It may also have been the type of computing that was going on. Pre-80s computing involved mainframes, data entry... secretarial tasks. Post-80s computing involved building businesses, getting heavily involved in the science, and generally hiding out in a dark room programming alone for weeks at a time.
What evidence is there for advertisement stereotyping being the causal factor? Wasn't advertisement stereotyping just as powerful in non-STEM fields like law? medicine? journalism? Why did they break the gender curse whereas STEM did not?
This reminds me of government R&D institutes in India. Where every one gets the title of a 'Scientist'. So there are levels like 'Scientist-1', 'Scientist-2', ...
In reality most of these levels have only clerical work to do. Sometimes things like writing down on a register that you issued somebody a Screw gauge. So basically book keeping, secretarial work, data entry etc etc.
But yeah, on the ground every gets to be called a 'Scientist'.
