Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
keeping track of response choice locations?
I am working on a comprehension experiment where participants see video recorded sentences and choose their response to the sentence with a button press. I would like to keep track of whether the response they choose is on the left of right side of the screen. 
Right now, I have button location randomized automatically. I can of course undo that and instead create pseudo randomization in my stimulus list by hand. But before I go that route, I thought I'd ask if there is a way keep the randomization and obtain information about response choice locations in the output?
Hi! We do not have an existing feature that supports this functionality, but a workaround is possible using the mouse_position response in tandem with the choice response. Response choice locations can be inferred on the basis of the mouse tracking data collected. 

What you would need to do is to add a mouse_position response to each trial containing one of these choice responses. Make sure that the sampling rate is relatively fine-grained to capture last minute mouse movements. Once the data are collected, you could identify the final position of the mouse before a selection was made and reclassify its position as 'left' or 'right'. If you set the reference_point to "center", then negative X coordinates will indicate selection of the choice on the left, and positive X coordinates will indicate selection of the choice on the right. As usual, "response_value" will record which choice was selected. 

For an example, see the below output from our Mouse-tracking Tutorial. The green row indicates the last mouse position recorded before a selection was made. The coordinates (-.25, .06) indicate that the choice that appeared on the left ("Mammal", next line) was selected. 


This method isn't 100% foolproof, as it is theoretically possible for a participant to jump move the mouse away from the button they select, and this information may be recorded for a few ms before the trial advances. However, we think this scenario is likely to be relatively rare. 

Please let me know if you have any questions about this. You might find the mouse tracking tutorial useful. It does caution against using locations = "random", but that is primarily because this information is not immediately discernible in the data file. Feel free to report back on whether this solution works for you! 


Forum Jump:

Users browsing this thread: 1 Guest(s)