0

I want to investigate if there is an interaction between "Group" and "Type" in my data. Type consists of "verbs" and "circumstantial information". "QCorrect" is performance.

Why does the model split Type up into verbs and circum (it is not supposed to do that)? Furthermore, what is the intercept? 

I1 = lmer(Qcorrect ~ Group * Type + (1|ID), data = Sdata1) 
summary(I1)

#Fixed effects:
#                Estimate Std. Error         df t value Pr(>|t|)    
#(Intercept)     8.095e-01  4.488e-02  4.140e+01  18.038   <2e-16 ***
#Group          -2.381e-02  6.347e-02  4.140e+01  -0.375    0.709    
#TypeVerb        1.190e-02  6.149e-02  3.200e+02   0.194    0.847    
#Group:TypeVerb -1.057e-17  8.696e-02  3.200e+02   0.000    1.000
  • It isn't splitting it into two, it is telling you that the variable is equal to 1 (TRUE) when the type is verbs and 0 (FALSE) when the type is circumstantial. What the intercept is depends on what your `Group` variable represents. – tbradley Dec 15 '17 at 18:34
  • 1
    For more threads about these issues, please see https://stats.stackexchange.com/search?q=+categorical+regression+intercept+%5BR%5D+answers%3A1. – whuber Dec 15 '17 at 21:49

0 Answers0