Millisecond Forums

A <surveypage> is a special type of <trial>. A <survey> is a special type of <block>.

/branch attributes *must* branch to an element of the same type as the element containing the /branch. I.e.,

- A <trial> can /branch to another <trial> (<openended>, <likert> and <surveypage> are types of <trial> elements). A <trial> *cannot* /branch to a <block>.
- A <block> can /branch to another <block>. It *cannot* /branch to a <trial>.

<expt>
/ blocks = [1=myblock]
</expt>

<block myblock>
/ trials = [1=mytrial]
/ branch = [if (values.myvalue) noreplace(block.a, block.b)]
</block>

<block a>
/ trials = [1=mytrial]
</block>

<block b>
/ trials = [1=mytrial]
</block>

<trial mytrial>
/ stimulusframes = [1=mytext]
/ validresponse = (57)
</trial>

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

<values>
/ myvalue = true
</values>

I feel like an idiot but I actually didn't know that. Thanks so much for clarifying! Now we should be able to get this right...

It's not necessarily obvious, there really isn't any reason to feel bad. Let me know if can't get it to work (note: you can attach files to a thread via the +Insert menu).

Thank you. Indeed another problem has come up, unfortunately:

The branch command now works after we placed it in the survey elements, but the branch doesn't lead where we need it to lead. Specifically: We have three shapes (circle, square, triangle) and three different picture stimuli (positive, negative and fear specific). Every participant is assigned to one specific condition where, say, circle is associated with positive stimuli, a square is associated with negative stimuli and a triangle is associated with fear specific stimuli. These associations are randomized, hence 6 different conditions + another 6 to vary the sizes of the shapes. This is made complicated by the questionaires that are presented to the participants after each shape/stimulus association has been tested through our decision task. Every participant sees the same questionaire, but afterwards they have to return to their respective condition. I.e. if they learned an association between circle and positive stimuli, they mustn't see an association between square and positive stimuli afterwards; for every participant all shapes need to have different stimuli attached to them so all stimuli occur once.

We had hoped to achieve this by assigning one of 12 numbers to each participant according to the condition they had been placed in. We used this values command:

<values>
/ Bedingungsvariable = true
</values>

<trial Zähler_1>
/timeout = 1
/ ontrialend= [values.bedingungsvariable = 1]
</trial>

<trial Zähler_2>
/timeout = 1
/ ontrialend= [values.bedingungsvariable = 2]
</trial>

<trial Zähler_3>
/timeout = 1
/ ontrialend= [values.bedingungsvariable = 3]
</trial>

<trial Zähler_4>
/timeout = 1
/ ontrialend= [values.bedingungsvariable = 4]
</trial>

<trial Zähler_5>
/timeout = 1
/ ontrialend= [values.bedingungsvariable = 5]
</trial>

<trial Zähler_6>
/timeout = 1
/ ontrialend= [values.bedingungsvariable = 6]
</trial>

<trial Zähler_7>
/timeout = 1
/ ontrialend= [values.bedingungsvariable = 7]
</trial>

<trial Zähler_8>
/timeout = 1
/ ontrialend= [values.bedingungsvariable = 8]
</trial>

<trial Zähler_9>
/timeout = 1
/ ontrialend= [values.bedingungsvariable = 9]
</trial>

<trial Zähler_10>
/timeout = 1
/ ontrialend= [values.bedingungsvariable = 10]
</trial>

<trial Zähler_11>
/timeout = 1
/ ontrialend= [values.bedingungsvariable = 11]
</trial>

<trial Zähler_12>
/timeout = 1
/ ontrialend= [values.bedingungsvariable = 12]
</trial>

The respective trial was then included into the first learning block of each condition. When branching from the questionaire two the second learning block, we used the code that I already quoted to assign them to the right block according to this value:

<survey DAS>
/ pages = [1=DAS1; 2=DAS2; 3=DAS3; 4=DAS4; 5=DAS5; 6=DAS6; 7=DAS7; 8=DAS8]
/ subcaptionfontstyle = ("Arial", 13pt, true, false, false, false, 5, 0)

