Reading results

[zoe_f]

Hi there,

I am making sure my results are read properly before I publish my study online. I notice two anomalies in the results files:

1) I have a KeyPress element in part of my experiment, that is logged. It appears that the keypress is logged fine (there is a column for this), but in the ‘comments’ column I see ‘failed key presses happened’, on every trial. Given that my key presses appear to be being logged fine, do I still need to be worried about this message? (I would guess that the message indicated the participant pressed a ‘non valid’ key during the trial, but it is in the comments for every single one of 144 trials, so I doubt this is what this message means).

2) In different parts of the experiment, I log different kinds of information (Keypress, rating, etc.). This has meant that in the results file, I see people’s data ending up in very strange columns. To give a concrete example, I have a questionnaire at the end with ratings. The ‘rating’ the participant gives ends up in a ‘keypress’ column for keypresses that I logged in trials earlier in the experiment – but not the questionnaire. While the ratings are being logged, so technically no information is lost, it would make my life a lot easier when I do the data analysis if logged results only ended up in the columnn where they are told to go (in the .log bit after the trial closing bracket).

Thanks!

Zoë

[Jeremy]

Hi Zoe,

1) The log command on a Key element (ie not the one on the closing parenthesis of newTrial), by default, reports a line in the results file for a keypress that validates a corresponding wait command—if you use log but you have no wait command, or if the Key element’s wait command was never validated (eg. because a Timer ended the trial early) then you’ll have a line in your results file that says “NA.” If the participant pressed a key anyway (hence without validating a wait command) you’ll have “(failed keypresses happened)” in the comments.
As long as your line reports the key that was actually pressed (and that you care about) then you shouldn’t worry about that

2) The newTrial().log commands append columns to the end of every result line for the corresponding trial, in the order in which they are used. If you have one newTrial with one log command, and another newTrial also with one log command, each trial’s result lines will append one additional column. If you want to use separate columns across trial, you need to keep track of your newTrial().log commands. E.g.:

newTrial( newKey("press","Enter").wait(),newVar("key").global().set(getKey("press")) )
  .log( "key" , getVar("key") )
  .log( "rating" , "NA" )

newTrial( newScale("scale", 10).print().wait(),newVar("rating").global().set(getScale("scale")) )
  .log( "key" , "NA" )
  .log( "rating" , getVar("rating") )

Jeremy

[Author]

Posts