Unwanted variation in presentation duration and delays related to number of times in a loop.

booradley

Hi folks,

I'm building a very straight forward paradigm with a practise block and 5 experimental blocks.

See here:

Four of the experimental blocks simply present text via Sketchpads sequentially with a fixed duration determined in keyboard_response.

I use Sketchpads instead of images of text so that we can make changes quickly etc if we need to:

I'm building the experiment using Windows 10 on a Surface Book 2.

I've noticed that when there a relatively few items in a block (such as the screenshot of Alpha_Block_Sequence above) and a key is pressed the next trial begins 'immediately'. However, when there are relatively more items as shown here:

there is a perceptible delay after a keypress before the next trial begins. Is there something I can do to avoid this happening?

Additionally, in the block below:

I have set each image to 1000ms (995 instructed) and to repeat each image 5 times randomly. The presentation order works fine but I notice a large variation in presentation time accuracy. Most are presented for approximately the requested 1000ms but random iterations of pictures are presented for longer than 1000ms. Is this likely to be an issue with my system?

Hope I've provided enough information/context on this, let me know if not.

Thanks for your time,

Boo.

eduard

Hi Boo,

Just wondering, how do you know that the presentation time is not accurate? Is it really that much off, or did you actually measure the presentation time? Not sure whether it helps, but instead of having 20+ sketchpads in a single loop, you can add another loop to that layer with a single sketchpad. This will at least release some of the memory necessary to store your experiment.

Does this make sense?

Eduard

booradley

Hi Eduard,

I haven't used any objective measure of the delays I described but they are well above perceptual threshold for detection.

I'm not sure I understand your solution to free up memory. Is it possible to have a text file with different text stimuli that can be called into a string variable that can be used in a single SketchPad?

Thanks for your help,

Boo.

EDIT: I guess I could use a Python script implementing lists or Panda dataframes. I'll look into that. Thanks for pointing me in the right direction!

eduard

Hi Boo,

I attach a little script that should illustrate what I meant earlier. Basically, the idea is to reduce the number of sketchpads that you have to a single one and define the content of that sketchpad dynamically across loop iterations. This works of course only if the sketchpads area conceptually all the same.