Millisecond Forums

Thanks a lot! Fast and efficient, as usual :)

Nevermind. Here's the problem:

<list randommean>
/ poolsize = 20
</list>

The list has 20 items. A <list>'s default behavior is to allow for *one* selection per trial, i.e. /selectionrate=trial. What this does then

<block CACB>
/ onblockbegin = [...; 
values.meanA=list.randommean.nextindex; values.meanA2=list.randommean.nextindex;values.meanA3=list.randommean.nextindex;values.meanA4=list.randommean.nextindex;values.meanA5=list.randommean.nextindex;values.meanA6=list.randommean.nextindex;values.meanA7=list.randommean.nextindex;values.meanA8=list.randommean.nextindex;values.meanA9=list.randommean.nextindex;values.meanA10=list.randommean.nextindex;
...
values.meanB=list.randommean.nextindex; values.meanB2=list.randommean.nextindex; values.meanB3=list.randommean.nextindex; values.meanB4=list.randommean.nextindex; values.meanB5=list.randommean.nextindex; values.meanB6=list.randommean.nextindex; values.meanB7=list.randommean.nextindex; values.meanB8=list.randommean.nextindex; values.meanB9=list.randommean.nextindex; values.meanB10=list.randommean.nextindex;
...;]

is effectively sample *a single item* from the list /onblockbegin. I.e. when the block starts actually running trials, there are only *19* unsampled items left in the list. The trials then sample from these 19 items without replacement, but since 20 trials sample from the list, the list has to reset: All 20 items are then available again for the final trial, and you are likely to sample an item that has been sampled previously.

Solution: Remove the unnecessary /onblockbegin sampling logic.

No worries! Here is a datafile. See line 252 and 262 (same picture as CS, with two different trialcode). The repeat option seemed to work.

Okay, thanks. Can you show me some actual data that shows "one CS [...] selected 2 times" in the respective block?

NB: This isn't supposed to imply that I don't believe you. I do believe you. I hope that this data will give me some more hints as to where to look for mistakes.

I don't understand your objection. <trial CAPic2> is among those 20 trials you pointed out, so of course this mistake matters:
<trial CAPic2>
/ ontrialbegin = [
values.csAPosition = list.csAPosition.nextvalue;
values.ucPosition = expressions.ucAPosition;
values.meanA2=list.randommean.nextindex; values.ucAPic2=list.randomucAPic.nextindex;
]
/ stimulustimes = [0=erase; 1=meanA2, ucAPic]
/ trialduration = values.exposureDuration
</trial>
/ trials = [1=blockinstructions; 2-21 = noreplace (CAPic,CAPic2,CAPic3,CAPic4,CAPic5,CAPic6,CAPic7,CAPic8,CAPic9, CAPic10, CBPic,CBPic2,CBPic3,CBPic4,CBPic5,CBPic6,CBPic7,CBPic8,CBPic9, CBPic10); 

Yes, but that would impact US selection and not CS selection (i.e. two different CSs may be paired with the same US, but there is no reason why one CS would be selected 2 times)

