I analyzed a conjoint experiment with the cregg R package. I have 7 attributes. And I computed a subgroup analysis (3 subgroups). My dependent variable is a continuous variable (respondents had to rate two options in a non-forced conjoint experiment on a 1-7 scale). Now I get an estimated AMCE that is higher than 1 (1.1072640) for one attribute level. How is that possible? (I checked my data and there is nothing noticeable wrong).
```
amce_rat_sub_sample <- cregg::cj(na.omit(conjoint_master_all),
rating ~ side_effects + effective + country + type + people +
duration + cost, id = ~ID, estimate = "amce", by =~sample)
```
rating is a continuous variable. people rated A and B item on a 7-point scale side_effects, effective.... are factors (attributes) sample is a factor with 3 levels (3 countries)
In the following output (shortened) you can see that there is an estimate > 1. I this possible? I thought amces should be between -1 and 1.
