ items = ZrWahl
/ hposition = list.hpos.nextvalue
</picture>

<picture NaWahl>
/ items = NaWahl
/ hposition = list.hpos.nextvalue
</picture>

<list hpos>
/ items = (30%, 70%)
/ selectionrate = always
</list>

<trial Suche>
/stimulustimes = [0=blank, fixation, ziellabel; 500=Zielreiz, ziellabel; 1500=blank, fixation; 2000=Suchbild; 6000=blank ;6001=Suchinstruktion]
/validresponse = ("t", "z", "g", "h", " ")
</trial>

<trial AltPresentation>
/stimulustimes = [0=Alternative, reset; 1000=blank, fixation; 1500=Alternative_2, reset; 5500=blank, warten]
/response = timeout (6500)
</trial>

(trials are still from the old version, so definitely not correct yet)

This is the current state. I am not sure if this is the best way to build up the whole thing, though. Basically, I have pairs of pictures - the A_(Z or N)_1 and B_1 version. It should randomly select out of the Zielreiz or Alternative Item which is linked to the SuchbildZ item if Zielreiz is selected and to the SuchbildN item if Alternative is selected. If it selects ZR1 (consecutively AZ1) it must always also select ZR2 as well (consecutively BZ1) but must not select NA1 and NA1k.
In the AltPresentation trial it should then always select the not-chosen version of ZR_ or NA_. So, if in the Such-trial ZR1 is chosen, it should choose AZ1 and ZR1k+BZ1 in the Such-trial as well, and then choose NA1 in the AltPresentation.

As I said, I am not sure if this choice of itemslists is the best way to do it, if it's better to pair ZR1 and NA1, ZR2 and NA2 etc. and then let inquisit choose between these pairs, consequtively linking them again to ZR1 or NA1, if there is a good way to do it at all or if pseudorandomizations with different groups where it is preselected if which SuchbilZ and which SuchbildN are randomly presented are way easier.

Sorry, I'm having trouble making sense of the description, particularly this part: "If it selects ZR1 (consecutively AZ1) it must always also select ZR2 as well (consecutively BZ1) but must not select NA1 and NA1k.
In the AltPresentation trial it should then always select the not-chosen version of ZR_ or NA_. So, if in the Such-trial ZR1 is chosen, it should choose AZ1 and ZR1k+BZ1 in the Such-trial as well, and then choose NA1 in the AltPresentation."

> "If it selects ZR1 (consecutively AZ1) [...] "
#1: What does "consecutively" mean here?

> "If it selects ZR1 [...] it must always also select ZR2"
#2: Why must ZR2 be selected if ZR1 has been selected?

> "it must always also select ZR2 as well (consecutively BZ1)"
#3: Why and in what way is ZR2 tied to BZ1? This doesn't seem to follow from the code at all.

#4: What is the purpose of these <picture> elements? They don't appear to be used in the code you provided.

<picture ZrWahl>
/ items = ZrWahl
/ hposition = list.hpos.nextvalue
</picture>

<picture NaWahl>
/ items = NaWahl
/ hposition = list.hpos.nextvalue
</picture>

> "In the AltPresentation trial it should then always select the not-chosen version of ZR_ or NA_. So, if in the Such-trial ZR1 is chosen, it should choose AZ1 and ZR1k+BZ1 in the Such-trial as well, and then choose NA1 in the AltPresentation"
#5: What is the relationship between your "Such-trial" and "AltPresentation"? Since there is nothing in the code that actually runs these trials (i.e. a <block>), it's not clear at all how those trials do or do not interrelate.

The shorter description you posted previously seems clearer ("[...] we have two versions of each Zielreiz and Suchbild - ZR and NA. It should randomly select from either ZR or NA item but if it selects item 1 from ZR it should never select item 1 from NA."), but I'm unable to reconcile the earlier description with the current one.

Basically, it's not at all obvious to me what is supposed to be paired to what. Please clarify.

Sorry for the confusing description. The code is only half-finished because I don't know how to continue, so it's definitely not clear from the code, you are right. I think I got a typo in there with #2 and #3. It should be: if ZR1 is selected, ZR1k must also always be selected (consecutively BZ1).

What I'm trying to do is the following:

Choose one item of either <item Zielreiz> or <Item Alternative>. If it chooses item 1 of Zielreiz (ZR1), it should then choose the according (numberwise) item of SuchbildZ, either version A (AZ1) or version B (BZ1), and the other, not chosen version (A or B), must be chosen at some point in the trial, too after ZR1k (which is just a copy of ZR1) was chosen. At the same time ZR1 should be eliminated as an option to choose from in <trial AltPresentation> and NA1 together with AN1 and BN1 should be eliminated as options in <trial Suche>. So it's always either ZR1 or NA1; ZR2 or NA2...
Later, in <trial Wahl> there should always be one picture which was chosen in <trial Suche> next to one picture which was chosen in <trial AltPresentation> but never two pictures which were chosen in the same trial. And I need to know which picture was chosen for which trial, so this must be possible to see in the data.

So, designwise following interrelations arise:

-  ZR1=ZR1k  (I did the copies because the same picture in <trial altpresentation> has to be shown twice (with a fixation cross in the middle) and it seemed the easiest solution I was actually able to programm at that time).
- ZR1=ZR1k are linked to AZ1 and BZ1. NA1=NA1k are linked to AN1 and BN1. NA2=NA2k are linked to AN2 and BN2, and so on. 
- If ZR1 is chosen in <trial Suche>, ZR1k must also be chosen in <trial Suche>.
- ZR1 and NA1 are counterparts - if ZR1 is chosen in <trial Suche>, NA1 should be eliminated for <trial Suche> but be presented in <trial AltPresentation>.
= Everything from <item Zielreiz> and <item Alternative> chosen for <trial Suche> should be eliminated as an option to choose from in <trial AltPresentation>
- In <trial Wahl> there must always be one of <item Zielreiz> or <item Alternative> which was chosen in <trial Suche> next to one of <item Zielreiz> or <item Alternative> chosen in <trial AltPresentation>.
- Only one of ZR1 or ZR1k should be chosen in <trial Wahl> (There's probably an easier way to show the same picture twice in <trial AltPresentation> where I can delete all the ZR_k copies I assume?)

So what I basically need is a "if...then not.." and work with differents lists, right? I'm totally lost since I'm an absolute beginner :S

The other option is always doing different groups where it is clear if ZR1 or NA1 is chosen in the <trial Wahl>, maybe that's just the better alternative here?

I'm having trouble with the terminology here, because it's potentially ambiguous:

"Choose one item of either <item Zielreiz> or <Item Alternative>. If it chooses item 1 of Zielreiz (ZR1), it should then choose the according (numberwise) item of SuchbildZ, either version A (AZ1) or version B (BZ1), and the other, not chosen version (A or B), must be chosen at some point in the trial, too after ZR1k (which is just a copy of ZR1) was chosen."

What do you consider to be a "trial" in the above?

At some point in the <trial Suche>, so if ZR1 or ZR1k is chosen in <trial Suche>, the other one (ZR1 or ZR1k) must also be chosen in <trial Suche>