Change the instruction to demand the participant to foveate the central fixation during the task

[lib314a]

I've made sure that I am using the latest version of the image from https://quay.io/vanessa/expfactory-builder.

The demand to keep fixation on the central cross (e.g. "You must fixate the central cross throughout the task") is missing in the instructions of the current task. However, previous studies, as those I read, all included this statement or the like, such as Fan et al. 2002:

Participants were instructed to focus on a centrally located fixation cross throughout the task, and to respond as quickly and accurately as possible

and Adolfsdottir et al. 2008:

... (the children) were told that sometimes the fish would appear alone, and other times it would swim together with other fishes. In all cases, they were told to concentrate on the fish in the middle...

etc.

[earcanal]

Just adding quote from Fan et al. (2002) on this:

"[However,] we did run a few subjects either under instruction to avoid eye movements, as in the full study described in this
paper, as well as with no instruction to maintain fixation. Under instruction to maintain fixation, eye movements were relatively rare (about 3–5% of all trials). When no instruction was given, eye movements were very common. RT appeared to be faster in eye-movement trials, however, the basic effects of cues and target looked very similar. The interactions found in the full study were probably not due to eye movements, but are very likely influenced a great deal by the acuity demands of the task."

[vsoch]

How can you be sure the instruction wasn’t verbally given?

[lib314a]

How can you be sure the instruction wasn’t verbally given?

I can't be sure and am totally fine with verbal instructions. However, if the aim of expfactory is to "provide open source behavioral experiments for reproducible science", leaving an critical instruction to the experimenters may not be the best idea

[vsoch]

The issue with hard coding it is that not all users might want to use it. The experiments should be reproducible, but I think this is advertised as a general network attention task and now exactly the one from that paper. How about adding this detail to the README? I’d be open and appreciate a PR for that if you’d like to contribute!

I did not design this original experiment, it was likely one of my old lab mates, and I suspect he chose to not include this instruction for a reason. You can try reaching out to him if you like, Ian Eisenberg.

[vsoch]

I would also argue that if a verbal instruction was given by changing the experiment and having it be a non-verbal instruction it would actually be making it more different than the original. The important point would be to make sure that documentation is provided alongside the experiment to explicitly state that the verbal instruction should be provided

[vsoch]

If you do want these instructions for your experiment, here is an example of how to do that: https://github.com/expfactory-experiments/breath-counting-task/blob/69bcaf412e75cedc62d782d504d578f5a0620129/experiment.js#L124

[earcanal]

I used participant variables to create a configurable ANT. You could do the same to have different instructions.

More generally, I think the way forward is to have configurable paradigms (ANT, Stroop etc.) so that the common variants get added without having to modify the core. Lots of forks for variants of a task is bad for replication.

Also, it's great that the conversations have reached this point!