NULL results, ignoring trials, and error messages

[meri]

Hi

Right now, I am working with a vowel-picture matching experiment. I am on my way, but I still have some questions about the technical issues:

1) Whenever I collect the results, the very last column in EVERY single line show the “NULL” result. Why do I get it?
2) What is the purpose with logging the buffer of the audio file? How important is it in my kind of experiment, where the focus of my experiment is the participants’ correspondence between sound and figure)?
3) Sometimes, when I opened my experiment in a new tab, some trials were ignored – this happened even if I did everything according to the instructions (I used the Basic and Advanced tutorials as a tool). Why does PCIbex sometimes ignore my trials?
4) Even though I made several types of mistakes, the only error message I get whenever I open the experiment in a new tab is “ERROR: There must be some items in the running order!”. Why do we get one kind of message for every different kind of mistake? This lack of specification makes it difficult to find out what problem should be solved and how…

Best regards,
Meri

[meri]

Here is the link 🙂

https://farm.pcibex.net/r/tYGUJl/

[Jeremy]

Hi Meri,

1) If you look at the comments in your results file (lines starting with #) you will see that the last column’s name is “Comments”: in your case, you get NULL in every line because there’s no comments to report. It’s not always the case: for example, if you use log on a Scale element without wait and the trial ends without the participant making any selection on the scale, you will get a line where that column reads “No selection happened”

2) One or multiple buffering events happening in the middle of a trial could indicate that the participant encountered difficulty playing back the audio, which could matter for some time sensitive tasks like primed lexical recognition. If that’s not relevant in your case, then you don’t have to log that information

3) I’m not sure what you are referring to. I tried opening your experiment multiple times, running as different groups, and the Sequence tab of the debugger always showed 25 trials. Maybe that used to happen in an earlier version of your project that had an extra value in the group column of the CSV table, or where there was a typo somewhere in the script that prevented part of the code from being evaluated?

4) This error message comes from the original Ibex: whenever an error prevents the script from filling the shuffleSequence variable (and there are many types of errors that can have that consequence, particularly syntax typos like missing a comma somewhere, which make the javascript parser crash) then Ibex complains that there’s no items in the running order. PennController adds a layer trying to catch as many errors as it can detect before Ibex kicks in, but some errors (again, particularly syntax errors) are harder to handle

Best,
Jeremy

[Author]

Posts