Reaction time hassles

I’ve realised a little too late that the new version of Superlab (Superlab 4.0) by default measures all reaction times from the presentation of the first event in the trial, instead of re-setting between events. Anyone have any suggestions about how to retrieve or calculate the actual reaction time to the single event?

Not if you’ve already run the experiment. Any extrapolation from your existing data file would effectively be guesswork, as there’s the question of how long it took to present your stimuli.

If you haven’t yet run your experiment, however, you can reset the timer after an event by checking the box at the bottom of the Input tab in the event editor.

RT measures

I have also realised too late that Superlab measures by default RTs from the presentation of the first event of the trial. My data are already collected and I have to try to save it (it represents a lot of work). Why do you mean by “it’s a guesswork, as there is the question of how long it takes for the stimulus presentation”?

In my experiments, a trial includes : (1) a cross (time duration: 300ms- no response) (2) a black screen (500 ms-no response) (3) the stimulus A that requires a response (free time to response) (4) a black screen (500 ms-no response) (5) the stimulus B that requires a response (free time to response) (5) black screen (1500 ms-no response)

I wanted to calculate the RT to the stimulus B by substracting the 500 ms (black screen, the fourth event) and the reaction time recorded from the stimulus A from the reaction time from this stimulus B.

RT (stimulus b)= RT (stimulus b) - 500 ms (black screen) - RT(stimulus a)

Is there any problem with that? Is there a link with the refresch cycle of the screen? If this is the case, what is approximativaly the range of the error measure? Does the timer measure RT from the first event of a trial or from the first event of the trial that required a response?

For information, stimuli are jpeg image (black ground with two white symbols). I have used a iBook G4 and a MAc Book iC2duo in one experiment. In a second one, I have connected a LCD monitor to the iBook G4.

]I have also realised too late that Superlab measures by default RTs from the presentation of the first event of the trial. My data are already collected and I have to try to save it (it represents a lot of work). Why do you mean by “it’s a guesswork, as there is the question of how long it takes for the stimulus presentation”?

In my experiments, a trial includes : (1) a cross (time duration: 300ms- no response) (2) a black screen (500 ms-no response) (3) the stimulus A that requires a response (free time to response) (4) a black screen (500 ms-no response) (5) the stimulus B that requires a response (free time to response) (5) black screen (1500 ms-no response)

I wanted to calculate the RT to the stimulus B by substracting the 500 ms (black screen, the fourth event) and the reaction time recorded from the stimulus A from the reaction time from this stimulus B.

RT (stimulus b)= RT (stimulus b) - 500 ms (black screen) - RT(stimulus a)

Is there any problem with that? Is there a link with the refresch cycle of the screen? If this is the case, what is approximativaly the range of the error measure? Does the timer measure RT from the first event of a trial or from the first event of the trial that required a response?

For information, stimuli are jpeg image (black ground with two white symbols). I have used a iBook G4 and a MAc Book iC2duo in one experiment. In a second one, I have connected a LCD monitor to the iBook G4.

Gaelle,

The difficulty in calculating RT(stimulus b) is that we don’t know exactly how long it took to draw the black screen or to draw stimulus b. I would like to think that it’s the same every time, but unfortunately, it’s never quite that simple. Whenever possible, redrawing the screen is synchronized with the screen’s internal refresh. So far, I haven’t been able to retrieve this information on any of Apple’s LCDs. On 10.4, the image itself is still synchronized with the refresh by the OS, but without access to the information, the actual onset time can’t be interpreted by SuperLab.

If you’re up for an overwhelming abundance of information (and are running at least SuperLab 4.0.3), I’ve attached a little helper application to launch SuperLab in verbose mode on a Mac. This was created with an open source prgram called Platypus, and is effectively just a wrapper around a script that launches SuperLab with the “–verbose” flag (SuperLab must be installed at /Applications/SuperLab 4.app in order for this to work). When launched this way, SuperLab will give you extended details about what happened when you ran your experiment with timestamps printed in milliseconds. This is currently a work in progress and is an obscure feature by design. I tried to make the output as human-readable as possible, but there are pieces of information that it spits out that have no written explanation anywhere. :wink:

Basically, you’ll need to rerun a small portion of your experiment on the original computer used, and then you can use this data to interpret what the additional delays probably looked like.

I hope this helps a little.

SuperLabVerbose.app.zip (50.6 KB)

log both event and trial RT?

Hi, I’m having a very similar problem to this one, but I’m only about half done runningmy subjects. Is there a way to log both the RT at the trial level and the RT at the event level for the rest of my subjects? That way I can compare the two times with the subjects I already have data for and decide whether I want to try to repair the first group’s trial-only RT measurements or try to recruit (and pay) replacements.

Thanks for any help.

Also, I tried this playtpus script and it doesn’t seem to work for me – it launches SuperLab and everything, but a dialog comes up complaining “catalog file for domain ‘superlab’ not found”. Is this because I don’t have write access to /Applications or something?

Catalog files

Could this be related to the Leopard download quarantine?

No, the warning that comes up is about translation catalogues that can’t be found. This is normal. They’re never found, but the message only comes up in verbose mode. It’s not an error, and you can safely ignore it.

If SuperLab launched and you got that warning, then it’s working correctly, actually. :slight_smile: