|
|
I need to fit categorization curves for the categorization of an 8-level
continuum of Dutch vowels /I/ to /i/ in each position and for each
participant in a logistic regression model, in order to obtain the 50%
cross-over points and to investigate any interactions between the F1 and
vowel duration and native language (Dutch and Spanish). Is there any way I
can do this in SPSS? I have tried to carry out a binary logistic regression
analysis, but I don't know how to find the 50% percent cross-over point, and
further, how to carry out logistic regression for each position and each
participant separately. Hopefully someone can help me out. Thanks in
advance!
Details:
F1 continuum: 8 levels.
F2: 3 levels.
Repetitions: 10.
Participants: 16 Dutch + 16 Spanish.
Method: Logistic regression.
Response: Binary (/I/ versus /i/).
|
Locked
