Add column for fixed vs infant-controlled trial length?

[christinabergmann]

I think it's an interesting distinction that however might vary more across than within fields...

[juliacarbajal]

Good point, I think it's not a bad idea to add it. Quite related to this and to a previous issue on eye-tracking, I've been thinking that it would be important to make a distinction between online- and offline-coding.

[christinabergmann]

To update this, I propose the following: column name: trial_control levels: fixed infant_training (= infant-controlled in habituation/familiarization phase but not test) infant_test (= infant-controlled in test phase but not in habituation/familiarization phase) infant_all (all trials were infant-controlled, also when there was only a test-phase)

[shotsuji]

On this now with Cecile. Decisions made:

• we add this variable for all studies, not only hab/fam. Thus, e.g. FC would be "infant_all", looking while listening would be "fixed"
• for hab/fam studies: We only consider a study infant controlled if the trial stops after infant looks away for x sec (thus, we do not consider it infant controlled because the sound starts playing once the infant looks)
• The above point is not trivial, since also trial/sound START can vary in contingency. See e.g. Chambers, Onishi, Fisher (2011) from AGL_phonotactics, where fam phase would start non-contingent, but test phase would start contingently.