Hello @ all FIMIX- experts,
I have identified in my model 2 segments (taking the best informations criteria and one of the best EN and taking into consideration the sense of the model), but when I look at variance in the segments it is lower than in the aggregated model (R2 in Segment1: 0,171 with Size1=0,630, R2 in Segment2: 0,696 with Size2=0,370 so 0,171*0,630+0,696*0,370= 0,365 and R2 in the aggregated model: 0,467)
Is it possible, or have I done something wrong by choosing the number of segments? Does it make sense to consider these 2 segments or should I better analyze the aggregated model?
I would appreciate if somebody could answer me. Thanks a lot in advance!
Best regards,
Joanna Ozga