<block CACB>
/ onblockbegin = [values.currentBlock += 1; 
values.meanA=list.randommean.nextindex; values.meanA2=list.randommean.nextindex;values.meanA3=list.randommean.nextindex;values.meanA4=list.randommean.nextindex;values.meanA5=list.randommean.nextindex;values.meanA6=list.randommean.nextindex;values.meanA7=list.randommean.nextindex;values.meanA8=list.randommean.nextindex;values.meanA9=list.randommean.nextindex;values.meanA10=list.randommean.nextindex;
values.ucAPic=list.randomucAPic.nextindex;values.ucAPic2=list.randomucAPic.nextindex;values.ucAPic3=list.randomucAPic.nextindex;values.ucAPic4=list.randomucAPic.nextindex;values.ucAPic5=list.randomucAPic.nextindex;values.ucAPic6=list.randomucAPic.nextindex;values.ucAPic7=list.randomucAPic.nextindex;values.ucAPic8=list.randomucAPic.nextindex;values.ucAPic9=list.randomucAPic.nextindex;values.ucAPic10=list.randomucAPic.nextindex;
values.meanB=list.randommean.nextindex; values.meanB2=list.randommean.nextindex; values.meanB3=list.randommean.nextindex; values.meanB4=list.randommean.nextindex; values.meanB5=list.randommean.nextindex; values.meanB6=list.randommean.nextindex; values.meanB7=list.randommean.nextindex; values.meanB8=list.randommean.nextindex; values.meanB9=list.randommean.nextindex; values.meanB10=list.randommean.nextindex;
values.ucBPic=list.randomucBPic.nextindex;values.ucBPic2=list.randomucBPic.nextindex;values.ucBPic3=list.randomucBPic.nextindex;values.ucBPic4=list.randomucBPic.nextindex;values.ucBPic5=list.randomucBPic.nextindex;values.ucBPic6=list.randomucBPic.nextindex;values.ucBPic7=list.randomucBPic.nextindex;values.ucBPic8=list.randomucBPic.nextindex;values.ucBPic9=list.randomucBPic.nextindex;values.ucBPic10=list.randomucBPic.nextindex;]
I don't quite understand why you perform those selections above /onblockbegin, only to perform new selections on a per-trial basis during the block, e.g.
<trial CAPic>
/ ontrialbegin = [
values.csAPosition = list.csAPosition.nextvalue;
values.ucPosition = expressions.ucAPosition;
values.meanA=list.randommean.nextindex; values.ucAPic=list.randomucAPic.nextindex;
]
/ stimulustimes = [0=erase; 1=meanA, ucAPic]
/ trialduration = values.exposureDuration
</trial>
I'm not sure this has anything to do with the problem under discussion, but I do have trouble grasping the reason (which is not to imply that there isn't a good reason).

Bad coding I'm afraid :/ but it didn't seem to affect the outcome. But I guess that if those two sets of instructions conflicted, everything would be randomized

Here's another thing you could maybe clarify for me:

<block CACB>
/ onblockbegin = [values.currentBlock += 1;
values.meanA=list.randommean.nextindex; values.meanA2=list.randommean.nextindex;values.meanA3=list.randommean.nextindex;values.meanA4=list.randommean.nextindex;values.meanA5=list.randommean.nextindex;values.meanA6=list.randommean.nextindex;values.meanA7=list.randommean.nextindex;values.meanA8=list.randommean.nextindex;values.meanA9=list.randommean.nextindex;values.meanA10=list.randommean.nextindex;
values.ucAPic=list.randomucAPic.nextindex;values.ucAPic2=list.randomucAPic.nextindex;values.ucAPic3=list.randomucAPic.nextindex;values.ucAPic4=list.randomucAPic.nextindex;values.ucAPic5=list.randomucAPic.nextindex;values.ucAPic6=list.randomucAPic.nextindex;values.ucAPic7=list.randomucAPic.nextindex;values.ucAPic8=list.randomucAPic.nextindex;values.ucAPic9=list.randomucAPic.nextindex;values.ucAPic10=list.randomucAPic.nextindex;
values.meanB=list.randommean.nextindex; values.meanB2=list.randommean.nextindex; values.meanB3=list.randommean.nextindex; values.meanB4=list.randommean.nextindex; values.meanB5=list.randommean.nextindex; values.meanB6=list.randommean.nextindex; values.meanB7=list.randommean.nextindex; values.meanB8=list.randommean.nextindex; values.meanB9=list.randommean.nextindex; values.meanB10=list.randommean.nextindex;
values.ucBPic=list.randomucBPic.nextindex;values.ucBPic2=list.randomucBPic.nextindex;values.ucBPic3=list.randomucBPic.nextindex;values.ucBPic4=list.randomucBPic.nextindex;values.ucBPic5=list.randomucBPic.nextindex;values.ucBPic6=list.randomucBPic.nextindex;values.ucBPic7=list.randomucBPic.nextindex;values.ucBPic8=list.randomucBPic.nextindex;values.ucBPic9=list.randomucBPic.nextindex;values.ucBPic10=list.randomucBPic.nextindex;]

I don't quite understand why you perform those selections above /onblockbegin, only to perform new selections on a per-trial basis during the block, e.g.

