[open] Response time accuracy

panos_uol

Hi all!

I am doing my PhD in alcohol attentional bias, and I used OpenSesame to build a simple alcohol-stroop task. The essence of that task is to record response time for various stimuli (how quickly participants will click on the correct button according to the colour of the word on their screens).

My question would be, how accurate is OpenSesame in recording response time, because this is an essential information I have to report in my paper. Has there been any previous study reporting on that?

I use the .droid extension and the task will run in smartphones with android (clarifying that in case it plays a role).

It would very helpful if someone could point me to the right direction to find this piece of info. Many thanks for your time!!

Panos

sebastiaan

Hi Panos,

There's extensive information about timing available here:

• http://osdoc.cogsci.nl/miscellaneous/timing/

However, there's not much Android-specific information, because most benchmarks have been done on regular PCs. You can take a look at this post though:

• http://forum.cogsci.nl/index.php?p=/discussion/comment/4568#Comment_4568

In general, I would say that the temporal precision on any reasonably modern Android device will be fine for reaction-time experiments.

Cheers!
Sebastiaan

rip

In general, I would say that the temporal precision on any reasonably modern Android >device will be fine for reaction-time experiments.

This is not true. In fact even modern Android devices will give you unacceptable reaction times with large variability between devices.

For example:
http://bgr.com/2013/09/20/iphone-android-touch-screen-responsiveness/

Please do some more research on Android response times before you conduct your study. The key problem is really the variability between devices especially if you are conducting a remote study.

sebastiaan

@rip Thanks for posting this information, that's very useful. I pasted the image below for reference.

Assuming that these results are representative (there are many devices out there!), I would say that these benchmarks actually show that reaction times are acceptable, because, usually, what really matters is the variation of response times on a single device. And this appears to be in the order of 20 ms, which is not too different from what you would get on a regular keyboard.

The fact that absolute reaction times vary wildly between devices is less important, assuming that your measures are 'within device'. It's the same situation as for between-subject variance: The fact that reaction times vary a lot between subjects doesn't matter too much if you're doing a within-subject study.

panos_uol

Thank you both for your comments! The study is within subjects so indeed inter-device variation would not be an issue. My real concern is how close the reaction time open sesame will give me as a result is to the real reaction time of the person. As far as I understood open sesame can pretty accurately estimate people's reaction times.