How to interpret the decrease of intensity from PsychoPy during a staircase?

Hello !

I created a 3 down and 1 up staircase design, including the following parameters:

  • nReps 30
  • start value = 1 (corresponding to 60dB)
  • min value = 0
  • step size = (1,1,0.5)
  • step type = db
  • 7 reversals

Participants have to identify two speech sounds during the experiment. The task runs correctly however, I am a little bit confused about the data file. More specifically, I am not sure how to interpret the column D corresponding to “staircase_loop.intensity” (see the file “example file”). example file.xlsx (14.6 KB)

→ Does the decrease of the intensity from PsychoPy corresponds to a decrease in percentage of the initial volume of the sound ? (see D69 in the excel file)

→ Does the decrease of the intensity from PsychoPy corresponds to a decrease of the power signal from P1 (start value =1) to P2 (identification threshold = D54 in the excel file) ? If so, can I use the logarithmic formula 10*log (P2/P1) to convert the decrease of the intensity from PsychoPy in dB ? (like I did in D57 in the excel file).

I don’t have any SPL meter with me, so it’s complicated to be sure about the identification threshold that I have…

Thank you in advance for your feedback !

Hi Sarah,

What’s the max value you have allowed in your staircase? since sound takes a value between 0 and 1 I am wondering if the selected step sizes are appropriate (1, 1, 0.5) because that will mean a fluctuation between min and max values in the staircase, which likely won’t converge on a threshold.

What I would suggest doing are the following steps:

  • Using an SPL meter (if you want a rough approximation whilst you are out of the lab most smartphones have SPL apps available, though I’de double check with a proper device in the lab), check what a step size (e.g. a volume change of 0.1) corresponds to in db terms on your computer. This is an important step for calibration, so that you know what values to expect on your setup. You can do this in PsychoPy, make a simple experiment with a sound component, where the volume of the sound increases by 0.1 on every iteration of the loop.
  • Once you know what a stepsize corresponds to, that will inform the step sizes you use in your staircase. The values in the staircase will be in “PsychoPy volume” terms, where 0 is min and 1 is max, but you will know what these are in dB. You’ll probably want something like; nReps = 50 (max of 50 trials), start value 0.5 (starting at a more central point, so long as it is suprathreshold), min value = 0, max value = 1 (or whatever is a comfortable/safe level on your device), step size = [0.2, 0.1, 0.5, 0.01] (i.e. much smaller values than you currently have, so you have a better chance of converging on a threshold.