<trial CAPic>
/ ontrialbegin = [
values.csAPosition = list.csAPosition.nextvalue;
values.ucPosition = expressions.ucAPosition;
values.meanA=list.randommean.nextindex; values.ucAPic=list.randomucAPic.nextindex;
]
/ stimulustimes = [0=erase; 1=meanA, ucAPic]
/ trialduration = values.exposureDuration
</trial>

I'm not sure this has anything to do with the problem under discussion, but I do have trouble grasping the reason (which is not to imply that there isn't a good reason).

I don't understand your objection. <trial CAPic2> is among those 20 trials you pointed out, so of course this mistake matters:
<trial CAPic2>
/ ontrialbegin = [
values.csAPosition = list.csAPosition.nextvalue;
values.ucPosition = expressions.ucAPosition;
values.meanA2=list.randommean.nextindex; values.ucAPic2=list.randomucAPic.nextindex;
]
/ stimulustimes = [0=erase; 1=meanA2, ucAPic]
/ trialduration = values.exposureDuration
</trial>

/ trials = [1=blockinstructions; 2-21 = noreplace (CAPic,CAPic2,CAPic3,CAPic4,CAPic5,CAPic6,CAPic7,CAPic8,CAPic9, CAPic10, CBPic,CBPic2,CBPic3,CBPic4,CBPic5,CBPic6,CBPic7,CBPic8,CBPic9, CBPic10);

Thanks for your reply. I did not suspect the problem would come from ucAPic2, since it is supposed to code for US selection and not CSs.

The conditioning block is the CACB block, postratings are termed "postratings" and the memorisation block is termed "Block Exclusion". Block CACB present 20 trials with, each corresponding to one specific CS paired with one specific US, then each of the 20 trial is repeated several times (repeatCA, etc.). But I don't get how Block_Exclusion and Postratings manage to display the same 20CS taken from the preratings (which is what I want), and the conditioning block makes this one, consistent mistake in CS selection.

Thats's why I suspect the error comes from the conditioning block, but nothing seems odd : 