/ fontstyle = ("Arial", 13pt, true, false, false, false, 5, 0)
/ responsefontstyle = ("Arial", 11pt, false, false, false, false, 5, 0)
/ itemspacing = 2%
/ showpagenumbers = false
/ showquestionnumbers = false
/ showbackbutton = false
/ nextlabel = "Weiter"
/ finishlabel = "Weiter"
/ branch = [if (values.bedingungsvariable == 1) noreplace (block.lernen1b, block.lernen2b)]
/ branch = [if (values.bedingungsvariable == 2) noreplace (block.lernen1b, block.lernen2b)]
/ branch = [if (values.bedingungsvariable == 3) noreplace (block.lernen3b, block.lernen4b)]
/ branch = [if (values.bedingungsvariable == 4) noreplace (block.lernen3b, block.lernen4b)]
/ branch = [if (values.bedingungsvariable == 5) noreplace (block.lernen5b, block.lernen6b)]
/ branch = [if (values.bedingungsvariable == 6) noreplace (block.lernen5b, block.lernen6b)]
/ branch = [if (values.bedingungsvariable == 7) noreplace (block.lernen7b, block.lernen8b)]
/ branch = [if (values.bedingungsvariable == 8) noreplace (block.lernen7b, block.lernen8b)]
/ branch = [if (values.bedingungsvariable == 9) noreplace (block.lernen9b, block.lernen10b)]
/ branch = [if (values.bedingungsvariable == 10) noreplace (block.lernen9b, block.lernen10b)]
/ branch = [if (values.bedingungsvariable == 11) noreplace (block.lernen11b, block.lernen12b)]
/ branch = [if (values.bedingungsvariable == 12) noreplace (block.lernen11b, block.lernen12b)]
</survey>

The problem is simply that it doesn't seem to work. The experiment runs through, but the test monkey sees conditions that shouldn't be seen in combination, i.e. circle/positive and then triangle/positive (when it should be triangle/negative or triangle/specific). I will attach our whole file this time; unfortunately the code is quite convoluted since we didn't have much time to build this experiment. Also, a lot of the names and comments are in German :/

In one of your <survey>s, the /branch attributes are simply wrong.

<survey STAIT>
/ pages = [1=STAIT1; 2=STAIT2]
/ subcaptionfontstyle = ("Arial", 13pt, true, false, false, false, 5, 0)
/ fontstyle = ("Arial", 13pt, true, false, false, false, 5, 0)
/ responsefontstyle = ("Arial", 11pt, false, false, false, false, 5, 0)
/ itemspacing = 2%
/ showpagenumbers = false
/ showquestionnumbers = false
/ nextlabel = "Weiter"
/ showbackbutton = false
/ finishlabel = "Weiter"
/ branch = [if (values.bedingungsvariable = 1) noreplace (block.lernen1c, block.lernen2c)]
/ branch = [if (values.bedingungsvariable = 2) noreplace (block.lernen1c, block.lernen2c)]
/ branch = [if (values.bedingungsvariable = 3) noreplace (block.lernen3c, block.lernen4c)]
/ branch = [if (values.bedingungsvariable = 4) noreplace (block.lernen3c, block.lernen4c)]
/ branch = [if (values.bedingungsvariable = 5) noreplace (block.lernen5c, block.lernen6c)]
/ branch = [if (values.bedingungsvariable = 6) noreplace (block.lernen5c, block.lernen6c)]
/ branch = [if (values.bedingungsvariable = 7) noreplace (block.lernen7c, block.lernen8c)]
/ branch = [if (values.bedingungsvariable = 8) noreplace (block.lernen7c, block.lernen8c)]
/ branch = [if (values.bedingungsvariable = 9) noreplace (block.lernen9c, block.lernen10c)]
/ branch = [if (values.bedingungsvariable = 10) noreplace (block.lernen9c, block.lernen10c)]
/ branch = [if (values.bedingungsvariable = 11) noreplace (block.lernen11c, block.lernen12c)]
/ branch = [if (values.bedingungsvariable = 12) noreplace (block.lernen11c, block.lernen12c)]
</survey>

You need to use == (a comparison operator) in the condition check, not = (the assignment operator).

Thanks again, that did the trick. There actually were some other minor mistakes in the code that kept us from finding the right solution when we played around with the =-operators prior to coming here. Good to know what = and == actually mean, we couldn't find that in the help file or the inquisit language reference.

Thanks a lot for your help, this whole day would have been a huge waste without you!

No problem! FYI: You'll find the meaning of =, == and other operators covered in the "Operators" topic in the Inquisit documentation.