search

kristabh / gaze-following-analysis

Multi-lab infant gaze following project
MIT License
2 stars 0 forks source link

Some reasons to wait on using handcoded data #40

Closed AlexisBlack2g closed 5 years ago

AlexisBlack2g commented 5 years ago

It does look like the data are different, likely a result of coding implementations in hand-coded data that do not match gazepath.

E.g. - look at the proportion of congruent first shift fixations:

FirstShiftProportions_CodingCheck

FirstLookProp_Coding.pdf

The frequency of looks to congruent as proportion of total shifts FreqProportion_Coding.pdf

AlexisBlack2g commented 5 years ago

whoops, didn't mean to make those diff. file types. pdfs are also images to check