<block CACB>
/ onblockbegin = [values.currentBlock += 1; values.meanA=list.randommean.nextindex; values.meanA2=list.randommean.nextindex;values.meanA3=list.randommean.nextindex;values.meanA4=list.randommean.nextindex;values.meanA5=list.randommean.nextindex;values.meanA6=list.randommean.nextindex;values.meanA7=list.randommean.nextindex;values.meanA8=list.randommean.nextindex;values.meanA9=list.randommean.nextindex;values.meanA10=list.randommean.nextindex;
values.ucAPic=list.randomucAPic.nextindex;values.ucAPic2=list.randomucAPic.nextindex;values.ucAPic3=list.randomucAPic.nextindex;values.ucAPic4=list.randomucAPic.nextindex;values.ucAPic5=list.randomucAPic.nextindex;values.ucAPic6=list.randomucAPic.nextindex;values.ucAPic7=list.randomucAPic.nextindex;values.ucAPic8=list.randomucAPic.nextindex;values.ucAPic9=list.randomucAPic.nextindex;values.ucAPic10=list.randomucAPic.nextindex;
values.meanB=list.randommean.nextindex; values.meanB2=list.randommean.nextindex; values.meanB3=list.randommean.nextindex; values.meanB4=list.randommean.nextindex; values.meanB5=list.randommean.nextindex; values.meanB6=list.randommean.nextindex; values.meanB7=list.randommean.nextindex; values.meanB8=list.randommean.nextindex; values.meanB9=list.randommean.nextindex; values.meanB10=list.randommean.nextindex;
values.ucBPic=list.randomucBPic.nextindex;values.ucBPic2=list.randomucBPic.nextindex;values.ucBPic3=list.randomucBPic.nextindex;values.ucBPic4=list.randomucBPic.nextindex;values.ucBPic5=list.randomucBPic.nextindex;values.ucBPic6=list.randomucBPic.nextindex;values.ucBPic7=list.randomucBPic.nextindex;values.ucBPic8=list.randomucBPic.nextindex;values.ucBPic9=list.randomucBPic.nextindex;values.ucBPic10=list.randomucBPic.nextindex;]>
/ trials = [1=blockinstructions; 2-21 = noreplace (CAPic,CAPic2,CAPic3,CAPic4,CAPic5,CAPic6,CAPic7,CAPic8,CAPic9, CAPic10, CBPic,CBPic2,CBPic3,CBPic4,CBPic5,CBPic6,CBPic7,CBPic8,CBPic9, CBPic10);
22-130 = noreplace(repeatCA,repeatCA,repeatCA,repeatCA,repeatCA2,repeatCA2,repeatCA2,repeatCA2,repeatCA3,repeatCA3,repeatCA3,repeatCA3,repeatCA4, repeatCA4,repeatCA4,repeatCA4,repeatCA5,
repeatCA5,repeatCA5,repeatCA5,repeatCA6,repeatCA6,repeatCA6,repeatCA6,repeatCA7,repeatCA7,repeatCA7,repeatCA7,repeatCA8,repeatCA8,repeatCA8,repeatCA8,repeatCA9,repeatCA9,repeatCA9,repeatCA9,repeatCA10, repeatCA10,repeatCA10,repeatCA10,repeatCA10,repeatCA10,repeatCA10,repeatCA10,repeatCA10,repeatCA10,
repeatCB,repeatCB,repeatCB,repeatCB,repeatCB2,repeatCB2,repeatCB2,repeatCB2,repeatCB3,repeatCB3,repeatCB3,repeatCB3,repeatCB4,repeatCB4,repeatCB4,repeatCB4,
repeatCB5,repeatCB5,repeatCB5,repeatCB5,repeatCB6,repeatCB6,repeatCB6,repeatCB6,repeatCB7,repeatCB7,repeatCB7,repeatCB7,repeatCB8, repeatCB8,repeatCB8,repeatCB8,repeatCB9,repeatCB9,repeatCB9,repeatCB9, repeatCB10,repeatCB10,repeatCB10,repeatCB10,repeatCB10,repeatCB10,repeatCB10,repeatCB10,repeatCB10,repeatCB10,
target,target,target,target,target,target,target,target,target,target,target,distractorPic,distractorPic,distractorPic,distractorPic,distractorPic,distractorPic,distractorPic,distractorPic,distractorPic,distractorPic)
]
</block>

It would be really helpful if you could reference the respective elements you are talking about by their names as they appear in the script. As it stands, it is relatively difficult to figure out what to look at. Just glancing over the code, I would suspect there are mistakes. For example:

<trial CAPic2>
/ ontrialbegin = [
values.csAPosition = list.csAPosition.nextvalue;
values.ucPosition = expressions.ucAPosition;
values.meanA2=list.randommean.nextindex; values.ucAPic2=list.randomucAPic.nextindex;
]
/ stimulustimes = [0=erase; 1=meanA2, ucAPic] <--- Shouldn't this be ucAPic2???
/ trialduration = values.exposureDuration
</trial>

<trial repeatCA2>
/ ontrialbegin = [
values.csAPosition = list.csAPosition.nextvalue;
values.ucPosition = expressions.ucAPosition;  ]
/ stimulustimes = [0=erase; 1= meanA2, ucAPic2]
/ validresponse = (57)
/ trialduration = values.exposureDuration
</trial>