How to detect dropped frame or lagging with movie stim in an online study

Online experiments
1

Hi all,

URL of experiment: no URL; a general question

Description of the problem:
I’ve a study where participants watch movies. Visual probes will sporadically appear on top of the movie stimuli. Participants are asked to press SPACE upon spotting the probes.

I noticed that with poor internet connection or slow computer, there seemed to have frame dropping or lagging issues with movie playing.

How to tell which frames are dropped or when a lagging happened? Also, will this problem affect rt accuracy? If a participant experienced this problem, should I discard part or all his/her data?

I saw a post on this subject but it seems to be for PsychoPy running locally. I’m not sure how to fix this problem when collecting data on Pavlovia.

Any suggestions are welcome! Many thanks in advance!

2

Until now, it was impossible to control the refresh rate because everyone had different computer devices

3

Hi fishpan,

From my understanding, as discussed in-depth on [this forum post] by Thomas Pronk, is that you cannot get the exact time the browser flips. Here, Thomas also discusses the implications for timing on RT, I highly recommend you read this post and their article (mental chronometry in the pocket).

I don’t know anything about movie presentations in Psychopy/JS (or at all tbh), however, my natural intuition is that if the movie is fully loaded by participants prior to the movie’s display, there should not be dropped frames. Are your video files quite big?

A random side-thought, have you explored whether the frame dropping is occurring on specific browsers? As not all browsers are equal, it maybe advantangeous to restrict and only allow participants to join on your preferred browser, if you have not done so already.

4

Hi Tim,

Thanks for your suggestions! I’ll read the post and the article.

I also feel that if the movies are loaded before experiment starts, people should not have lagging issues. But it happened during a pilot test when the participant’s laptop had several apps running at the same time. I will instruct participants to close other apps beforehand, but there is no guarantee that they will actually follow the instruction. That’s why I I asked for a solution :slight_smile:

The browser idea is interesting! I’ll look more into it.