langcog / metalab-archive

meta-analyses of language acquisition phenomena
http://metalab.stanford.edu
10 stars 2 forks source link

Distinguish different eye movement measures in "response_mode" #115

Open christinabergmann opened 7 years ago

christinabergmann commented 7 years ago

Hi,

A source of repeated confusion is that we code hand-coded looking data and automated eye tracking the same way in "response_mode" (used to be eye-tracking, is now according to specs.yaml oculomotor [even though the tabular data still use the old coding] -> ???). I would suggest eye-tracking for automated measures and eye-movement for hand-coded.

Any objections? I know it would require some recoding, but that could be done based on "method".

mcfrank commented 7 years ago

I agree that it's confusing but I'm not sure that "tracking" and "movement" convey what you want...

Mike

On Tue, Feb 7, 2017 at 4:31 AM, Christina Bergmann <notifications@github.com

wrote:

Hi,

A source of repeated confusion is that we code hand-coded looking data and automated eye tracking the same way in "response_mode" (used to be eye-tracking, is now according to specs.yaml oculomotor [even though the tabular data still use the old coding] -> ???). I would suggest eye-tracking for automated measures and eye-movement for hand-coded.

Any objections? I know it would require some recoding, but that could be done based on "method".

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/langcog/metalab/issues/115, or mute the thread https://github.com/notifications/unsubscribe-auth/ABGul-hl4xRsMI0MBOBtA_tVI76ALMv6ks5raGQ-gaJpZM4L5bir .

christinabergmann commented 7 years ago

Sure, but in general the split might be useful. Wording suggestions welcome. I thought because eye-tracking is so closely linked with automated, it would be good to continue using this term for the respective subset and add a second descriptor. I am also curious how oculomotor happened.

christinabergmann commented 7 years ago

I just want to bring this issue back up, it would be great to somehow distinguish automatically tracked results from those based on hand-coding. Maybe even in a separate column that is optional.