Millisecond Forums

A simple /timeout in <trial recall_square> won't do because said trial element is repeated over and over until the subject clicks exit. Look at the /branch:

/ branch = [if(trial.recall_square.response=="exit")trial.square_feedback else trial.recall_square]

What you need to do is something like summing up the latencies for each run of the trial and then branch along the lines of the above if the sum exceeds a certain threshold. In any case, please make sure you first work through the original script and develop a sufficient understanding of its mechanics prior to introducing any modifications.

Regards,

~Dave

FWIW, if you wanted to limit the max. recall time allotted per each square (i.e., each run of <trial recall_square>) instead of limiting the overall recall time allotted for a given set of squares, the modifications are quite simple:

<trial recall_square>
/ ontrialbegin = [values.recallcount+=1; values.squarenumber+=1]
/ ontrialbegin = [if(values.recallcount==1)values.recalldelay=500 else values.recalldelay=0]
/ pretrialpause = values.recalldelay
/ stimulusframes = [1=WhiteScreen, matrixbg, R1C1, R1C2, R1C3, R1C4, R2C1, R2C2, R2C3, R2C4,
    R3C1, R3C2, R3C3, R3C4, R4C1, R4C2, R4C3, R4C4, _, clear, exit, recallprompt, squarestrings]
/ validresponse = (R1C1, R1C2, R1C3, R1C4, R2C1, R2C2, R2C3, R2C4, R3C1, R3C2, R3C3, R3C4, R4C1, R4C2, R4C3, R4C4, _, clear, exit)
/ monkeyresponse = (R1C1, R1C2, R1C3, R1C4, exit)
/ ontrialend = [if(trial.recall_square.response!="exit" && trial.recall_square.response!="clear")
    {item.RecalledSquares.item=trial.recall_square.response}]
/ ontrialend = [if(trial.recall_square.response=="clear"){clear(item.RecalledSquares); values.squarenumber=0;
    text.R1C1.textbgcolor=(white); text.R1C2.textbgcolor=(white); text.R1C3.textbgcolor=(white);
    text.R1C4.textbgcolor=(white); text.R2C1.textbgcolor=(white); text.R2C2.textbgcolor=(white);
    text.R2C3.textbgcolor=(white); text.R2C4.textbgcolor=(white); text.R3C1.textbgcolor=(white);
    text.R3C2.textbgcolor=(white); text.R3C3.textbgcolor=(white); text.R3C4.textbgcolor=(white);
    text.R4C1.textbgcolor=(white); text.R4C2.textbgcolor=(white); text.R4C3.textbgcolor=(white);
    text.R4C4.textbgcolor=(white);
    text.R1C1.item.1=" "; text.R1C2.item.1=" "; text.R1C3.item.1=" "; text.R1C4.item.1=" ";
    text.R2C1.item.1=" "; text.R2C2.item.1=" "; text.R2C3.item.1=" "; text.R2C4.item.1=" ";
    text.R3C1.item.1=" "; text.R3C2.item.1=" "; text.R3C3.item.1=" "; text.R3C4.item.1=" ";
    text.R4C1.item.1=" "; text.R4C2.item.1=" "; text.R4C3.item.1=" "; text.R4C4.item.1=" "}]
/ ontrialend = [if(trial.recall_square.response=="R1C1"){text.R1C1.item.1=values.squarenumber; text.R1C1.textbgcolor=(red)};
    if(trial.recall_square.response=="R1C2"){text.R1C2.item.1=values.squarenumber; text.R1C2.textbgcolor=(red)};
    if(trial.recall_square.response=="R1C3"){text.R1C3.item.1=values.squarenumber; text.R1C3.textbgcolor=(red)};
    if(trial.recall_square.response=="R1C4"){text.R1C4.item.1=values.squarenumber; text.R1C4.textbgcolor=(red)};
    if(trial.recall_square.response=="R2C1"){text.R2C1.item.1=values.squarenumber; text.R2C1.textbgcolor=(red)};
    if(trial.recall_square.response=="R2C2"){text.R2C2.item.1=values.squarenumber; text.R2C2.textbgcolor=(red)};
    if(trial.recall_square.response=="R2C3"){text.R2C3.item.1=values.squarenumber; text.R2C3.textbgcolor=(red)};
    if(trial.recall_square.response=="R2C4"){text.R2C4.item.1=values.squarenumber; text.R2C4.textbgcolor=(red)};
    if(trial.recall_square.response=="R3C1"){text.R3C1.item.1=values.squarenumber; text.R3C1.textbgcolor=(red)};
    if(trial.recall_square.response=="R3C2"){text.R3C2.item.1=values.squarenumber; text.R3C2.textbgcolor=(red)};
    if(trial.recall_square.response=="R3C3"){text.R3C3.item.1=values.squarenumber; text.R3C3.textbgcolor=(red)};
    if(trial.recall_square.response=="R3C4"){text.R3C4.item.1=values.squarenumber; text.R3C4.textbgcolor=(red)};
    if(trial.recall_square.response=="R4C1"){text.R4C1.item.1=values.squarenumber; text.R4C1.textbgcolor=(red)};
    if(trial.recall_square.response=="R4C2"){text.R4C2.item.1=values.squarenumber; text.R4C2.textbgcolor=(red)};
    if(trial.recall_square.response=="R4C3"){text.R4C3.item.1=values.squarenumber; text.R4C3.textbgcolor=(red)};
    if(trial.recall_square.response=="R4C4"){text.R4C4.item.1=values.squarenumber; text.R4C4.textbgcolor=(red)};]
/ ontrialend = [if(trial.recall_square.response=="exit" || trial.recall_square.response=="0"){
    text.R1C1.textbgcolor=(white); text.R1C2.textbgcolor=(white); text.R1C3.textbgcolor=(white);
    text.R1C4.textbgcolor=(white); text.R2C1.textbgcolor=(white); text.R2C2.textbgcolor=(white);
    text.R2C3.textbgcolor=(white); text.R2C4.textbgcolor=(white); text.R3C1.textbgcolor=(white);
    text.R3C2.textbgcolor=(white); text.R3C3.textbgcolor=(white); text.R3C4.textbgcolor=(white);
    text.R4C1.textbgcolor=(white); text.R4C2.textbgcolor=(white); text.R4C3.textbgcolor=(white);
    text.R4C4.textbgcolor=(white);
    text.R1C1.item.1=" "; text.R1C2.item.1=" "; text.R1C3.item.1=" "; text.R1C4.item.1=" ";
    text.R2C1.item.1=" "; text.R2C2.item.1=" "; text.R2C3.item.1=" "; text.R2C4.item.1=" ";
    text.R3C1.item.1=" "; text.R3C2.item.1=" "; text.R3C3.item.1=" "; text.R3C4.item.1=" ";
    text.R4C1.item.1=" "; text.R4C2.item.1=" "; text.R4C3.item.1=" "; text.R4C4.item.1=" "}]
/ responsemessage = (clear, clickclear, 150)
/ responsemessage = (exit, clickexit, 150)
/ responsemessage = (_, click_, 150)
/ branch = [if(trial.recall_square.response=="exit" || trial.recall_square.response=="0")trial.square_feedback else trial.recall_square]
/ recorddata = true
/ timeout = 5000
</trial>

Regards,

~Dave

Hey Dave,

Could you provide an example to start off with regarding summing up latency then branching? Would a good way be to create a new value which would update on every trial?

Another solution I considered would be to branch if after a certain amount of time (trial.sometrial.latency >= 10000) the subject hasn't clicked exit yet. Would it be possible to code an automatic exit click after 10000ms if the subject fails to? 

Yes. See e.g. for a somewhat related example: http://www.millisecond.com/forums/Topic6989.aspx#6989

That's unnecessary. The /timeout attribute alone already does that.