Why Would You Suggest That? Human Trust in Language Model Responses