Millisecond Forums

<trial ab>
...
/ ontrialend = [values.blocked_1 = 3; values.blocked_2 = 4]
...
</trial>

as in

<block myblock>
/ trials = [1-24=learningtrialselection]
</block>

<trial learningtrialselection>
/ontrialbegin = [
values.counttrials_phase1 += 1;
values.nexttrial = list.trialselection_phase1.nextvalue;
]
/branch = [
if (values.nexttrial == 1) trial.ab
else if (values.nexttrial == 2) trial.ba
else if (values.nexttrial == 3) trial.bc
else if (values.nexttrial == 4) trial.cb
else if (values.nexttrial == 5) trial.cd
else if (values.nexttrial == 6) trial.dc
else if (values.nexttrial == 7) trial.de
else if (values.nexttrial == 8) trial.ed
]
/ trialduration = 0
/ recorddata = false
</trial>

<list trialselection_phase1>
/ items = (1, 2, 3, 4, 5, 6, 7, 8)
/ poolsize = 24
/ not = (values.blocked_1)
/ not = (values.blocked_2)
</list>

<values>
/ nexttrial = 0
/ counttrials_phase1 = 0
/ blocked_1 = 0
/ blocked_2 = 0
</values>

- trials bc and cb never follow trials ab or ba,
<trial ab>
/ ontrialend = [values.blocked_1 = 3; values.blocked_2 = 4]
/ stimulusframes = [1=mytext]
/ validresponse = (0)
/ trialduration = 100
</trial>

<trial ba>
/ ontrialend = [values.blocked_1 = 3; values.blocked_2 = 4]
/ stimulusframes = [1=mytext]
/ validresponse = (0)
/ trialduration = 100
</trial>

- for trials cd and dc to never follow trials bc and cb,
<trial bc>
/ ontrialend = [values.blocked_1 = 5; values.blocked_2 = 6]
/ stimulusframes = [1=mytext]
/ validresponse = (0)
/ trialduration = 100
</trial>

<trial cb>
/ ontrialend = [values.blocked_1 = 5; values.blocked_2 = 6]
/ stimulusframes = [1=mytext]
/ validresponse = (0)
/ trialduration = 100
</trial>

- and for trials de and ed to never follow trials cd and dc.
<trial cd>
/ ontrialend = [values.blocked_1 = 7; values.blocked_2 = 8]
/ stimulusframes = [1=mytext]
/ validresponse = (0)
/ trialduration = 100
</trial>

<trial dc>
/ ontrialend = [values.blocked_1 = 7; values.blocked_2 = 8]
/ stimulusframes = [1=mytext]
/ validresponse = (0)
/ trialduration = 100
</trial>

<trial de>
/ ontrialend = [values.blocked_1 = 0; values.blocked_2 = 0]
/ stimulusframes = [1=mytext]
/ validresponse = (0)
/ trialduration = 100
</trial>

<trial ed>
/ ontrialend = [values.blocked_1 = 0; values.blocked_2 = 0]
/ stimulusframes = [1=mytext]
/ validresponse = (0)
/ trialduration = 100
</trial>

<text mytext>
/ items = ("<%script.currenttrial%>")
</text>

Hi Dave,

Thanks for your help. Currently the experiment is set up to select numbers randomly from the trialselection_phase1 list, and then specifies the type of trial based on those numbers.

<trial learningtrialselection>
/ontrialbegin = [
values.counttrials_phase1 += 1;
values.nexttrial = list.trialselection_phase1.nextvalue;
]
/branch = [
if (values.nexttrial == 1) trial.ab
else if (values.nexttrial == 2) trial.ba
else if (values.nexttrial == 3) trial.bc
else if (values.nexttrial == 4) trial.cb
else if (values.nexttrial == 5) trial.cd
else if (values.nexttrial == 6) trial.dc
else if (values.nexttrial == 7) trial.de
else if (values.nexttrial == 8) trial.ed
]

Do you know how I could specify the conditions for trial ab specifically like you mentioned above?

This example might be worth looking at: https://www.millisecond.com/forums/FindPost20200.aspx

While it concerns stimulus selection (opposed to trial selection as in your case) and the constraint is slightly different, the general approach -- storing categories "blocked" from selection in <values>, using /not constraints -- is the same.

Thanks for the clarification. Then I would recommend specifying two /not constraints in your <list>

<list trialselection_phase1>
/ items = (1, 2, 3, 4, 5, 6, 7, 8)
/ poolsize = 24
/ not = [values.blocked_1]
/ not = [values.blocked_2]
</list>

and setting values.blocked_1 and values.blocked_2 to the numerical trial identifiers that are impermissible to follow a given <trial>-type. I.e.

<trial ab>
...
/ ontrialend = [values.blocked_1 = 3; values.blocked_2 = 4]
...
</trial>

meaning the values representing trial.bc (3) and trial.cb (4) would be blocked from being selected for the next trial and so forth. Does this make sense or need I elaborate more?

Hi Dave,

Thanks for your reply - I meant to ask about Inquisit 5.

Thanks.

Given that you use <parameters> in your <list>'s /poolsize, you are designing this in Inquisit 5 apparently. Since you posted this to the Inquisit 4 forum, however, I have to ask whether you plan on administering the script via Inquisit 5 or Inqusit 4. This is importatnt because (a) Inquisit 4 syntax does not have a <parameters> element, and (b) what viable solution(s) may exist for your actual trial-selection problem will depend on whether you are using Inquisit 4 or Inquisit 5 as well.

Would you please clarify? Thanks.