Hi all,
I just noticed a bug with the slidertrial element that is affecting a planned study. I played around a bit and was able to reduce it to what I believe is a minimal reproducible example:
<expt debug>
/ blocks = [1 = nobug; 2 = bugrepro; 3 = bugrepro2]
</expt>
<slidertrial slider>
/ labels = ("Definitively false", "Neutral", "Definitively true")
</slidertrial>
<text trial>
/ items = ("I am a TRIAL")
</text>
<trial trial>
/ validresponse = (" ")
/ stimulustimes = [0 = trial]
</trial>
<block nobug>
/ trials = [1 = slider; 2 = trial; 3 = slider]
</block>
<text survey>
/ items = ("I am a SURVEYPAGE")
</text>
<surveypage survey>
/ stimulustimes = [0 = survey]
</surveypage>
<block bugrepro>
/ trials = [1 = slider; 2 = survey; 3 = slider]
</block>
<block bugrepro2>
/ trials = [1 = slider; 2 = survey; 3-10 = trial; 11 = slider]
</block>
On my system, the blocks bugrepro and bugrepro2 show the slider on the first presentation of slidertrial.slider, but they do not show it on its second presentation. I believe this is because a surveypage element is displayed between the two instances of slidertrial.slider, which is supported by the lack of an issue in the nobug trial.
Note that the bug persists even if the surveypage element is completely empty (no stimuli or questions); the text elements are only there for ease of understanding.
Is this a known issue? If not, is it likely to be fixed given that the current version is Inquisit 7? In any case, it might be helpful for others to know about this.
A workaround appears to be using a different slider element after displaying the surveypage, although this does not fix the issue for the original slider:
<block bugnorepro>
/ trials = [1 = slider; 2 = survey; 3 = slider2; 4 = slider]
</block>
<slidertrial slider>
/ labels = ("Definitively false", "Neutral", "Definitively true")
</slidertrial>
<slidertrial slider2>
/ labels = ("Definitively false", "Neutral", "Definitively true")
</slidertrial>
<text survey>
/ items = ("I am a SURVEYPAGE")
</text>
<surveypage survey>
/ stimulustimes = [0 = survey]
</surveypage>
As can be seen, slider2 is displayed correctly, but slider isn't the second time. The workaround may not be practical for structures that repeat a certain order of questions often; perhaps using a surveypage with a slider element instead of the slidertrial could be an effective alternative for those case (I haven't tried it out).