Elsevier journal to retract widely debunked masks study whose author claimed a Stanford affiliation