it is interesting to investigate if different constraints indeed impacting the results of theta in table 1 and 2.
and note that why only the bias of slope parameter increased in table 2?
in real data, other models like irtcg hybrid and irtgpc should be compared as well.